Fuel Injection

Fuel Injection
edited by
Daniela Siano
SCIYO
Fuel Injection
Edited by Daniela Siano
Published by Sciyo
Janeza Trdine 9, 51000 Rijeka, Croatia
Copyright 2010 Sciyo
All chapters are Open Access articles distributed under the Creative Commons Non Commercial Share
Alike Attribution 3.0 license, which permits to copy, distribute, transmit, and adapt the work in any
medium, so long as the original work is properly cited. After this work has been published by Sciyo,
authors have the right to republish it, in whole or part, in any publication of which they are the author,
and to make other personal use of the work. Any republication, referencing or personal use of the work
must explicitly identify the original source.
Statements and opinions expressed in the chapters are these of the individual contributors and
not necessarily those of the editors or publisher. No responsibility is accepted for the accuracy of
information contained in the published articles. The publisher assumes no responsibility for any
damage or injury to persons or property arising out of the use of any materials, instructions, methods
or ideas contained in the book.

Publishing Process Manager Ana Nikolic
Technical Editor Goran Bajac
Cover Designer Martina Sirotic
Image Copyright Megastocker, 2010. Used under license from Shutterstock.com
First published September 2010
Printed in India
A free online edition of this book is available at www.sciyo.com
Additional hard copies can be obtained from publication@sciyo.com
Fuel Injection, Edited by Daniela Siano
p. cm.
ISBN 978-953-307-116-9
SCIYO.COM
WHERE KNOWLEDGE IS FREE
free online editions of Sciyo
Books, Journals and Videos can
be found at www.sciyo.com
Chapter 1
Chapter 2
Chapter 3
Chapter 4
Chapter 5
Chapter 6
Chapter 7
Chapter 8
Chapter 9
Chapter 10
Preface VII
Gasoline direct injection 1
Mustafa Bahattin elik and Blent zdalyan
Liquid Sprays Characteristics in Diesel Engines 19
Simn Martnez-Martnez, Fausto A. Snchez-Cruz,
Vicente R. Bermdez and Jos M. Riesco-vila
Experimental Cells for Diesel Spray Research 49
Simn Martnez-Martnez, Miguel Garca Yera and Vicente R. Bermdez
Experimental study of spray generated by
a new type of injector with rotary swinging needle 65
Hubert Kuszewski and Kazimierz Lejda
Effect of injector nozzle holes on diesel engine performance 83
Semin and Abdul Rahim Ismail
Accurate modelling of an injector for common rail systems 95
Claudio Dongiovanni and Marco Coppo
The investigation of the mixture formation
upon fuel injection into high-temperature gas ows 121
Anna Maiorova, Aleksandr Sviridenkov and Valentin Tretyakov
Integrated numerical procedures for the design,
analysis and optimization of diesel engines 143
Daniela Siano, Fabio Bozza and Michela Costa
Hydrogen fuelled scramjet combustor - the impact of fuel injection 167
Wei Huang, Zhen-guo Wang, Mohamed Pourkashanian, Lin Ma, Derek
B.Ingham, Shi-bin Luo and Jun Liu
Plasma ame sustained by microwave
and burning hydrocarbon fuel: Its applications 183
Yongcheol Hong and Han Sup Uhm
Contents
V
Chapter 11
Chapter 12
The blast furnace trazability by helium 211
Rafael Barea, Ramn Martn D, I. Ruiz Bustinza and Javier Mochn
Experimental investigations into the production behavior
of methane hydrate in porous sediment under ethylene
glycol injection and hot brine stimulation 227
Xiao-Sen Li and Gang Li
IueI Injeclion is a key process characlerising lhe conluslion deveIopnenl vilhin Spark-
Ignilion (SI) and Conpression Ignilion (CI) InlernaI Conluslion Lngines (ICLs). IueI Injeclion
and Spray ehaviour in facl IargeIy conlroI lhe fueI-air nixing, conluslion process efhciency,
slaliIily, lhe produclion of noxious species, lhe radialed noise, elc.The proper design of lhe
fueI injeclion syslen requires lhe enpIoynenl of lolh experinenlaI and nunericaI lechniques,
sonelines coupIed for oplinisalion procedures.
Research and deveIopnenl of lhe fueI injeclion syslen is nol Iiniled lo ICLs. A proper design
of lhis syslen is required in nany induslriaI appIicalions, invoIving differenl ruIes and
requiring very differenl design largels.
The chaplers in lhis look ain lo presenl lhe slale of lhe arl of lhe experinenlaI and nunericaI
nelhodoIogies appIied lo deepen lhe underslanding of fueI injeclion syslen lehaviour,
for lolh gasoIine and dieseI engines. Chapler 1 descriles lhe polenliaI of a CasoIine
Direcl Injeclion (CDI) for a SI-ICL, vhiIe chaplers 2 lo 4 are devoled lo lhe presenlalion of
experinenlaI anaIyses on spray lehaviour in a dieseI engine. Chaplers 5 lo 7 are indeed
focused on lhe nodeIIing of lhe fueI injeclion syslen, and anaIyse ils inpacl on engine
perfornance, vhiIe chapler 8 puls logelher experinenlaI and nunericaI lechniques for an
overaII syslen oplinisalion under lhe poinl of viev of lolh engine perfornance, noxious
enission and radialed noise.
Chaplers 9 lo 12 focus on non-engine appIicalions and give an oulIook of lhe differenl
requirenenls lhal a nodeI fueI injeclion syslen needs lo ensure in various induslriaI
appIicalions.
Ldilor
DanIc!a 5Iann
|ns|i|u|c Mc|cri - CNR,
||a|q
Preface
X

GasoIine direct injection

Muslafa ahallin eIik
*
and Ienl zdaIyan
**

Kara|u| Unitcrsi|q, |nginccring |acu||q
Kara|u| Unitcrsi|q, Tccnnc|cgq |acu||q
Tur|cq

1. Introduction
The lasic goaIs of lhe aulonolive induslry, a high pover, Iov specific fueI consunplion,
Iov enissions, Iov noise and leller drive conforl. Wilh increasing lhe vehicIe nunler, lhe
roIe of lhe vehicIes in air poIIulion has leen increasing significanlIy day ly day. The
environnenl proleclion agencies have dravn dovn lhe enission Iinils annuaIIy.
Iurlhernore, conlinuousIy increasing price of lhe fueI necessilales inproving lhe engine
efficiency. Since lhe engines vilh carlurelor do nol hoId lhe air fueI ralio cIose lo lhe
sloichionelric al differenl vorking condilions, calaIylic converler cannol le used in lhese
engines. Therefore lhese engines have high enission vaIues and Iov efficiency. LIeclronic
conlroIIed Iorl IueI Injeclion (III) syslens inslead of fueI syslen vilh carlurelor have leen
used since 198Os. In fueI injeclion syslens, induced air can le nelered preciseIy and lhe
fueI is injecled in lhe nanifoId lo air anounl. y using lhe Ianlda sensor in exhausl syslen,
air/fueI ralio is heId of slalIe vaIue. IueI syslens vilhoul eIeclronic conlroIIed il is
inpossilIe lo conpIy vilh lhe increasingIy enissions IegisIalion.

If porl fueI injeclion syslen is conpared vilh carlurelor syslen, il is seen lhal has sone
advanlages. These are,

1. Lover exhausl enissions.
2. Increased voIunelric efficiency and lherefore increased oulpul pover and lorque.
The carlurelor venluri prevenls air and, in lurn, voIunelric efficiency decrease.
3. Lov specific fueI consunplion. In lhe engine vilh carlurelor, fueI cannol le
deIivered lhe sane anounl and lhe sane air/fueI ralio per cycIe, for each cyIinder.
4. The nore rapid engine response lo changes in lhrollIe posilion. This increases lhe
drive conforl.
5. Ior Iess rolalion conponenls in fueI injeclion syslen, lhe noise decreases
(Heyvood, 2OOO, Ierguson, 1986).

Though lhe porl fueI injeclion syslen has sone advanlages, il cannol le neel conlinuousIy
increased lhe denands aloul perfornance, enission IegisIalion and fueI econony, al lhe
presenl day (Slone, 1999). The eIeclronic conlroIIed gasoIine direcl injeclion syslens vere
slarled lo le used inslead of porl fueI injeclion syslen since 199Os.
1
Fuel njection 2

The CasoIine Direcl Injeclion (CDI) engines give a nunler of fealures, vhich couId nol le
reaIized vilh porl injecled engines: avoiding fueI vaII fiIn in lhe nanifoId, inproved
accuracy of air/fueI ralio during dynanics, reducing lhrollIing Iosses of lhe gas exchange ly
slralified and honogeneous Iean operalion, higher lhernaI efficiency ly slralified operalion
and increased conpression ralio, decreasing lhe fueI consunplion and CO
2
enissions, Iover
heal Iosses, fasl healing of lhe calaIysl ly injeclion during lhe gas expansion phase,
increased perfornance and voIunelric efficiency due lo cooIing of air charge, leller coId-
slarl perfornance and leller lhe drive conforl (Zhao el aI., 1999, KaranangiI, 2OO4, Snilh el
aI., 2OO6).

2. The Performance and Exhaust Emissions of The GasoIine Direct Injection
(GDI) Engine
2.1 Performance of the GDI Engine
The paranelers lhal have lhe grealesl infIuence on engine efficiency are conpression ralio
and air/fueI ralio. The effecl of raising conpression ralio is lo increase lhe pover oulpul
and lo reduce lhe fueI consunplion. The naxinun efficiency (or nininun specific fueI
consunplion) occurs vilh a nixlure lhal is veaker lhan sloichionelric (eIik, 2OO7).
ecause lhe porl fueI injeclion engines vork al sloichionelric air/fueI ralio, il is inpossilIe
lo see nore inprovenenl in lhe fueI econony. In lhese engines, lhe conpression ralio is
aloul 9/1-1O/1. To prevenl lhe knock, lhe conpression ralio cannol le increased nore. Ior
lhe sane engine voIune, lhe increasing voIunelric efficiency aIso raises lhe engine pover
oulpul.

CDI engine operale vilh Iean nixlure and unlhrollIed al parl Ioads, lhis operalion provide
significanlIy inprovenenls in fueI econony. Al fuII Ioad, as lhe CDI engine operales vilh
honogeneous charge and sloichionelric or sIighlIy rich nixlure, lhis engine gives a leller
pover oulpul (Spicher el aI., 2OOO). In CDI engine, fueI is injecled inlo cyIinder lefore spark
pIug igniles al Iov and nediun Ioads. Al lhis condilion, Air/IueI (A/I) ralio in cyIinder
vary, lhal is, nixlure in fronl of spark pIug is rich, in olher pIaces is Iean. In aII cyIinder A/I
ralio is Iean and A/I ralio can access unliI 4O/1. In honogeneous operalion, fueI slarls
injecling inlo cyIinder al inlake slroke al fuII Ioads (AIger el aI., 2OOO, inar, 2OO1). The fueI,
vhich is injecled in lhe inlake sloke, evaporales in lhe cyIinder. The evaporalion of lhe fueI
cooIs lhe inlake charge. The cooIing effecl permits higher conpression ralios and increasing
of lhe voIunelric efficiency and lhus higher lorque is ollained (Muoz el aI., 2OO5). In lhe
CDI engines, conpression ralio can gain unliI 12/1 (Kune, 1996). The knock does nol occur
lecause onIy air is conpressed al Iov and nediun Ioads. Al fuII Ioad, since fueI is injecled
inlo cyIinder, lhe charge air cooI and lhis, in lurn, decreases knock lendency.

Since lhe vehicIes are used usuaIIy in urlan lraffic, sludies on inproving lhe urlan driving
fueI econony have increased. Lngines have run usuaIIy al parl Ioads (Iov and nediun
Ioads) in urlan driving. VoIunelric efficiency is Iover al parl Ioads, so engine effeclive
conpression ralio decreases (e.g. fron 8/1 lo 3/1-4/1), engine efficiency decreases and fueI
consunplion increases. The urlan driving fueI econony of lhe vehicIes is very high (eIik,
1999). Dislinclion lelveen lhe highvay fueI econonies of vehicIes is very IillIe. As najorily

of lhe Iife line of lhe vehicIes pass in lhe urlan driving, lhe ovners of lhe vehicIes prefer lhe
vehicIes of vhich lhe urlan driving fueI econony is Iov.

Al fuII Ioad, as lhe CDI engine operale vilh lhrollIe, onIy a snaII reduclion of fueI
consunplion can le ollained lo lhe III engine. There is lhe nore fueI econony polenliaI al
parl Ioad. Al conpression slroke, since air is given lhe cyIinders vilhoul lhrollIe for
slralified charge node, punping Iosses of lhe CDI engine is nininun al parl Ioads, Iig.1
(aungarlen, 2OO6). The inprovenenls in lhernaI efficiency have leen ollained as a resuIl
of reduced punping Iosses, higher conpression ralios and furlher exlension of lhe Iean
operaling Iinil under slralified conluslion condilions al Iov engine Ioads. In lhe DI
gasoIine engines, fueI consunplion can le decreased ly up lo 2O, and a 1O pover oulpul
inprovenenl can le achieved over lradilionaI III engines (Ian el aI., 1999).

Iig. 1. Reduclion of lhrollIe Iosses in lhe slralified-charge conluslion (aungarlen, 2OO6).

The CO
2
enissions, vhich are one of lhe gases, lring aloul lhe gIolaI varning. To decrease
CO
2
enilled fron vehicIes, il is required lo decrease fueI consunplion. Dovnsizing
(reduclion of lhe engine size) is seen as a najor vay of inproving fueI consunplion and
reducing greenhouse enissions of spark igniled engines. In lhe sane veighl and size,
significanl decreases in CO
2
enissions, nore pover and higher lreak nean effeclive
pressure can le ollained. CDI engines are very suilalIe for lurlocharger appIicalions. The
use of CDI engine vilh lurlocharger provides aIso high engine knock resislance especiaIIy
al high Ioad and Iov engine speed vhere III lurlocharged engines are sliII Iiniled
(Lecoinle & Monnier, 2OO3, SloffeIs, 2OO5). Turlocharged CDI engines have shoved greal
polenliaI lo neel lhe conlradiclory largels of Iover fueI consunplion as veII as high lorque
and pover oulpul (KIeelerg, 2OO6).

Fuel njection 4

In CDI engine, ly using lvin charging syslen drive conforl, engine lorque and pover can
le increased for lhe sane engine size. Ior exanpIe, VoIksvagen (VW) has used lhe duaI
charging syslen in TSI (lvin charged slralified injeclion) engine. The syslen incIudes a
rools-lype supercharger as veII as a lurlocharger. The supercharger is lasicaIIy an air
conpressor. A nechanicaI device driven off lhe engine's crankshafl, il enpIoys rolaling
vanes vhich spin in opposile direclions lo conpress air in lhe engine's inlake syslen. The
high and conslanl lorque is ollained al vide range speed ly aclivale supercharger al Iov
speeds and lurlo charger al high speeds (Anon, 2OO6).

In TalIe 1, il is given specificalions of lhe lvo differenl engines leIonging lo lhe 2OO9 nodeI
VW Iassal vehicIe, for exanpIe. TSI engine urlan driving fueI econony is 18 Iover lhan
lhal of III engine. CO
2
enission is 12 Iover lhan lhal of III engine. AIlhough TSI engine
svepl voIune is Iover lhan III engine, pover and lorque is higher ly 2O and 35,
respecliveIy (TalIe 1). As engine lorque is naxinun al inlervaI 15OO-4OOO 1/nin, shifling is
nol necessary al lhe acceIeralion and lhus drive conforl increase (Anon, 2OO9).

Lngine
Type
Svepl
voIune
Max. Iover
Max.
Torque
Mixlure
fornalion
syslen
IueI econony
(urlan driving)
L/1OOkn
IueI econony
(highvay
driving)
L/1OOkn
CO2
enission
g/kn
CasoIine
engine
1.6 L
75 kW
56OO 1/nin
148 Nn
38OO 1/nin
III (porl fueI
injeclion)
1O,5 6,O 179
TSI
gasoIine
engine
1.4 L
9O kW
5OOO - 55OO
1/nin
2OO Nn
15OO - 4OOO
1/nin
CDI (CasoIine
direcl injeclion)
8,6 5,5 157
TalIe 1. Conparison of lhe CDI and III engines (Anon, 2OO9).

2.2 Exhaust Emissions of the GDI Engine
CO enission is very Iov in CDI engine. CO varies depending on air /fueI ralio. CO is high
al rich nixlures. Since CDI engines operale vilh Iean nixlure al parl Ioads and
sloichionelric nixlure al fuII Ioad, CO is nol a prolIen for lhese engines. In CDI engine,
due lo lhe velling of lhe pislon and lhe cyIinder vaIIs vilh Iiquid fueI, HC enission can
increase. Hydrocarlon (HC) enissions are a funclion of engine lenperalure and, lherefore il
can rise during coId slarl. The coId slarls characlerislics vary depending on lhe fueI
dislrilulion characlerislics, lhe in-cyIinder air nolion, fueI vaporizalion, and fueI-air nixing
(Candhi el aI., 2OO6).

During coId-slarl of a CDI engine, honogeneous operalion can le enpIoyed due lo a higher
exhausl gas lenperalure resuIling in a shorler line for calaIysl Iighl-off, and Iover engine
oul HC enissions (Candhi el aI., 2OO6). CasoIine engines do nol enil sool enission
nornaIIy. Sool enission can occur al very rich nixlures. Hovever, lhe CDI engines enil
sool al slralified-charge operalion, as in-cyIinder can le areas vilh very rich nixlures. In
addilion, in CDI engine, if nixlure fornalion do nol reaIize al fuII Ioads due lo rich nixlure,
lhe sool enission can increase. NOx enission is naxinun al high cyIinder lenperalures
and al =1.1. As lorque oulpul rises, lenperalures rise and, in lurn, lhe engine-oul NOx
enissions dispIay an increase. NOx enissions increase especiaIIy al fuII Ioad.


2.3 The Emission ControI in GDI Engine
LnvironnenlaI IegisIalion delernines lhe Iinils for exhausl enissions in lhe spark ignilion
engines. Il is required lhe lrealnenl of lhe exhausl gases lo neel lhese Iinils. The lhree-vay
calaIylic converler shov high perfornance for converling lhe CO, HC and NOx in lhe
engines vilh operalion al =1.O. ul, NOx cannol le conpIeleIy converled harnIess gases al
Iean nixlure operalion. Therefore, engines vilh Iean nixlure aIso require a NOx slorage
lype calaIylic converler lo converl lhe NOx.

The lvo calaIylic converlers are successiveIy used in CDI engine exhausl syslen. The one is
Ire-calaIylic converler (Three Way Converler -TWC). This converler has IillIe voIune and is
connecled cIose lo lhe engine. The olher is nain calaIylic converler vhich conlines a NOx
calaIysl and a TWC. This converler has higher voIune lhan lhe pre-calaIylic converler and
is connecled nol cIose lo lhe engine. The Ire-calaIylic converler converl lhe CO, HC and
NOx lo harnIess gases (CO
2
, H
2
O and N
2
) al =1.O. Hovever, vhen engine operales al
slralified node vilh Iean nixlure, NOx cannol le converled lo nilrogen. In such cases, NOx
is senl lo nain calaIylic converler (Anon, 2OO2).

In lhe NOx slorage lype calaIylic converler, lhe conponenls such as a and Ca are used for
NOx conversion al Iean nixlures. These conponenls provide NOx lo slorage. Al =1.O, lhe
operalion of lhe NOx converler resenlIes lhree vay converler. Al Iean nixlures, NOx
conversion is reaIized in lhree slages: NOx accunuIalion, NOx reIease and conversion.
Nilrogen oxides reacls chenicaIIy vilh lariun oxide (aO) and lhus lariun nilrale
(a(NO
3
)
2
forns. (NOx slorage slage). Then, lo converl, engine is operaled nonenlariIy in
lhe rich honogeneous node. Thanks lo rich nixlure, lhere is CO in exhausl syslen. The
lariun nilrale reacls chenicaIIy vilh CO and, as a resuIl of lhis CO
2
, aO and NO arise
(NOx reIease slage). And lhen, NO reacls chenicaIIy vilh CO and, N
2
and CO
2
forn
(conversion slage). NOx slorage converler can slorage lhe NOx al lenperalures of 25O-5OO`C
(Anon, 2OO2, auer, 2OO4). An exhausl gas recircuIalion syslen is necessary, as lhe NOx
aflerlrealnenl syslens do nol reach lhe conversion rales of = 1 concepls. Wilh lhe
exceplion al lhe highesl Ioads, exhausl gas recircuIalion (LCR) is used exlensiveIy lo conlroI
NOx enissions (AIkidas, 2OO7).

To neel lhe vaIid enission Iinils and diagnose lhe pre and nain calaIysl fauIls, and
provide oplinun engine operalion 4 sensors (3 Ianlda sensor and 1 exhausl gas
lenperalure sensor) are used in lhe exhausl syslen. The vide land Ianlda sensor
upslrean of pre-calaIysl delernines residuaI oxygen vaIue in exhausl gas. The required
for honogeneous Iean operalion can le conlroIIed ly lhis sensor. Ior each calaIylic
converler lvo Ianlda sensors (upslrean and dovnslrean sensor) are used. The fauIls of lhe
pre and nain converlers can le diagnosed ly signaI of duaI sensors. The lenperalure sensor
is used lo delernine lhe lenperalure of lhe NOx calaIysl (KseII el aI., 1999).

Fuel njection 6

3. The Mixture Formation and Operation Modes in The GDI Engine
3.1 The Mixture Formation
The air-fueI nixlure in lhe gasoIine engines is prepared in-cyIinder and oul-cyIinder. WhiIe
lhe nixlure in lhe engine vilh carlurelor and porl fueI injeclion is prepared oul-cyIinder,
nixlure in lhe gasoIine direcl injeclion engines is prepared in-cyIinder, Iigure 2.

Iig. 2. The nixlure fornalion syslens in lhe gasoIine engines.

In pIace of III engines vhere lhe fueI is injecled lhrough lhe porl, in CDI engines, lhe fueI is
injecled direclIy inlo cyIinders al a high pressure. During lhe induclion slroke, onIy lhe air
fIovs fron lhe open inlake vaIve and il enlers inlo lhe cyIinder. This ensures leller conlroI
of lhe injeclion process and parlicuIarIy provides lhe injeclion of fueI Iale during lhe
conpression slroke, vhen lhe inlake vaIves are cIosed (Sercey el aI., 2OO5). The acling of lhe
inlake syslen as a pre-vaporizing chanler is an advanlage in lhe III engines (Rolondi,
2OO6). As lhe Iack of line lo fueI vaporize in CDI engines, lhe fueI is injecled inlo lhe
cyIinder al a very high pressure lo heIp lhe alonizalion and vaporizalion process. The
duralion for injeclion lining is IillIe, advanced injeclion lining causes pislon velling and
relarded injeclion lining decrease sufficienl line for fueI-air nixing (Candhi el aI., 2OO6). In
lhe III engine, a Iiquid fiIn is forned in lhe inlake vaIve area of lhe porl, vhich causes
deIayed fueI vaporizalion. LspeciaIIy during coId slarl, il is necessary lo increase fueI
anounl for lhe ideaI sloichionelric nixlure. This overfueIing Ieads lo increasing HC
enissions during coId slarl. AIlernaliveIy, injecling lhe fueI direclIy inlo lhe conluslion
chanler avoids lhe prolIens such as increasing HC and giving lhe excess fueI lo engine
(HenlscheI, 2OOO).

To lhe CDI engines, il is inpIenenled lhe lvo lasic charge nodes, slralified and
honogeneous charge. Al lhe parliaI Ioad condilions, slralified charge (Iale injeclion) is used,
lhal is, fueI is injecled during lhe conpression slroke lo suppIy lhe slralified charge. The
engine can le operaled al an air-fueI ralio exceeding 1OO and fuIIy unlhrollIed operalion is
possilIe, lul lhe engine is lhrollIed sIighlIy in lhis zone and lhe air-fueI ralio is conlroIIed lo
range fron 3O lo 4O in order lo inlroduce a Iarge quanlily of Lxhausl Cas RecircuIalion
(LCR) and lo suppIy lhe vacuun for lhe lrake syslen. A honogeneous charge (earIy
injeclion) is preferred for lhe higher Ioad condilions, lhal is, fueI is injecled during lhe inlake

slroke so as lo provide a honogeneous nixlure. In nosl of lhis node, lhe engine is operaled
under sloichionelric or a sIighlIy rich condilion al fuII Ioad. In lhe Iovesl Ioad condilions in
lhis node, lhe engine is operaled al honogeneous Iean condilions vilh a air-fueI ralio of
fron 2O lo 25 for furlher inprovenenl of fueI econony (Kune, 1996). During operalion vilh
honogeneous charge lhe adjuslnenl of engine Ioad is done ly lhrollIing vhiIe during
operalion vilh slralified charge lhe engine runs vilh unlhrollIed condilions and engine Ioad
is adjusled ly fueI/air-equivaIence ralio (Spicher el aI., 2OOO). Iig.3 shovs lhe honogeneous
(earIy injeclion) and slralified-charge nodes (Iale injeclion).

Iig. 3. Honogeneous and slralified-charge node.

In lhe slralified operalion, lhree conluslion syslens are used lo forn an ignilalIe nixlure
near spark pIug al lhe inslanl ignilion. These are lhe vaII-guided, air-guided and spray-guided
conluslion syslens, Iig. 4. The dislinclion lelveen lhe differenl concepls is lhe used nelhod
vilh vhich lhe fueI spray is lransporled near lhe spark pIug (Orlnann el aI., 2OO1).

Iig. 4. The vaII-guided, air-guided and spray-guided conluslion syslens al slralified
charge (Slefan, 2OO4).
Fuel njection 8

Wa!!-GuIdcd cnmbustInn systcm: The fueI is lransporled lo lhe spark pIug ly using a
speciaIIy shaped pislon surface. As lhe fueI is injecled on lhe pislon surface, il cannol
conpIeleIy evaporale and, in lurn, HC and CO enissions, and fueI consunplion increase.
To use lhis syslen aIone is nol efficienl.

AIr-GuIdcd cnmbustInn systcm: The fueI is injecled inlo air fIov, vhich noves lhe fueI
spray near lhe spark pIug. The air fIov is ollained ly inIel porls vilh speciaI shape and air
speed is conlroIIed vilh air laffIes in lhe nanifoId. In lhis lechnique, fueI does nol vel lhe
pislon and cyIinder. Mosl of slralified-charge CDI engines use a Iarge-scaIe air nolion (svirI
or lunlIe) as veII as speciaIIy shaped pislon a surface in order lo keep lhe fueI spray
conpacl and lo nove il lo lhe spark pIug (aungarlen, 2OO6). In lhe air-guided and vaII-
guided conluslion syslens lhe injeclor is pIaced renole lo lhe spark pIug.

VW direcl injeclion conluslion syslen is a conlinalion of lvo syslens- vaII guided and
air guided -ly lunlIe fIov. This syslen is Iess sensilive againsl lhe cycIic varialions of
airfIov. This conluslion syslen shovs advanlages as veII in lhe slralified and in lhe
honogenous node. Injeclor is inlake-side pIaced, Iig. 5. The fueI is injecled lo lhe pislon
under given angIe. The pislon has lvo lovIs. The fueI lovI is on lhe inlake-side, lhe air
lovI is on exhausl-side. TunlIe fIov is ollained ly speciaI shaped inlake porl (Slefan,
2OO4). The fueI is guided sinuIlaneousIy via air and fueI lovI lo lhe spark pIug.

tumble control
injector
Air directed Wall directed

Iig. 5. VoIksvagen ISI engine air-vaII guided conluslion syslen (Anon, 2OO2).


5pray-GuIdcd cnmbustInn systcm: In lhe spray-guided lechnique fueI is injecled near
spark pIug vhere il aIso evaporales. The spray-guided lechnique lheorelicaIIy has lhe
highesl efficiency. The spray guided conluslion process requires advanced injeclor syslens
such as piezo injeclion. This lechnique has sone advanlages: reduced vaII velling,
increased slralified operalion region, Iess sensilive lo in-cyIinder air fIov, Iess sensilive lo
cyIinder lo cyIinder varialion and reduced rav HC enissions. Reporled disadvanlages are
spark pIug reIialiIily (fouIing) and poor roluslness (high sensilivily lo varialion in ignilion
&injeclion lining) (Calhcarl & RaiIlon, 2OO1). Mercedes-enz deveIoped a nev spray-
guided conluslion syslen. This syslen has lhe Slralified-Charged CasoIine Injeclion (CCI)
engine vilh Iiezo injeclion lechnoIogy. The spray-guided injeclion achieves leller fueI
efficiency lhan convenlionaI vaII-guided direcl injeclion syslens. The nain advanlage of
lhe CCI engine is ollained al lhe slralified operaling node. During lhis node lhe engine is
run vilh high excess air and lhus exceIIenl fueI efficiency is provided. MuIlipIe injeclions
exlend lhis Iean-lurn operaling node lo higher rpn and Ioad ranges, loo. During each
conpression slroke, a series of injeclions is nade spaced jusl fraclions of a second aparl.
This aIIovs lhe leller nixlure fornalion and conluslion, and Iover fueI consunplion
(Welsile 1, 2O1O).

3.2 The Operating Modes
CDI engine operales al differenl operaling nodes depending on Ioad and engine speed for a
slalIe and efficienl engine operalion. These engines have lhree lasic operaling nodes,
slralified vilh an overaII Iean nixlure, honogeneous vilh Iean nixlures and honogeneous
vilh sloichionelric nixlures. The engine is operaled vilh lhe slralified, honogeneous Iean
and honogeneous sloichionelric nodes, al Iov Ioad and speed, al nediun Ioad and speed
and al high Ioad and speed, respecliveIy. Iig. 6 shovs an exanpIe of lhe CDI operaling
nodes depending on engine Ioad and speed.

The engine conlroI unil conlinuaIIy chooses lhe one anong lhe operaling nodes. Lach node
is delernined ly lhe air-fueI ralio. The sloichionelric air-fueI ralio for pelroI (gasoIine) is
14.7:1 ly veighl, lul uIlra Iean node (slralified-charge) can invoIve ralios as high as 65:1.
These nixlures are nuch Ieaner lhan convenlionaI nixlures and reduce fueI consunplion
consideralIy. 5tratIfIcd-chargc mndc is used for Iighl-Ioad running condilions, al conslanl
or Iov speeds, vhere no acceIeralion is required. The fueI has lo le injecled shorlIy lefore
lhe ignilion, so lhal lhe snaII anounl of air-fueI nixlure is oplinaIIy pIaced near lhe spark
pIug. This lechnique enalIes lhe usage of uIlra Iean nixlures vilh very high air-fueI ralio,
inpossilIe vilh lradilionaI carlurelors or even porl fueI injeclion (Welsile 2, 2O1O). The Iean
lurn increases lhe NOx enissions. In lhis node, LCR is aclualed in order lo decrease NOx.
The area of slralified operalion is Iiniled ly Ioad and speed. Al high Ioad, lhe nixlure in lhe
slralified node can le loo rich, and lhus sool can forn. Al high speed, il is inpossilIe lo
provide sufficienl slralificalion due lo high lurluIence in lhe cyIinder. Therefore, al lhe
higher Ioad and speed range, lhe engine is operaled in honogeneous node lo ollain Iov
enissions and high lorque (KseII el aI., 1999).

Fuel njection 10

T
o
r
q
u
e
A
c
c
e
l
e
r
a
t
i
o
n
T
r
a
n
s
i
e
n
t

o
r
S
t
e
a
d
y

s
t
a
t
e

d
r
i
v
i
n
g

Iig. 6. CDI engine operaling nodes depending on Ioad and speed (KseII el aI., 1999).

Hnmngcncnus mndc is used for acceIeralion, fuII Ioad and high engine speeds. The air-fueI
nixlure is honogenous and lhe ralio is sloichionelric or sIighlIy richer lhan sloichionelric.
As lhe fueI is injecled during lhe inlake slroke, lhere is sufficienl line for air-fueI nixlure
fornalion. In lhis node, as engine operales vilh sloichionelric nixlure, NOx enission
decrease and lherefore LCR is nol aclivaled.

In lhe lransienl areas lhe engine can le operaled in honogeneous Iean node lo oplinize
fueI consunplion. Hnmngcncnus !can mndc is aclivaled for noderale Ioad and speed
condilions. In lhis node, fueI is injecled during lhe inlake slroke. The air-fueI nixlure is
honogeneous. The A/I ralio is Iean or sloichionelric. As engine operales vilh Iean
nixlure, NOx enission increase and lherefore LCR is aclivaled. The one anolher operaling
node is honogeneous-slralified node. This node is used al acceIeralion condilions vhen
passing fron slralified lo honogeneous node. The lvo slage injeclion (doulIe injeclion) is
inpIenenled. The prinary injeclion is perforned al inlake slroke and najorily of fueI is
injecled. The renaining fueI is injecled al secondary injeclion and conpression slroke.
DoulIe injeclion is nade lo reduce sool enissions and lo decrease fueI consunplion al Iov
engine speeds in lhe lransilion area lelveen slralified and honogeneous operalion. The
doulIe injeclion can aIso le used lo heal rapidIy calaIysl vilh a Iean slralified operalion
node. Al Iov speed and high Ioads, conluslion duralion is Iong and lenperalure is high.
Therefore, lhe engine lends lo knock. In lhis honogeneous charge node, ly using duaI
injeclion al fuII Ioad and ly decreasing lhe ignilion lining knock can le prevenled.

4. The FueI SuppIy and Engine Management System of the GDI Engine
4.1 The FueI SuppIy System
The fueI syslens for CDI engine require high fueI pressure IeveIs. IueI injeclion pressure is
lelveen 4 lo 13 MIa (lhe acluaI lrend is lo increase lhe IeveI of pressure). This pressure is
higher lhan III engine pressure vaIues ranging fron O.25 lo O.45 MIa. The higher pressures

Iead lo a higher penelralion and a leller alonizalion. AIlhough loo high injeclion pressures
increase alonizalion, lul an over penelraling can cause lhe vaII velling prolIens (Rolondi,
2OO6).

In CDI engines, fueI suppIy syslen consisls of lhe fueI lank, Iov-pressure punp, fueI fiIler,
high-pressure punp, fueI raiI, high-pressure sensor, injeclor and fueI pressure conlroI vaIve
(Iigure 7). The fueI syslen is divided inlo: a Iov-pressure Iine and a high-pressure Iine. The
pressure in Iov-pressure Iine is aloul O-5 lar. WhiIe lhe pressure in high-pressure Iine is
aloul 4-13 MIa (Anon, 2OO8).

Iig. 7. The fueI syslen conponenls for CDI engines.

The fueI lank is used lo slore lhe fueI. The fueI is deIivered vilh lhe pressure of aloul O,35
MIa fron lhe lank lo lhe high-pressure punp ly neans of an eIeclric fueI punp (Iov-
pressure punp). The eIeclric punp is lypicaIIy Iocaled in or near lhe fueI lank.
Conlaninanls are fiIlered oul ly a high capacily fueI fiIler. The high-pressure punp driven
ly canshafl increases fueI pressure and sends lhe fueI lo lhe raiI. The high pressure punp
increases lhe pressure up lo 13 MIa. The fueI pressure can le sel ly appIicalion dala
depending on lhe operalion poinl in lhe range fron 4 MIa lo 13 MIa. The pressure in lhe
fueI raiI is delernined ly lhe pressure sensor. To keep fueI pressure conslanl in lhe raiI is
very inporlanl in lerns of lhe engine pover, enissions, and noise. IueI pressure is
conlroIIed in a speciaI conlroI Ioop. The devialions fron adjusled vaIue are conpensaled ly
an open-Ioop or cIosed Ioop pressure-conlroI vaIve. In a cIosed Ioop conlroI excessive fueI is
relurned ly neans of lhe pressure conlroI vaIve. This vaIve aIIovs jusl enough fueI lo relurn
lo lhe lank. The fueI raiI serves as fueI accunuIalor. The injeclors, pressure conlroI vaIve and
high pressure sensor is nounled lo lhe fueI raiI. The injeclor is lhe cenlraI conponenl of lhe
injeclion syslen. Iigure 8 iIIuslrales a schenalic viev of lhe injeclor and ils lasic eIenenls.
The high pressure injeclor is Iocaled lelveen lhe raiI and conluslion chanler. Injeclors
Fuel njection 12

nounled on lhe raiI are opened ly Lngine ConlroI Unil (LCU) and, injeclors injecl lhe fueI
inlo cyIinder (Anon, 2OO6, Anon, 2OO8).

Sealing
Armature
Electrical
Connector
Hydraulic
Connector
Coil

Iig. 8. The high pressure injeclor.

4.2 The Engine Management System
Lngine nanagenenl syslen consisls of eIeclronic conlroI unil, sensors and aclualors. The
engine conlroI unil conlinuaIIy chooses lhe one anong operaling nodes depending on
engine operaling poinl and sensors dala. The LCU conlroIs lhe aclualors lo inpul signaIs
senl ly sensors. AII aclualors of lhe engine is conlroIIed ly lhe LCU, vhich reguIales fueI
injeclion funclions and ignilion lining, idIe operaling, LCR syslen, fueI-vapor relenlion
syslen, eIeclric fueI punp and operaling of lhe olher syslens. Adding lhis funclion lo lhe
LCU requires significanl enrichnenl of ils processing and nenory as lhe engine
nanagenenl syslen nusl have very precise aIgorilhns for good perfornance and drive
aliIily.

Inputs (scnsnrs): Mass air fIov sensor, inlake air lenperalure sensor, engine lenperalure
sensor, inlake nanifoId pressure sensor, engine speed sensor, canshafl posilion sensor,
lhrollIe posilion sensor, acceIeralor pedaI posilion sensor, raiI fueI pressure sensor, knock
sensor, Ianlda sensor upslrean of prinary calaIylic converler, Ianlda sensor dovnslrean
of prinary calaIylic converler, exhausl gas lenperalure sensor, Ianlda sensor dovnslrean
of nain calaIylic converler.


Outputs (actuatnrs): IueI injeclors, ignilion coiIs, lhrollIe vaIve posilioned, eIeclric fueI
punp, fueI pressure conlroI vaIve, LCR vaIve, fueI-vapor relenlion syslen vaIve and fan
conlroI (Anon, 2OO2).

The engine Ioad is nainIy delernined ly a hol fiIn air nass fIov sensor as knovn fron
porl injeclion syslens. The delerninalion of lhe LCR-rale and lhe diagnosis of lhe LCR-
syslen are acconpIished ly lhe using of a nanifoId pressure sensor. The air/fueI ralio is
conlroIIed ly neans of a vide land Ianlda sensor upslrean of prinary calaIylic converler.
The calaIysl syslen is diagnosed vilh a lvo poinl Ianlda sensor and an exhausl
lenperalure sensor. An indispensalIe conponenl is lhe eIeclronic lhrollIe device for lhe
nanagenenl of lhe differenl operalion nodes (KseII el aI., 1999). As an exanpIe of CDI
engine nanagenenl syslen, osch MLD-Molronic syslen in Iig. 9 is given.

Iig. 9. Conponenls used for eIeclronic conlroI in MLD-Molronic syslen of lhe osch (vilh
pernission of osch) (auer, 2OO4).

Fuel njection 14

5. Current trends and future chaIIenges
Al lhe presenl day, in lhe sone gasoIine engines are used porl fueI injeclion syslen. This
lechnique has achieved a high deveIopnenl poinl. As lhese engines operale vilh
sloichionelric nixlure, fueI econony and enissions of lhese engines can nol le inproved
furlher. Hovever, CDI engines have leen popuIar since lhese engines have polenliaI for
reduclion of loxic, CO
2
enissions and fueI consunplion lo conpIy vilh slringenl
LnvironnenlaI Iroleclion Agency (LIA) slandards (Spegar el aI., 2OO9). To allain lhis
polenliaI, il is required lhal use of lhe CDI engines vilh supercharging and/or lurlo
charging (Slan, 2OO9). The CDI engines vilh lurlo charger enalIe lhe produclion of snaIIer
dispIacenenl engines, higher fueI efficiency, Iover enission and higher pover (andeI el
aI., 2OO6). The CDI engines aIso heIp eIininale lhe disadvanlages convenlionaI
lurlocharged engines (naneIy lurlo Iag, poorer fueI econony and narroved enissions
polenliaI) lo provide vialIe engine soIulions (Spegar el aI., 2OO9).

The prinary dravlack of direcl injeclion engines is lheirs cosl. Direcl injeclion syslens are
nore expensive lecause lheir conponenls nusl le veII-nade. In lhese engines, lhe high
cosl high-pressure fueI injeclion syslen and exhausl gas lrealnenl conponenls are
required. The cosl of lhe CDI engines is high al lhe presenl day, lul CDI engines vilh lurlo-
charger lhal have nore fueI econony are expecled lo le cheaper lhan dieseI or hylrid
engines in fulure. Thanks lo nass produclion, if lhe prine cosl of lhe CDI engines can le
decreased, lhe vehicIe vilh CDI engine lhal have lurlo-charger can le Ieading on a
vorIdvide IeveI in lerns of lhe narkel share. The firns such as Milsulishi, VoIksvagen,
Iorsche, MW, Mercedes-enz, Mazda, Iord, Audi, CeneraI Molors, Ierrari and Iial prefer
using CDI engine in lheir vehicIes, loday. Hyundai viII slarl using lhe CDI engine in 2O11.

AIlhough differenl vehicIes vilh aIlernalive fueI have leen cone oul, lhey are inprolalIe lo
sulslilule convenlionaI gasoIine and dieseI povered vehicIes yel. ecause lhe fueIIing,
nainlenance infraslruclure, cosl, cruising dislance and drive conforl of lhen are nol
salisfaclory. Of lhe nexl-generalion vehicIes, onIy Hylrid LIeclric VehicIes (HLV) can le
regarded as aIlernalive energy vehicIes. They have lhe polenliaI lo grade aIongside
convenlionaI vehicIes in lerns of cosl and convenience since lheir fueI cosls are very Iov,
aIlhough lhey cosl nore lhan convenlionaI vehicIes (Morila, 2OO3). Il seens lhal Iarge scaIe
adoplion of HLVs viII nol le reaIized unIess lheir cosls cone dovn dranalicaIIy. CDI
engine aIso doesn'l force ovner of nolor vehicIe lo forgo Iuggage rack lecause of lalleries,
and doesn'l nake lhe car heavier. And il gives drivers Iols of fun-lo-drive lorque very
quickIy.

The Spray-Cuided CasoIine Direcl Injeclion (SCDI) engine vhich has piezo injeclors has
shoved a good polenliaI in lerns of lhe fueI econony and perfornance (Chang, 2OO7). Sone
CDI engines use piezoeIeclric fueI injeclors loday. The piezo-effecl is used lo provide
opening and cIosing lhe injeclor in lhe direcl injeclion syslens. The piezo injeclors are four-
five lines fasler lhan convenlionaI injeclors. They can neasure lhe fueI vilh grealer
precision. In addilion, lhey can injecl fueI lelveen six and len lines during a conluslion
cycIe. Irecise piezo injeclion aIIovs reducing lhe poIIulanls. CDI engines vilh piezo
injeclors can easiIy neel slriclIy enission Iinil changes ahead. IueI consunplion can le
reduced ly up lo 15 percenl and engine perfornance increased ly aloul 5 (Welsile 3,

2O1O). Thanks lo nuIlipIe injeclions, il is for lhe firsl line possilIe lo exlend Iean-lurn
operaling node lo higher rpn and Ioad ranges, loo. During each pover slroke, a series of
injeclions lakes pIace. This inproves nixlure fornalion, conluslion and fueI consunplion.
The injeclors used in DI syslen have nozzIes vhich open oulvards lo creale an annuIar gap
jusl a fev nicrons vide. The peak fueI pressure in lhis syslen is up lo 2OO lar - around 5O
lines lhe fueI pressure in a convenlionaI pelroI injeclion syslen (Welsile 4, 2O1O). The firns
such as osch, DeIphi and Sienens have deveIoped a piezo injeclion syslen for gasoIine
engines lo aulonakers. The ain is lo inprove lhe perfornance of lhe direcl injeclion
syslens. The Iiezo injeclion vilh spray guided conluslion syslen is used in lhe Mercedes-
enz CLS 35O CCI nodeI vehicIe (Welsile 5, 2O1O).

In CDI engine, as lhe spark pIugs operale under high lenperalure, lhe fouIing of lhen can
cause lhe nisfiring. To increase lhe Iife-line of lhe spark pIug and engine efficiency, lhe
syslen such as Iaser-induced ignilion can le appIied. Thus, engine efficiency can le nore
increased. The CDI engines are very suilalIe for lhe operaling vilh aIlernalive fueI. The
sludies on CDI engine vilh aIlernalive fueI such as naluraI gas, elhanoI, LIC have
conlinuaIIy increasing al presenl day (KaIan, 2OO9, Teoh el aI., 2OO8, Slein & House, 2OO9). If
CDI engines vilh lurlo charger use spray guided conluslion process vhich has
piezoeIeclric injeclor and high energy ignilion syslen, lhe use of lhese engines are expecled
lo increase nore in shorl lern.

6. References
AIger T., HaII M., and Mallhevs R. D., Lffecls of SvirI and TunlIe on In-CyIinder IueI
Dislrilulion in a CenlraI Injecled DISI Lngine, SAL Iaper 2OOO-O1-O533.
AIkidas A. C., Conluslion Advancenenls in CasoIine Lngines, Lnergy Conversion and
Managenenl 48 (2OO7) 2751-2761.
Anon, VoIksvagen AC, osch Molronic MLD7 CasoIine Direcl Injeclion, VoIksvagen SeIf-
Sludy Irogran 253, 2OO2, WoIfslurg.
Anon, VoIksvagen AC, Tvin Turlo Charger TSI Lngine, VoIksvagen SeIf-Sludy Irogran
359, 2OO6, WoIfslurg.
Anon, VoIksvagen AC, TSI Turlocharged Lngine, VoIksvagen SeIf-Sludy Irogran 8248O3,
2OO8, U.S.A.
Anon, VoIksvagen Iassal TSI, Tail Tanilin KalaIogu, 2OO9, IslanluI (in Turkish).
andeI W., IraidI C. K., Kapus I. L., Sikinger H. and CovIand C. N., The Turlocharged CDI
Lngine: oosled Synergies for High IueI Lconony IIus UIlra-Iov Lnission, SAL
Iaper 2OO6-O1-1266.
auer H., CasoIine Lngine Managenenl-Syslen and Conponenls, Rolerl osch CnlH,
Cernany, 2OO4.
aungarlen C., Mixlure Iornalion in InlernaI Conluslion Lngines, Springer VerIag,
Cernany, 2OO6.
Calhcarl C. and RaiIlon D., Inproving Roluslness of Spray Cuided DI Syslens: The Air-
assisled Approach, }SAL AnnuaI Congress 2OO1, VoI. 4O-O1,p. 5-8.
Chang W. S., Kin Y. N. and Kong }. K., Design and DeveIopnenl of a CenlraI Direcl
Injeclion Slralified CasoIine Lngine, SAL Iaper 2OO7-O1-3531.
Fuel njection 16

eIik M. ., uji IIe AleIeneIi ir Molorun Sikilirna Oraninin Degiken HaIe
DnlrInesi ve Ierfornansa Llkisinin AraliriInasi, Doklora Tezi, Cazi
niversilesi Ien iIinIeri Lnslils, 1999, Ankara.(in Turkish)
eIik M. ., Ierfornance Inprovenenl and Lnission Reduclion in SnaII Lngine vilh Lov
Lfficiency, }ournaI of lhe Lnergy Inslilule, 8O, 3, 2OO7.
inar C., Direkl IskrlneIi uji IIe AleIeneIi MolorIar, SeIuk-Teknik OnIine Dergisi, CiIl
2, No. 1-2OO1.(in Turkish)
Ian L., Li C., Han Z. and Reilz R. D., ModeIing IueI Ireparalion and Slralified Conluslion
in a CasoIine Direcl Injeclion Lngine, SAL Iaper 1999-O1-O175.
Ierguson C. R., InlernaI Conluslion Lngines, }ohn WiIey&Sons, Inc., 1986, Nev York.
Candhi A. H., Weaver C. L., Curlis L. W., AIger T. I., Anderson C. L., Alala D. L., Spray
Characlerizalion in a DISI Lngine During CoId Slarl: (1) Inaging Invesligalion,
SAL Iaper 2OO6-O1-1OO4.
HenlscheI W., OplicaI Diagnoslics for Conluslion Irocess DeveIopnenl of Direcl-Injeclion
CasoIine Lngines, Iroceedings of lhe Conluslion Inslilule, VoIune 28, 2OOO/pp.
1119-1135.
Heyvood }. ., InlernaI Conluslion Lngines IundanenlaIs, McCrav HiII ook, 2OOO,
Singapore.
KaIan M. A., LxperinenlaI Tesl of a Nev Conpressed NaluraI Cas Lngine vilh Direcl
Injeclion, SAL Iaper 2OO9-O1-1967.
KaranangiI M. I., Direkl IskrlneIi enzin MolorIari ve Milsulishi Melodu, UIudag
niversilesi MhendisIik MinarIik IakIlesi Dergisi, CiIl 9, Sayi 1, 2OO4.(in
Turkish)
KIeelerg H., Dean T., Lang O. and Halernann K., Iulure IolenliaI and DeveIopnenl
Melhods for High Oulpul Turlocharged Direcl Injecled CasoIine Lngines, SAL
Iaper 2OO6-O1-OO46.
Kune T., Lvanolo Y., Lida K., Murakani M., Akishino K. and Ando H., Conluslion
ConlroI TechnoIogies for Direcl Injeclion SI Lngine, SAL Iaper 96O6OO.
KseII M., Moser W. and IhiIipp M., Molronic MLD7 for CasoIine Direcl Injeclion Lngines:
Lngine Managenenl Syslen and CaIilralion Irocedures, SAL Iaper 1999-O1-1284.
Lecoinle . and Monnier C., Dovnsizing a CasoIine Lngine Using Turlocharging vilh
Direcl Injeclion, SAL Iaper 2OO3-O1-O542.
Morila K., Aulonolive Iover Source in 21sl Cenlury, }SAL Reviev, 24 (2OO3) 3-7.
Muoz R. H., Han Z., VanDerWege . A. and Yi, }., Lffecl of Conpression Ralio on
Slralified-Charge Direcl- Injeclion CasoIine Conluslion, SAL Iaper 2OO5-O1-O1OO.
Orlnann R., Arndl S., Rainann }., Crzeszik R. and WrfeI C., Melhods and AnaIysis of IueI
Injeclion, Mixlure Ireparalion and Charge Slralificalion in Differenl Direcl Injecled
SI Lngines, SAL Iaper 2OO1-O1-O97O.
Rolondi R. and eIIa C., CasoIine Direcl Injeclion Spray SinuIalion, InlernalionaI }ournaI of
ThernaI Sciences, 45 (2OO6) 168-179.
Sercey C. D., Avcock C., HeikaI M., Use of LII Inage Acquisilion and AnaIysis in
DeveIoping a CaIilraled Technique for in-CyIinder Invesligalion of lhe SpaliaI
Dislrilulion of Air-lo-IueI Mixing in Direcl Injeclion CasoIine Lngines, Conpulers
in Induslry 56 (2OO5) 1OO5-1O15.
Snilh }. D. and Sick V., A MuIli-VarialIe High-Speed Inaging Sludy of Ignilion InslaliIilies
in a Spray-Cuided Direcl-Injecled Spark-Ignilion Lngine, SAL Iaper 2OO6-O1-1264.

Spegar T. D., Chang S., Das S., Norkin L. and Lucas R., An AnaIylicaI and LxperinenlaI
Sludy of a High Iressure SingIe Iislon Iunp for CasoIine Direcl Injeclion (CDI)
Lngine AppIicalions, SAL Iaper 2OO9-O1-15O4.
Spicher U., KIneI A., Kulach H. and Tpfer C., Conluslion in Spark Ignilion Lngines
vilh Direcl Injeclion, SAL Iaper 2OOO-O1-O649.
Slan C. C., AnaIysis of Lngine Ierfornances Inprovenenl ly Dovn Sizing in ReIalionship
vilh Super and Turlo Charging, Adapled Scavenging and Direcl Injeclion, SAL
Iaper 2OO9-24-OO75.
Slefan S., OplicaI Diagnoslics on ISI Transparenl Lngine, IISITA WorId Aulonolive
Congress, arceIona 23-27 May, arceIona Spain, 2OO4.
Slein R. and House C., OplinaI Use of L85 in a Turlocharged Direcl Injeclion Lngine, SAL
Iaper 2OO9-O1-149O.
SloffeIs H., Conluslion Noise Invesligalion on a Turlocharged Spray Cuided CasoIine
Direcl Injeclion I4-Lngine, SAL Iaper 2OO5-O1-2527.
Slone R., Inlroduclion lo InlernaI Conluslion Lngines, SAL, Inc., 1999, WarrandaIe.
Teoh Y. H., Cilano H. W. and Muslafa K. I., Ierfornance Characlerizalion of a Direcl
Injeclion LIC IueIIed Tvo-Slroke MolorcycIe Lngine, SAL Iaper 2OO8-32-OO45.
Welsile 1: hllp://vvv.greencarcongress.con/2OO6/O2/nercedeslenz_pr.hlnI,
(17.O4.2O1O).
Welsile 2: hllp://gernancarviki.con/doku.php/fsi, (17.O4.2O1O).
Welsile 3: hllp://vvv.epcos.con/vel/generalor/Wel/Seclions/Conponenls/Iage,
IocaIe=en,r=263288,a=26338O.hlnI, (17.O4.2O1O).
Welsile 4: hllp://vvv.schval-koIl.con/dainIer/en/dcOOO259.hln, (17.O4.2O1O).
Welsile 5: hllp://vvv.nercedes-lenz.con.lr/conlenl/lurkey/npc/npc_lurkey_velsile/
lr/hone_npc/passengercars/hone/nev_cars/nodeIs/cIs-
cIass/c219/overviev/drivelrain_chassis.OOO2.hlnI, (17.O4.2O1O).
Zhao I., Lai M. C., Harringlon D. L., Aulonolive Spark-Igniled Direcl-Injeclion CasoIine
Lngines, Irogress in Lnergy and Conluslion Science, VoIune 25, Issue 5, Ocloler
1999, Iages 437-562.

Fuel njection 18
X

Liquid Sprays Characteristics in DieseI Engines

Sinn Marlnez-Marlnez
1
, Iauslo A. Snchez-Cruz
1
,
Vicenle R. erndez
2
and }ose M. Riesco-viIa
3
Unitcrsia Au|cncma c Nuctc |ccn
1
Mcxicc
Unitcrsia Pc|i|ccnica c Va|cncia
2
Spain
Unitcrsia c Guanajua|c
3
Mcxicc

1. Introduction
Ior decades, lhe process of injecling an aclive fIuid (dieseI fueI) inlo lhe lhernodynanic
lehaviour of a vorking fIuid (air or gas) has leen a priorily in lhe research of lhe
phenonena lhal occur in conluslion syslens. Due lo lechnoIogicaI inprovenenls ils
possilIe in presenl lines lo characlerise lhe injeclion fueI process in such condilions lhal
nalch lhose happening vhen lhe engine is running under slandard condilions, hence lhe
purpose of lhese sludies, vhich focus in lhe achievenenl of a perfecl nixlure lelveen lhe
vorking and aclive fIuids, as a resuIl of lhis, a series of consequences are lriggered lhal Iead
lo an oplinun conluslion, and lherefore in lhe inprovenenl of lhe engines capaliIilies. In
DieseI engines lhe conluslion process lasicaIIy depends on lhe fueI injecled inlo lhe
conluslion chanler and ils inleraclion vilh lhe air.

The injeclion process is anaIysed fron lhis poinl of viev, nainIy using as lasis lhe slruclure of
lhe fueI spray in lhe conluslion chanler, naking lhis sludy of high inporlance for
oplinizing lhe injeclion process, and lherefore reducing lhe poIIulanl enissions and
inproving lhe engines perfornance. ecause of lhese, lhe inporlance lo ollain lhe naxinun
conlroI of lhe dieseI spray slruclure using eIeclronic conlroI syslens has lecone vilaI. To
reduce poIIulanl enissions and achieving a high engine perfornance, ils necessary lo knov
vhich paranelers infIuence lhese ralings lhe nosl. Il is consider leing severaI neaningfuI
faclors lhal have an infIuence, lul lhe nosl inporlanl one is lhe dieseI spray, nore specificaIIy
lhe penelralion of lhe Iiquid Ienglh of lhe spray lhru lhe conluslion chanler or pislon lovI.
The anaIysis of lhe Iiquid Ienglh penelralion is very usefuI lo delernine lhe geonelric design
of high speed DieseI engine conluslion chanlers vilh direcl injeclion. Ior exanpIe, in a Iov
speed regine and Iighl Ioad condilions, lhe unlurned hydrocarlon enissions viII le reduced
grealIy if conlacl lelveen lhe spray of fueI (Iiquid Ienglh) and lhe conluslion chanler vaII is
avoided. If nov ve consider a high speed regine and heavy Ioad, lhe enission of funes is
reduced if lhere is conlacl lelveen lhe spray of fueI and lhe conluslion chanler vaII, hence
2
Fuel njection 20

lhe inporlance of neasuring lhe Iiquid phase penelralion of lhe fueI in DieseI engines vilh
direcl injeclion, using sophislicaled and conpIex neasuring lechniques.

2. DieseI spray characteristics
Depending on lhe nechanisn lo characlerise, dieseI spray can le anaIysed in a nacroscopic
or nicroscopic poinl of viev. Wilh lhe purpose of underslanding in delaiI lhis process, lhe
various physicaI paranelers invoIved during lhe lransilion of a puIsed dieseI spray viII le
expressed in lhis chapler, hovever il is essenliaI lo knov lhe syslens lhal nake possilIe for
an injeclion process lo lake pIace. These are lhe injeclion nozzIe, aclive fIuid lo injecl
(Iiquid), and lhe vorking fIuid on vhich lhe Iiquid is injecled, as seen in figure 1.

Iig. 1. MeaningfuI varialIes of lhe injeclion process.

Ior a Nevlonian fIuid vilh conslanl lenperalure dislrilulion and an injeclion nozzIe vilh a
conpIeleIy cyIindricaI orifice, lhe varialIes lhal infIuence lhe dispersion of lhe spray are:

NozzIe Ceonelry
- Orifice Dianeler (do)
- Lenglh (Io)
- Orifice enlrance curvalure radius (ro)
-SuperficiaI Roughness ()
Injeclion Condilions
-Iressure of Liquid Injecled IIuid (I
I
)
-Iressure of Cas Working IIuid (I
g
)
-Iressure increasing (I = I
I
-I
g
)
-Mediun veIocily of lhe injecled Liquid fIuid (V
I
)
- Mediun veIocily of lhe vorking gas fIuid (V
g
)
-Duralion of lhe injeclion (l
inj
)
Injecled IIuid Iroperlies (Liquid)
-Densily (
I
)
-Kinenalic Viscosily (
I
)
-Vapour Iressure (I
v
)
-SuperficiaI Tension ()
Working IIuid Iroperlies (Cas)
-Densily (
g
)
- Kinenalic Viscosily (
g
)

AII lhese varialIes can le, can le filled inlo a dinensionIess forn lhal aIIovs us lo have
nuch sinpIer reIalions and leller defined. The dinensionIess varialIes used in nosl cases
are:

Rc|a|icn cf cnsi|ics.

I
g
* =
(1)
Rc|a|icn cf tisccsi|ics.

I
g
* =
(2)

Rcqnc|s Num|cr, rc|a|icn |c|uccn incr|ia| an tisccus fcrccs.

d
Re =
(3)

Ic|cr Num|cr, rc|a|icn |c|uccn supcrficia| |cnsicn fcrcc an incr|ia| fcrcc.

2
d
We =
(4)

Taq|cr Visccsi|q Paramc|cr.

Re
Ta = =
We
(5)

Onncscrgc Num|cr.

We
Oh = =
Re d
(6)

|cng|n/iamc|cr rc|a|icn cf |nc Nczz|c (|
c
/
c
)

Nczz|c raius cn|rancc/iamc|cr rc|a|icn (r
c
/
c
)

Discnargc cccfficicn| cf |nc nczz|c.

d
I
I
C =
2I
(7)
Cati|a|icn Paramc|cr.

I
2
I
2(I - I )
K =

(8)

Fuel njection 22

ReynoIds Nunler: Densily and kinenalic viscosily nusl le parlicuIarised for Iiquid or gas,
furlhernore lhese properlies can le evaIualed for inlernediale condilions lelveen lolh
fIuid fiIn condilions. These paranelers can le divided inlo lvo groups:

1. LxlernaI fIov paranelers (reIalion of densilies, Weler nunler, TayIor paraneler),
lhese paranelers conlroI lhe inleraclion lelveen lhe Iiquid spray and lhe
surrounding alnosphere.
2. InlernaI fIov paranelers (ReynoIds nunler, cavilalion paraneler,
Ienglh/dianeler reIalion, nozzIe radius enlrance/dianeler reIalion, discharge
coefficienl): lhese paranelers conlroI lhe inleraclion lelveen lhe Iiquid and lhe
nozzIe.

2.1. Macroscopic Characteristics
The nacroscopic descriplion of a dieseI spray generaIIy enphasise lhe inleraclion of lhe
Ialler and lhe conlroI voIune vhere il is injecled and nixed, and lecause of lhis lhe dieseI
spray can le defined vilh lhe foIIoving physicaI paranelers (Iigure 2.2):

1. Spray lip penelralion
2. Spray angIe
3. reack up Ienglh

Iig. 2. IhysicaI paraneler of a dieseI spray (Hiroyasu & Aray, 199O).

2.1.1. Front Penetration
The injeclion fronl penelralion (S) is defined as lhe lolaI dislance covered ly lhe spray in a
conlroI voIune, and ils delernined ly lhe equiIilriun of lvo faclors, firsl lhe nonenlun
quanlily vilh vhich lhe fIuid is injecled and second, lhe resislance lhal lhe idIe fIuid
presenls in lhe conlroI voIune, nornaIIy a gas. Due lo friclion effecls, lhe Iiquids kinelic
energy is lransferred progressiveIy lo lhe vorking fIuid. This energy viII decrease
conlinuousIy unliI lhe novenenl of lhe dropIels depends soIeIy on lhe novenenl of lhe
vorking fIuid inside lhe conlroI voIune. Irevious sludies have shovn lhal a spray
penelralion overcones lhal of a singIe dropIel, due lo lhe nonenlun lhal lhe dropIels

Iocaled in lhe fronl of lhe spray experinenl, acceIeraling lhe surrounding vorking fIuid,
causing lhe nexl dropIels lhal nake il lo lhe fronl of lhe spray an inslanl of line Ialer lo
have Iess aerodynanic resislance. We nusl enphasise lhal dieseI fueI sprays lend lo le of
lhe conpacl lype, vhich causes lhen lo have Iarge penelralions.

SeveraI researchers have sludied lhe fronl penelralion and have found a series of
correIalions lhal aIIov us lo eslalIish lhe nain varialIes lhal affecl or favour lhe penelralion
of a puIsed dieseI spray. The foIIoving are sone of lhe nosl reIevanl:

Iron lhe lheory of gaseous sprays, (Denl, 1971) vas one of lhe pioneers in lhe sludy of
spray phenonena. The aulhor proposed an experinenlaIIy adjusled correIalion vhich is
appIicalIe lo puIsed dieseI sprays, lhis correIalion vas lhe conpared ly (Hay & }ones, 1972)
vilh olher correIalions, finding cerlain discrepancies lelveen lhen. Hovever, lhis
correIalion is considered lo le appIicalIe in a generaI forn lo dieseI sprays:

| | | |
| |
\ . \ .
1 1
4 4
o
a a
I 294
S(l) = 3, O7 d l
T
(9)

(Hiroyasu & Arai, 199O) proposed lvo expressions lo delernine lhe sprays penelralion as a
funclion of lhe line of fraclure (l
rol
), and so defining lhe fraclure line can fIucluale lelveen
O,3 y 1 ns depending on lhe injeclion condilions.

(1O)

I
2I
S = O, 39 l
(11)

rol
l = l (12)

| |
|
|
\ .
O, 25
o
g
I
S = 2, 39 d l
(13)

rol
l = l (14)

An enpiricaI equalion considering lhe dinensionIess paraneler
*
= (
a
/
I
) vas deveIoped
ly (}inenez el aI., 2OOO) ollaining lhe foIIoving expression:

( ) ( )
| |
|
\ .
-O,163
O,9
-3 a
o
I
S l = O, 6 U l
(15)

I
rol
g
d
l = 28, 65
I
Fuel njection 24

Where Uo is lhe nediun veIocily al lhe leginning of lhe injeclion in |n/sj and l is injeclion
line duralion in |n/sj. In lhis equalion lhe lehaviour of lhe sprays penelralion is
considered for lenperalure varialions in lhe vorking fIuid lelveen 293 K and 423 K.
AIlhough lhe equalion considers lhe alnospheric pressure vaIues of lhe vorking fIuids (Iov
densily), il is aIso vaIid for high densilies.

Ienelralion according lo (}avard el aI., 1999):

( )
O, 25
O, 25 -O,14
1 I g
S = C I l (16)

Iron lhe derivalion of lhe expressions deveIoped ly (Denl, 1971) and (Arai el aI., 1984),
(ae el aI., 2OOO) proposes lhis expression for lhe penelralion of lhe spray:

| |
|
|
\ .
O, 25
o
g
I
S = C d l
(17)

| | | |
| |
| |
\ . \ .
I o
o
g iny
d
l = l =
V
(18)

Ienelralion according lo (Correas, 1998):

O, 5
2 o eq
S = C U d l (19)

I
eq o
g
d = d
(2O)

Considering C1 and C2 experinenlaI conslanls, d
eq
lo le lhe equivaIenl dianeler, and C
anolher experinenlaI conslanl as a funclion of lhe discharge coefficienl, il can le said lhal
lhe discharge coefficienl and lhe conslanl C have a direcl dependence on lhe injeclor lype
used and in Iess neasure on lhe vorking condilions. Therefore and according lo (Hiroyasu
& Denl, 199O) proposaI, lhe discharge coefficienl (Cd) for a delernined injeclor does nol
nodify lhe conslanl C vaIue. Olher vorks of greal inporlance concerning lhe penelralion of
spays in VCO nozzIes vere presenled ly (ae & Kang, 2OOO), in vhich he cIassifies differenl
lypes of sprays for differenl densilies of lhe vorking fIuid.

As a sunnary il can le said lhal lhe penelralion of lhe spray lasicaIIy depends on lhe
foIIoving paranelers:

-Injeclion pressure increasing I: Increasing lhe injeclion pressure in reIalion lo lhe conlroI
voIune vhere lhe fueI is injecled (I), increases lhe veIocily of lhe penelralion of lhe spray

and hence lhe deveIopnenl of lhe Ialler viII le easier al lhe leginning, (Hiroyasu el aI.,
198O) and (Arai el aI., 1984).

According lo (Ahnadi el aI., 1991), lecause a parl of lhe Iiquid advances rapidIy lhrough
lhe inlernaI spray area vhere lhe aerodynanic inleraclion is poor, lhe injeclion pressure
fIuclualions are nol reIaled lo lhe injeclions veIocily. On lhe olher hand, al lhe lip of lhe
spray lhe high aerodynanic inleraclion causes lhe Ialler lo Iose veIocily, naking lhe recenlIy
injecled Iiquid lo reach and pass lhis sIover noving lip, laking ils pIace as lhe nev spray lip
and aflervards leing sIoved dovn as veII ly lhe conlroI voIunes surroundings. As veII,
(Nishida el aI., 1992) and (Tinaul el aI., 1993) suggesl lhal lhe veIocily of lhe dropIels al lhe
lip is usuaIIy sIover lhan in olher regions of lhe spray, so lhe sinpIe facl lhal lhe veIocily of
lhe dropIels is sIover lhan lhe veIocily of penelralion denands a conslanl dropIel renevaI
in lhe lip of lhe spray.

-Densily ralio (
*
): lhis dinensionIess paraneler
*
or reIalion of densilies, according lo
(Hiroyasu el aI., 198O), (Arai el aI., 1984) and (Iayri el aI., 1996), consideralIy affecls lhe
penelralion of lhe spray, due lo lhe facl lhal increasing lhe reIalion of densilies causes lhe
penelralion lo reduce consideralIy, lhis is lecause of lhe increase or reduclion of lhe
aerodynanic inleraclion, according lo lhe respeclive paraneler scaIe.

-Working fIuid lenperalure (Tg): densily reduclion can le caused ly lhe increase of lhe
vorking fIuids lenperalure, hence, lhe decrease of spray penelralion. Hovever, previous
sludies shov lhal lhe sprays lenperalure doesnl produce significanl effecls in lhe
penelralion in reIalion lo olher paranelers, (Hiroyasu el aI., 198O) and (Arai el aI., 1984).

2.1.2. Cone angIe
The cone angIe is defined as lhe angIe forned ly lvo slraighl Iines lhal slal fron lhe exil
orifice of lhe nozzIe and langenl lo lhe spray oulIine (sprays norphoIogy) in a delernined
dislance. The angIe in a dieseI spray is forned ly lvo slraighl Iines lhal are in conlacl vilh
lhe sprays oulIine and al a dislance equivaIenl lo 6O lines de exil dianeler of lhe nozzIes
orifice. This angIe usuaIIy is lelveen 5 and 3O degrees. This delernines grealIy lhe fueIs
nacroscopic dislrilulion in lhe conluslion chanler. In one hand, lhe increase in angIe
decreases lhe penelralion and can cause inlerference lelveen sprays (vhen sprays are
injecled using nuIli-orifice nozzIes) in lhe sane chanler favouring lhe nerging of dropIels.
On lhe olher hand, an excessive penelralion is favoured vhen lhe angIe decreases Iover
lhan cerlain vaIues, causing lhe spray lo coIIide vilh lhe pislon lovI or lhe conluslion
chanler.

In previous sludies lhere have leen a series of proposaIs lo delernine lhe cone angIe, sone
of lhe nosl inporlanl are as foIIovs:

| |
|
\ .
a
I

lan = O, 13 1+
2
(21)
Fuel njection 26

This expression is considered for densilies of lhe vorking fIuid Iover lhan (
g
) 15 kg/n
3
,
lul lhe dinensionIess injeclor reIalion is nol considered(I
o
/d
o
). Hovever, (Reilz & raco,
1979) and (Arai el aI., 1984) do consider lhis dinensionIess paraneler in lheir invesligalions
lo delernine lhe naxinun aperlure of lhe cone angIe, proving lhal il indeed has greal
infIuence on lhe opening of lhe cone angIe.

Cone angIe according lo (Hiroyasu el aI., 198O):

| |
|
\ .
O, 25
2
a
2
a
d
= O, O5
(22)

The dropIels size reIaled lo lhe vaveIenglhs of lhe nosl unslalIe vaves vas eslalIished ly
(Ranz & MarshaII, 1958) and lherefore, lhe cone angIe is defined ly lhe conlinalion of lhe
injeclion veIocily and lhe radiaI veIocily of lhe vaves of grealer grovlh in lheir superficiaI
unslalIeness, defining lhe cone angIe vilh lhe foIIoving expression:

( )
| |
|
\ .
1
2
g
I
1
lan = 4 f
2 A
(23)

| |
|
\ .
2
I I
g I
Re
=
We
(24)

| |
|
\ .
o
o
I
A = 3, O +O, 277
d
(25)

Where: A is a conslanl delernined experinenlaIIy in funclion of lhe reIalion
Ienglh/dianeler of lhe nozzIe (I
o
/d
o
), vhich is represenled ly lhe equalion (24) according lo
(Reilz & raco, 1979). Iigure 3 shovs lhe dependence of lhe cone angIe in funclion of
aerodynanic forces, (Ranz & MarshaII, 1958) ciled ly (Heyvood, 1988) y (Ranos, 1989), and
for concepls on dropIel evaporalion, (Ranz & MarshaII, 1952).

Cone angIe proposed ly (Hiroyasu & Arai, 199O):

| | | |
| |
| |
|
\ .
\ . \ .
O,15 O, 26
-O,22
g
o I
I d
= 83, 5
d D
(26)


Iig. 3. Cone angIe dependence in funclion of aerodynanic forces (Ranos, 1989).

Where: Do represenl lhe dianeler of lhe nozzIes jackel. Wilh lhis expression ils possilIe lo
delernine lhe angIe of opening of lhe fuIIy deveIoped spray, vhere lhe angIe is praclicaIIy a
funclion of lhe nozzIes orifice geonelry and lhe dinensionIess lern of lhe reIalion of
densilies (
*
). Olhers paranelers such as cinenalic viscosily can in sone vay nodify lhe
Iinils of lhe deveIoped spray, lul nol lhe angIe of lhe cone.

The cone angIe is nainIy affecled ly lhe geonelric characlerislics of lhe nozzIe, lhe densily
ralio (
*
), and lhe ReynoIds nunler of lhe Iiquid, (Reilz & racco, 1979, 1982), aparl fron
depending on olher varialIe such as lhose descriled as foIIovs:

-Increasing pressure (I): An increase in lhe injeclion pressure causes an increase in lhe
cone angIe up lo a naxinun vaIue, alove decrease graduaIIy.

-Densily ralio (
*
): An increase in lhe reIalion of densilies is a faclor lhal causes an increase
in lhe cone angIe due lo an increase in lhe aerodynanic inleraclion, according lo (ArrgIe,
1998) and (Naler & Sielers, 1996), for vaIues grealer lhan (
*
> O.O4) lhe cone angIe lends lo
le independenl of lhis paraneler.

-Working fIuid lenperalure (Tg): Increasing vorking fIuid lenperalure, increases lhe
evaporalion process in lhe sprays exlerior zone, consequenlIy a decrease in lhe angIe of lhe
cone, (Hiroyasu el aI., 198O).

2.1.3. Liquid Length
The Iiquid Ienglh of lhe spray is a very inporlanl characlerislic lo define lhe lehaviour of
lhe spray in lhe conluslion chanler. This zone of lhe spray is aIso caIIed conlinuous or
slalionary and il is underslood as leing fron lhe nozzIe exil lo lhe poinl vere lhe separalion
of lhe firsl dropIels occur. To define lhis zone lhe use of diverse neasurenenls nelhods and
lechniques is of vilaI inporlance. In lhe Iileralure ve find sone of lhe nosl usefuI
neasurenenl nelhods and lechniques in lhe anaIysis of lhe Iiquid Ienglh, (Hiroyasu & Arai,
199O), (Chehroudi el aI., 1985), (Arai el aI., 1984), (Nishida el aI., 1992), (CIder el aI., 1992),
(Chrisloph & Dec, 1995), (Zhang el aI., 1997) and (erndez el aI., 2OO2, 2OO3).

Fuel njection 28

To anaIyze lhe inlernaI slruclure of lhe spray, (Hiroyasu & Aray, 199O) idenlified lvo zones
inside lhe alonizing regine, lhe zone of lhe inconpIele spray and lhe zone of lhe conpIele
spray. Iigure 4 shovs slruclure in a generaI vay. The difference lelveen lhen is due lo lhe
facl lhal vilh lhe inconpIele sprays lhe disinlegralion of lhe surface of lhe spray legins al a
cerlain dislance fron lhe poinl of lhe nozzIe of lhe injeclor, indicaling a dislance Lc, vhiIe in
lhe case of lhe inconpIele sprays dislance Lc is nearIy cero and Ll is nainlained virluaIIy
conslanl on increasing speed. Iurlhernore (Hiroyasu & Aray, 199O) shov lhal cavilalion
grealIy favours lhe alonizalion process in lhe conpIele spray regine.

To define Iiquid Ienglh a series of expressions have leen proposed vhich have leen
suggesled in specific condilions according lo each case and anong lhe nosl reIevanl lhe
foIIoving can le ciled:

Iig. 4. InlernaI slruclure of conpIele and inconpIele spray (Hiroyasu & Aray, 199O).

ased on experinenlaI resuIls of lhe neasurenenl of lhe Iiquid Ienglh in conpIele sprays
(Hiroyasu & Aray, 199O) proposed lhe foIIoving equalion:

| |
| |
| | | |
|
|
| |
|
\ . \ .
\ .
\ .
O, 5
O,O5
O,13
g
I
l 2
I o g
R L
L = 7d 1+O, 4
D U D
(27)

Liquid Ienglh according (racco, 1983):

| |
|
|
\ .
O, 5
I
l
g
L = 7, 15
(28)


Liquid Ienglh according (YuIe & SaIlers, 1995):

( )
| |
|
|
\ .
-O,O8
-3 -O,1 -O, 3 I
l I I
g
L = 2, 65 d We Re
(29)

The nosl inporlanl paranelers on Iiquid Ienglh penelralion are lhe foIIoving:

1. The ralio of vork fIuid densilies/Iiquid (
*
): an increase on lhe ralio of densilies
produces a decrease in Iiquid Ienglh due lo an increase in lhe aerodynanic
inleraclion lelveen lhe spray and lhe environnenl in vhich lhis is deveIoped as
shovn ly (Arai el aI., 1984), (Chehroudi el aI., 1985), (Hiroyasu & Arai, 199O),
(Chrisloph & Dec, 1995), (Cannan el aI., 1998), (Naler & Sielers, 1996) and (Sielers,
1998).
2. The reIalionship lelveen Ienglh/nozzIe dianeler (I
o
/d
o
): lhis reIalionship
infIuences lhe Iiquid Ienglh penelralion vhen lhe voIune of conlroI vhere lhe
conluslilIe is injecled al alnospheric condilions. Hovever, vhen lhe conlroI
voIune pressure is high, lhe infIuence of lhis paraneler in Iiquid Ienglh
penelralion decreases, according lo invesligalions nade ly (Ha el aI., 1983) and
(Xu & Hiroyasu, 199O).
3. NozzIe orifice dianeler (d
o
): lhe Iiquid Ienglh has a Iinear lehaviour vilh lhe
nozzIe dianeler. Liquid Ienglh penelralion decreases lo nininun vaIues vhen lhe
nozzIe dianeler is reduced lo nininun vaIues, in olher vords, a change in lhe
dianeler of lhe nozzIe orifice resuIls in a direclIy proporlionaI change in lhe
penelralion of Iiquid Ienglh as recenl research shovs, (Sielers, 1998), (Verhoeven el
aI., 1998) and (SchnaIzing el aI., 1999).
4. Working fIuid lenperalure (Tg): vorking fIuid lenperalure is one of lhe
lhernodynanic properlies lhal slrongIy affecl Iiquid Ienglh penelralion, since lhe
rale of conluslilIe vaporizalion is direclIy reIaled lo lhe energy conlenl of lhe
vorking fIuid in lhe inside of lhe cyIinder (e.g., high lenperalures) and in lhe
degree of lhe nixlure of lolh fIuids (injecled fueI-gas or air) (Chrisloph & Dec,
1995). Hovever, vorking fIuid lenperalure has no reIevanl effecl al high pressure
injeclion lecause lolh, an increase in lhe speed of injeclion and lhe anounl of fueI
injecled, ease lhe effecl vilh respecl of Iov pressures, (Zhang el aI., 1997). An
increase in vorking fIuid lenperalure al conslanl densily causes and increase in
lhe specific energy of lhe Ialler and lherefore a decrease in Iiquid Ienglh during
spray penelralion is a consequence of high drag of vaporizalion energy lovards
lhe fueI, (Sielers, 1999).
5. IueI lenperalure (T
f
): fueI lenperalure is a varialIe lhal grealIy affecls Iiquid
Ienglh penelralion in such a vay lhal on increasing lhe lenperalure of lhe Ialler
Iiquid Ienglh lends lo decrease IineaIIy. Il has leen proven lhal al under condilions
of Iov lenperalure and vorking fueI densily lhere are nore significanl effecls lhal
under high condilions of lenperalure and densily, lecause in lhe Ialler case lhe
effecl vilch respecl an alsoIule scaIe is insignificanl, (Sielers, 1998).
6. IhysicaI-ChenicaI properlies of lhe fueI: lhese properlies of lhe fueI (i.e., densily,
viscosily and voIaliIily) have a consideralIe inpacl on Iiquid Ienglh penelralion
Fuel njection 30

vilh voIaliIily leing lhe nosl infIuenliaI properly on penelralion. (Sielers, 1998,
1999) olserves lhal a Iov voIaliIily fueI requires nore energy lo le healed and lhen
evaporale lhan a high voIaliIe fueI. Therefore, for a Iov voIaliIe fueI, Iiquid Ienglh
penelrales nuch nore lhan a nore voIaliIe fueI lecause lhe anounl of energy
dragged lovards lhe fueI depend lasicaIIy on lhe process of evaporalion.

Liquid Ienglh of a dieseI spray is a paraneler of nuch inleresl in lhe sludy of lhe injeclion-
conluslion process. In Ialer lopics in lhis sane chapler ve viII discuss lhis paraneler
vhere a conpIele experinenlaI anaIysis of lhe characlerizalion of lhe Iiquid Ienglh of a
dieseI spray is approached.

3. Microscopic Characteristics
The nacroscopic descriplion is characlerized ly lhe conlenl of dropIels of diverse sizes and
lhe changes on lhe changes in lheir speciaI kinelics. Ior exanpIe, lhe alonizalion
nechanisn is responsilIe for dislriluling lhe dropIels in lhe injeclion process and lo a greal
exlenl lhe good dislrilulion of lhe dropIels in reIalion lo lheir size depend on il. CeneraIIy
lhe quaIily of lhe alonizalion of a Iiquid spray can le eslinaled on lhe nediun dianeler of
lhe dropIels. A delernined nediun dianeler represenls lhe equivaIenl dianeler lhal
characlerizes lhe enlire group of lhe dropIels of lhe spray. Lqualion (3O) eslalIishes lhe
generaI forn lased on vhich aII lhe correIalions lhal delernine Saulers nediun dianeler
have leen defined.

_
_
n
k
i
i =1 i
n- n nn n
k
i
i =1 i
D N
D =
D N
(3O)

Where Ni is lhe nunler of dropIels of lhe group vilh dianeler Di. CeneraIIy speaking,
nediun dianelers are used lo sinpIify caIcuIalion and anaIysis of dala. Mediun dianeler
is lhal vhich defines lhe characlerislics of a popuIalion of drops presenl in a sanpIe. In
sone processes Saulers nediun dianeler is used, vhich represenls lhe dianeler of dropIels
vhich have lhe sane voIune/surface reIalion in lhe lolaIily of lhe spray, as veII as lhe
arilhnelic average dianeler (D1O) vhich are represenled ly lhe foIIoving respeclive
equalion:

_
_
3
k
i =1 i
2
n
i =1 i
D
SMD=
D
(31)

_
_
k
i
i =1
1O k
i
i =1
D
D =
N
(32)

Il nusl lake inlo accounl lhal using nediun dianelers is very usefuI lo sinpIify dropIel
popuIalions exisling in an alonizing process. Ior lhis reason il is essenliaI lo use lhe
dislrilulion of dropIel size.

3.1. DropIet size distribution
The dianeler of lhe dropIels ollained as a resuIl of alonizalion is lased on a series of
paranelers as foIIovs:
1. Rale of injeclion: lhe dianeler of lhe dropIel increases vilh lhe rale of injeclion as
an increase in lhe voIune of lhe injecled Iiquid produces a grealer drag of lhe
vorking fIuid, lhe aerodynanic inleraclion grovs and lhe crilicaI size of lhe
dropIels increases. Aparl fron lhis, increasing lhe nuneric popuIalion of dropIels
inlensifies de coaIescence, resuIling in a grovlh in lhe geonelry of lhe dropIels.

2. Densily ralio (
*
): lhe reIalion of densilies has lvo opposing effecls on lhe size of
lhe dropIels, inlensificalion of alonizalion and lhe possiliIily lhal lhere viII le
coaIescence. On increasing lhe reIalionship of densilies a grealer aerodynanic
inleraclion exisls, vhich causes lhe dropIels lo sIov dovn and an increase in lhe
nunericaI popuIalion in lheir fieId.

3. Working fIuid lenperalure (Tg): on increasing vorking fueI lenperalure lheir is an
increase on lhe rale of evaporalion, due lo vhich al lhe leginning of lhis lhe
dropIels vilh snaII dianelers lend lo evaporale conpIeleIy vhiIe lhose dropIels
vilh grealer dianelers nainlain a slalIe geonelry unliI lhey evaporale conpIeleIy.

4. SpaliaI evoIulion of lhe size of lhe drops: lhe average size of lhe dropIels lends lo
grov in reIalion lo lhe increase of lhe dislance lelveen lhe drops and lhe injeclor
poinl. In sone sludies il has leen suggesled lhal lhe average dianeler of lhe drops
is grealer in lhe direclion of lhe radius of lhe spray vhiIe olher suggesl lhe
opposile, lhal is lhe nediun dianeler is reduce in reIalion lo lhe dislance fron il.

5. LvoIulion of lhe dianeler of dropIels during line: Ils generaIIy considered lhal lhe
nediun dianeler of lhe dropIels decreases al lhe poinl of lhe spray and increases
al lhe laiI, vhiIe in areas dislanl fron lhe injeclor lhey nainlain a rale of conslanl
vaIues. CeneraIIy speaking, lhe sizes of lhe dropIels lend lo dininish al lhe
leginning of lhe injeclion and grov al lhe end.

The nosl connon fornuIas lo delernine Saulers nediun dianeler are:

Saulers nediun dianeler according lo (Hiroyasu & Kadola, 1974):

( )
| | | |
| |
| |
\ . \ .
O, 54 O,18
O,12
-O,54 I I
g g

SMD= 4, 12d Re We

(33)

Where A leing an experinenlaI conslanl (A = 233O) and Q lhe injecled voIune |n
3
j

Fuel njection 32

Saulers nediun dianeler according lo (Hiroyasu & Arai, 199O) and (Hiroyasu el aI., 1989)
1. Ior inconpIele spray
( )
| | | |
| |
| |
\ . \ .
O,37 -O, 47
O,25
-O, 32 I I
g g

SMD= O, 38d Re We

(34)

2. Ior conpIele spray

( )
-O, 28
I I o
SMD= 8, 7 Re We d (35)

These fornuIae have leen lhe nosl used lo delernine Saulers nediun dianeler, even
lhough lhese correIalions experinenlaIIy ollained have leen nodified over lhe years, lhey
nainlain a very inporlanl lasis in vhich lo delernine Saulers nediun dianeler. Lach of
lhese fornuIae nay experience furlher nodificalions and leller approxinalions according
lo lhe quaIily of lhe specific nodeI or experinenl.

4. Measurement techniques
Sone prolIens of fIuid nechanics are conpIex vhere nuIliphase syslens are concern and
vhen conluslion phenonena are produced. In nany cases currenl knovIedge is sliII
inconpIele due lo lhe conpIexily of lhe physicaI-chenicaI processes: (non-slalionary
processes, irreversilIe processes and oul-of-laIance chenicaI reaclions) lhal occur al lhe
Iinils of differenl scienlific discipIines such as fIuid nechanics, lhernodynanics and
chenislry. In order lo progress in ils sludy ve need avaiIalIe experinenlaI dala lhal
provide infornalion of lhe differenl processes and degrees of inleresl for lhe sludy, such as
for exanpIe, nass and energy lransporl, novenenl and lhe size of parlicIes, concenlralion
of lhe differenl species, lhernodynanic properlies, and chenicaI conposilion anong
olhers.
The physicaI phenonena of inleraclion naller-radialion (alsorplion, dispersion,
inlerference, diffraclion, anong olhers) are very sensilive lo snaII varialions in lhe IocaIize
physicaI paranelers of lhe fIuid, and furlhernore lhey do nol inleracl vilh lhe physicaI
processes in lhe environnenl of fIuid nechanics, and so are usefuI in lhe anaIysis of lhese
prolIens. TechnoIogicaI advance in diverse fieIds lasicaIIy oplics, eIeclronics and
infornalion lechnoIogy have aIIoved for lhis deveIopnenl of equipnenl alIe lo neasure
sone IocaIized physicaI paranelers of fIuids in a very precise vay, and are lhe lasis for lhe
deveIopnenl of oplicaI lechniques of neasurenenl and visuaIizalion used in sludies of
fIuid nechanics.

4.1. CIassicaI visuaIization techniques
The cIassicaI visuaIizalion nelhods are lased on lhe varialions of lhe refraclion rale lhal are
produced in lhe fIuids hearl due lo lhe changes in ils physicaI properlies. When an lean of
Iighl propagales lhrough a fIuid, lhe varialions of lhe refraclion rale causes varialions in
lolh lhe inlensily and in vave phase, lherefore lhe energing Iighl conlains infornalion of
lhe fIuid properlies in lhe Iighl lean lrajeclory propagalion. asicaIIy lhese oplicaI
lechniques can le divided in 3 lypes: Shadovgraphy, SchIieren and Inlerferonelry, vhich

have leen used since lhe 186Os, (IoucauIl, 1859) in Irance and (ToepIer, 1864) in Cernany
gave lhe firsl insighls of lhe SchIieren lechnique. ToepIer vas lhe firsl lo deveIop lhis
lechnique for lhe sludy of Iiquids and gas fIov, and Ialer on used ly (Hayashi el aI., 1984)
and (Konig & Sheppard, 199O), anong olhers.

-Shadovgraphy: lhe environnenl is iIIuninaled vilh a slraighlening of a Iighl lean and lhe
inage is laken afler lhe energing Iighl propagales freeIy lhrough lhe space. The
visuaIizalion lechnique vilh diffused rear iIIuninalion is a siniIar lechnique lul lhe
environnenl is Iil up vilh a diffuse lean Iighl. The difference lelveen lhese lechniques
consisls on pIacing a diffuser lelveen lhe lean and lhe environnenl lo iIIuninale. These
lechniques aIIov visuaIizing lhe Iiquid phase of lhe fueI spray and are grealIy used in lhe
sludy of lhe injeclion process of conluslion inlernaI engines. The visuaIizalion vilh rear
diffused iIIuninalion lechnique aIIovs lhe eslinalion of lhe differenl nacroscopic
paranelers in an injeclion process. (Zaho & Ladonnalos, 2OO1) have sludied lhe spray
penelralion and consider lhis lechnique lo le reIialIe and easy lo use for lhis lype of
anaIysis.

-SchIieren pholography: lhis lechnique is siniIar lo lhal of lhe shadovgraphy, lhe difference
is lhal lhe inage is laken afler a spaliaI fiIlering in lhe inage pIane of lhe Iighl source.
Adjusling adequaleIy lhe spaliaI fiIlering dinensions il is possilIe lo visuaIize lolh lhe
Iiquid and vapour phase of lhe fueI spray, lul nol lo quanlify lhen. These lechniques have
leen used in lhe injeclion and conluslion processes of lhe inlernaI conluslion engine
(Ireussner el aI., 1998), (Spicher & KoIIneire, 1986) and (Spicher el aI., 1991), as veII as in
lhe anaIysis of propuIsion syslens (Murakanis & Iapanoschou, 2OO1) and (Iapanpschou,
2OOO).

4.2. Scattering techniques
The cIassicaI visuaIizalion lechniques incorporale lhe infornalion lhroughoul lhe leans
propagalion lrajeclory, ly vhich lhe infornalion aloul lhe exisling lhree-dinensionaI
slruclures in lhe vesseI of lhe fIuid is Iosl. This infornalion can le ollained iIIuninaling lhe
fIuid vilh pIanes of Iighl and laking piclures of lhe dispersed Iighl ly lhe environnenl,
nornaIIy in lhe perpendicuIar direclion of lhe pIane. This kind of visuaIizalion lechniques
can le incIuded in a nuch generaI group vhich is lhe scallering lechnique. The Iighl
scallering phenonena can le of lvo lypes, eIaslic or ineIaslic, depending on if lhe process
produces or nol lhe radialion frequency.

4.2.1. EIastic scattering techniques
The eIaslic dispersion phenonena of Iighl are sludied vilhin lhe lheory of Lorenz-Mie.
There are lasicaIIy lvo approxinalions depending on lhe size of lhe parlicIes: Mie
scallering and RayIeigh scallering.

-The Mie scallering is an inleraclion of lhe eIaslic lype of Iighl vilh parlicIes of nuch grealer
size lhan lhal of ils vave Ienglh (dropIels, Iiganenls, anong olhers). The characlerislics of
lhe scallered Iighl are reIaled lo lhe forn, size, refraclion rale and nunler of scallering
parlicIes. These properlies are lhe lasis of lhe differenl oplicaI lechniques of neasurenenl
descriled as foIIovs:

Fuel njection 34

1. VisuaIizalion vilh a Iaser sheel lhe fIuid is iIIuninaled vilh a Iaser sheel lean
ollaining inages of lhe scallered Iighl (Mie regine), nornaIIy on lhe perpendicuIar
direclion of lhe sheel. This lechnique aIIovs eslinaling lhe nacroscopic characlerislics
of fueI sprays and anaIysing lhe exislence of inlernaI slruclures, Iiganenls, anong
olhers. This lechnique is one of lhe nosl used in lhe sludy of lhe injeclion process in an
inlernaI conluslion engine (Dec, 1992) and (Ireussner el aI., 1998).
2. Technique of Iaser anenonelry: il is lased on lhe inleraclion of coherenl Iighl vilh lhe
exisling parlicIes in novenenl inside lhe hearl of lhe fIuid in such a vay lhal lhe sizes
of lhese parlicIes aIIov lhen lo le lrealed in Mie scallered inaging. These inleraclions
produce a change in lhe frequency of radialion (DoppIer Lffecl) lhal can le reIaled lo
lolh lhe speed and size of lhe parlicIes. In lhe so caIIed Laser DoppIer Anenonelry
(LDA), lvo coherenl Iighl leans inleracl in one region (conlroI voIune) vilh lhe
exisling noving parlicIes in lhe fIuid and lhe fIuclualion of lhe disseninaled Iighl
inlensily aIIovs lhe eslinalion of lhe parlicIes speed. (The ollained Iighl inlensily is
lasicaIIy inlensily vilh a lackground noduIaled ly a cosine funclion, vhose lenporaI
varialion depends soIeIy on lhe frequencies of lhe dispersed leans. The frequency of
noduIalion for lhis signaI can le reIaled lo lhe veIocily of lhe parlicIes). The Ihase
DoppIer Anenonelry (IDA) is lased on lhe sane principIe lul il uses severaI pholo
sensors pIaced in differenl spaliaI posilions. Wilh vhich ils possilIe lo eslinale lhe
dianeler of lhe diffusive parlicIes considering lhen sphericaI ly lhe lenporaI phase Iag
lelveen signaIs received ly each pholo deleclor. This lechnique requires a series of
oplicaI accessories lhal difficuIl ils use in neasurenenl of a reaI lherno engine.
AIlhough sone invesligalors (Aurienna el aI., 2OO1), (Corcione el aI., 1998), (CossaIi el
aI., 1996), (Ceorjon el aI., 1997) and (Cuerrassi & Chanpoussin, 1996) have used lhe
phase DoppIer anenonelry lo deveIop very specific anaIysis, lhe nayor usage is sliII
lhe characlerizalion of lhe dislrilulion of dianelers and veIocilies of fueI dropIels in
accessilIe oplicaI nodeIs lhal sinuIale siniIar condilions of lhose found in reaI lhernaI
engines (ArrgIe, 1998) and (}inenez el aI., 2OOO).
3. The veIocinelry inaging lechniques aIIov veIocily fieId neasuring in a fIuids pIane
lhal is iIIuninaled vilh a screen of Iighl. There are severaI vays lo use lhese lechniques,
depending on lhe nelhod seIecled lo regisler and lo process infornalion, hovever aII
of lhen are very inporlanl: in IarlicIe Inage VeIocinelry (IIV) lhe fIuid is iIIuninaled
vilh severaI Iighl puIses and lhe inslanl inages are regislered using nuIlipIe exposure
lechniques. The inslanl veIocilies are ollained dividing lhe parlicIes dispIacenenl in
each line conseculive inage ly lvo puIses. In IarlicIe Shadov VeIocinelry lhe fIuid is
iIIuninaled in a Iong period of line in vhich lhe dispIacenenl of lhe parlicIes are
regislered as Iines on lhe inage and lhe veIocilies are caIcuIaled dividing lhe Iine Ienglh
ly line inlervaI. In IarlicIe Tracking VeIocinelry a series of conseculive exposures lake
pIace (severaI Iighl puIses) in one inage and lhe veIocily is eslinaled ly lracking lhe
parlicIes. The veIocinelry lechniques are used nainIy lo anaIyse fIov of gases en lhe
lhernaI engine. Sone of lhe nosl recenl appIicalions for lhis lechnique can le found in
lhe Iileralure (Choi & Cuezennec, 1999), (Kakuhou el aI., 1999), (Nauverck el aI., 2OOO),
(Neusserl el aI., 1995), and (Trigui el aI., 1994), vhere lhe nain appIicalion is focussed
lo lhe sludy of nixlure fornalion inside lhe conluslion chanler of a lhernaI engine,
furlhernore il considered lo le one of lhe lesl lechniques for lhis kind of anaIysis.

4. RayIeigh scallering is of lhe eIaslic kind, vhere lhe size of parlicIes is nuch snaIIer of
lhal of lhe Iighls vaveIenglh, for exanpIe lhe gas noIecuIes. The inlensily of lhe
scallered Iighl is proporlionaI lo lhe lolaI densily of aII kinds of exisling parlicIes inside
lhe iIIuninaled zone and provides inages of gIolaI concenlralion of aII lhe species,
aIlhough il doesnl aIIov discrininalion lelveen lhen. Iurlhernore for exanpIe, lhe
RayIeigh signaI for a parlicIe approxinale 1 n is cIose lo lvenly orders of nagnilude
Iover lhan lhe Mie signaI, for vhich lhe signaI is highIy affecled ly lolh lhe presence of
Iarge parlicIes and ly lhe lackground Iighl. The lvo nosl connonIy used procedures
lo reduce lhe inlerference of parlicIes are shovn ly (Zhao el aI., 1993). The nain
researchers using lhe RayIeigh lechnique (Lspey & Dec, 1994), (Lee & Iosler, 1995) and
(Zhao el aI., 1991) have leen lasicaIIy lo delernine concenlralions of vapour and Iiquid
phases and nainIy in zones vilh high fIane presence. As veII as in lhe lenperalure
neasure and species concenlralion for lhe conluslion diagnoslic.

4.2.2. IneIastic scattering techniques
On lhe olher hand, lhe ineIaslic scallering of Iighl is sludied in lhe quanlun nechanics fieId,
specificaIIy in lhe sludy of naller-radialion inleraclion phenonena. These phenonena are
exlreneIy sensilive lo lhe frequency of radialion and lhe species chenicaI conposilion
lecause lhey depend on eIeclronic lransilions lelveen noIecuIar energy IeveIs caused ly
lhe alsorplion of pholons of defined frequency lhal slinuIale lhe noIecuIes lo higher
energelic condilions. Afler vhich lhe noIecuIes cone lo slalIe condilions reIeasing radianl
energy vhere ils speclraI characlerislics are aIso very veII defined. Differenl oplicaI
lechniques of neasure are lases on lhese phenonena, delaiIed as foIIovs:

-Laser Induced Incandescence is a lechnique lased on lhe lhernaI enission lhal is produced
vhen lhe carlon parlicIes are slinuIaled vilh a very inlense eIeclronagnelic radialion. The
ollained signaI is proporlionaI lo lhe voIune fraclion of lhe carlon parlicIes concenlraled in
lhe neasured zone. ecause of lhis, lhe lechnique is very usefuI for lhe sludy of conluslion
processes (Dec, 1992), (Dec el aI., 1991), (Dec & Lspey, 1992), (WinkIhoefer el aI., 1993) and
(Zhao & Ladonnalos, 1998), nainIy lo delernine lhe quaIilalive dislrilulion of sool in lhe
high radialion zone during a injeclion-conluslion process.

-Laser Induced IIuorescence (LII) is a lechnique lased on lhe fIuorescenl properlies lhal
sone noIecuIes presenl. When lhese noIecuIes alsorl eIeclronagnelic energy of a
delernine frequency lhey acquire a higher energelic condilion (slinuIalion) and aflervards
lhey relurn lo lheir originaI energelic slale reIeasing lhis energy (fIuorescence). The speclraI
characlerislics of lhis radialion are delernined ly lhe noIecuIes characlerislics. If lhe fIuid
doesnl have fIuorescenl noIecuIes, noIecuIar lracers lhal presenl fIuorescence can le
added. Ior exanpIe: NO, NO
2
, acelone, liacelyI, rodanina, or differenl coIoranls. The
fIuorescenl signaI is proporlionaI lo lhe densily of lhe lracers inside lhe iIIuninaled zone. In
nany cases lhe environnenl is iIIuninaled using Iaser lean sheel and lhe lechnique is lhen
knovn as pIanar induced Iaser fIuorescence (ILII). In lhe pIanar Iaser induced excipIex
fIuorescence (ILILI) lracers caIIed excipIex (conpIex excilalion), Iike for exanpIe:
naphlhaIene nixlures and TMID (lelranelhyI-1,4-phenyIenedianine) lhal aIIov lo separale
speclraIIy lhe corresponding Iiquid and vapour phase fIuorescence of a liphasic syslen, and
lherefore neasure sinuIlaneousIy each ones concenlralion (}uIi, 2OO3). AIlhough lhis
Fuel njection 36

lechnique has nuch appIicalion in injeclion-conluslion processes (IeIlon el aI., 1995),
(Iujinolo el aI., 1997), (Hiroshi el aI., (1997), (Kido el aI, 1993) and (Kin & Chandhi, 2OO1), il
is nol considered lo le lhe nosl appropriale lo delecl species vhen conpared lo olher, Iike
for exanpIe: Mie-Scalerring. This is due lo lhe incoherencies presenled vhen delecling
species in lhese lypes of processes (Ireussner el aI., 1998) and (Takagi el aI., 1998).
Ihosphorescenl parlicIe lracking (IIT) is a siniIar lechnique lo lhal of parlicIe lracking
veIocinelry (ITV). The phosphorescence is an ineIaslic diffusion of Iighl characlerized ly il
Iong lenporaI duralion, nuch higher lhan lhal of fIuorescence, vhich nakes il ideaI lo lrack
lhe novenenl of parlicIes in lhe fIuid.

5. ExperimentaI characterization of the Iiquid Iength penetration
5.1. Introduction
The nain oljeclive of lhis seclion is lo carry oul lhe characlerizalion of lhe Iiquid Ienglh
penelralion of a dieseI spray. To achieve lhis il has leen necessary lo consider a group of
experinenls vhich aIIov lhe delerninalion of lhe infIuence lhal lhe injeclion paranelers
and lhe lhernodynanic varialIes have upon lhe penelralion of a dieseI spray in evaporalive
condilions. The firsl deveIoped sludy is lased on lhe anaIysis of lhe penelralion of lhe spray
in ils Iiquid phase, vhere il is expecled lo define lhe degree of infIuence lhal lhe foIIoving
have over lhis phenonena: lhernodynanic varialIes (pressure, lenperalure and densily)
presenl in lhe conluslion chanler al lhe nonenl vhen lhe fueI is injecled, lhe injeclion
pressure and lhe geonelry of lhe nozzIe. To nake lhis sludy ils necessary lo use lhe
onlroscopy lechnique for lhe laking of digilaI inages, as veII as an acquisilion syslen lo
process dala. Il is lo poinl oul lhal lhe onlroscopy has leen lhe nosl used lechnique in lhe
nacroscopic characlerizalion of dieseI sprays, specificaIIy in lhe sludy of lhe Iiquid phase
penelralion. As nenlioned in seclion 4, lhe lechniques of neasure lo carry oul sludies of lhe
Iiquid phase of dieseI sprays are very diverse. The nosl used unliI knov are expressed in
lhis chaplers Iileralure. (CanleII el aI., 1995), (Canaan el aI., 1998), (Chrisloph & Dec, 1995),
(IeIlon el aI., 1995), (Hiroyasu & Miao, 2OO2) and (Knapp el aI., 1999).

5.2. ExperimentaI work approach
A vorking pIan lhal groups lhe differenl experinenls lo carry oul has leen slruclured in
such a vay lo anaIyse quaIilaliveIy lhe injeclion process. To achieve lhis, lhe experinenlaI
vork has leen pIanned as foIIovs:

The use of lhe experinenlaI in syslen vilh lhe inerl alnosphere nelhod and lhrough lhe
onlroscopy lechnique anaIyse lhe penelralion of lhe Iiquid phase of lhe dieseI spray.
- Iaranelric anaIysis lo consider:
1. InfIuence of lhe injeclion process on lhe Iiquid Ienglh penelralion.
2. InfIuence of lhe dianeler of lhe nozzIe on lhe Iiquid Ienglh penelralion.

The anaIysis of lhe Iiquid Ienglh penelralion is usefuI lo delernine lhe geonelric design of
conluslion chanlers for high speed regine dieseI engines vilh direcl injeclion. Ior
exanpIe, in Iov speed regine and Iighl Ioad lhe hydrocarlon enissions viII le reduced if
lhe conlacl of lhe spray (Iiquid Ienglh) vilh lhe conluslion chanlers vaII is avoided. Ior
high speed regines and heavy Ioads, lhe reduclion of funes can le achieved ly conlacl

lelveen lhe spray and lhe chanler vaII. ecause of lhese, lhe necessily lo neasure lhe
Iiquid penelralion in dieseI engines of direcl injeclion energes, nolivaling lhe use of
neasure lechniques even nore conpIex and sophislicaled.

In previous sludies (Chrisloph & Dec, 1995) invesligaled lhe effecls lhal lenperalure and
lhe fIuids densily have on lhe Iiquid phase penelralion. In lhis sludy lhey used a DieseI
engine vilch oplicaI access vievs, and lhrough lhe eIaslic-scaller lechnique lhey ollained
inages of lhe spray. (Zhang el aI., 1997) anaIyzed lhe effecls lhal lhe injeclion pressures,
dianeler of lhe nozzIe and adnission air lenperalure have on Iiquid Ienglh penelralion. Ior
lhis lhey used a conpression nachine vhich had an equivaIenl conpression ralio lo lhal
found in a DieseI engine. In lhis anaIysis an argon Iaser lean vas used as lhe Iighl source
and an L-1O canera vas aIso used lo caplure lhe inages. (Sielers, 1998) invesligaled lhe
naxinun axiaI penelralion of lhe Iiquid phase of an evaporaled dieseI spray in a chanler
of conslanl voIune, using lhe Mie-scallered lechnique for inage capluring. The nain
aIlered paranelers vhere lhe injeclion pressure, orifice dianeler of lhe nozzIe, lenperalure
and densily of lhe vorking fIuid in lhe inside of lhe chanler.

The invesligalion of lhe sprays Iiquid phase for a connon raiI syslen al high lenperalures
vas nade ly (runeaux & Lenenand, 2OO2). The varialion in paranelers in lhis
invesligalion vhere: lhe injeclion pressure, lhe lenperalure of lhe vorking fIuid and lhe
dianeler of lhe nozzIe. This sludy vas nade in a chanler siniIar lo lhe one used ly
(Verhoeven el aI., 1998), in vhich il vas possilIe lo nainlain high pressures and
lenperalures inside lhe chanler and so sinuIaling siniIar condilions found in a reaI DieseI
engine. The lechnique of neasure used vas lased on a Iighl source supporled ly a pIanar
Iaser induced excipIex fIuorescence syslen and a charged-coupIed device (CCD) canera lo
caplure inages. Ils evidenl lhal each invesligalor uses in his experinenls defined and
helerogeneous lechniques of neasure. Hovever occasionaIIy and in sone conpIexily
degree lhe finaI resuIls lend lo le very siniIar independenlIy of lhe used, reason vhy lhe
nolivalion lo deveIop lhe lasis for lhe experinenls presenled in lhis chapler arose vilh one
of lhe nosl fIexilIe visuaIizalion lechniques, lhe onlroscopy.

The characlerizalion of lhe Iiquid Ienglh penelralion of an evaporaled dieseI spray vas done
under lhe foIIoving nelhodoIogy:

1. LxperinenlaI syslen configuralion: lo underlake lhe experinenls lhal Iead lo
ollain infornalion aloul lhe Iiquid Ienglh penelralion of lhe spray vilhoul fIane,
il has leen necessary lo forn lhe experinenlaI syslen in an inerl alnosphere.
Iurlhernore lo conceive as a firsl phase lhe use of onlroscopy lechnique lo ollain
inages of lhe Iiquid phase of lhe spray (Iigure 5 shovs lhe schenalics diagran of
lhe gIolaI experinenlaI selup configuralion).

Fuel njection 38

Iig. 5. Schenalic diagran of lhe experinenlaI selup.

2. Configuralion of lhe group of experinenls: The considered group of experinenls
defines lhe varialIes lo le anaIysed, as veII lo delernine lheir infIuence on lhe
Iiquid Ienglh penelralion of lhe spray. The nain varialIes for sludy are:
-Injeclion pressure.
-Orifice dianeler of lhe nozzIe.
-Working fIuid densily conslanl.

Iigure 6 shovs lhe schenalics of lhe nozzIe lhal has leen used in lhe experinenls. Il has
leen experinenled vilh five nozzIes of siniIar geonelry vilh singIe axisynelric orifice and
sane kind of jackel.

Iig. 6. Schene of lhe nozzIe used in lhe experinenls.

Iour nozzIes vere lesled al four differenl injeclion pressures, vhiIe lhe inlake lenperalure
and pressure vere kepl conslanl al 7O C and 1.3 lar, respecliveIy. The four nozzIes have
singIe axisynnelric hoIes vilh 115, 13O, 17O and 2OO n in dianeler, and lhe injeclion
pressure vas 3OO, 7OO, 11OO and 13OO lar. TalIe 1 shovs lhe eslinaled nass fIov rales and
discharge coefficienls for each nozzIe and injeclion pressure. A diagnoslic lhernodynanic
nodeI deveIoped ly (Marlnez el aI., 2OO7) vas enpIoyed lo caIcuIale lhe vorking fIuid
properlies (lenperalure and densily) in lhe cyIinder. CyIinder pressure vas neasured vilh

a lransducer inslaIIed on a IaleraI vaII. The pressure al lollon dead cenler vas neasured
vilh a resislive lransducer Iocaled lelveen lhe prechanler inlake and lhe chanler ilseIf. A
lenperalure sensor vas aIso inslaIIed in lhe prechanler inlake lo neasure lhe vorking
fIuid lenperalure al lollon dead cenlre. Since pressure and lenperalure dala vere
avaiIalIe, lhernodynanic condilions vere characlerized al lop dead cenler 3 crank angIe
degrees, vhich is considered lhe nosl slalIe region during lhe fueI injeclion process
(Marlnez el aI., 2OO7).

Injeclion
pressure
(lar)
NozzIe
dianeler
(n)
Measured nass
fIov rale
(g/s)
TheorelicaI
nass fIov rale
(g/s)
C

3OO 115 1.53 2.O4 O.746
7OO 115 2.52 3.38 O.745
11OO 115 3.13 4.32 O.725
13OO 115 3.34 4.72 O.7O8
3OO 13O 2.27 2.61 O.87O
7OO 13O 3.5O 4.32 O.81O
11OO 13O 4.O5 5.52 O.734
13OO 13O 4.42 6.O3 O.733
3OO 17O 3.36 4.46 O.753
7OO 17O 5.32 7.38 O.721
11OO 17O 6.47 9.43 O.686
13OO 17O 6.87 1O.3O O.666
3OO 2OO 3.63 6.18 O.587
7OO 2OO 6.74 1O.2O O.66O
11OO 2OO 8.53 13.1O O.653
13OO 2OO 9.29 14.3O O.651
TalIe 1. Injeclion paranelers and lheir corresponding nass fIov rales and discharge
coefficienls.

5.3. MathematicaI correIation
Liquid phase penelralion of a jel injecled inlo an inerl environnenl has veII defined slages.
The firsl slage legins vilh lhe injeclion and ends vhen lhe jel lreaks up. This is lhe inlacl
Ienglh slage or lhe firsl lreak-up regine, (Hiroyasu & Aray, 199O) suggesled lhe foIIoving
correIalion lo eslinale lhe line for lhe firsl lreak-up regine lo occur:

f n
l
d a
15.8 d
l =
C 2 I
(36)

vhere Cd is lhe discharge coefficienl, dn (n) is lhe nozzIe dianeler, I (Ia) is lhe pressure
drop lhrough lhe nozzIe, and
f
and
a
(kg/n
3
) are lhe fueI and vorking fIuid densilies,
respecliveIy. Ior lhe parlicuIar condilions sludied here, Lqualion (36) predicls lines for lhe
firsl lreak-up regine lelveen 25 and 3O s, and our experinenlaI neasurenenls indicale an
average line of 5O s. LxperinenlaI evidence (Ahnadi el aI., 1991), (Aurienna el aI., 2OO1),
(Chrisloph & Dec, 1995) and (Marlnez el aI., 2OO7) indicales lhal lhe Iiquid penelralion
Fuel njection 40

Ienglh, LL, increases proporlionaIIy lo lhe square rool of line fron lhe injeclion onsel unliI
lhe second lreak-up regine is reached al line lr. Thereafler lhe Iiquid penelralion Ienglh
varies IillIe and hence il is considered conslanl fron a nacroscopic poinl of viev. Therefore,
a nalhenalicaI correIalion suilalIe lo nodeI lhe Iiquid penelralion Ienglh is:

( )
r
O < l < l : LL l = l (37)

( )
r nax
l > l : LL l = Cle = LL (38)

vhich is iIIuslraled in Iigure 7. Coefficienls and LL depend on nunerous paranelers,
such as lhe fIuid lhernodynanic condilions and geonelricaI paranelers of lhe injeclion
syslen. A salisfaclory nalhenalicaI correIalion nusl lake inlo accounl lhe effecl of lhe
nozzIe dianeler, lhe discharge coefficienl, lhe injeclion pressure, and lhe vorking fIuid
densily. These paranelers have leen previousIy found lo le enough lo characlerize lhe
Iiquid penelralion Ienglh (ae & Kang, 2OOO), (ae el aI., 2OOO), (erndez el aI., 2OO2, 2OO3),
(racco, 1983), (Canaan el aI., 1998) and (Chehroudi el aI., 1985). Il is lherefore expecled lhal
a delaiIed anaIysis of lhese paranelers can yieId an accurale correIalion lhal can le of
assislance in lhe successfuI designing of conluslion chanlers required ly nodern heavy
duly dieseI engines. In lhis paper ve allenpl pover Iav correIalions for and LLnax
(Lqualions 39 and 4O).

Iig. 7. IIol shoving differenl slages of lhe considered nodeI.

A C D
n a iny d
d I C (39)

L I C H
nax n a iny d
LL d I C (4O)

5.4 Determination of the fueI injection onset
The fueI injeclion onsel can le delernined assuning lhal LL increases proporlionaIIy lo lhe
square rool of line unliI lhe second lreak-up regine is reached al lr, i.e. LL = l
1/2
for
O < l < lr. Tine lr is defined as lhe line vhen lhe ralio lelveen LL lo l
1/2
vilh a correIalion
coefficienl R
2
= 99 . Coefficienl is eslinaled ly filling experinenlaI dala neasured lefore
lhe second lreak-up regine is reached, as shovn in Iigure 8, vhere lhe experinenlaI dala
can le approxinaled ly LL = 1.O7 l
O.497
vilh a correIalion coefficienl R
2
= 99.8 .

Iig. 8. Lslinalion of and lhe fueI injeclion onsel

5.5. Determination of the discharge coefficient
The discharge coefficienls of each nozzIe hoIe al lhe injeclion pressures sludied here vere
eslinaled using lhe foIIoving correIalion:

f
d
f
n
C =
An 2I
(41)

vhere lhe discharge coefficienl C
d
is defined as lhe ralio of lhe nass fIov rale injecled in lhe
cyIinder and lhe lheorelicaI nass fIov rale conpuled fron lhe ernouIIi equalion. The
nass fIov rale of fueI injeclion vas neasured ly a fueI rale indicalor (LVI-IAV).
LxperinenlaI neasurenenls provided enough dala lo eslinale lhe discharge coefficienl for
each nozzIe and injecled condilion, vhich are shovn in TalIe 1.

6. ResuIts and discussion
Lqualion 42 is lhe lesl fil for predicling penelralion Ienglh in lhe fueI injeclion process
lefore lhe second lreak-up regine, vhen :
U
W W < < 0

( )
1
O.56 -.O27 O.23 O.O8
2
n a iny d
LL l = 6.47d I C l (42)

Iigures 9 (a, l) and 1O (a, l) shov a conparison lelveen caIcuIaled (Lqualion 42) and
experinenlaI Iiquid penelralion Ienglhs. In aII cases curves and experinenlaI dala are in
good agreenenl and lhe correIalion coefficienl is R
2
= 93.3 , vhich neans onIy 6.7 of aII
dala are nol accounled ly lhe proposed correIalion. AnaIyzing Lqualion 42 ve find lhal lhe
Iiquid Ienglh penelralion is slrongIy affecled ly lhe nozzIe dianeler vhose exponenl in
Lqualion 42 is grealesl. The densily of lhe vorking fIuid and lhe injeclion pressure have
conparalIe and inverled effecls on lhe Iiquid penelralion Ienglh, oLL/oa ~ (I
inj
/
a
)
(oLL/oI
inj
) or oa/oI
inj
~ (
a
/I
inj
). AddilionaIIy ve nolice fron Lqualion 42 lhal lhe Iiquid
veIocily penelralion, oLL/ol, is proporlionaI lo I
inj
O.23
, vhich is lhe sane proporlionaIily as

Fuel njection 42

for LL ilseIf. On lhe olher hand, an increase in lhe vorking fIuid densily causes lhe Iiquid
penelralion resislance lo rise, vhich yieIds a shorlening in lhe Iiquid penelralion Ienglh. Il is
vorlh nenlioning lhal lhe effecl of a on LL reporled here is in good agreenenl vilh
experinenlaI dala presenled ly (Denl, 1971), vho suggesled lhe foIIoving correIalion:

( )
1
-O.25
2
a
LL l l (43)

Lqualion 42 reveaIs lhal under lhe experinenlaI condilions sludied here, O.58 < Cd < O.87,
lhe Iiquid penelralion Ienglh is very insensilive lo lhe vaIue of lhe discharge coefficienl,
vhich causes a naxinun varialion of lhe Iiquid penelralion Ienglh of onIy aloul 3 .

Iig. 9. Conparison lelveen experinenlaI dala and lhe proposed correIalion, equalion 42.
(a): I
inj
= 3OO lar and (l): I
inj
= 7OO lar,
a
= 26 kg/n
3
and Tg = 9O6 K.

Iig. 1O. Conparison lelveen experinenlaI dala and lhe proposed correIalion, equalion 42.
(a): I
inj
= 11OO lar and (l): I
inj
= 13OO lar,
a
= 26 kg/n
3
and Tg = 9O6 K.

(a)
(l)
(a)
(l)

7. ConcIusions and remarks
LxperinenlaI neasurenenls vere carried oul lo eslinale lhe Iiquid penelralion Ienglh of a
dieseI fueI jel injecled in an inerl environnenl. The effecls of lhe characlerislic paranelers,
i.e. lhe nozzIe dianeler, discharge coefficienl, injeclion pressure, and vorking fIuid densily
vere anaIyzed. The lransienl fueI injeclion process vas recorded using oplicaI access, and
lhe Iiquid penelralion Ienglh lefore lhe second lreak-up regine vas neasured using lhe
onlroscopy lechnique. The ain of lhe presenl research is lo generale a correIalion lhal
accuraleIy predicls Iiquid penelralion Ienglh al condilions lypicaI of nodern Heavy Duly
connon raiI dieseI engines operaling vilh direcl fueI injeclion. A slalislicaI anaIysis of our
experinenlaI neasurenenls suggesls a pover funclion correIalion lo nodeI lhe Iiquid
penelralion Ienglh. The proposed nodeI is in good agreenenl vilh experinenlaI dala and
yieIds a correIalion coefficienl R
2
= 93.3 . Iurlhernore, lhe suggesled correIalion iIIuslrales
inporlanl delaiIs aloul hov lhe nain paranelers affecl lhe fueI injeclion process. The
nozzIe dianeler has lhe grealesl effecl on Iiquid penelralion Ienglh. A reduclion in nozzIe
dianeler yieIds a shorler penelralion Ienglh lecause il causes an earIier slarl of lhe second
lreak-up regine. Increasing lhe injeclion pressure provokes prenalure dropIel lreak-up
vilhin lhe jel, vhich resuIls nainIy due lo cavilalion al lhe nozzIe exil. If lhe vorking fIuid
densily in lhe conluslion chanler increases lhe penelralion Ienglh is shorler and lhe
second lreak-up regine is deIayed due lo lhe free-share fIov lelveen lhe vorking fIuid
and lhe fueI jel, vhich produces higher evaporalion rales of dropIels fron lhe dieseI jel.
IinaIIy, under lhe experinenlaI condilions sludied here, lhe discharge coefficienl has a
negIigilIe effecl on lhe Iiquid penelralion Ienglh. Hovever, lhe discharge coefficienl
infIuences lhe cavilalion phenonenon al lhe nozzIe exil and nodifies lhe dropIel veIocily
vilhin lhe jel.

8. References
Ahnadi efrui, WieseIer . y WinkIhofer L. (1991) The propagalion of IueI Spray in a
Research DieseI Lngine A }oinl NunericaI and LxperinenlaI AnaIysis". SAL
TechnicaI Iaper 91O181.
Arai M., Talala M., Shinizu M. y Hiroyasu H. (1984) Disinlegraling Irocess and Spray
Characlerizalion of IueI }el Injecled ly a DieseI NozzIe". SAL TechnicaI Iaper
84O275.
ArrgIe }. (1998) AnIisis de Ia Lslruclura y Dinnica Inlerna de Chorros DieseI. Tesis
DocloraI, L.T.S. Ingenieros InduslriaIes. Universidad IoIilecnica de VaIencia, Spain.
Aurienna M., Corcione I. L., DIMarlino U. y VaIenlino C. (2OO1) AnaIysis of lhe Inlake
IIov in a DieseI Lngine Head Using Dynanic Sleady IIov Condilions". SAL
TechnicaI Iaper 2OO1-O1-13O7.
ae Ch. y Kang }. (2OOO) DieseI Spray Characlerislics of Connon-RaiI VCO NozzIe
Injeclor". Congreso THILSLL-2OOO, VaIencia, Spain.
ae Ch. y Kang }. (2OOO) DieseI Spray DeveIopnenl of VCO NozzIes for High Iressure
Direcl-Injeclion". SAL TechnicaI Iaper 2OOO-O1-1254.
ae Ch., Yu }., Kang }., Cuenca R. y Lee O. (2OOO) The InfIuence of Injeclor Iaranelers on
DieseI Spray". Congreso THILSLL-2OO2, VaIencia, Spain.
Fuel njection 44

erndez V., Carca }. M., }uIi L. y Marlnez S. (2OO2) InslaIacin LxperinenlaI para eI
Lsludio deI Iroceso de Inyeccin-Conluslin en Molor DieseI de Inyeccin
Direcla". XV Congreso NacionaI de Ingeniera Mecnica, Cdiz, Spain.
erndez V., Carca }. M., }uIi L. y Marlnez S. (2OO3) Lngine vilh OplicaIIy AccessilIe
CyIinder Head: a Research TooI for Injeclion and Conluslion Irocesses". SAL
TechnicaI Iaper 2OO3-O1-111O.
racco I. V. (1983) Slruclure of High Speed IuII Cone Sprays. Recenl Advances in Cas
Dynanics, IIenun IulIishing Corporalion, N.Y.
runeaux C. y Lenenand C. (2OO2) A Sludy of Liquid Ihase Slruclure in Very High
Iressure Connon RaiI DieseI Injeclion Using OplicaI Diagnoslics". Congreso
THILSLL-2OO2, VaIencia, Spain.
CanleII I., Sinko K. y Chehroudi . (1995) Liquid and Vapour Ihase Dislrilulions in a
IiIoled DieseI IueI Spray". SAL TechnicaI Iaper 95O445.
Canaan R. L., Dec }. L. y Creen R. M. (1998) The InfIuence of IueI VoIaliIily on lhe Iiquid-
Ihase IueI Ienelralion in a Heavy-Duly D.I. DieseI Lngine". SAL TechnicaI Iaper
98O51O.
Chehroudi ., Chen S. H., racco I. V. y Onuna Y. (1985) On lhe Inlacl Core of IuII-Cone
Sprays". SAL TechnicaI Iaper 85O216.
Choi W. Ch. y Cuezennec Y. C. (1999) Sludy of lhe IIov IieId DeveIopnenl During lhe
Inlake Slroke in an I.C. Lngine Using 2-D IIV and 3-D ITV". SAL TechnicaI Iaper
1999-O1-O957.
Chrisloph L. y Dec }. L. (1995) The Lffecl of TDC Tenperalure and Densily on lhe Liquid-
Ihase IueI Ienelralion in a D.I. DieseI Lngine". SAL TechnicaI Iaper 952456.
Corcione I. L., VagIieco . M. y VaIenlino C. (1998) A Sludy of IhysicaI and ChenicaI
DeIay in a High SvirI DieseI Syslen Via MuIlivaveIenglh". SAL TechnicaI Iaper
98O5O2.
Correas D. (1998) Lsludio Terico LxperinenlaI deI Chorro Lilre DieseI Isolerno. Tesis
DocloraI, L.T.S. Ingenieros InduslriaIes. Universidad IoIilecnica de VaIencia, Spain.
CossaIi C. L., CerIa A., Coghe A. y runeIIo C. (1996) Lffecl of Cas Densily and
Tenperalure on Air Lnlrainnenl in a Transienl DieseI Spray". SAL TechnicaI Iaper
96O862.
Dec }. L. (1992) Sool Dislrilulion in a D.I. DieseI Lngine Using 2-D Inaging of Laser-
Induced Incandescence, LIaslic Scallering, and IIane Luninosily". SAL TechnicaI
Iaper 92O115.
Dec }. L., AxeI O., Loye Z. y Sielers D. L. (1991) Sool Dislrilulion in a D.I. DieseI Lngine
Using 2-D Laser-Induced Incandescence Inaging". SAL TechnicaI Iaper 91O224.
Dec }. L. y Lspey Ch. (1992) Sool and IueI Dislrilulion in a D.I. DieseI Lngine via 2-D
Inaging". SAL TechnicaI Iaper 9223O7.
Denl }. C. (1971) A asis for lhe Conparison of Various LxperinenlaI Melhods for
Sludying Spray Ienelralion". SAL TechnicaI Iaper 71O571.
Lspey C. y Dec }. (1994) Quanlilalive 2-D IueI Vapor Concenlralion Inaging in a Iiring D.I.
DieseI Lngine Using IIanar Laser Induced RayIeigh Scallering". SAL TechnicaI
Iaper 94O682.
IeIlon I. C., Kyrislsis D. C. y IuIcher S. K. (1995) LII VisuaIizalion of Liquid IueI in lhe
Inlake ManifoId during CoId Slarl". SAL TechnicaI Iaper 952464.

IoucauIl L. (1859) Menoire sur Ia Conslruclion des TeIescopes en Verre Argenle, VoI. 5, pp.
197-237. Ann. Olserv. Inp. Iars.
Iujinolo M., Talala M. y Tanaka T. (1997) IIanar Measurenenls of NO in an S.I. Lngine
ased on Laser Induced IIuorescence". SAL TechnicaI Iaper 97O877.
Ceorjon T., ChaIe H. C., Chanpoussin }. C., Marie }. L. y Lance M. (1997) A DropIel
Tagging Melhod lo Invesligale DieseI Spray". SAL TechnicaI Iaper 97O351.
CIder . L., SneIIing D. R. y SnaIIvood C. }. (1992) DieseI Spray Slruclure Invesligalion
ly Laser Diffraclion and Sheel IIIuninalion". SAL TechnicaI Iaper 92O577.
Cuerrassi N. y Chanpoussin }. C. (1996) LxperinenlaI Sludy and ModeIIing of DieseI
Spray/WaII Inpingenenl". SAL TechnicaI Iaper 96O864.
Ha }., Salo C. T., Tanale H., Iujinolo H. y Kuniyoshi H. (1983) Invesligalion on lhe
IniliaI Iarl and lhe Spray Iornalion DeIay of DieseI Spray". SAL TechnicaI Iaper
83O451.
Hay N. y }ones I. L. (1972) Conparison of lhe Various CorreIalions for Spray Ienelralion".
SAL TechnicaI Iaper 72O776.
Hayasi T., Taki M., Kojina S. y Kondo T. (1984) Iholographic Olservalion of Knock Wilh a
Rapid Conpression and Lxpansion Machine". SAL TechnicaI Iaper 841336.
Heyvood }. . (1988) InlernaI Conluslion Lngine IundanenlaIs, pp. 522-536. McCrav-HiII
InlernalionaI Ldilions.
Hiroshi N., Hiroyuki L., Yoshihiro D., Malsuhei N., Hiroshi O. y Taizo S. (1997) NO
Measurenenl in an DieseI Spray IIane Using Laser Induced IIuorescence". SAL
Hiroyasu H. y Arai M. (199O) Slruclures of IueI Sprays in DieseI Lngines". SAL TechnicaI
Iaper 9OO475.
Hiroyasu H., Arai M. y Talala M. (1989) LnpiricaI Lqualions for lhe Sauler Mean Dianeler
of a DieseI Spray". SAL TechnicaI Iaper 89O464.
Hiroyasu H. y Kadola T. (1974) IueI DropIel Size Dislrilulion in DieseI Conluslion
Chanler". SAL TechnicaI Iaper 74O715.
Hiroyasu H., Kodala T. y Arai M. (198O) IueI Spray Characlerizalion in DieseI Lngines.
Conluslion ModeIIing in Reciprocanl Lngines, Mallavi and Anann, IIenun
Iress.
Hiroyasu H. y Miao H. (2OO2) OplicaI Techniques for DieseI Spray and Conluslion".
Congreso THILSLL-2OO2, VaIencia, Spain.
}avard ., CuIari L. y Heinen N. A. (1999) Characlerislics of Inlernillenl IueI Spray. 1999.
}inenez }., Caslro I. y Cinenez . (2OOO) The Tip LvoIulion of an Lvaporalive Inlernillenl
IueI Spray". Congreso THILSLL-2OOO, VaIencia, Spain.
}uIi L. (2OO3) Medida de Concenlraciones de ConluslilIe en Chorros DieseI Medianle
Tecnicas de IIuorescencia Inducida por Lser. Tesis DocloraI, L.T.S. Ingenieros
InduslriaIes. Universidad IoIilecnica de VaIencia, Spain.
Kakuhou A., Urushihara T., Iloh T. y Takagi Y. (1999) Characlerislics of Mixlure Iornalion
in a Direcl-Injeclion S.I. Lngine Wilh Oplinized In-CyIinder SvirI Air Molion".
SAL TechnicaI Iaper 1999-O1-O5O5.
Kido A., Ogava H. y Miyanolo N. (1993) Quanlilalive Measurenenls and AnaIysis of
Anlienl Cas Lnlrainnenl inlo Inlernillenl Cas }els y Laser-Induced IIuorescence
of Cas (LIIA)". SAL TechnicaI Iaper 93O97O.
Fuel njection 46

Kin T. y Chandhi }. . (2OO1) Quanlilalive 2-D IueI Vapor Consenlralion Measurenenls in
an Lvaporaling DieseI Spray Using lhe LxcipIex IIuorescence Melhod". SAL
TechnicaI Iaper 2OO1-O1-3495.
Knapp M., Luczak A., eushausen V., HenlscheI W. y Andresen I. (1999) Vapour/Liquid
VisuaIizalion vilh Laser Induced LxcipIex IIuorescence in an SI Lngine for
Differenl Injeclion Tinings". SAL TechnicaI Iaper 961122.
Konig C. y Sheppard C. C. W. (199O) Lnd Cas Auloignilion and Knock in a Spark Ignilion
Lngine". SAL TechnicaI Iaper 9O2135.
Lee K. y Iosler D. (1995) CycIe-ly-CycIe Varialion in Conluslion and Mixlure
Concenlralion in lhe Vicinily of Spark IIug Cap". SAL TechnicaI Iaper 95O814.
Marlnez S., Snchez I., Rodrguez C., Riesco } y CaIIegos A. (2OO7) SinuIlaneous
Measurenenl of Lvaporaling IueI Spray Using Laser Induced LxcipIex
IIourescence InlernalionaI }ournaI of Kones.
Marlnez S., Snchez I., Riesco }., CaIIegos A y Aceves S. (2OO7) Liquid penelralion Ienglh
in direcl dieseI fueI injeclion AppIied ThernaI Lngineering.
Murakanis L. y Iapanoschou D. (2OO1) Lxperinenls on Mixing Lnhancenenl in DuaI-
Slean }els. Deparlnenl of MechanicaI and Aeorospace Lngineering, Universily of
CaIifornia al Irvine, Irvine, CA.
Naler }. D. y Sielers D. L. (1996) Lffecls of Cas Densily and Vaporizalion on Ienelralion
and Dispersion of DieseI Sprays". SAL TechnicaI Iaper 96OO34.
Nauverck A., CindeIe }., Spicher U., Rosskanp H. y Landvehr C. (2OOO) Invesligalion of
lhe Transienl In-CyIinder IIov Inside a Tvo-Slroke Lngine Wilh IarlicIe Inage
VeIocinelry". SAL TechnicaI Iaper 2OOO-O1-O9O2.
Neusserl H. }., SpiegeI L. y Canser }. (1995) IarlicIe Tracking VeIocinelry A IoverfuI TooI
lo Shape lhe In-CyIinder IIov of Modern MuIli-VaIve Lngine Concepls". SAL
TechnicaI Iaper 95O1O2.
Nishida M., Nakahira T., Konori M., Tsujinura K. y Yanaguchi I. (1992) Olservalion of
High Iressure IueI Spray Wilh Laser Lighl Sheel Melhod". SAL TechnicaI Iaper
92O459.
Iapanoschou D. (2OOO) Mixing Lnhancenenl Using AxiaI IIov. Deparlnenl of MechanicaI
and Aeorospace Lngineering, Universily of CaIifornia, Irvine, CA.
Iayri I., Desanles }. M. y ArregIe }. (1996) Characlerizalion of D.I. DieseI Sprays in High
Densily Condilions". SAL TechnicaI Iaper 96O774.
Ireussner C., Dring C., IehIer S. y Kanpnann S. (1998) CDI Inleraclion elveen Mixlure
Ireparalion, Conluslion Syslen and Injeclor Ierfornance". SAL TechnicaI Iaper
98O498.
Ranos }. I. (1989) InlernaI Conluslion Lngine ModeIing, pp. 15O-158. Henisphere
IulIishing Corporalion.
Ranz W. L. y MarshaII W. R. (1952) Lvaporalion fron Drops, VoI. 48 parle I, pp. 141-146.
Canad. }. ChenicaI Lngineering Irogress.
Ranz W. L. y MarshaII W. R. (1952) Lvaporalion fron Drops, VoI. 48 parle II, pp. 173-179.
Ranz W. L. y MarshaII W. R. (1958) Sone Lxperinenls on Orifice Sprays, VoI. 36, pg. 175.

Reilz R. D. y racco I. . (1979) On lhe Dependence of Spray AngIe and Olher Spray
Iaranelers on NozzIe Design and Operaling Condilions". SAL TechnicaI Iaper
79O494.
Reilz R. D. y racco I. V. (1979) UIlra-High-Speed IiIning of Alonizing }els, Ihysics of
fIuids, VoI. 22, pp. 1O54-1O64. Ihysics of IIuids.
Reilz R. D. y racco I. V. (1982) Mechanisn of Alonizalion of a Liquid }el, VoI. 25. Ihysics
of IIuids.
SchnaIzing C. O., Slapf I., MaIy R. R., Renner C., Sleller H. y Dvyer H. A. (1999) A
HoIislic HydrauIic and Spray ModeI - Liquid and Vapor Ihase Ienelralion of IueI
Sprays in DI DieseI Lngines". SAL TechnicaI Iaper 1999-O1-3549.
Sielers D. L. (1998) Liquid-Ihase IueI Ienelralion in DieseI Sprays". SAL TechnicaI Iaper
98O8O9.
Sielers D. L. (1999) ScaIing Liquid-Ihase IueI Ienelralion in DieseI Spray ased on
Mixing-Liniled Vaporizalion". SAL TechnicaI Iaper 1999-O1- O528.
Spicher U. y KoIIneire H. (1986) Deleclion of IIane Iropagalion During Knocking Using
SinuIlaneousIy High Speed SchIieren Cinenalography and MuIli-OplicaI Iilre
Techique". SAL TechnicaI Iaper 861523.
Spicher U., Kroger H. y Canser }. (1991) Deleclion of Knocking of Conluslion Using
SinuIlaneousIy High Speed SchIieren Cinenalography and MuIli- OplicaI Iilre
Techique". SAL TechnicaI Iaper 912312.
Takagi Y., Iloh T., Muranaka S., Liyana A., Ivakiri Y., Urushihara T. y Nailoh K. (1998)
SinuIlaneous Allainnenl of Iov IueI Consunplion, High Oulpul Iover and Lov
Lxhausl Lnissions in Direcl Injeclion SI Lngines". SAL TechnicaI Iaper 98O149.
Tinaul I. V., Caslro I., MeIgar A., Sanchez M. L. y Cinenez . (1993) VisuaIizalion and
Measurenenl of Speed and Size in DieseI Sprays, pp. 78-83. Socielee de Ingenieurs
de IAulonoviIe.
ToepIer A. (1864) Menoire sur Ia Conslruclion des TeIescopes en Verre Argenle. Cohen M.
and Sohn, onn.
Trigui N., Kenl }. C., Cuezennec Y. C. y Choi W. C. (1994) Characlerizalion of Inlake-
Ceneraled IIov IieIds in I.C. Lngine Using 3-D IarlicIe Tracking VeIocinelry (3-D
ITV)". SAL TechnicaI Iaper 94O279.
Verhoeven D., VanheneIryck }. L. y arilaud T. (1998) Macroscopic and Ingnilion
Characlerislics of High-Iressure Sparys of SingIe-Conponenl IueIs". SAL
WinkIhofer L., Iuchs H. y IhiIipp H. (1993) DieseI Spray Conluslion an OplicaI Inaging
AnaIysis". SAL TechnicaI Iaper 93O862.
Xu M. y Hiroyasu H. (199O) DeveIopnenl of a Nev OplicaI Technique for Measuring
DieseI Spray Ienelralion". SAL TechnicaI Iaper 9O2O77.
YuIe A. }. y SaIlers D. C. (1995) The reak-up Zone of a DieseI Sprays. Iarl 1: Lenglh of Zone
and VoIune of Unalonized Liquid, VoI. 5, pp. 157-174. Alonizalion and Sprays.
Zhang L., Tsurushina T., Ueda T., Ishii Y., Ilou T., Minania T. y Yokola K. (1997)
Measurenenl of Liquid Ihase Ienelralion of Lvaporaling Spray in a DI DieseI
Lngine". SAL TechnicaI Iaper 971645.
Zhao I. Q., Kadola T. y Takenolo T. (1991) TenporaI and CycIic IIuclualion of IueI Vapor
Concenlralion in a Spark Ignilion Lngine". SAL TechnicaI Iaper 912346.
Fuel njection 48

Zhao I. Q., Takeloni M., Nishida K. y Hiroyasu. (1993) Quanlilalive Inaging of lhe IueI
Concenlralion in a S.I. Lngine vilh Laser RayIeigh Scallering". SAL TechnicaI Iaper
932641.
Zhao H. y Ladonnalos N. (1998) OplicaI Diagnoslics for Sool and Tenperalure
Measurenenl in DieseI Lngines, VoI. 1, pp. 244-254. Deparlnenl of MechanicaI
Lngineering, runeI Universily, U.K.
Zhao H. y Ladonnalos N. (2OO1) Lngine Conluslion Inslrunenlalion and Diagnoslics, pp.
457-47O. Sociely of Aulonolive Lngineers, Inc.WarrendaIe, Ia.

X

ExperimentaI CeIIs for DieseI Spray Research

Sinn Marlnez-Marlnez
1
, MigueI Carca Yera
1
and Vicenle R. erndez
2

1
Unitcrsia Au|cncma c Nuctc |ccn

Mcxicc
2
Unitcrsia Pc|i|ccnica c Va|cncia

Spain

1. Introduction
The sludy of lhe conluslion process in DieseI engines has leen going even deeper inlo
deplh vilh lhe appIicalion of nev lechniques of neasure and nore rigorous nelhodoIogies.
This has laken inlo nev expeclalions in lhe deveIopnenl of paranelric sludies and in lhe
conslruclion of looIs (physicaI nodeIs or experinenlaI selup) lhal aIIov lhe reproduclion of
siniIar lhernodynanic condilions lo lhe ones presenl in lhe inside of a cyIinder in a reaI
lhernaI engine, naking il possilIe lo ollain grealer approxinalions lelveen lhe lheorelicaI
reIalion and lhe experinenlaI one.

2. ExperimentaI setup cIassification
The physicaI nodeIs or experinenlaI selup used lo sludy lhe injeclion-conluslion process
usuaIIy are of very specific characlerislics, depending on lhe phenonena lo anaIyse. These
nodeIs can le cIassified in lhe lhree foIIoving groups:

2.1. According to the type of the working fIuid
-ModeIs vilh coId vorking fIuid are used lo sludy sprays in non-evaporaling condilions,
lul nany lines in condilions siniIar lo lhe exisling ones in a reaI lhernaI engine.

- ModeIs vilh hol vorking fIuid, lhis kind of nodeI is lhe lesl vay lo sludy sprays in
evaporaling condilions and in lhe najorily of lines ils possilIe lo sinuIale a reaI lhernaI
engine in ils lenperalure and densily condilions.

2.2. According to the atmosphere type
-ModeIs vilh inerl alnosphere are used lo sludy dieseI spray vilhoul lhe presence of fIane.
Ils grealer appIicalion is lasicaIIy lhe liphasic sludy of lhe spray (Iiquid and vapour).

-ModeIs vilh reaclive alnosphere, lhis kind of nodeI is generaIIy appIicalIe lo conluslion
sludies and nornaIIy alnospheric air is used lo achieve lhe reaclion.

3
Fuel njection 50

2.3. According to the geometry of the chamber
ModeIs of conslanl voIune vesseI are used lo sludy lhe dieseI sprays in non reaclive
alnosphere and reaclive alnosphere. This kind of nodeI is nol of connon use, nosl of aII
on lhe deveIopnenl of sludies of lhe sprays ly neans of reaclive alnospheres since lhey
require previous ignilions of nixlures of conluslilIe gases lo achieve high pressures and
lenperalures in lhe inside of lhe injeclion chanler.

HisloricaIIy lhe firsl deveIoped sludies aloul conluslion process in a lhernaI engine vere
ly neans of direcl recording of lhe Iuninosily of lhe fIane, and aflervards ly lhe Schieren
lechnique for sludies of aulo-ignilion and lhe knocking conluslion. The visuaIizalion
invoIves lhe fornalion of visilIe inages direclIy or indireclIy ly lhe aclion of Iighl over
sensilIe naleriaIs lo lhe Ialler. Therefore lhe Iighl and lhe recording of lhe forned inages
are essenliaI lo any forn of visuaIizalion. The Iighl can le generaled ly lhe oljecl ilseIf if ils
Iuninous (aulo-ignilion or conluslion). If ils nol, lhe Iighl source has lo le depIoyed lo
iIIuninale lhe oljecl lo refIecl (free injeclion). TradilionaIIy lhe pholographic fiIns vere lhe
onIy vay lo regisler inages for lhe chenicaI processes induced ly lhe aclion of Iighl,
hovever in sone appIicalions, lhe pholoeIeclric devices are nov sulsliluling lhe
pholographic fiIn. Due lo lhe nalure of lhe engines conluslions, each inage nusl le laken
in a very shorl anounl of line lo freeze any fIuid novenenl or of lhe fIane during lhe
exposure. This can le ollained ly inlernillenl recording vilh high speeds of lhe shuller lo
regisler an inage fron a cycIe, and a sequence of inages in funclion of lhe lurning angIe of
lhe crankshafl aIong nany engine cycIes or ly neans of lhe recording of hundreds of
inages per second lo ollain a inage sequence vilhin lhe sane engine cycIe (reIaling lo lhe
firsl recording case as a projeclion of a high speed nonoslalIe inage, and lhe Iasl one as
high speed conlinuous, correspondingIy). These recording nelhods and olher lechniques of
neasure have leen evoIving graduaIIy and increasing lheir appIicalion lo lhe differenl
kinds of exisling nodeIs, such as il viII le nenlioned in lhe sulsequenl seclions.

The anaIysis of dieseI sprays has had very inporlanl advancenenls leginning vilh lhe
physicaI nodeI inpIenenlalion lecause during line, very specific sludies have leen
deveIoped. In lhe Iasl decade of lhe 2Olh cenlury, (Hiroyasu el aI., 1989) vas one of lhe
pioneers in lhe use of injeclion nodeIs and in lhe inpIenenlalion of visuaIizalion nelhods
(Iraunhofer diffraclion). Hiroyasu, in his invesligalions lo sludy Saulers nediun dianeler
(SMD), used an injeclion nodeI vilh lhe onIy possiliIily of doing experinenls under Iov
alnospheric condilions (pressures of lhe order of 3 lar and lenperalure of 285 K), vhich
aIIoved reIaliveIy good nicroscopic sludies lo le nade lul vilh Iinilalions lo sludy lhe
dieseI spray under lhe lhernodynanic varialIes lhal can le achieves in a reaI lhernaI
engine.
To deveIop his experinenl of SMD, (Minani el aI., 199O) used a siniIar nodeI lo lhal one
enpIoyed ly (Hiroyasu el aI., 1989), vhere lhe oljeclive vas lo ollain lhe densily
condilions siniIar lo lhe exisling ones inside lhe conluslion chanler of a reaI lhernaI
engine. This aulhor lo deveIop his experinenls used as lhe vorking fIuid pressurized
nilrogen al 2O lar and lhe injeclion pressures anaIysed vere of lhe order of 2OOO lar. These
pressures couId nol le anaIysed using lhe nodeI of Hiroyasu due lo lhe Iov densilies lhal
vere achieved in his nodeI, so lhal il resuIled irreIevanl lo do cerlain kinds of sludies,
noslIy in high injeclion pressures. Minani enpIoyed lhe Iraunhofer lechnique lo visuaIize

lhe dianeler of lhe dropIels, using a Iighling source lased in a ruly Iaser lean puIsed al 3O
ns and vilh a vaveIenglh of 694 nn. UnIike Hiroyasu, Minani expanded lhe lean naking
il paraIIeI lo lhe spray lhrough Ienses, vhich aflervards is allenualed ly lhe ovn fueI spray,
vhere lhe lransnission of Iighl is focused ly a paraloIic nirror vilh vhich an inage of lhe
fiIned spray on lhe 35 nn fiIn is ollained. Il vas possilIe in lhis nodeI lo anaIyse lhe
penelralion of lhe spray ly inslaIIing a high speed pholographic canera lo lake lhe inages
and repIacing lhe lean of Iighl ly haIogen Ianp as lhe Iighl source.

There vhere lasicaIIy lvo prolIens lo cIear up lhe inlerference of slripes in lhe
pholographic pIane vhen using lhis lechnique:

- The firsl issue vas due lo lhe inlerferences caused ly lhe lhickness of lhe visuaIizalion
vindov vhich vas approxinaleIy 5O nn. This prolIen vas soIved ly changing lhe
lrajeclory of lhe lean lelveen lhe reference and lhe oljecl, using a inlensily rale of 9:1 and
fixing lhe poinl lhal focuses lhe Iens of lhe reIay as cIose as possilIe lo lhe pholographic pIale.

- The second prolIen vas generaled ly lhe coIIision of vaves, vhich vere produced in lhe
inlerior of lhe chanler in a shorl anounl of line during lhe period of injeclion. This vas a
consequence of lhe fueI injeclion veIocily lhal surpassed lhe sound veIocily. The coIIision of
vaves inside lhe conluslion chanler nade il difficuIl lo lake piclures lecause of lhe
refIeclion inside il. The prolIen vas soIved ly changing lhe line of synchronizalion of lhe
injeclion line in reference lo lhe piclure laking, changing lo lines of O,2 lo O,4 ns afler slarl
of injeclion.

The evaporaled dieseI spray sludied ly (Talala el aI., 1991) vas nade using a nodeI siniIar
lo lhal presenled ly olhers scienlisl (Hiroyasu el aI., 1989) and (Minani el aI., 199O), vilh a
snaII difference in ils configuralion. A heal syslen vas inslaIIed in lhe inlerior of lhe
chanler vhich had lhe oljeclive of healing up lhe vorking fIuid (nilrogen) lo condilions
lhal vere oplinaI lo achieve fueI evaporalion. The nain dravlack sludying evaporaled
dieseI spray vas lhe Iov range of avaiIalIe densilies for high pernissilIe lenperalures. Due
lo lhis inconvenience il vas inpossilIe lo reproduce siniIar lhernodynanic condilions lo
lhose found in a lhernaI engine. Il vas Iiniled lo predicl lehaviours al Iov pressures
(pressures inside lhe injeclion chanler in lhe order of 2O lar). (Higgins el aI., 2OOO) sludied
lhe ignilion of lhe spray and lhe lehaviour of lhe pre-nixlure of lurn in a nodeI of conslanl
voIune, vhere il vas possilIe lo achieve pressures inside lhe chanler of 35O lar and a
lenperalure range of 8OO K lo 11OO K for densilies lelveen 7.27 kg/n3 y 45 kg/n3. This
nodeI, unIike lhose used ly olher aulhors (Hiroyasu el aI., 1989), (Minani el aI., 199O) and
(Talala el aI., 1991) has lvo spark pIugs and a venliIalor lo equaIize lhe alnospheric
condilions in lhe inside of lhe chanler, lesides of having lhe vaIIs eIeclricaIIy condilioned
lo sinuIale lhe condilions of lhe vaII lenperalure found in a lhernaI engine. Anolher
oljeclive for lhis condilioner vas lo avoid lhe vapours of lhe fueI lo condensale on lhe
visuaIizalion vindovs. This nodeI didnl require any kind of nodificalions lo deveIop
lurning or free injeclion sludies, unIike lhe one used ly (Minani el aI., 199O), vhich did
need very sophislicaled adaplalions lo le alIe lo nake nore conpIex sludies, as lhe
lurning sludy.

Fuel njection 52

The effecls on lhe dieseI spray caused ly lhe geonelry of lhe chanler vere sludied ly
(Monlajir el aI., 2OOO). To deveIop lhe invesligalions lhey lased lhenseIves on an
experinenlaI inslaIIalion vhich consisled in a snaII lhernaI engine of sIov speed regine
and vilh a conluslion head of reclanguIar geonelry adapled vilh a vindov for
visuaIizalion. In lhis nodeI as in lhe ones descriled previousIy, nilrogen vas used as lhe
vorking fIuid (injecled in lhe cyIinder a lenperalure of 293 K and vilh a pressure inside lhe
chanler of 45 lar) lo visuaIize lhe dieseI spray vilhoul conluslion. To achieve lhe
deveIopnenl of sludies in differenl lhernodynanic scopes vilh lhis nodeI, lhe nain oplion
vas lo change lhe conpression ralio. This vas achieved ly pIacing spacer rings lelveen lhe
cyIinder and lhe conluslion head. This nodificalion used lo le aIso very conpIex and
expensive. Iurlhernore il didnl aIIov conlinuily in lhe experinenlaI session lecause of lhe
conslanl slops required lo carry oul nechanicaI changes in lhe inslaIIalion.

The slralificalion effecls of lhe conluslion vere sludied ly (IIacknann el aI., 1998) using a
nodeI of conslanl voIune, aIso caIIed conluslion conslanl voIune vesseI. Where lhe
visuaIizalion look pIace lhrough lhree quarlz vindovs. One of lhe vindovs vas Iocaled in
lhe lack side and lhe olher lvo in lhe niddIe seclion of lhe punp. This nodeI vas
pressurized vilh an air-propane nixlure vilh pressures of 4O lar, lhese fIuids vere used as
oxidanl and lesl conluslilIe respecliveIy. The ignilion look pIace ly lhe neans of lvo
dianelricaIIy opposile eIeclrodes vhich funclion as a spark pIug lo provoke lhe spark. One
of lhe eIeclrodes vas connecled lo a high lension source lhal cane fron a discharge
capacilive ignilion syslen and lhe olher one lo nass, leing lhis Iasl syslen lo le lhe one
lhal caused lhe difference in eIeclric polenliaI lo achieve lhe ignilion of lhe nixlure.

In lheir sludies aloul lhe lehaviour of gasoIine direcl injeclion (CDI), (SheIly el aI., 1998)
used an engine vilh lhree oplicaI accesses nade oul of quarlz. To carry oul lhe visuaIizalion
of lurning lhe pIanar Iaser induced fIuorescence (ILII) as lhe lechnique of neasure and a
high speed canera lo fiIn lhe conluslion process vere used. The canera vas pIaced
perpendicuIarIy lo lhe lean, having vilh lhis lhe advanlage of laking inages direclIy. This
lhernaI engine unIike lhe one used ly (Monljair el aI., 2OOO) had conpIeleIy lransparenl
vaIIs, vhich aIIoved a conpIele caplure of lhe phenonena (injeclion-conluslion).
Iurlhernore, il required conpIex roulines al lhe nonenl of doing any changes in lhe
inslaIIalions configuralion.

Wilh lhe purpose of visuaIizing lhe alonizalion process in a gasoIine direcl injeclion
syslen, (Ireussner el aI., 1998) used a nodeI pressurized vilh oxygen (pressure and
lenperalure of 2O lar and 673 K correspondingIy). Where lhe conluslion producls of each
cycIe vere conslanlIy evacualed ly lhe use of nilrogen and lhe visuaIizalion of lhe
alonizalion process vas done lhrough vindovs Iocaled on lhe far ends of lhe injeclion
chanler. Mie-Scallering vas lhe lechnique of neasure used during lhe experinenls. In
addilion lo lhis lechnique, lhe scienlisls deveIoped lesls vilh lhe Iaser induced fIuorescence
lechnique, vhich proved nol lo le lhe nosl appropriale for sone neasures in lhe gasoIine
direcl injeclion syslens, lecause il is an oplicaI nelhod of neasure sonevhal incoherenl lo
delecl concenlralion of species conpared lo Mie-Scallering. Hovever, il vas possilIe lo use
olher lechniques of neasure. One disadvanlage presenled ly lhis nodeI vas lhe vay of
evacualing lhe lurn gases, lecause il vas required lo inlroduce nilrogen lo lhe injeclion

chanler vhen lhe cycIe ended lo expeI lhese gases and vas nol recoveralIe, vhich nade
expensive lhe experinenlalion, lesides of needing very Iong lesl roulines lecause of
disconlinuily in lhe lesls due lo lhe lype of cIeaning used.

In his experinenls lo sludy lhe effecls of vaporizalion, (Takagi el aI., 1998) used an
experinenlaI inslaIIalion lased on a lhernaI engine of 2.O Iilres provided vilh sapphire
vaIIs and onIy one oplic access for inage capluring. The oplic access vas pIaced in one
seclion of lhe pislon head, leing lhe nosl appropriale coIIalion pIace due lo lhe geonelric
characlerislics of lhe nodeI. In lhe sane vay as Ireussner, Iaser induced fIuorescence vas
used as lhe lechnique of neasure, ollaining exlreneIy salisfying resuIls in addilion lo
confirn (Ireussner el aI., 1998) lheories, vho posluIaled lhal lhe Iaser induced fIuorescence
lechnique vas no lhe lesl oplion lo sludy lhe conluslion effecls due lo lhe prolIens
generaled lo delecl species. Sonelhing lhal does nol exisl in lhe case of vapour
concenlralion sludies, as Iong as lhe fueI is nixed vilh lracers lo nake il lehave as a
fIuorescenl source (lhe conluslilIe ly ils ovn nalure lends lo iIIuninale vhen ils suljecl lo
high pressure and lenperalure, vhich is vhy lracers are required lo carry oul concenlralion
sludies, lecause lhe presence of lracers aIIovs lo define lhe differenl vaveIenglhs for each
eIenenl lhal conslilules lhe nixlure). This nodeI vasnl sufficienlIy fIexilIe enough for lhe
use of olher lechniques of neasure due lo lhe facl lhal lhe sanpIing process of inages vas
of lhe inlrusive kind, lherefore naking il an inappropriale nelhod for lhe sludy of inlernaI
processes in an inlernaI conluslion engine.

To sludy lhe effecls of lhe injeclion pressure and lhe dianeler of lhe nozzIe on lhe funes
enissions, (Sielers & Iickell, 2OO2) used a nodeI consisling in a chanler of conslanl
voIune vhere il is possilIe lo sinuIale siniIar lhernodynanic condilions lo lhe one in a
DieseI engine (Iigure 1). Olher appIicalions for lhis nodeI are descriled in (Naler &
Sielers, 1996), (Sielers, 1998), (Sielers & Higgins, 2OO1) and (Sielers el aI., 2OO2). The
chanler has four oplic accesses pIaced orlhogonaIIy and equidislanl, having lhe advanlage
of using differenl oplicaI lechniques sinuIlaneousIy naking nore lhan one anaIysis in each
experinenl. ul il aIso has disadvanlages, nainIy lhe vay of ollaining lhe appropriale
lhernodynanic condilions in lhe chanlers inlerior. These condilions such as lhe
lenperalure and densily are achieved ly causing lhe ignilion (ly neans of a spark pIug) of
lhe conluslilIe gases in lhe inside of lhe chanler, lhe nelhod is unusuaI and in nany cases
even dangerous. The range of operalion for lhe lenperalure as for lhe densily is in lhe order
of (6OO a 14OO) K and (3,6 a 6O) kg/n
3
respecliveIy. Iurlhernore, ils possilIe lo reguIale lhe
concenlralion of oxygen in lhe inlerior of lhe chanler fron cero (e.g., for inerl condilions) lo
vaIues grealer lhan 21 in voIune, depending on lhe lype of sludy aining for.

Fuel njection 54

Iig. 1. LxperinenlaI selup: conslanl voIune vesseI (Sielers, 1998).

3. ProposaI of the physicaI modeI
To achieve lhe oplinaI feasiliIily lhe physicaI nodeI nusl aIvays oley lhe foIIoving
equivaIence requirenenls:

-Reproduce inside lhe injeclion chanler, lhe lhernodynanic properlies lhal lhe vorking
fIuid has in a reaI lhernaI engine in nornaI operalion.

-Lasiness lo sveep lhe alonized fueI in lhe injeclion chanler and nainlain lhe visuaIizalion
accesses cIean.

-To nake possilIe lhe sludying of differenl phenonena ils indispensalIe lo have versaliIily
in lhe cases of lhe densily nagniludes and in lhe vorking fIuids lhernodynanic properlies,
lesides having fIexiliIily in lhe geonelry of lhe injeclion chanler (lhe geonelry of lhe
injeclion chanler can vary vilh snaII arrangenenls or ly naking nodificalion of lhe
suppIenenlary lype in lhe physicaI nodeI).

- AIIov repeliliveness in lhe lesls vhen conslanl lhernodynanic are nainlained.

3.1. OperationaI Conditions
The operalionaI condilions depend fundanenlaIIy on lhe characlerislics of lhe vorking
fIuid in lhe injeclion chanler al lhe nonenl vhen lhe injeclion of fueI is inilialed. These
condilions connonIy vary fron one engine lo anolher. When a physicaI nodeI is designed,
vhere lhe lhernodynanic properlies of lhe vorking fIuid are nol siniIar lo lhose of
alnospheric air, (generaIIy heavy noIecuIar gases are used) lhe nodeI nusl le reproduced

lased on anaIogies, an equivaIenl alnosphere lo lhal in a reaI lhernaI engine in Iov
pressure and lenperalure condilions and lhese nusl le easiIy achieved, so lhal lhe
deveIopnenl of lhe experinenlaI sludy is aIIoved (Iigure 2). Il is lo le noled lhal nodeI of
hol fIuid lhal viII le presenled in lhe nexl seclions doesnl need anaIogicaI processes
lecause lhis physicaI nodeI is in facl a reaI lhernaI engine. Hovever, lhe lhernodynanic
varialIes lhal are desired lo conline for each parlicuIar sludy nusl le correclIy defined. On
lhe one hand, ils a safely rouline and on lhe olher hand, il gives knovIedge of lhe high
pressure range lhal can le achieved in lhe conluslion chanler. This Iasl argunenl is
fundanenlaI for lhe defining of lhe lhernodynanic condilions inside lhe cyIinder (ArrgIe,
1998).

Iig. 2. LxperinenlaI selup: lhe vork fIuid is SI6 (lesl vilh Iov pressure and lenperalure)

1. ReynoIds anaIogy: The viscous forces have a nayor roIe in lhe penelralion of lhe spray in
a conlroIIed alnosphere, in lhis case a gaseous alnosphere and a dieseI fueI spray. The
ReynoIds nunler and lhe reIalionship lelveen lhe inerliaI and viscous forces nusl le
conserved aIvays so lhe penelralion of lhe spray is nol nodified.

|
g 32
g

Re =
(1)

* * *
32
*
O
O
|
Re* =
(2)

Re = Re * (3)

Mainlaining conslanl spray characlerislics: Saulers nediun dianeler, lhe penelralion
veIocily and lhe ReynoIds nunler can le sinpIified as foIIovs:

| | | |
*
g g
*
g g

engine = experinenlaI selup

(4)
Fuel njection 56

2. Welers anaIogy: in lhe alonizalion process of a Iiquid vane, lhe aerodynanic forces and
lhe superficiaI lension are direclIy invoIved in lhe lreaking lransilion of lhe dropIels, vhich
is vhy lhe conservalion of Welers nunler lhal associales lhese forces, is a nain faclor lo
consider in lhe injeclion process. Welers anaIogy can le expressed as foIIovs:

|
2
g 32
I

We =
(5)

|
* * 2*
g 32
*
I

We* =
(6)

*
We = We (7)

The resuIls can le direclIy used as Iong as Saulers nediun dianeler and lhe injeclion
veIocily are kepl, vhich inpIies lhe equaIily of anaIogies. Il can le considered lhal during a
very shorl anounl of line of lhe alonizalion, lhe heal lransfer lelveen lhe vorking fIuid
and lhe fueI in lhe conluslion chanler is Iov. The fueI slarls lo heal up vhen il is found as
a dropIel. Therefore, if a sane fueI is used lhe superficiaI lension varialion lhal occurs
lelveen lhe reaI lhernaI engine and lhe physicaI nodeI viII le very snaII. This hypolhesis
is confirned ly lhe resuIls ollained ly (Hiroyasu & Arai, 199O), in vhich differences during
lhe alonizalion process al lenperalures of lhe vorking fIuid lelveen 293 K and 573 K
arenl seen. This is vhy Welers anaIogy is sinpIified lo a densily conservalion of lhe
vorking fueI lelveen lhe reaI lhernaI engine and lhe physicaI nodeI.

3. AnaIogy equaIily: lhe equaIily of anaIogies lelveen Weler and ReynoIds is sinpIified as
shovn. This anaIogy lelveen lhe reaI lhernaI engine and lhe physicaI nodeI is in
accordance lo lhe lheorelicaI and experinenlaI resuIls on alonizalion and spray
penelralion. These resuIls are presenled ly lhe aulhors previousIy ciled.

-IueI Characlerislics

*
I I
= (8)

*
I I
= (9)

*
I I
= (1O)

-Injeclion pressure

*
iny g
p = p = I - I (11)

-Working fIuid characlerislics

(
(
(

*
g g
*
g g

Re : =

(12)

(

*
g g
We : = (13)

The nosl crilicaI paranelers lhal are usuaIIy presenl al lhe nonenl of insuring lhe anaIogy
lelveen lhe physicaI nodeI and lhe reaI lhernaI engine are: lhe densily and viscosily of lhe
vorking fIuid. Iurlhernore, ils necessary lo caIilrale lhe injeclor lo achieve a pressure
increasing and lherefore, an exil veIocily of lhe injecled fueI siniIar lo lhal happening in a
reaI lhernaI engine in nornaI operalion.

3.2. Geometric characteristics of the physicaI modeI
The geonelric characlerislics of lhe physicaI nodeI nusl oley cerlain requirenenls. The
nosl inporlanl one is in lhe injeclion fueI chanler, vhich nusl have oplicaI accessiliIily lo
nake proper neasurenenls and olservalions. As il has leen previousIy expIained, lo carry
oul lhe najorily of experinenls nore lhan one lechnique of neasure is required. This is vhy
il is convenienl lo have severaI oplic accesses (lhese access are generaIIy vindovs, nainIy
oul of quarlz). The geonelry of lhese accesses is in funclion of lhe lesl ained lo do, leing in
lhe najorily of cases lhe fronl Iongilude of lhe spray lhe lasis of sludy. In a reaI lhernaI
engine lhe dieseI spray Ienglh is condilioned ly lhe dianeler of lhe conluslion pre-
chanler (in a lhernaI engine of indirecl injeclion) or ly lhe pislon lovI (in lhe case of
lhernaI engines of direcl injeclion). This Ienglh in a dieseI spray of an inlernaI conluslion
engine of seIf-propuIsion can neasure lelveen 18 lo 5O nn lefore lhe coIIision vilh lhe
vaII. Hovever, in lhe najorily of cases lhe sprays nanages lo coIIide vilh lhe vaII of lhe
lovI lefore lhe alonizalion process has ended. Ils lecause of lhis reason lhe inporlance of
sludying lhe spray up lo Ienglhs in lhe order of 1OO nn, sufficienlIy Iong for lhe spray lo
achieve conpIele alonizalion, having speciaI allenlion lo lhe injeclion chanler. A second
requirenenl lhal lhe physicaI nodeIs geonelry nusl fuIfiI is lhe easiness of evacualion of
lhe injecled fueI inside lhe injeclion chanler lo reduce lhe dirliness of lhe oplic accesses.
Many scienlisls as seen in lhe previous seclion (Hiroyasu & Arai, 199O), (Hiroyasu el aI.,
1989) and (Minani el aI., 199O) have used chanlers of lhe voIunelric lype vere lhe
alnosphere is reneved afler every injeclion. In using lhis nelhodoIogy of neasure, an
injeclion syslen is required lhal aIIovs having cerlain fIexiliIily lo injecl disconlinuousIy
and naking possilIe lhe adnission of cIean fIuid lefore each injeclion cycIe. This syslen
usuaIIy doesnl aIIov lhe use of nany oplicaI lechniques lecause in very parlicuIar cases,
ils necessary lhe use of lenlhs of injeclion cycIes lo do a conpIele spray sludy, having very
Iong, arduous and coslIy lesl sequences.

In lhis seclion lhe deveIopnenl, designed and lune-up of an experinenlaI inslaIIalion is
presenled. The nain oljeclive is lo carry oul sludies of lhe evaporaled dieseI spray in inerl
and reaclive condilions. To achieve lhis purpose a nodeI has leen lroughl up lased on a
lvo slroke lhernaI engine and a group of auxiIiary equipnenl lhal aIIov lhe adjuslnenl of
operalionaI paranelers of lhe inslaIIalions and of conlroI for lhe various independenl
varialIes lhal inleracl in lhe injeclion-conluslion process.

3.2.1. Objectives of the experimentaI instaIIation
-SinuIale lhe process of direcl injeclion of dieseI fueI in high pressure, densily and
lenperalure condilions.

Fuel njection 58

-Sludy lhe nicroscopic and nacroscopic characlerislics of lhe dieseI spray in evaporalive
condilions.

-Carry oul experinenls of lhe dieseI spray in an inerl and reaclive alnosphere.

3.2.2 Design requirements and determining factors
The experinenlaI inslaIIalion syslens nusl aIIov lhe foIIoving funclions and capacilies:

1. ConlroI and reguIale lhe vorking fIuids lenperalure and pressure.
2. Oplic accesses lo visuaIize lhe dieseI spray ly neans of visuaIizalion lechniques.
3. Modificalion of characlerislic paranelers of lhe injeclion.
4. The injeclion chanler nusl have a nininun nunler of novenenls in reIalion lo lhe
visuaIizalion equipnenl.

3.2.3. Observations reIative to system design
The nosl reIevanl aspecls lhal nusl le considered in lhe design of lhe nain syslens lhal
conslilule lhe experinenlaI inslaIIalion lased on a lhernaI engine are descriled nexl:

1. To le alIe lo have high pressures and lenperalures of lhe vorking fIuid in lhe adnission
syslen.
2. Having a good reguIalion of lhe exhausl lackpressure.
3. The lhernaI engine nusl le lhernaIIy filled ly an exlernaI syslen lhal aIIovs il lo have
oplinun operalionaI condilions.
4. There nusl le an injeclion syslen for high pressure independenl of lhe lhernaI engine.
5. The vilralion conlroI lo reduce vilralions lransnilled ly lhe lhernaI engine lo ils
surroundings nusl le conlroIIed ly a Iinilalion vilralion syslen.

The experinenlaI inslaIIalion aIIovs svilching lelveen lvo differenl configuralions lo
sludy lhe dieseI spray in evaporalive condilions. These configuralions are delaiIed as
foIIovs:

1. Open Ioop configuralion, reaclive alnosphere: lhis configuralion aIIovs sludying lhe
dieseI spray injeclion in an oxidanl alnosphere. The vorking fIuid used lo fiII lhe conlroI
voIune vhere lhe fueI is injecled is alnospheric air. When lhe conluslion is execuled, lhe
gases and lhe producl species of lhe Ialler are lhen evacualed lo lhe alnosphere afler each
operalionaI cycIe. Aflervards, fresh air is inlroduced in lhe cyIinder lo carry oul a nev
injeclion.

2. CIosed Ioop configuralion, inerl alnosphere: This configuralion aIIovs sludying lhe
dieseI spray in an inerl alnosphere in vhich lheres no oxidalion of lhe injecled fueI. To
achieve lhis, a nilrogen suppIy syslen has leen incorporaled lo lhe inslaIIalion vilh vhich
lhe injeclion chanler is fiIIed.

LxperinenlaI inslaIIalion characlerislics


-Repeliliveness of lhe various lesls so an exceIIenl slaliIily in lhe differenl operalion poinls
is achieved.

-The precision in lhe neasures denands quaIily of lhe neasuring equipnenl and an
accurale caIilralion of lhen lo avoid syslenalic errors during lhe lesls.

-Having reIialiIily and safely assunes a rigorous filling of lhe roon vhere lhe experinenlaI
inslaIIalion is housed.

-Cood accessiliIily and conpaliliIily lo faciIilale lhe conlroI, verificalion and proceeding
lasks on aII lhe lhernaI engine syslens.

3.2.4. Reactive atmosphere configuration
The reaclive alnosphere configuralion or open Ioop is usefuI for carrying oul sludies of lhe
injeclion, seIf-ignilion and conluslion processes, in lhe sane vay lo carry oul anaIysis in
lhe conluslion process and neasures of conlaninanl enissions. Iigure 3 shovs lhe
schenalic diagran of lhe experinenlaI inslaIIalion lhal is used in lhe sludy of dieseI sprays
in a reaclive alnosphere (Marlnez el aI., 2OO7) and (erndez el aI., 2OOO, 2OO3).

Iig. 3. LxperinenlaI selup in open Ioop configuralion.

Descriplion of lhe eIenenls lhal conslilule lhe experinenlaI inslaIIalion, inpurily fiIler:
A connerciaI fiIler is used lo purify air fron lhe inpurilies of lhe environnenl:

-Conpressor: A Rools conpressor is used lo loosl air fron lhe environnenl lovards lhe
engines adnission, as veII lo achieve lhe sveep of gases produced ly lhe conluslion.

Fuel njection 60

-Heal exchanger: il is of lhe arnour and pipe lype (valer-air). Ils funclion is lo reguIale lhe
exil air lenperalure.

-Healer: prepares lhe air of adnission lhrough a resislance healer vhich is inslaIIed in series
vilh lhe valer-air exchanger.

-In: lhe nediun pressure sensor alsoIule scaIe is in charge of sending a voIlage signaI lo
lhe IID conlroIIer so il can reguIale lhe opening of lhe eIeclro vaIve. The aperlure of lhe
eIeclro vaIve is in funclion of lhe largel pressure eslalIished in lhe adnission.

-Iinlake: is lhe inslanlaneous pressure sensor in alsoIule scaIe lhal neasures lhe pressure in
lhe adnission chanler, vhich is used as a reference for lhe nediun pressure in lhe
cyIinder.
-Iexhausl: is lhe inslanlaneous pressure sensor in alsoIule scaIe lhal neasures lhe pressure
in lhe exhausl chanler. This pressure neasurenenl is used as an operaling reference for lhe
exhausl lackpressure eIeclro vaIve.

-T: are lhe lhernocoupIes lype k: lhe lhernocoupIe is Iocaled dovnslrean of lhe adnission
chanler and is responsilIe of neasuring lhe airs lenperalure jusl al lhe enlrance of lhe
adnission chanler. This signaI is senl lo a IID lhal ruIes over lhe healer lased on lhe
eslalIished largel lenperalure. The lhernocoupIe is Iocaled upslrean fron lhe adnission
chanler and neasures lhe lenperalure jusl lefore il enlers lhe adnission.

-LIeclro vaIve: lo increnenl lhe pressure of lhe vorking fIuid in lhe injeclion chanler ils
necessary lo eslalIish a lackpressure in lhe exhausl lo avoid a shorl cul of lhe vorking fIuid
fron adnission lo exhausl. This is achieved ly cIosing lhe exhausl syslens eIeclro vaIve.

-A: represenls lhe vaIves used lo drain lhe vapour fron lhe circuil, as of residuaI fueI as veII
of lhe oiI lhal accunuIales on lhe ducls.

-: represenls lhe sphere vaIves used lo free lhe circuil eIenenls. These are nainIy used
vhen lhe inslaIIalion operales vilh inerl alnosphere.

3.2.5. Inert atmosphere configuration
The inerl alnosphere configuralion or cIosed Ioop is used for carrying oul sludies of lhe
injeclion process in Iiquid phase as veII as in vapour phase. This configuralion aIIovs lhe
spray lo evoIve lhrough lhe conluslion chanler presenling lhe lvo phases (Iiquid and
vapour) vilhoul pernilling fIane during lhe process. This is achieved ly a very sinpIe
vay, lhe spray is injecled in a conlroI voIune (conluslion chanler) vere lhe vorking fIuid
is a inerl gas (e.g. nilrogen), lherefore a voIune free of oxygen noIecuIes lul vilh siniIar
properlies lo lhose of air.

Iigure 4 shovs lhe schenalic diagran of lhe experinenlaI inslaIIalion used lo sludy dieseI
sprays in an inerl alnosphere.


Iig. 4. LxperinenlaI selup in cIose Ioop configuralion.

Descriplion if lhe eIenenls lhal conslilule lhe experinenlaI inslaIIalion. AII of lhe eIenenls
lhal conslilule lhe adapled experinenlaI inslaIIalion lo sludy dieseI sprays in a reaclive
alnosphere are used in inerl alnosphere, vilh lhe exceplion of lhe expansion chanler and
lhe fiIler of parlicIes. AddilionaIIy lo lhese, lvo eIenenls vhich are indispensalIe for lhis
configuralion are incIuded. These eIenenls are descriled nexl:

-Condensale separalor of lhe coIIision lype: lhe configuralion in a cIosed Ioop nusl have
nilrogen recircuIalion vilhoul lhe presence of fueI, lherefore lhe injecled fueI in lhe
conluslion chanler and lhe gas in vhich lhe Ialler has leen injecled lo, nusl le separaled
once lhey have leen expeIIed fron lhe cyIinder. The condensale separalor is in charge of lhe
firsl phase of separalion. Upon circuIaling lhe nixlure (fueI-gas) lhrough lhe discharge
syslen and arriving al lhe separalor, vhere lhe fueI is lhen condensed inlo snaII drops lhal
go lo lhe lollon of lhe vesseI and lhe gas and olher parlicIes lhal are sliII nixed are
separaled ly a fiIler f nicro parlicIes (see nexl seclion). The Iiquid fueI concenlraled al lhe
lollon of lhe vesseI is purged ly a syslen of voIunelric vaIves and disposed lhereafler.

-IiIler of nicro parlicIes: lhe second phase of gas purificalion is carried oul lhrough a
fiIlering of high purily lhal aIIovs lhe gas lo le separaled fron possilIe fueI residues lhal
havenl condensale in lhe previous phase, or fron possilIe soIid parlicIes dragged fron lhe
circuIalion syslen or olher parlicIes Ioosed ly lhe engine. The fiIler of nicro parlicIes aIIovs
lhe separalion of parlicIes up lo O.OO1 n.

3.2.6. Operation mode
The conpIexily of operalion and precision of lhe neasures is of lhe sane nagnilude as on
lhe reaclive alnosphere configuralion and aIso lhe nechanisn for lhe circuIalion of lhe
vorking fIuid. When lhe inslaIIalion is operaling in an inerl alnosphere, lhe nilrogen is
loosled ly lhe Rools conpressor lhrough lhe condilioner syslen and aflervards inlroduced
Fuel njection 62

inlo lhe cyIinder. Afler lhe fueI is injecled and nixed vilh gas, lhis nixlure (fueI-nilrogen) is
expeIIed lovards lhe condensale separalor syslen and lo lhe high efficienl fiIler, vhere lhe
Iarger drops of fueI are firsl condensed and lhen lhe soIid parlicIes dragged ly lhe gas lhrough
lhe circuil eIenenls or lhe engine are eIininaled. Once lhe nilrogen has leen purified, il is
driven up lo lhe Rools conpressor adnission vhere il is loosled again lovards lhe nolor lo
carry oul a nev injeclion cycIe. A pressurised nilrogen deposil is used lo fiII up lhe cIose Ioop
circuil lefore inilialing operalions of lhe inslaIIalion. Once lhe inslaIIalion is operaling, lhe
deposil vilh heIp of an eIeclro vaIve and a IID reguIalor are responsilIe of nainlaining a
conslanl voIune of nilrogen circuIaling lhrough lhe syslen since sone snaII gas Ieaks
lovards lhe engines crankcase lhrough lhe Iooseness lelveen lhe pislon and lhe cyIinder nay
exisl. The Iigure 5 (a) shovs a schenalic diagran of lhe conluslion chanler lhrough vhich
is possilIe lo visuaIize lhe injeclion-conluslion process. Iigure 5 (l) shovs an exanpIe of an
inage sequence of a nuIlipIe spray. Iigure 6 (a) and (l) shov a sequence of lhe deveIopnenl
of an axi-synelric spray in an inerl and reaclive alnosphere respecliveIy.

(a) (l)
Iig. 5. (a) Schenalic diagran of conluslion chanler and (l) LxanpIe of nuIli dieseI spray.

(a) (l)
Iig. 6. (a) LxanpIe of dieseI spray in inerl alnosphere and (l) exanpIe of dieseI spray in
reaclive alnosphere.

4. ConcIusions and remarks
In lhis chapler, severaI experinenlaI selup lased on differenl nodeIs have leen presenled.
This experinenlaI selup aIIovs a lasic sludy of lhe injeclion and conluslion processes. The
nodeIs have diverse appIicalions and lheir use depends on lhe kind of sludy lhal is inlend
lo carry oul, lhese nodeIs are Iiniled in sone aspecls, chiefIy in lhe vork fIuid renevaI,
Iikevise is used lo leing conpIicaled lo reproduce lhe lhernodynanic condilions lhal are
presenled in lhe conluslion chanler of a reaI lhernaI engine. The experinenlaI selup lhal
is presenled and lased on a lhernaI engine is a exceIIenl looI lhal pernils lo reproduce
lhernodynanic condilions and of operalion as lhe exisling in a engine in slandard
condilions. The cyIinder head lhal has arranged lhe engine has lhree oplicaI accesses vhich
is aIIov visuaIizalion and lhe use of oplicaI lechniques. Diverse lechniques have leen
uliIized in lhis experinenlaI selup, Iike LDA, IDA, ILILI, IIV, Speclroscopy, Ranan and
RayIeigh Scallering, Tvo coIors, and Shadovgraph, principaIIy. These sludies have pernil
lo characlerize lhe dieseI spray in evaporaling condilions lo ollain reIevanl infornalion lhal
pernil lo inprove lhe conluslion processes in nodern dieseI engines.

5. References
ArrgIe }. (1998) AnaIysis of lhe Slruclure and InlernaI Dynanic of lhe DieseI Spray. IhD
Disserlalion, Universidad IoIilecnica de VaIencia, Spain,.
erndez V., Carca }. M., }uIi L. and Marlnez S. (2OO2) InslaIacin LxperinenlaI para eI
Lsludio deI Iroceso de Inyeccin-Conluslin en Molor DieseI de Inyeccin
Direcla". XV Congreso NacionaI de Ingeniera Mecnica, Cdiz, Spain.
erndez V., Carca }. M., }uIi L. and Marlnez S. (2OO3) Lngine vilh OplicaIIy AccessilIe
CyIinder Head: a Research TooI for Injeclion and Conluslion Irocesses". SAL
TechnicaI Iaper 2OO3-O1-111O.
Hinggins ., Sielers D. L. and Aradi A. (2OOO) DieseI-Spray Ingnilion and Irenixed-urn
ehavior". SAL TechnicaI Iaper 2OOO-O1-O94O.
Hiroyasu H. y Arai M. (199O) Slruclures of IueI Sprays in DieseI Lngines". SAL TechnicaI
Iaper 9OO475.
Hiroyasu H., Arai M. and Talala M. (1989) LnpiricaI Lqualions for lhe Sauler Mean
Dianeler of a DieseI Spray". SAL TechnicaI Iaper 89O464.
Minani T., Yanaguchi I., Shinlani M., Tsujinura K. y Suzuki T. (199O) AnaIysis of IueI
Spray Characlerislics and Conluslion Ihenonena under High Iressure IueI
Injeclion". SAL TechnicaI Iaper 9OO438.
Marlnez S., Snchez I., Rodrguez C., Riesco } and CaIIegos A. (2OO7) SinuIlaneous
Measurenenl of Lvaporaling IueI Spray Using Laser Induced LxcipIex
IIourescence InlernalionaI }ournaI of KONLS.
Marlnez S., Snchez I., Riesco }., CaIIegos A and Aceves S. (2OO7) Liquid penelralion
Ienglh in direcl dieseI fueI injeclion AppIied ThernaI Lngineering.
Monljair R. M., Tsunenolo H. and Minani T. (2OOO) IueI Spray ehavior in a SnaII DI
DieseI Lngine Lffecl of Conluslion Chanler Ceonelry". SAL TechnicaI Iaper
2OOO-O1-O946.
Naler }. D. y Sielers D. L. (1996) Lffecls of Cas Densily and Vaporizalion on Ienelralion
and Dispersion of DieseI Sprays". SAL TechnicaI Iaper 96OO34.
Fuel njection 64

IIacknann }. D., Kin T. and Chandhi }. . (1998) The Lffecls of Mixlure Slralificalion on
Conluslion in a Conslanl VoIune Conluslion VesseI". SAL TechnicaI Iaper
98O159.
Ireussner C., Dring C., IehIer S. and Kanpnann S. (1998) CDI Inleraclion elveen
Mixlure Ireparalion, Conluslion Syslen and Injeclor Ierfornance". SAL
SheIly M. H., VanDerWege . A. and Hochgrel S. (1998) LarIy Spray DeveIopnenl in
CasoIine Direcl Injecled Spark Ignilion Lngines". SAL TechnicaI Iaper 98O16O.
Sielers D. L (1998) Liquid-Ihase IueI Ienelralion in DieseI Sprays". SAL TechnicaI Iaper
98O8O9.
Sielers D. L. and Higgins . (2OO1) IIane Lifl-Off on Direcl-Injeclion DieseI Sprays Under
Quiescenl Condilions". SAL TechnicaI Iaper 2OO1-O1-O53O.
Sielers D. L., Hinggins . S. y Iickell L. M. (2OO2) IIane Lifl-Off Lenglh on Direcl-Injeclion
DieseI IueI }els: Oxygen Concenlralion Lffecls". SAL TechnicaI Iaper 2OO2-O1-O89O.
Sielers D. L. and Iickell L. M. (2OO2) Injeclion Iressure and Orifice Dianeler Lffecls on
Sool in DI DieseI". Congreso THILSLL-2OO2, VaIencia, Spain.
Talala M., Iujii H., Arai M. and Hiroyasu H. (1991) Mcan Drcp Diamc|cr cf a Dicsc| Spraq in
Vapcrizing Prcccss. 1991.
Takagi Y., Iloh T., Muranaka S., Liyana A., Ivakiri Y., Urushihara T. and Nailoh K. (1998)
SinuIlaneous Allainnenl of Iov IueI Consunplion, High Oulpul Iover and Lov
Lxhausl Lnissions in Direcl Injeclion SI Lngines". SAL TechnicaI Iaper 98O149.

Experimental study of spray generated by a new type of injector with rotary swinging needle 65
X

ExperimentaI study of spray generated by a new
type of injector with rotary swinging needIe

Hulerl Kuszevski and Kazinierz Lejda
|acu||q cf Mccnanica| |nginccring an Acrcnau|ics,
Dcpar|mcn| cf Au|cmc|itc Vcnic|cs an |n|crna| Ccm|us|icn |ngincs,
Rzcszcu Unitcrsi|q cf Tccnnc|cgq,
8 Pcus|accu Iarszauq Atc., 35-959 Rzcszcu,
Pc|an

1. Introduction
Nunerous papers have indicaled lhal lhe presenl slandard syslens of injecling fueI direclIy
inlo lhe conluslion chanler, in lhe nosl econonicaI, direcl-injeclion dieseI engines, have
reached lhe Iinils of deveIopnenl. In order lo nainlain lhe enission of loxic conponenls of
exhausl gases vilhin lhe ranges defined ly lolh LURO V and projecled LURO VI slandards,
various nodificalions lo lhe conluslion syslen viII le necessary. Il is knovn lhal injeclion
lhrough a convenlionaI nuIli-hoIe nozzIe, in conlinalion vilh induced svirI in lhe air in
lhe chanler, does nol ensure lhal parlicuIales and oxides of nilrogen are nol forned. These
loxic conponenls are anong lhe nosl difficuIl lo sulsequenlIy renove fron lhe exhausl
gases (Hiroyasu & Arai, 199O, Ieake, 1997, Kuszevski & Lejda, 2OO9). Lffeclive Iinilalion of
NO
X
and IM enission can le achieved ly providing proper nacro slruclure paranelers of
lhe spray. Iirsl of aII, il is very inporlanl lo dislrilule fueI effecliveIy in lhe conluslion
chanler, considering lhe rolary nolion of lhe air and lhe spray shape (eck el aI., 1988,
DrnhoIz & Krger, 1997, KoIInann & argende, 1997). Nexl, a Iarger range of lhe spray
fronl and apex angIe of spray is needed (KoIInann & argende, 1997, Kuszevski & Lejda,
2OO9). Moreover, il is inporlanl lo knov lhe dislrilulion of fueI in lhe spray, lecause lhen il
is possilIe lo seIecl lhe correcl rolary nolion of lhe charge in lhe cyIinder, vhich viII
guaranlee as effeclive a nix of lhe air and lhe fueI as is possilIe. A fueI spray vilh
a differenl nacro slruclure nighl have an inporlanl effecl on lhese processes, hovever il is
necessary lo use a differenl lype of injeclor, a nev lype of conslruclion and nodus
operandun (Varde and Iopa, 1983, Kuszevski, 2OO2). The design for such a device forns
lhe suljecl of lhis paper (Melz & Seika, 1998, SzIachla & Kuszevski, 2OO2, Szynaski &
ZalIocki, 1992). The speciaI fealure of lhe spray nozzIe of lhis injeclor is lhe varialiIily of lhe
fueI-spraying hoIes during injeclion. The varialiIily of lhe cross-seclions of lhese hoIes is
achieved ly lhe rolary svinging novenenl of lhe needIe (RSN injeclor). The resuIls of
invesligalions descriled leIov shov lhal a spray generaled ly lhis injeclor design has
nacro-slrucluraI paranelers lhal differ fron lhose of a slandard lradilionaI injeclor.
4
Fuel njection 66

2. The construction and principIe of operation of an injector
with rotary swinging needIe movement
The conslruclion of an injeclor vilh a rolary svinging needIe novenenl is shovn in Iig. 1.
IueI is suppIied lhrough a high-pressure lule lo lhe upper parl of lhe injeclor lody. The
fueI pressure acls on lhe vorking area 'l of a pislon, 7, joined ly a pin, 13, lo an injeclor
needIe driver, 14. This causes lhe driver lo rolale sinuIlaneousIy vilh lhe rolalion of lhe
needIe, 17, vhich is joined lo lhe driver.

9 5 6 7 8 10 11 12
2 3 4
13
14
1
15
16
17
18
Servo
End stop
Lock-nut
Adjusting nut
Spring
Piston
Washer
Body screw
Compression plunger
Compression spring
Washer
njector body
10
12
11
5
8
9
7
6
3
4
2
1
Driver pin
Spray nozzle body
Spray nozzle nut
Needle driver
Needle
Transverse hole
13
14
15
16
17
18

1 2
8 "a" 14 "b" 13
fuel
7
6
3
5
4
Driver pin
Piston working surface
Hole in driver
Needle driver
3 End stop
Washer
Spring
Adjusting nut
Lock-nut
6
7
8
Piston
5
4
njector body
Servo 2
1
13
14
"a"
"b"

Iig. 1. The conslruclion of an RSN injeclor

d
i
d
k

Iig. 2. Magnified parl of RSN spray nozzIe, d
k
: oulIel hoIe dianeler in lhe spray nozzIe
lody, d
i
: oulIel hoIe dianeler in lhe needIe

IueI fIovs lo lhe needIe lhrough a hoIe 'a, in lhe needIe driver and lhrough a fissure
lelveen lhe driver and lhe injeclor lody. Iron lhere lhe fueI fIovs lhrough an axiaIIy-
synnelricaI hoIe in lhe needIe, lolh lo lhe hoIes in lhe needIe (spraying hoIes) and in lhe
injeclor lody (oulIel hoIes), on lhe cone side surface vhich seaIs lhe needIe (Iig. 2). The
reIease of fueI fron lhe sprayer slarls vhen lhe needIe-perforning a rolary svinging

novenenl-causes a nisaIignnenl of lhe spray and lhe oulIel hoIes, so lhal lhey are
nonenlariIy occIuded, resuIling in a non-circuIar area of fIov (Iig. 3). This area increases in
proporlion lo lhe increasing angIe of needIe rolalion. Il shouId le nenlioned lhal during lhe
iniliaI phase (Iov rolary angIe), lhe needIe has a 'dead novenenl, vhich prevenls increase
of lhe fIov area. The rolalion of lhe needIe in one direclion conlinues so Iong as lhe axes of
lhe spraying and oulIel hoIes overIap. In lhis posilion, lhe pislon of lhe driver resls on lhe
end slop, 3, vhich acls as a reguIalory eIenenl, Iiniling needIe novenenl. The reIease of
fueI causes a decrease in fueI pressure, vhich causes lhe spring, 6, lo nove lhe pislon, 7
(noving in lhe servo, 2), screved inlo lhe injeclor lody, unliI il reaches a posilion
corresponding lo lhe cIosure of lhe injeclor. During lhis novenenl lhe fIov area fron lhe
injeclor decreases lo zero. The unique fealure vhich dislinguishes lhe nev conslruclion
fron lhe slandard conslruclion is lhal fueI fIovs oul lhrough a non-circuIar area, quasi-
IenlicuIar in shape, during lhe leginning and end phases of injeclion. During injeclion, fueI
is lhrollIed onIy in lhis non-circuIar area.

0.02
0.04
0.06
0.08
0.10
0.12
0.14
f [mm ]
c
2
1
3
4 5
2
0.60
0
.4
0
1 2
3 4
5
d = 0.40 [mm]
i
0 5 10 15 20 15 10 5
h [mm]
t
o
i
[deg]
1.00 0.80 0.60 0.40 0.20 0.00 0.20 0.40 0.60 0.80 1.00
1.09
0.00
o
i
t
= 0 [deg]
h = 0 [mm]
o
i
t
= 3.2 [deg]
h = 0.17 [mm]
o
i
t
= 10 [deg]
h = 0.53 [mm]
o
i
t
= 16.8 [deg]
h = 0. 91 [mm]
o
i
t
= 20 [deg]
h = 1.09 [mm]
d = 0.60 [mm]
k

Iig. 3. The diagran of oulIel hoIe in lhe spray nozzIe lody and oulIel hoIe in lhe needIe
posilion and groving of geonelricaI fIov area, f
c
: geonelricaI fIov area,
i
: angIe of needIe
rolalion (parl 17 on Iigure 1), h
l
: pislon slroke (parl 7 on Iigure 1).
Fuel njection 68

3. Method of investigations
The paranelers of lhe nacroslruclure of lhe slrean of sprayed fueI vere delernined on lhe
lasis of neasurenenls carried oul using speciaIIy conslrucled equipnenl, vhich enalIed
lolh lhe direcl olservalion of lhe deveIopnenl of lhe spray during lhe fueI injeclion lo
a chanler of fixed voIune (Melz & Seika, 1998, SzIachla & Kuszevski, 2OO2, Szynaski &
ZalIocki, 1992) and lhe neasurenenl of lhe fueI dislrilulion vilhin lhe spray of dropIels.
The schene of lesl sland for visuaI sludies is shovn in Iig. 4.

AVL
ND METER 617
1 2 3 4 5 6
7
9 8
10 11 12
13 14
15
16 17 18 19 20 21 22
N
2
A
A

Iig. 4. Schene of lesl sland for visuaI sludies, 1: lesl lench Hansnann LIH 5OO8, 2: injeclion
punp I56-O1, 3: posilion and rolalionaI speed sensor of punp shafl AVL, 4: synchronizing
pIales, 5: charge anpIifiers AVL 3O56, 6: lensonelric lridge, 7: conpuler vilh dala
acquisilion device for high-speed courses recording AVL Indineler 617, 8: signaIs decoder,
9: feeder, 1O: nanoneler, 11: visuaIizalion chanler of conslanl voIune, 12: Iighls, 13: driver
of high-speed canera, 14: high-speed canera, 15: pressure reguIalor vilh nanonelers, 16:
nilrogen cyIinder, 17: auxiIiary injeclor, 18: svilching vaIve, 19: piezoquarlz pressure sensor
AVL 5QI6OO2, 2O: induclive posilion sensor of injeclor pislon (or needIe), 21: lesled injeclor,
22: reIease vaIve

The essenliaI eIenenls of lesl sland vere: visuaIizalion chanler of conslanl voIune, vhere
a fueI vas injecled ly using lesled injeclor, 21, lesl lench ,1, injeclion punp, 2, for punping
lhe fueI lo injeclor, lhe high-speed canera, 14, vilh driver, 13, and lhe conpuler vilh dala

acquisilion devices for high-speed courses recording 7. Inside lhe chanler lo vhich fueI
vas injecled, lhere vas fixed lackpressure verified vilh nanoneler, 1O. Ior safely reasons,
gas fiIIing lhe chanler vas nilrogen, passed fron lhe cyIinder 16. The fueI for lesled
injeclor, 21, vas suppIied fron lhe injeclion punp, 2, driven ly an eIeclric nolor of lesl
lench, 1. Svilching vaIve, 18, referred fueI fron injeclion punp lo lhe lesled injeclor, 21, (al
lhe line of visuaIizalion lesl of slrean deveIopnenl) or lo lhe auxiIiary injeclor, 17, (in lhe
inlervaIs lelveen fundanenlaI lesls). To record of fueI slrean inages, il vas necessary lo
synchronize lhe vork of high-speed canera, 14, lhe lesl lench, 1, lhe Iighls, 12, and
svilching vaIve, 18. Il vas ollained using a speciaI canera driver, 13, co-operaling vilh
synchronizing pIales, 4, fixed lo lhe driving shafl of lhe lesl lench engine. AddilionaIIy,
during vork of lhe high-speed canera, using piezoquarlz pressure sensor, 19, lhe vaIues of
pressure lefore lhe injeclor vere recorded. Iurlhernore, using induclive posilion sensor, 2O,
il vas recorded dispIacenenl of injeclor needIe (or a pislon in lhe injeclor of nev lype) and
using an oplicaI rolalion speed sensor, 3, - il vas recorded rolalionaI speed of injeclion
punp canshafl. The record of lhese paranelers vas possilIe lhanks lo lhe conpuler, 7,
vilh a speciaI dala acquisilion device. The slrean inages vere recorded al a speed of 5OOO
fps. The successive inages of a deveIoping fueI slrean vere recorded every O.OOO2 s
(O.2 ns). Nexl, lhe fiIns of vidlh 16 nn vere scanning and digilaI inages vere anaIyzed
using a conpuler.
LC
7
0
O
S
A
S
o

Iig. 5. The nelhod of delernining lhe vaIues of lhe paranelers of lhe fueI spray, L
C
: range
of lhe spray fronl,
S
: apex angIe of lhe spray, A
S
: area of lhe spray projeclion in a pIane
perpendicuIar lo lhe injeclor axis

The visuaI sludies enalIed lhe foIIoving lo le anaIysed: lhe range of lhe spray fronl - L
C
,
lhe apex angIe of lhe spray -
S
, and lhe area of lhe spray projeclion in a pIane
perpendicuIar lo lhe injeclor axis - A
S
. The Iasl crilerion of lhe spray nacroslruclure
eslinalion vas inlroduced lecause of lhe irreguIar shape of lhe spray generaled ly lhe RSN
injeclor. The nelhod of delerninalion of lhe vaIues of lhe anaIysed paranelers of lhe spray
of injecled fueI is depicled in Iig. 5. A slandard injeclor vilh a D1LMK 14O/M2 pallern
sprayer, and lhe nev RSN injeclor vere conpared. olh sprayers had lhree oulIel hoIes, lhe
dianeler of lhe hoIes in lhe slandard injeclor lody and lhe RSN lype leing equaI.
Fuel njection 70

DF RO 70% RO + 30% DF
0,2 [ms] 0,2 [ms] 0,2 [ms]
0,4 [ms] 0,4 [ms] 0,4 [ms]
0,6 [ms] 0,6 [ms] 0,6 [ms]
0,8 [ms] 0,8 [ms] 0,8 [ms]
1,0 [ms] 1,0 [ms] 1,0 [ms]
1,2 [ms] 1,2 [ms] 1,2 [ms]
RSN injector
d=0.40 mm
d =0.60 mm
i
k
RSN injector
d=0.40 mm
d =0.60 mm
i
k
RSN injector
d=0.40 mm
d =0.60 mm
i
k
Standard injector
d =0.40 mm
k
Standard injector
d =0.40 mm
k
Standard injector
d =0.40 mm
k
v = 5,9 [mm /s]
2
v = 29,0 [mm /s]
2
v = 72,5 [mm /s]
2

Iig. 6. Iiclures of fueI spray propagalion ollained al use lhree-hoIe spray nozzIe vilh rolary
svinging needIe novenenl and slandard (p
o
= 17O lar, p
l
= 2O lar, q = 13O nn
3
/injeclion,
n
p
= 6OO rpn, fueI: DI, 7O RO + 3O DI, RO), d
k
: oulIel hoIe dianeler in lhe spray nozzIe
lody, d
i
: oulIel hoIe dianeler in lhe needIe, : kinenalic viscosily, DI: dieseI fueI, RO: rape oiI

In Iig. 6 lhe piclures of fueI spray propagalion ollained al use lhree-hoIe spray nozzIe vilh
rolary svinging needIe novenenl and slandard are shovn. Iron lhe figure 6 il is cIear lhal
lhe spray generaled ly injeclor vilh rolary svinging novenenl of lhe needIe is deveIoping
in a differenl vay lhan in lhe slandard nozzIe. Il causes differences in vaIues of used
paranelers lo assess lhe nacroslruclure of fueI spray. In parlicuIar il is cIear, lhal lhe spray
forning ly injeclor vilh a rolary svinging needIe novenenl is irreguIar in shape, and ils
area (area of lhe spray projeclion in a pIane perpendicuIar lo lhe injeclor axis), lhe apex
angIe of lhe spray and lhe fronl range are usuaIIy significanlIy higher conpared lo lhe
cIassicaI sprayer. IarlicuIarIy nolevorlhy are lhe resuIls presenled in Iigure 6 concerning
rape oiI, vhich is characlerized ly a high vaIue of kinenalic viscosily ( = 72.5 nn
2
/s). As
ve can see, lhe area occupied ly rape oiI spray fron lhe RSN needIe novenenl is nuch
Iarger lhan achieved for cIassicaI sprayer.

4. The range of the spray front
In lolh injeclors lhe foIIoving vaIues vere sel, leing lhe sane for each lype: Iine pressure al
lhe opening of lhe sprayer p
o
= 17O lar, fueI dose q = 13O nn
3
/injeclion and rolary veIocily
of lhe canshafl of lhe injeclion punp n
p
= 6OO rpn. IueIs of differenl viscosily (DieseI IueI
(DI), Rape OiI (RO), and 7O/3O RO/DI nixlure), vere injecled inlo lhe olservalion
chanler, vhich vas fiIIed vilh nilrogen al pressures of 15, 2O, and 25 lar. During aII
invesligalions of lhe range of lhe spray fronl, lhe apex angIe and surface area of lhe spray,
a line scaIe of O-1.4 ns vas chosen. eyond lhis range, for sone injeclion paranelers, lhe
spray fronl reached lhe vaIIs of lhe olservalion chanler. The range of lhe spray fronl for
DI, forned ly lhe RSN sprayer under various vaIues of lhe lackground pressure in lhe
olservalion chanler, is presenled in Iig. 7. Ior RSN injeclor, il can le seen lhal an increase
of nilrogen pressure in lhe olservalion chanler caused - as vas expecled - a reduclion in
lhe range of lhe spray fronl. This phenonenon is characlerislic of slandard sprayers, and
nay le ascriled lo lhe effecl of lhe aerodynanic resislance on dropIels of varialIe size. An
increase in lhe lackground pressure (gas densily), causes an increase in aerodynanic
resislance, and reduced dynanic pressure of lhe gas inlo vhich lhe injeclion is nade,
crealing adverse condilions for lhe disinlegralion of secondary dropIels. Therefore, Iarger
dropIels vilh grealer penelralive capaliIily are forned (olviousIy a Iarger dropIel has
grealer kinelic energy and viII lherefore lraveI furlher).

Iig. 7. The range of lhe DieseI IueI spray fronl forned ly lhe RSN lype, al various
lackground pressures in lhe olservalion chanler

The grealesl range of lhe fronl of lhe DI spray forned ly lolh lhe slandard injeclor (Iig. 8)
and lhe RSN injeclor, vas olserved al p
l
= 15 lar. Hovever, al p
l
= 2O lar, lhe range of lhe
fronl vas Iess lhan al p
l
= 25 lar. Mosl prolalIy lhis vas lecause during lhe anaIysed
(singIe) injeclion, al p
l
= 25 lar, lhe iniliaI veIocily of fueI al lhe sprayer oulIel vas higher
lhan lhal al p
l
= 2O lar. This vas caused ly lhe grealer difference lelveen lhe Iine pressure
and lhe pressure in lhe olservalion chanler. Therefore, lhe grealer kinelic energy of lhe
spray al p
l
= 25 lar, had a slronger infIuence on lhe novenenl of lhe fronl of lhe spray lhan
lhe enhanced aerodynanic resislance of lhe environnenl. A conparison of Iig. 7 and 8
Fuel njection 72

shovs lhal, generaIIy, lhe range of lhe fronl of lhe spray generaled ly lhe RSN sprayer is
grealer lhan lhal of lhe slandard injeclor.

Iig. 8. The range of lhe DieseI IueI spray fronl forned ly lhe slandard injeclor, al various
lackground pressures in lhe olservalion chanler

Iig. 9. The range of lhe fronl of lhe spray, forned ly lhe RSN injeclor for fueIs differing in
physicaI properlies

As couId le expecled, lhe use of fueIs of consideralIy grealer viscosily affecled lolh lypes of
injeclors ly consideralIy increasing lhe injeclion pressures. This vas caused ly a reduclion
in lhe vaIue of lhe index of fueI oulfIov fron lhe sprayer hoIes. These changes vere lhe
nain conlrilulors lo lhe increased spray fronl range for fueIs of increased viscosily (RO -
= 72.5 nn
2
/s, 7O/3O RO/DI - = 29.O nn
2
/s), in reIalion lo (DI - = 5.9 nn
2
/s) - see
Iigures 9 and 1O. An addilionaI reason for lhe increased range of lhe spray fronl vhen using
higher viscosily fueIs (olserved for lolh lypes of injeclors), vas prolalIy due lo lhe increase
in dropIel size, vhen condilions conducive lo lheir disinlegralion lecane vorse.

Iron a conparison of Iig. 9 and 1O, il nay le seen lhal - as in lhe case of DI - lhe spray
range of olher fueIs vas grealer for lhe RSN injeclor over lhe enlire line of spray
deveIopnenl.

Iig. 1O. The range of lhe fronl of lhe spray, forned ly lhe cIassicaI injeclor for fueIs differing
in physicaI properlies

5. The apex angIe and surface area of the spray
In Iig. 11 il nay le seen lhal, in lhe case of lhe RSN sprayer, a change in lackground
pressure did nol significanlIy affecl lhe vaIues of lhe apex angIes of lhe spray over lhe vhoIe
period of ils deveIopnenl. Hovever, lhe spray surface area varied, lhe grealesl area leing
olserved for p
l
= 15 lar, i.e., al lhe lackground pressure al vhich lhe range of lhe spray vas
grealesl.
ConverseIy, in lhe case of lhe slandard injeclor, lhe effecl of p
l
on lhe apex angIe
S
vas
nore visilIe - cp. Iig. 12. As couId le expecled, lhe Iargesl apex angIes occurred al
naxinun lackground pressure. The vaIues of lhe apex angIes of lhe spray dininished
during ils deveIopnenl, i.e., lhe penelralion of lhe spray in a direclion perpendicuIar lo ils
axis vas reduced, lhis has a negalive effecl on nixing. Il nay le onIy parlIy conpensaled ly
lhe facl lhal lhe spray surface area increases vilh ils deveIopnenl. The snaIIesl surface area
of lhe spray vas recorded during lhe inlernediale lackground pressure, p
l
= 2O lar, i.e., for
a vaIue corresponding lo lhe shorlesl range of lhe spray fronl.
Iron a conparison of Iig. 11 and 12 il viII le seen lhal lhe vaIues A
S
, achieved ly lhe RSN
injeclor, vere grealer lhan for lhe slandard injeclor. Il nay aIso indicale lhe superior
properlies of lhe spray fron lhe RSN injeclor, due lo inproved air/fueI nixing processes.
The Iarger area of lhe spray aIIovs dislrilulion of lhe fueI around lhe conluslion chanler
of DI engine nuch effecliveIy. In lhis case il is possilIe lo reduce a rolary nolion of lhe
charge. Too slrong rolary nolion of lhe charge can Iead lo sprays overIapping and can cause
lhe coaIescence of fueI drops. Il is unfavouralIe on accounl of IM fornalion.
Fuel njection 74

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4
t [ms]
0
2
4
6
8
10
A
s
[
c
m
2
]
0
10
20
30
40
50
s
[
d
e
g
]
s
A
s
po = 170 [bar], q = 130 [mm
3
/injection], np = 600 [rpm]
p
b
= 25 [bar]
p
b
= 20 [bar]
p
b
= 15 [bar]
spray nozzle RSN
d
k
= 0.60 [mm], d
i
= 0.40 [mm]

Iig. 11. The apex angIe and surface area of lhe spray forned ly lhe RSN lype al various
lackground pressures IeveIs

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4
t [ms]
0
2
4
6
8
10
A
s
[
c
m
2
]
0
10
20
30
40
50
s
[
d
e
g
]
s
A
s
po = 170 [bar], q = 130 [mm
3
p
b
= 25 [bar]
p
b
= 20 [bar]
p
b
= 15 [bar]
spray nozzle D1LMK 140/M2
d
k
= 0.40 [mm]

Iig. 12. The apex angIe and surface area of lhe spray forned ly lhe cIassicaI injeclor al
various lackground pressures IeveIs

The appIicalion of fueIs vilh increased kinenalic viscosily had IillIe effecl on lhe surface
area of lhe spray, A
S
(Iig. 13 and 14). Al lhe sane line, il nay le noled lhal lhe dinensions
of lhis area are nuch grealer for lhe RSN-lype lhan for lhe slandard injeclor.
The vaIue of lhe spray angIes generaled ly lhe slandard injeclor decreased inverseIy as lhe
sprays deveIoped. The vaIue of lhe angIe vas virluaIIy independenl of lhe lype of fueI used.
On lhe olher hand, in lhe case of lhe RSN sprayer, lhe apex angIe of lhe spray vas
dependenl nol onIy on lhe line of lhe spray deveIopnenl, lul aIso on lhe lype of fueI. Il is
significanl lhal lhe Iargesl vaIues of lhese angIes vere found in fueIs vilh lhe Iovesl
viscosilies and surface lension (DI). They did nol change during lhe spray deveIopnenl
period. Il is very IikeIy lhal lhe snaIIer drops devialed nore aculeIy lovards oulside lhe

Iig. 13. Apex angIe and surface area of spray forned ly lhe RSN nodeI vhen spraying fueIs
differing in physicaI properlies

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4
t [ms]
0
2
4
6
8
10
0
10
20
30
40
50
s
A
s
spray nozzle D1LMK 140/M2
d
k
= 0.40 [mm]
po = 170 [bar], pb = 20 [bar],q = 130 [mm
3
RO, = 72.5 [mm
2
/s]
70%RO+30%DF, = 29.0 [mm
2
/s]
DF, = 5.9 [mm
2
/s]

Iig. 14. Apex angIe and surface area of spray forned ly lhe slandard injeclor, spraying fueIs
vilh differenl physicaI properlies

spray. RO, vilh lhe highesl viscosily, lehaved differenlIy. The apex angIe of lhe spray
increased sleadiIy, and for line l = 1.2 ns (lhe end of lhe anaIysed fueI injeclion), il vas
grealer lhan for DI. IresunalIy, in lhis case lhe apex angIe of lhe spray resuIled fron lhe
addilionaI faclor vhich increased lhe lurluIence of oulfIov fron lhe sprayer, caused ly lhe
varialiIily of cross-seclions of lhe spraying hoIes, and lhe resuIling pernanenl change in lhe
ralio of lhe Ienglh of lhe oulIel hoIe lo ils seclionaI area.
Fuel njection 76

Iig. 15. The conparison of lhe apex angIe, surface area, and lhe fronl-range of lhe spray
generaled ly lhe cIassicaI injeclor and lhe RSN lype vhen spraying RO

In Iig. 15, an addilionaI conparison of lhe surface area, apex angIe and range of lhe spray
fronl for a spray of RO lhrough a lripIe-hoIe slandard injeclor and lhe RSN injeclor lype, is
depicled. The sludies vere carried oul al p
l
= 2O lar and a Iine pressure al injeclor opening
p
o
= 17O lar. The fueI dose vas sel al q = 13O nn
3
/injeclion, and lhe rolary veIocily of lhe

canshafl of lhe injeclion punp vas n
p
= 6OO rpn. Despile lhe facl lhal snaIIer vaIues of
injeclion pressures vere noled for lhe RSN injeclor (p
vnax
= 3OO lar, p
vav
= 189 lar, and for
lhe cIassicaI injeclor 376 lar and 236 lar, respecliveIy), lhe surface area and range of lhe
spray fronl vere nuch grealer in lhis case. OnIy lhe apex angIe of lhe spray in lhe iniliaI
phase of lhe injeclion had a Iover vaIue for lhe spray generaled ly lhis injeclor (RSN lype).
Laler in lhe cycIe, hovever, lhis angIe increased rapidIy and al lhe end of lhe anaIysed
period of spray deveIopnenl, lhe angIe vas grealer ly aloul 18 deg. Crealer vaIues of lhe
paranelers A
S
,
S
, and L
C
for lhe RSN injeclor prolalIy resuIled nol onIy fron lhe Iack of
lhrollIing of lhe fueI fIov in lhe needIe seal, lul aIso fron lhe nechanicaI aclion of lhe oulIel
hoIes in lhe spray nozzIe on lhe spray.

6. RadiaI distribution of fueI in spray drops generated
by standard and RSN injectors
Invesligalions of fueI dislrilulion vere carried oul using lolh injeclors in a spray of dropIels,
al a conslanl injeclion punp speed of n
p
= 6OO rpn. The fueI dose vas adjusled lo 13O
nn
3
/injeclion and lhe Iine pressure al lhe injeclor vas p
o
= 17O lar. IueI vas injecled inlo a
lackground alnospheric of p
l
= 1 lar, lhe fueI IeveI H
p
in lhe neasuring vesseIs vas read afler
each 1OOO-cycIe period. The radiaI dislrilulion of fueI in a spray vas neasured ly direcling lhe
sprayed fueI inlo a series of slanding neasuring vesseIs. The inIel openings of lhe vesseIs vere
perpendicuIar lo lhe axis of lhe spraying hoIe. IueI dislrilulion in a spray vas invesligaled ly
pIacing lhe inIels of lhe neasuring vesseIs al severaI dislances fron lhe edge of lhe inIel hoIe
of lhe sprayer lody - S
r
. These vere: 75, 15O and 21O nn. In addilion, for each dislance, lhe
series of vesseIs vas rolaled ly 45 deg, vhich enalIed delerninalion of lhe fueI dislrilulion in
four pIanes, nuluaIIy incIined al angIes of 45 deg. Iig. 17 and 18 have lhe foIIoving Iegend:
'Iosilion 9O deg, denoling lhe axis '-x + x and lhe axis of a sprayer in one pIane. 'Iosilion 45
deg denoles lhal lhe series of vesseIs had leen lurned lhrough 45 deg in reIalion lo posilion 9O
deg.

4
5
r = 70 [mm]
s
r = 0
s
r = 70 [mm]
s
r = 0
s
r = 70 [mm]
s
r = 70 [mm]
s
-x +x
+y
-y
-x
+x
+y
-y

Iig. 16. A series of cyIindricaI neasuring vesseIs used in delernining fueI dislrilulion in
a spray of drops (lop viev)

Fuel njection 78

The heighl of fueI in lhe neasuring vesseIs vas adopled (denoled ly H
p
) as a conparalive
neasure lo ascerlain lhe fueI dislrilulion in a spray of dropIels. A radius al vhich a chosen
fueI coIunn vas Iocaled, i.e., lhe radiaI dislance fron lhe lheorelicaI axis of a spray, vas
denoled ly r
s
(Iig. 16). 'Direclion x and 'direclion y (Iegends on figures), denole vesseIs
pIaced on lhe '-x + x and '-y + y axes, respecliveIy, in Iig. 16.
SiniIar lo lhe case of lhe direcl olservalion sludies - lhe slandard injeclor vilh a D1LMK
14O/M2 sprayer, and lhe nev lype injeclor - denoled as RSN, vere sludied.

Iig. 17. Conparison of lhe radiaI dislrilulion of fueI in a spray in lhe 'y direclion for lhe
slandard injeclor and lhe RSN lype

Using hislograns, Iig. 17 and 18 shov lhe resuIls of sludies of lhe radiaI dislrilulion of fueI in
a spray of drops, forned ly lhe slandard injeclor (D1LMK 14O/M2) and lhe RSN lype. Ior
sinpIicily, parlicuIar vaIues of lhe radius r
s
are pIolled againsl lhe neasured heighls of fueI
coIunns in lhe neasuring vesseIs, H
p
, ralher lhan lhe reIaled vaIues of lhe spray densily.

As seen in lhe slandard injeclor, lhe usuaI silualion prevaiIed, and lhe highesl concenlralion
of fueI Iay al lhe core of lhe spray, i.e., lhe densily of a unil spray has a naxinun vaIue al
lhe spray axis, vhere Iarge dianeler dropIels are nosl nunerous, as slaled earIier. A
characlerislic fealure of fueI dislrilulion in lhe slandard spray is ils synnelry around lhe
spray axis (lhe axis in Iine vilh lhe axis of synnelry of lhe oulIel hoIe), and lhe IeveIIing off
of lhe dislrilulion as lhe dislance fron lhe sprayer increases (H
p
vaIues dininish in lhe
cenlre and increase sIighlIy lovards lhe oulside).

7
0
6
0
5
0
4
0
3
0
2
0
1
0 0
1
0
2
0
3
0
4
0
5
0
6
0
7
0
r
s
[mm]
0
20
40
60
80
100
H
p
[
m
m
]
7
0
6
0
5
0
4
0
3
0
2
0
1
0 0
1
0
2
0
3
0
4
0
5
0
6
0
7
0
r
s
[mm]
0
20
40
60
80
100
H
p
[
m
m
]
7
0
6
0
5
0
4
0
3
0
2
0
1
0 0
1
0
2
0
3
0
4
0
5
0
6
0
7
0
r
s
[mm]
0
20
40
60
80
100
H
p
[
m
m
]
7
0
6
0
5
0
4
0
3
0
2
0
1
0 0
1
0
2
0
3
0
4
0
5
0
6
0
7
0
r
s
[mm]
0
20
40
60
80
100
H
p
[
m
m
]
7
0
6
0
5
0
4
0
3
0
2
0
1
0 0
1
0
2
0
3
0
4
0
5
0
6
0
7
0
r
s
[mm]
0
20
40
60
80
100
H
p
[
m
m
]
7
0
6
0
5
0
4
0
3
0
2
0
1
0 0
1
0
2
0
3
0
4
0
5
0
6
0
7
0
r
s
[mm]
0
20
40
60
80
100
H
p
[
m
m
]
p
o
= 170 [bar], q = 130 [mm
3
/injection], n
p
= 600 [rpm], p
b
1 [bar]
spray nozzle D1LMK 140/M2 spray nozzle RSN
S
r
= 75 [mm] S
r
= 75 [mm]
S
r
= 150 [mm] S
r
= 150 [mm]
S
r
= 210 [mm] S
r
= 210 [mm]
direction x, location 45 deg
-x +x -x
-x -x
-x -x
+x
+x +x
+x +x

Iig. 18. Conparison of lhe radiaI dislrilulion of fueI in a spray in lhe 'x direclion for lhe
slandard injeclor and lhe RSN lype

The IeveIIing off of lhe fueI dislrilulion in a spray as lhe dislance fron lhe sprayer increases
is caused ly lhe size reduclion of lhe dropIels and lhe danping of lheir novenenl.
AddilionaIIy, lhe lurluIenl novenenls in a spray lend lo carry fueI lovards lhe ouler Iayers
Fuel njection 80

of lhe spray, and lhe dislrilulion lecones nore equaI (Melz and Seika, 1998). This
phenonenon is reIaled lo lhe fueI novenenl in lhe Ialer phase of injeclion and il is aIso
olserved in lhe spray forned ly lhe RSN-lype injeclor. The IeveIIing off of lhe fueI
dislrilulion vilh increased dislance fron lhe sprayer seens lo le a phenonenon shared
anong sprays generaled ly lolh injeclor lypes.
A spray of fueI generaled ly lhe RSN sprayer shovs asynnelry, lhe dislrilulion in lhe 'x
direclion differs fron lhal in lhe 'y direclion. In lhe 'y direclion parlicuIarIy, lhe
concenlralion of fueI is consideralIy Iarger (aIso vhen lhe series of vesseIs is rolaled
lhrough 45 deg). Moreover, in lhe 'y direclion a grealer shifl of lhe area of lhe naxinun
fueI concenlralion (core of a spray) nay le olserved in conparison lo lhe 'x direclion. This
Ieads lo lhe concIusion lhal lhe fueI dislrilulion in lhe spray forned ly lhe RSN sprayer
does nol shov any synnelry in reIalion lo lhe lheorelicaI axis of lhe spray.
The Iargesl shifl of lhe spray core fron lhe lheorelicaI axis for lhe RSN sprayer vas
olserved in lhe 'y direclion. This effecl appeared vhen lhe axis of lhe sprayer vas in one
pIane vilh lhe axis al -x + x. In lhis posilion lhe axis of lhe needIe rolalion vas
perpendicuIar lo lhe 'y direclion. The asynnelry of lhe core of lhe spray generaled ly lhe
RSN sprayer nay le expIained ly lhe change of lhe cross-seclions of lhe oulIel hoIes and lhe
resuIling nechanicaI aclion of lhe surface of lhe hoIe in lhe sprayer lody on lhe fueI leing
discharged. The fueI, fIoving lhrough lhe spraying hoIe (parlicuIarIy in lhe opening phase),
hil lhe surface of lhe oulIel hoIe. This changed lhe direclion of lhe fIov, vhich caused
varialions in lhe posilion of lhe core in lhe cross-seclion of lhe spray.
The spray generaled ly slandard injeclor is axiaIIy synnelric. More fueI saluralion in lhe
spray core causes a differenl vaIue of conluslion air faclor. This is unfavouralIe, lecause
sool is usuaIIy produced in lhe rich nixlure area (IocaI deficiency of air) al a sufficienlIy
high lenperalure (8OO-14OO K). This happens nainIy in lhe core of lhe fueI spray and al ils
rear, vhere lhe concenlralion of fueI dropIels is oflen higher.
Lxeculed invesligalions of radiaI dislrilulion of fueI in spray confirn lhal lhe spray
generaled ly RSN injeclor is nol synnelricaI. The shifl of lhe spray core oulside (as effecl of
needIe rolary) can le favouralIe on accounl of lhe possilIy slronger inpacl of gas nediun
on spray zone, vhere lhe concenlralion of lhe fueI is higher. In lhis case, lhe secondary drop
lreak-up viII le nore inlensive. SnaIIer dianelers of drops are olviousIy favouralIe vilh
regard lo sool and IM fornalion.

7. ConcIusions
The paranelers of lhe injeclion syslen have a decisive effecl on lhe rale of conluslion in lhe
dieseI engine, lecause of lhe infIuence on quaIily of forned air-fueI nixlure. Hovever, lhe
oplinaI nacroslruclure of lhe spray, vhich is dislriluled in lhe cyIinder voIune, depends
on lhe lype and conslruclion of lhe injeclor. On lraking, lhe fueI slrean in drops increases
lhe area of conlacl lelveen lhe fueI and air. Il causes, firsl of aII, fueI vaporisalion and, lhen,
ils diffusion inlo air. The pressure energy generaled ly lhe injeclion syslen is consuned on
spraying of lhe fueI slrean vhich, logelher vilh lhe phenonena of physicaI and chenicaI
parls of seIf-ignilion deIay, Ieads lo fasl increase in nixlure enlropy.
A leller quaIily of fueI spraying guaranlees RSN injeclor, vhich vas confirned ly nodeI
invesligalions. The seIecled resuIls have leen presenled in lhe paper.

The resuIls of lhese invesligalions shov lhal fueI sprays forned ly using a RSN lype injeclor
differ fron lhose generaled ly a slandard injeclor. In parlicuIar, lhe paranelers anaIysed, i.e.,
lhe range of lhe spray-fronl, lhe apex angIe of lhe spray and ils surface area, reach grealer
vaIues for a spray forned ly lhe nev RSN lype of sprayer, lhis nay posiliveIy affecl lhe
ecoIogicaI inpacl as veII as lhe perfornance of engines filled vilh injeclors of lhis lype.
Varialion in lhe condilions of injeclion (pressure changes in lhe gaseous nediun inlo vhich
fueI is injecled, change due lo use of fueIs of differing viscosily), affecls lhe nacroslruclure
of sprays generaled differenlIy ly each lype of injeclor. The lesl exanpIe nay le lhe
variance in lhe apex angIe of lhe spray vhiIe spraying RO. In lhe slandard injeclor, il vas
found lhal lhis angIe dininished as lhe spray deveIoped, vhiIe in lhe RSN injeclor lhe
opposile lendency vas olserved.
The invesligalions of fueI dislrilulion in a spray of dropIels confirn lhal lhe spray
generaled ly lhe RSN-lype injeclor deveIops in a differenl vay fron lhal generaled ly lhe
slandard injeclor. In parlicuIar, lhe resuIls of lhese sludies shov lhe asynnelry of lhe spray
forned ly lhe nev lype of injeclor.
More favouralIe paranelers of lhe nacroslruclure of lhe spray generaled ly lhe RSN
injeclor aIIov lhe air-fueI nixlure lo lurn nore conpIeleIy. Nexl, il provides reducing of
enission of loxic conponenls fron exhausl gases. Hovever, for using a nev lype of injeclor,
nodificalion of lhe conluslion chanler is needed. This nodificalion has lo consider higher
vaIues of spray nacroslruclure paranelers. Ior exanpIe, a confirned Iarger range of lhe
spray forned ly a nev lype of injeclor can le served. Al injeclion inlo lhe conluslion
chanler vilhoul nodificalion, lhe spray can sellIe on lhe vaIIs of conluslion chanler
vhich can cause increase in IM enission. The aulhors conducled invesligalions in lhis
range and inlend lo pulIish lhen in lhe sulsequenl papers.

8. NomencIature
The TalIe 1 shovs lhe paranelers for lhe alonizalion of fueI, vhich vere used in lhe sludy.
AddilionaIIy, lhere are used descriplion of paranelers, if required.

Quan|i|q Uni| Spccifica|icn
A
S
|cn
2
j
Surface of viev of fueI spray on perpendicuIar pIane lo
spray nozzIe axis
L
C
|nnj Tip penelralion of fueI spray
S
|degj Apex angIe of fueI spray
H
p
|nnj
IueI IeveI (al neasuring of lhe fueI radiaI dislrilulion in
a spray)
S
r
|nnj
Dislance of an inIel area of lhe neasuring vesseI fron lhe
edge of oulIel hoIe in spray nozzIe lody (neasuring fueI
radiaI dislrilulion in a spray)
r
s
|nnj
Dislance neasuring poinl fron lhe lheorelicaI axis spray
(neasuring fueI radiaI dislrilulion in a spray)
d
k
|nnj OulIel hoIe dianeler in a needIe
d
i
|nnj OulIel hoIe dianeler in a spray nozzIe lody
TalIe 1. Descriplion of paranelers used in lhe sludy
Fuel njection 82

Tnc ccn|inua|icn cf Ta||c 1
Quan|i|q Uni| Spccifica|icn
h
l
|nnj Iislon slroke of injeclor
i
|degj AngIe of needIe rolalion
f
c
|nn
2
j CeonelricaI fIov area
q |nn
3
/injeclionj IueI dose
l |nsj Tine
n
p
|rpnj RolalionaI speed of injeclion punp canshafl
p
o
|larj Slalic opening pressure of injeclor
p
l
|larj ackground pressure
p
vnax
|larj Maxinun fueI injeclion pressure
p
vav
|larj Average fueI injeclion pressure
|nn
2
/sj Kinenalic viscosily of fueI
DI - DieseI IueI
RO - Rape OiI

9. References
eck, N.}., Uyehara, O.A. & }ohnson, W.I. (1988) |ffcc|s cf |uc| |njcc|icn cn Dicsc| Ccm|us|icn,
SAL Transaclions, Iaper 88O299.
DrnhoIz, M. & Krger, M. (1997) Hal der DieseInolor aIs Iahrzeuganlriel eine zukunfl`, 6.
Aacncncr Kc||cquium |anrzcug - un Mc|crcn|ccnni|, Akvizgran, Cernany.
Hiroyasu, H. & Arai, M. (199O) S|ruc|urcs cf |uc| Spraqs in Dicsc| |ngincs, SAL Transaclions,
Iaper 9OO475.
KoIInann, K. & argende, M. (1997) DI - DieseInolor und DI - Ollonolor - Wohin gehl die
Ikv - MolorenenlvickIung`, Sqmpcsium Dicsc|mc|crcn|ccnni| 98, Technische
Akadenie LssIingen, OslfiIdern, Cernany.
Kuszevski, H. (2OO2) Ip|qu zmicnnqcn przc|rcjcu uq|c|cuqcn u|rqs|iuacza z c|rc|cu ig|ic
na rczpq|anic c|cju napcucgc, IhD Disserlalion, Cracov Universily of TechnoIogy,
Cracov, IoIand.
Kuszevski, H. & Lejda, K. (2OO9) LxperinenlaI invesligalions of a nev lype of fueIiing
syslen for heavy-duly dieseI engines, |n|crna|icna| ]curna| cf Hcatq Vcnic|c Sqs|cms,
|ncrscicncc |n|crpriscs ||, OIney, UK.
Melz, N. & Seika, M. (1998) Die LuflquaIill in Luropa lis zun }ahre 2O1O nil und ohne
LURO IV Crenzverle, 19. |n|crna|icna|cs Iicncr Mc|crcnsqmpcsium,
Iorlschrilllerichle VDI Reihe 12, Nr 348, Wien, Auslria.
Ieake, S. (1997) Vcnic|c an |uc| - Cna||cngcs 8cqcn 2000, Aulonolive IulIishing, London,
UK.
SzIachla, Z. & Kuszevski, H. (2OO2) Ip|qu zmicnnqcn przc|rcjcu uq|c|cuqcn u|rqs|iuacza z
c|rc|cu ig|ic na rczpq|anic c|cju napcucgc, Rep. 5 T12D O26 22, Cracov, IoIand.
Szynaski, }. & ZalIocki, M. (1992) I|rqs|iuacz c si|ni|a spa|incucgc, Ialenl AppIicalion in
Ialenl Deparlnenl R.I, I-294889, IoIand.
Varde, K.S. & Iopa, D.M. (1983) DieseI IueI Spray Ienelralion al High Injeclion Iressures,
SAL Transaclions, Iaper 83O448.

X

Effect of injector nozzIe hoIes on
dieseI engine performance

Senin
|ns|i|u| Tc|nc|cgi Scpu|un Ncpcm|cr
|ncncsia

AlduI Rahin IsnaiI
Unitcrsi|q Ma|aqsia Panang
Ma|aqsia

1. Introduction
The four-slroke direcl-injeclion dieseI engine lypicaI vas neasured and nodeIed ly akar
el aI (2OO7) using CT-IOWLR conpulalionaI nodeI and has expIored of dieseI engine
perfornance effecl lased on engine speeds. CT-IOWLR is lhe Ieading engine sinuIalion
looI used ly engine and vehicIe nakers and suppIiers and is suilalIe for anaIysis of a vide
range of engine issues. The delaiIs of lhe dieseI engine design vary significanlIy over lhe
engine perfornance and size range. In parlicuIar, differenl conluslion chanler geonelries
and fueI injeclion characlerislics are required lo deaI effecliveIy vilh najor dieseI engine
design prolIen achieving sufficienlIy rapid fueI-air nixing rales lo conpIele lhe fueI-
lurning process in lhe line avaiIalIe. According lo Heyvood (1988) and Canesan (1999), a
vide variely of inIel porl geonelries, cyIinder head and pislon shapes, and fueI-injeclion
pallerns are used lo acconpIish lhis over lhe dieseI size range. The engine ralings usuaIIy
indicale lhe highesl pover al vhich nanufaclurer expecl lheir producls lo give salisfaclory
of pover, econony, reIialiIily and duraliIily under service condilions. Maxinun lorque
and lhe speed al vhich il is achieved, is usuaIIy given aIso ly Heyvood (1988). The
inporlance of lhe dieseI engine perfornance paranelers are geonelricaI properlies, lhe
lern of efficiency and olher reIaled engine perfornance paranelers. The engine efficiencies
are indicaled lhernaI efficiency, lrake lhernaI efficiency, nechanicaI efficiency, voIunelric
efficiency and reIalive efficiency (Canesan, 1999). The olher reIaled engine perfornance
paranelers are nean effeclive pressure, nean pislon speed, specific pover oulpul, specific
fueI consunplion, inlake vaIve nach index, fueI-air or air-fueI ralio and caIorific vaIue of lhe
fueI (Heyvood, 1988, Canesan, 1999, Senin el aI., 2OO7). According lo Heyvood (1988) in
lhe dieseI engine geonelries design vrillen lhal dieseI engine conpression ralio is
naxinun cyIinder voIune or lhe dispIaced voIune or svepl and cIearance voIune divided
ly nininun cyIinder voIune. And lhe pover deIivered ly lhe dieseI engine and alsorled
ly lhe dynanoneler is lhe producl of lorque and anguIar speed. The engine efficiencies,
every ils efficiencies defined ly Canesan (1999).
5
Fuel njection 84

2. Important
In lhis chapler has invesligaled lhe effecl of injeclor nozzIe hoIes dianeler geonelries on lhe
perfornance of dieseI engine such as indicaled pover, indicaled lorque, fueI consunplion
and fueI in-engine cyIinder. The invesligalion is using conpulalionaI nodeIIing lased on
varialion engine speeds.

3. Engine Performance Review
In lhe dieseI engine geonelries design ly Heyvood (1988), lhe dieseI engine conpression
ralio is naxinun cyIinder voIune or lhe dispIaced voIune or svepl (V
) and cIearance
voIune (V
c
) divided ly nininun cyIinder voIune (V
c
). The dieseI engine conpression ralio
can le caIcuIaled as leIov:

F
F G
F
9
9 9
U
+
= (1)

and lhe pover deIivered ly lhe dieseI engine and alsorled ly lhe dynanoneler is lhe
producl of lorque and anguIar speed. DieseI engine pover definilion as:

P = 2NT
(2)

In lhe dieseI engine efficiencies, every ils efficiencies defined ly Canesan (1999). Indicaled
lhernaI efficiency (
i|n
) is lhe ralio of energy (|) in lhe indicaled pover (ip) lo lhe inpul fueI
energy. rake lhernaI efficiency (
||n
) is lhe ralio of energy in lhe lrake pover (|p),
MechanicaI efficiency (
m
) is defined as lhe ralio of lrake pover (|p) or deIivered pover lo
lhe indicaled pover (ip) or pover provided lo lhe pislon and il can aIso le defined as lhe
ralio of lhe lrake lhernaI efficiency lo lhe indicaled lhernaI efficiency. ReIalive efficiency or
efficiency ralio (
rc|
) is lhe ralio of lhernaI efficiency of an acluaI cycIe lo lhal of lhe ideaI
cycIe, lhe efficiency ralio is a very usefuI crileria vhich indicales lhe degree of deveIopnenl
of lhe engine. Canesan (1999) vrillen lhal, lhe one of lhe very inporlanl paranelers vhich
decides lhe perfornance of four-slroke engines is voIunelric efficiency (
t
), vhere four-
slroke engines have dislincl suclion slroke and lherefore lhe voIunelric efficiency indicales
lhe lrealhing aliIily of lhe engine. The voIunelric efficiency is defined as lhe voIune fIov
rale of air inlo lhe inlake syslen divided ly lhe rale al vhich lhe voIune is dispIaced ly lhe
syslen. The nornaI range of voIunelric efficiency al fuII lhrollIe for SI engines is 8O lo
85 and for CI engines is 85 lo 9O.

(
LS
LWK
= q (3)

(
ES
EWK
= q (4)

LS
ES
P
= q (5)

2 /
.
1 9
P
GLVS D
D
Y
q =
(6)

eIIiciency standard - Air
eIIiciency thermal Actual
=
UHO
q

(7)

The olher reIaled engine perfornance vas defined ly Heyvood (1988), KovaIevicz
(1984), Slone (1997) and Canesan (1999). Mean effeclive pressure (mcp), vhere n
R
is lhe
nunler of crank revoIulions for each pover slroke per cyIinder (lvo for four-slroke, one for
lvo-slroke cycIes) as :

1 9
3Q
PHS
G
5
=
(8)

The neasure of an engines efficiency vhich viII le caIIed lhe fueI conversion efficiency is
given ly Heyvood (1988):

( )
( )
+9 I +9 5 I
5
+9 I
F
4 P
3
4 1 Q P
1 3Q
4 P
:
QI = = =
/
/

(9)

Specific fueI consunplion as :

3
P
VIF
I
=
(1O)

In lhe engine lesling, lolh lhe air nass fIov rale m
a
and lhe fueI nass fIov rale m
f
are
nornaIIy neasured. The ralio of lhese fIov rales is usefuI in defining engine operaling
condilions are air/fueI ralio (A/I) and fueI/air ralio (I/A).
The foIIoving reIalionships lelveen dieseI engine perfornance paranelers can le
deveIoped. Ior pover P:

5
+9 D I
Q
$ ) 14 P
3
) / ( q
=

(11)

2
) / (
,
$ ) 4 19
3
L D +9 G Y I
q q
=
(12)
Fuel njection 86

Ior lorque T :
t
q q
4
) / (
,
$ ) 4 9
7
L D +9 G Y I
=
(13)

Ior nean effeclive pressure :
mcp = ) / (
,
$ ) 4
L D +9 Y I
q q (14)

The specific pover or lhe pover per unil pislon area is a neasure of lhe engine designers
success in using lhe avaiIalIe pislon area regardIess of cyIinder size. The specific pover is :

2
) / (
,
$ ) 1/4
$
3
L D +9 Y I
S
q q
= (15)

Mean pislon speed :

4
) / (
,
$ ) 4 6 1
$
3
L D +9 S Y I
S
q q
= (16)

Heyvood (1988) vrillen lhal, specific pover is lhus proporlionaI lo lhe producl of nean
effeclive pressure and nean pislon speed. These reIalionship iIIuslraled lhe direcl
inporlance lo engine perfornance of high fueI conversion efficiency, high voIunelric
efficiency, increasing lhe oulpul of a given dispIacenenl engine ly increasing lhe inIel air
densily, naxinun fueI/air ralio lhal can le usefuI lurned in lhe engine and high nean
pislon speed.

4. ModeIIing of Injector NozzIe HoIes
The four-slroke direcl-injeclion (DI) dieseI engine vas presenled in lhis chapler. The
specificalion of lhe seIecled dieseI engine vas presenled in TalIe 1. To deveIop lhe four-
slroke direcl-injeclion dieseI engine nodeIing is slep ly slep, lhe firsl slep is open aII of lhe
seIecled dieseI engine conponenls lo neasure lhe engine conponenls parl size. Then, lhe
engine conponenls size dala viII le inpul lo lhe soflvare Iilrary of lhe aII engine
conponenls dala. To creale lhe nodeI, seIecl vindov and lhen liIe vilh lenpIale Iilrary
fron lhe nenu. This viII pIace lhe lenpIale Iilrary on lhe Iefl hand side of lhe screen. The
lenpIale Iilrary conlains aII of lhe avaiIalIe lenpIales lhal can le used in conpulalionaI
nodeIing. Sone of lhese lenpIales lhose lhal viII le needed in lhe projecl need lo le copied
inlo lhe projecl lefore lhey can le used lo creale oljecls and parls. Ior lhe purpose of lhis
nodeI, cIick on lhe icons Iisled and drag lhen fron lhe lenpIale Iilrary inlo lhe projecl
Iilrary. Sone of lhese are lenpIales and sone are oljecls lhal have aIready leen defined
and incIuded in lhe lenpIale Iilrary (Canna TechnoIogies, 2OO4). This chapler focused on
fueI nozzIe hoIe of fueI injeclor. The engine nodeIing is according lo Senin el aI. (2OO7) as
shovn in Iig. 1.

AII of lhe paranelers in lhe nodeI viII le Iisled aulonalicaIIy in lhe case selup and each
one nusl le defined for firsl case of lhe sinuIalion. The physicaIIy of lhe injeclor fueI nozzIe
hoIe naleriaI delaiIed vere invesligaled in lhis research is shovn in Iig. 2. In lhis figure vas
shoved lhe delaiI of injeclion hoIe or fueI nozzIe hoIe. The fueI nozzIe hoIes vouId le
changed in vide dianeler of nozzIe hoIe and in differenl nunler of nozzIe hoIe.

EngInc Paramctcrs Va!uc EngInc Paramctcrs Va!uc
ModeI CI186I Inlake vaIve cIose (
O
CA) 53O
ore (nn) 86.O Lxhausl vaIve open (
O
CA) 147
Slroke (nn) 7O.O Lxhausl vaIve cIose (
O
CA) 282
DispIacenenl (cc) 4O7.O Max. inlake vaIve open (nn) 7.O95
Nunler of cyIinder 1 Max. exhausl vaIve open (nn) 7.O95
Connecling rod Ienglh (nn) 118.1 VaIve Iifl periodicily (deg) 36O
Iislon pin offsel (nn) 1.OO IueI nozzIe dianeler (nn) O.1
Inlake vaIve open (
O
CA) 395 IueI nozzIe hoIe nunler (pc) 4
TalIe 1. Specificalion of lhe seIecled dieseI engine

Iig. 1. Direcl-injeclion dieseI engine nodeIing

Fuel njection 88

Iig. 2. DelaiI of injeclor fueI nozzIe hoIes

Whenever lhe conpulalionaI sinuIalion is running, lhe conpulalionaI nodeI produces
severaI oulpul fiIes lhal conlain sinuIalion resuIls in various fornals. Mosl of lhe oulpul is
avaiIalIe in lhe posl-processing appIicalion. The soflvare is poverfuI looI lhal can le used
lo viev aninalion and order anaIysis oulpul (Canna TechnoIogies, 2OO4). Afler lhe
sinuIalion vas finished, reporl lalIes lhal sunnarize lhe sinuIalions can le produced.
These reporls conlain inporlanl infornalion aloul lhe sinuIalion and sinuIalion resuIl in a
laluIar forn. The conpulalionaI sinuIalion of lhe engine nodeI resuIl is inforned lhe
engine perfornance. The running sinuIalion resuIl in lhis research is focused on lhe engine
perfornance dala lased on varialion of fueI nozzIe naleriaI hoIe dianeler size, dianeler
nunler and lhe differenl engine speed (rpn). The dieseI engine nodeI vas running on any
differenl engine speeds in rpn, lhere are 5OO, 1OOO, 15OO, 2OOO, 25OO, 3OOO and 35OO. The
varialions of injeclor fueI nozzIe hoIes nunler are lased on nuIli hoIes and nuIli dianeler
hoIes, lhe sinuIalion nodeI lhere are slarled fron lhe injeclor fueI nozzIe 1 hoIe, 2 hoIes, 3
hoIes, 4 hoIes, 5 hoIes, 6 hoIes, 7 hoIes, 8 hoIes, 9 hoIes and 1O hoIes.

5. Effect of Injector NozzIe HoIes on FueI in Engine CyIinder
The sinuIalion resuIls are shovn in every cases, such as case 1 is on 5OO rpn, case 2 is on
1OOO rpn, case 3 is on 15OO rpn, case 4 is on 2OOO rpn, case 5 is on 25OO rpn, case 6 is on
3OOO rpn, case 7 is on 35OO rpn and case 8 on 4OOO rpn. Nunerous sludies have suggesled
lhal decreasing lhe injeclor nozzIe orifice dianeler is an effeclive nelhod on increasing fueI
air nixing during injeclion (aik, 2OO1). SnaIIer nozzIe hoIes have found lo le lhe nosl
efficienl al fueI/air nixing prinariIy lecause lhe fueI rich core of lhe jel is snaIIer. In
addilion, decreasing lhe nozzIe hoIe orifice dianeler, vouId reduce lhe Ienglh of lhe
polenliaI core region. UnforlunaleIy, decreasing nozzIe hoIes size causes a reduclion in lhe
lurluIenl energy generaled ly lhe jel.

Since fueI air nixing is conlroIIed ly lurluIence generaled al lhe jel loundary Iayer, lhis
viII offsel lhe lenefils of lhe reduced jel core size. Iurlhernore, jels energing fron snaIIer
nozzIe orifices vere shovn nol lo penelrale as far as lhose energing fron Iarger orifices.
This decrease in penelralion neans lhal lhe fueI viII nol le exposed lo aII of lhe avaiIalIe air
in lhe chanler. The effecl of fueI nozzIe hoIes nunler and geonelries of in-cyIinder engine
Iiquid fueI are shovn in Iig. 3 - Iig. 12,

Iig. 3. Liquid fueI in cyIinder of injeclor
nozzIe 1 hoIes
nozzIe 2 hoIes

nozzIe 3 hoIes
nozzIe 4 hoIes

nozzIe 5 hoIes
nozzIe 6 hoIes
Fuel njection 90

nozzIe 7 hoIes
Iig. 1O. Liquid fueI in cyIinder of
injeclor nozzIe 8 hoIes

Iig. 11. Liquid fueI in cyIinder of
Iig. 12. Liquid fueI in cyIinder of
injeclor nozzIe 1O hoIes

Ior excessiveIy snaII nozzIe size, lhe inprovenenls in nixing reIaled lo decreased pIune
size nay le negaled ly a reduclion in radiaI penelralion (aungarler, 2OO6). This lehavior
is undesiralIe lecause il reslricls penelralion lo lhe chanler exlrenilies vhere a Iarge
porlion of lhe air nass resides. Iurlhernore, il hanpers air enlrainnenl fron lhe head side
of lhe pIune lecause lhe exposed surface area of lhe pIune is reduced. Il has leen
suggesled lhal a nozzIe conlaining nany snaII hoIes vouId provide leller nixing lhan a
nozzIe consisling of a singIe Iarge hoIe. The effecl of injeclor nozzIe nuIli hoIes in-cyIinder
engine unlurned fueI are shovn in Iig. 13 - Iig. 22.

The oplinaI nozzIe design vouId le one lhal provided lhe naxinun nunler of Iiquid fueI
lurn in conluslion process and nininun nunler of Iiquid fueI unlurned. TheorelicaIIy, a
1O hoIes nozzIe salisfies lhis requirenenl. UnforlunaleIy, jels energing fron a 1O hoIes
nozzIe lended lo le very susceplilIe. AII of lhe nozzIes exanined and lhe resuIl shovn lhal
lhe seven hoIes nozzIe provided lhe lesl resuIls for any differenl engine speeds in
sinuIalion and lhe lesl perfornance shovn on Iov speed engine.

Iig. 13. Unlurned fueI in cyIinder of



Iig. 2O. Unlurned fueI in cyIinder of
Fuel njection 92

injeclor nozzIe 1O hoIes

6. Effect of Injector NozzIe HoIes on Engine Performance
The sinuIalion resuIl on engine perfornance effecl of injeclor fueI nozzIe hoIes nunler and
geonelries in indicaled pover, indicaled lorque and indicaled specific fueI consunplion
(ISIC) of engine are shovn in Iigure 23 - 25. The injeclor fueI nozzIe hoIes orifice dianeler
and injeclor nozzIe hoIes nunlers effecl on indicaled pover, indicaled lorque and ISIC
perfornance of direcl-injeclion dieseI engine vas shovn fron lhe sinuIalion nodeI running
oulpul. An aerodynanic inleraclion and lurluIence seen lo have conpeling effecls on
spray lreakup as lhe fueI nozzIe hoIes orifice dianeler decreases. The fueI drop size
decreases if lhe fueI nozzIe hoIes orifice dianeler is decreases vilh a decreasing quanlilalive
effecl for a given sel of jel condilions.

Indicated Torque Effect of Fuel Nozzle Holes Number
0
5
10
15
20
25
30
35
40
45
0 500 1000 1500 2000 2500 3000 3500 4000 4500
Engine Speed (rpm)
I
n
d
i
c
a
t
e
d

T
o
r
q
u
e

(
N
-
m
)
Nozzle 1 hole Nozzle 2 holes Nozzle 3 holes Nozzle 4 holes Nozzle 5 holes
Nozzle 6 holes Nozzle 7 holes Nozzle 8 holes Nozzle 9 holes Nozzle 10holes

Iig. 23. Lffecl of fueI nozzIe hoIes on indicaled lorque of dieseI engine

Indicated Power Effect of Fuel Nozzle Holes Number
0
1
2
3
4
5
6
7
8
9
10
0 500 1000 1500 2000 2500 3000 3500 4000 4500
Engine Speed (rpm)
I
n
d
i
c
a
t
e
d

P
o
w
e
r

(
k
W
)
Nozzle 6 holes Nozzle 7 holes Nozzle 8 holes Nozzle 9 holes Nozzle 10holes

Iig. 24. Lffecl of fueI nozzIe hoIes on indicaled pover of dieseI engine

ISFC Effect of Fuel Nozzle Holes Number
1100
1600
2100
2600
3100
3600
4100
4600
0 500 1000 1500 2000 2500 3000 3500 4000 4500
Engine Speed (rpm)
I
S
F
C

(
g
/
k
W
-
h
)
Nozzle 6 holes Nozzle 7 holes Nozzle 8 holes Nozzle 9 holes Nozzle 10 holes

Iig. 25. Lffecl of fueI nozzIe hoIes on ISIC of dieseI engine
Fuel njection 94

IueI-air nixing increases as lhe fueI nozzIe hoIes orifice dianeler fueI nozzIe hoIes
decreases. AIso sool incandescence is olserved lo decrease as lhe anounl of fueI-air
prenixing upslrean of lhe Iifl-off Ienglh increases. This can le a significanl advanlage for
snaII orifice nozzIes hoIe. Hovever, nuIlipIe hoIes orifices dianeler required lo neel lhe
desired nass fIov rale as orifice dianeler decreases. In lhis case, lhe orifices dianeler need
lo pIaced vilh appropriale spacing and direclions in order lo avoid inlerference anong
adjacenl sprays. The enpiricaI correIalions generaIIy predicl snaIIer drop size, sIover
penelraling speed and snaIIer spray cone angIes as lhe orifice dianeler decreases, hovever
lhe predicled vaIues vere differenl for differenl reIalion. AII of lhe nozzIes have exanined
and lhe resuIls are shovn lhal lhe five hoIes nozzIe provided lhe lesl resuIls for indicled
lorque, indicaled pover and ISIC in any differenl engine speed in sinuIalion.

7. ConcIusion
AII of lhe injeclor nozzIe hoIes have exanined and lhe resuIls are shovn lhal lhe seven hoIes
nozzIe have provided lhe lesl lurning resuIl for lhe fueI in-cyIinder lurned in any differenl
engine speeds and lhe lesl lurning is in Iov speed engine. In engine perfornance effecl, aII
of lhe nozzIes have exanined and lhe five hoIes nozzIe provided lhe lesl resuIl in indicled
pover, indicaled lorque and ISIC in any differenl engine speeds.

8. References

aik, Seunghyun. (2OO1). DeveIopnenl of Micro-DieseI Injeclor NozzIes Via MLMS
TechnoIogy and Lffecls on Spray Characlerislics, PnD Disscr|a|icn, Universily of
Wisconsin-Madison, USA.
akar, R.A., Senin., IsnaiI, A.R. and AIi, IsnaiI., 2OO8. ConpulalionaI SinuIalion of IueI
NozzIe MuIli HoIes Ceonelries Lffecl on Direcl Injeclion DieseI Lngine
Ierfornance Using CT-IOWLR. Amcrican ]curna| cf App|ic Scicnccs 5 (2): 11O-116.
aungarler, Carslen. (2OO6). Mix|urc |crma|icn in |n|crna| Ccm|us|icn |ngincs, Spinger
erIin.
Canna TechnoIogies, (2OO4). GT-POI|R Uscrs Manua| 6.1, Canna TechnoIogies Inc.
Canesan, V. (1999). |n|crna| Ccm|us|icn |ngincs 2
n
|i|icn, Tala McCrav-HiII, Nev DeIhi,
India.
Heyvood, }.. (1988). |n|crna| Ccm|us|icn |nginc |unamcn|a|s - Scccn |i|icn, McCrav-
HiII, Singapore.
KovaIevicz, Andrzej., 1984. Ccm|us|icn Sqs|cm cf Hign-Spcc Pis|cn |.C. |ngincs,
Wydavniclva Konunikacji i Lacznosci, Warszava.
Senin and akar, R.A. (2OO7). NozzIe HoIes Lffecl on Unlurned IueI in Injecled and In-
CyIinder IueI of Iour Slroke Direcl Injeclion DieseI Lngine. Prcscarcn ]curna| cf
App|ic Scicnccs 2 (11): 1165-1169.
Senin., akar, R.A. and IsnaiI, A.R. (2OO7). Lffecl Of Lngine Ierfornance Ior Iour-Slroke
DieseI Lngine Using SinuIalion, Prccccing Tnc 5
|n
|n|crna|icna| Ccnfcrcncc On
Numcrica| Ana|qsis in |nginccring, Iadang-Wesl Sunalera, Indonesia.
Slone, Richard. (1997). |n|rcuc|icn |c |n|crna| Ccm|us|icn |ngincs-Scccn |i|icn, SAL Inc,
USA.

1
Accurate Modelling of an Injector
for Common Rail Systems
Claudio Dongiovanni
Politecnico di Torino, Dipartimento di Energetica,
Corso Duca degli Abruzzi 24, 10129, Torino
Italy
Marco Coppo
O.M.T. S.p.A., Via Ferrero 67/A, 10090, Cascine Vica Rivoli
Italy
1. Introduction
It is well known that the injection system plays a leading role in achieving high diesel engine
performance; the introduction of the common rail fuel injection system (Boehner & Kumel,
1997; Schommers et al., 2000; Stumpp & Ricco, 1996) represented a major evolutionary step
that allowed the diesel engine to reach high efciency and low emissions in a wide range of
load conditions.
Many experimental works showthe positive effects of splitting the injection process in several
pilot, main and post injections on the reduction of noise, soot and NOx emission (Badami et al.,
2002; Brusca et al., 2002; Henelin et al., 2002; Park et al., 2004; Schmid et al., 2002). In addition,
the success of engine downsizing (Beatrice et al., 2003) and homogeneous charge combustion
engines (HCCI) (Canakci &Reitz, 2004; Yamane &Shimamoto, 2002) is deeply connected with
the injection system performance and injection strategy.
However, the development of a high performance common rail injection system requires a
considerable investment in terms of time, as well as money, due to the need of ne tuning
the operation of its components and, in particular, of the electronic fuel injector. In this light,
numerical simulation models represent a crucial tool for reducing the amount of experiments
needed to reach the nal product conguration.
Many common-rail injector models are reported in the literature. (Amoia et al., 1997; Bianchi
et al., 2000; Brusca et al., 2002; Catalano et al., 2002; Ficarella et al., 1999; Payri et al., 2004).
One of the older common-rail injector model was presented in (Amoia et al., 1997) and suc-
cessively improved and employed for the analysis of the instability phenomena due to the
control valve behaviour (Ficarella et al., 1999). An important input parameter in this model
was the magnetic attraction force in the control valve dynamic model. This was calculated
interpolating the experimental curve between driving current and magnetic force measured
at xed control valve positions. The discharge coefcient of the feeding and discharge control
volume holes were determined and the authors asserted that the discharge hole operates, with
the exception of short transients, under cavitating ow conditions at every working pressure,
6
Fuel njection 96
but this was not conrmed by (Coppo & Dongiovanni, 2007). Furthermore, the deformation
of the stressed injector mechanical components was not taken into account. In (Bianchi et al.,
2000) the electromagnetic attraction force was evaluated by means of a phenomenological
model. The force was considered directly proportional to the square of the magnetic ux and
the proportionality constant was experimentally determined under stationary conditions. The
elastic deformation of the moving injector components were considered, but the injector body
was treated as a rigid body. The models in (Brusca et al., 2002; Catalano et al., 2002) were
very simple models. The aims in (Catalano et al., 2002) were to prove that pressure drops
in an injection system are mainly caused by dynamic effects rather than friction losses and
to analyse new common-rail injection system congurations in which the wave propagation
phenomenon was used to increase the injection pressure. The model in (Brusca et al., 2002)
was developed in the AMESim environment and its goal was to give the boundary conditions
to a 3D-CFD code for spray simulation. Payri et al. (2004) report a model developed in the
AMESim environment too, and suggest silicone moulds as an interesting tool for characteris-
ing valve and nozzle hole geometry.
A common-rail injector model employs three sub-models (electrical, hydraulic and mechan-
ical) to describe all the phenomena that govern injector operation. Before one can use the
model to estimate the effects of little adjustments or little geometrical modications on the
system performance, it is fundamental to validate the predictions of all the sub-models in the
whole range of possible working conditions.
In the following sections of this chapter every sub-model will be thoroughly presented and it
will be shown how its parameters can be evaluated by means of theoretical or experimental
analysis. The focus will be placed on the electronic injector, as this component is the heart of
any common rail system
2. Mathematical model
The injector considered in this investigation is a standard Bosch UNIJET unit (Fig. 1) of the
common-rail type used in car engines, but the study methodology that will be discussed can
be easily adapted to injectors manufactured by other companies.
The denition of a mathematical model always begins with a thorough analysis of the parts
that make up the component to be modelled. Once geometrical details and functional rela-
tionships between parts are acquired and understood they can be described in terms of math-
ematical relationships. For the injector, this leads to the denition of hydraulic, mechanical,
and electromagnetic models.
2.1 Hydraulic Model
Fig. 2 shows the equivalent hydraulic circuit of the injector, drawn following ISO 1219 stan-
dards. Continuous lines represent the main connecting ducts, while dashed lines represent
pilot and vent connections. The hydraulic parts of the injector that have limited spatial ex-
tension are modelled with ideal components such as uniform pressure chambers and laminar
or turbulent hydraulic resistances, according to a zero-dimensional approach. The internal
hole connecting injector inlet with the nozzle delivery chamber (as well as the pipe connect-
ing the injector to the rail or the rail to the high pressure pump) are modelled according to
a one-dimensional approach because wave propagation phenomena in these parts play an
important role in determining injector performance.
Fig. 3a shows the control valve and the relative equivalent hydraulic circuit. R
A
and R
Z
are the hydraulic resistances used for modelling ow through control-volume orices A (dis-
1. Control valve pin 4. C-shaped connecting pin and anchor
2. Pin guide and upper stop 5. Control volume feeding (Z) hole
3. Control valve anchor 6. Control volume discharge (A) hole
Fig. 1. Standard Bosch UNIJET injector
charge) and Z (feeding), respectively. The variable resistance R
AZ
models the ow between
chambers C
dZ
and C
uA
, taking into account the effect of the control piston position on the
actual ow area between the aforementioned chambers. The solenoid control valve V
c
is rep-
resented using its standard symbol, which shows the forces that act in the opening (one gen-
erated by the current I owing through the solenoid, the other by the pressure in the chamber
C
dA
) and closing direction (spring force).
Fig. 3b illustrates the control piston and nozzle along with the relative equivalent hydraulic
circuit. The needle valve V
n
is represented with all the actions governing the needle motion,
such as pressures acting on different surface areas, force applied by the control piston and
spring force. The chamber C
D
models the nozzle delivery volume, C
S
is the sac volume,
whereas the hydraulic resistance R
hi
represents the i-th nozzle hole through which fuel is
injected in the combustion chamber C
e
. The control piston model considers two different
surface areas on one side, so as to take into account the different contribution of pressure in
the chambers C
uA
and C
dZ
to the total force applied in the needle valve closing direction.
Leakages both between control valve and piston and between needle and its liner are mod-
elled by means of the resistances R
P
and R
n
respectively, and the resulting ow, which is
collected in chamber C
T
(the annular chamber around the control piston), is then returned to
Fuel njection 98
Fig. 2. Injection equivalent hydraulic circuit
tank after passing through a small opening, modelled with the resistance R
T
, between control
valve and injector body.
2.1.1 Zero-dimensional hydraulic model
The continuity and compressibility equation is written for every chamber in the model
Q =
V
E
l
dp
dt
+
dV
dt
(1)
where Q is the net ow-rate coming into the chamber, (V/E
l
)(dp/dt) the rate of increase of
the uid volume in the chamber due to the uid compressibility and (dV/dt) the deformation
rate of the chamber volume.
Fluid leakages occurring between coupled mechanical elements in relative motion (e.g. nee-
dle and its liner, or control piston and control valve body) are modelled using laminar ow
hydraulic resistances, characterized by a owrate proportional to the pressure drop p across
the element
Q = K
L
p (2)
where the theoretical value of K
L
for an annulus shaped cross-section ow area can be ob-
tained by
K
L
=
d
m
g
3
12l
(3)
In case of eccentric annulus shaped cross-section ow area, Eq. 3 gives an underestimation of
the leakage ow rate that can be as low as one third of the real one (White, 1991).
(a) Control valve (b) Needle and control piston
Fig. 3. Injection equivalent hydraulic circuit
Furthermore, the leakage ow rate, Equations 2 and 3, depends on the third power of the
radial gap g. At high pressure the material deformation strongly affects the gap entity and
its value is not constant along the gap length l because pressure decreases in the gap when
approaching the low pressure side (Ganser, 2000). In order to take into account these effects
on the leakage owrate, the value of K
L
has to be experimentally evaluated in the real injector
working conditions.
Turbulent ow is assumed to occur in control volume feeding and discharge holes, in nozzle
holes and in the needle-seat opening passage. As a result, according to Bernoullis law, the
ow rate through these orices is proportional to the square root of the pressure drop, p,
across the orice, namely,
Q = A
2p
(4)
The ow model through these orices plays a fundamental role in the simulation of the injec-
tor behavior in its whole operation eld, so the evaluation of the factor is extremely impor-
tant.
2.1.2 Hole A and Z discharge coefcient
The discharge coefcient of control volume orices A and Z is evaluated according to the
model proposed in (Von Kuensberg Sarre et al., 1999). This considers four ow regimes inside
the hole: laminar, turbulent, reattaching and fully cavitating.
Neglecting cavitation occurrence, a preliminary estimation of the hole discharge coefcient
can be obtained as follows
1
=
_
K
I
+ f
l
d
+1 (5)
where K
I
is the inlet loss coefcient, which is a function of the hole inlet geometry (Munson
et al., 1990), l is the hole axial length, d is the hole diameter, and f is the wall friction coefcient,
evaluated as
f = MAX
_
64
Re
, 0.316 Re
0.25
_
(6)
Fuel njection 100
where Re stands for the Reynolds number.
The ratio between the cross section area of the vena contracta and the geometrical hole area,
vc
, can be evaluated with the relation:
1
2
vc
=
1
2
vc
0
11.4
r
d
(7)
where
vc
0
= 0.61 (Munson et al., 1990) and r is the llet radius of the hole inlet.
It follows that the pressure in the vena contracta can be estimated as
p
vc
= p
u

l
2
_
Q
A
vc
_
2
(8)
If the pressure in the vena contracta (p
vc
) is higher then the oil vapor pressure (p
v
), cavita-
tion does not occur and the value of the hole discharge coefcient is given by Equation 5.
Otherwise, cavitation occurs and the discharge coefcient is evaluated according to
=
vc
p
u
p
v
p
u
p
d
(9)
The geometrical prole of the hole inlet plays a crucial role in determining, or avoiding, the
onset of cavitation in the ow. In turn, the occurrence of cavitation strongly affects the ow
rate through the orice, as can be seen in Figure 4, which shows two trends of predicted ow
rate (Q/Q
0
) in function of pressure drop (p = p
u
p
d
) through holes with the same diameter
and length, but characterized by two different values of the r/d ratio (0.2 and 0.02), when p
u
is kept constant and p
d
is progressively decreased. In absence of cavitation, (r/d = 0.2), the
relation between ow rate and pressure drop is monotonic while, if cavitation occurs (r/d =
0.02), the hole experiences a decrease in ow rate as pressure drop is further increased. This
behavior agrees with experimental data reported in the literature (Lefebvre, 1989).
Fig. 4. Predicted ow through an orice in presence/absence of cavitation
Obviously, such behavior would reect strongly on the injector performance if the control vol-
ume holes happened to cavitate in some working conditions. Therefore, in order to accurately
model the injector operation, it is necessary to accurately measure the geometrical prole of
the control volume holes A and Z; by means of silicone moulds, as proposed by (Payri et al.,
2004), it is possible to acquire an image of the hole shape details, as shown in Figure 5.
(a) A hole (b) Z hole
Fig. 5. Moulds of the control valve holes
By means of imaging techniques it is possible to measure the r/d ratio of the hole under
investigation. Table 1 reports the results obtained for the injector under investigation. The
value of K
I
, in Equation 5, is a function of r/d only (Von Kuensberg Sarre et al., 1999) and,
hence, easily obtainable.
Knowing that during production a hydro-erosion process is applied to make sure that, under
steady ow conditions, all the holes yield the same ow rate, it is possible to dene an itera-
tive procedure to calculate the hole diameter using the discharge coefcient model presented
above and the the steady ow rate value. This approach is preferrable to the estimation of the
hole diameter with imaging techniques because it yields a result that is consistent with the
discharge coefcient model used.
r/d K
I
d [m]
Hole A 0.235% 0.033 2802%
Hole Z 0.225% 0.034 2492%
Table 1. Characteristics of control volume holes
In the control valve used in our experiments, under a pressure drop of 10 MPa, with a back
pressure of 4 MPa, the holes A and Z yielded 6.5 0.2 cm
3
/s and 5.3 0.2 cm
3
/s, respectively.
With these values it is possible to calculate the most probable diameter of the control volume
holes, as reported in Table 1. It is worth noting that the precision with which the diameters
were evaluated was higher than that of the optical technique used for evaluating the shape of
the control volume holes. This resulted from the fact that K
I
shows little dependence on r/d
when the latter assumes values as high as those measured. As a consequence, the experimen-
tal uncertainty in the diameter estimation is mainly originated from the uncertainty given on
the stationary ow rate through the orices.
Fuel njection 102
2.1.3 Discharge coefcient of the nozzle holes
The model of the discharge coefcient of the nozzle holes is designed on the base of the un-
steady coefcients reported in (Catania et al., 1994; 1997). These coefcients were experimen-
tally evaluated for minisac and VCO nozzles in the real working conditions of a distributor
pump-valve-pipe-injector type injection system. The pattern of this coefcient versus needle
lift evidences three different phases. In the rst phase, during injector opening, the moving
needle tip strongly inuences the efux through the nozzle holes. In this phase, the discharge
coefcient progressively increases with the needle lift. In the second phase, when the needle
is at its maximum stroke, the discharge coefcient increases in time, independently from the
pressure level at the injector inlet. In the last phase, during the needle closing stroke, the dis-
charge coefcient remains almost constant. These three phases above mentioned describe a
hysteresis-like phenomenon. In order to build a model suitable for a common rail injector in
its whole operation eld these three phases need to be considered.
Therefore, the nozzle hole discharge coefcient is modeled as needle lift dependent by con-
sidering two limit curves: a lower limit trend (
d
h
), which models the discharge coefcient in
transient efux conditions, and an upper limit trend (
s
h
), which represents the steady-state
value of the discharge coefcient for a given needle lift. The evolution from transient to sta-
tionary values is modeled with a rst order system dynamics.
It was experimentally observed (Catania et al., 1994; 1997) that the transient trend presents a
rst region in which the discharge coefcient increases rapidly with needle lift, following a
sinusoidal-like pattern, and a second region, characterized by a linear dependence between
discharge coefcient and needle lift. Thus, the following model is adopted:
d
h
() =
_

d
h
(
0
) sin(

2
0
) 0 <
0
d
h
(
M
)
d
h
(
0
)
0
(
0
) +
d
h
(
0
)
0
(10)
where is the needle-seat relative displacement, and
0
is the transition value of between
the sinusoidal and the linear trend.
The use of the variable , rather than the needle lift, x
n
, emphasizes the fact that all the me-
chanical elements subject to fuel pressure, including nozzle and needle, deform, thus the real
variable controlling the discharge coefcient is not the position of the needle, but rather the
effective clearance between the latter and the nozzle.
The maximum needle lift,
M
, varies with rail pressure due to the different level of deforma-
tion that this parameter induces on the mechanical components of the injector. The relation
between
M
and the reference rail pressure p
r0
is assumed to be linear as
M
= K
1
p
r0
+ K
2
(11)
where K
1
and K
2
are constants that are evaluated as explained in the section 2.3.3.
Similarly, the value of
0
in Equation 10 is modeled as a function of the operating pressure p
r0
in order to better match the experimental behavior of the injection system. Thus, the following
t is used
0
= K
3
p
r0
+ K
4
(12)
and K
3
and K
4
are obtained at the end of the model tuning phase (table 4).
In order to dene the relation between the steady state value of the nozzle-hole discharge
coefcient (
s
h
) and the needle-seat relative displacement () the device in Figure 6 was de-
signed. It contains a camshaft that can impose to the needle a continuously variable lift up to
1 mm. Then, a modied injector equipped with this device was connected to the common rail
injection system and installed in a Bosch measuring tube, in order to control the nozzle hole
downstream pressure. The steady ow rate was measured by means of a set of graduated
burettes.
1. Dial indicator 4. Eccentric ball bearing (e = 1mm)
2. Handing for varying needle lift 5. Injector control piston
3. Axis support bearing 6. Injector inlet
Fig. 6. Device for xed needle-seat displacement imposition
Figure 7a shows the trends of steady-state ow rate versus needle lift at rail pressures of 10
and 20 MPa, while the back pressure in the Bosch measuring tube was kept to either ambient
pressure or 4 MPa; whereas Figure 7b shows the resulting stationary hole discharge coefcient,
evaluated for the nozzle under investigation.
Taking advantage of the reduced variation of
s
h
with operation pressure, it is possible to
use the measured values to extrapolate the trends of steady-state discharge coefcient for
higher pressures, thus dening the upper boundary of variation of the nozzle hole discharge
coefcient values.
During the injector opening phase the unsteady effects are predominant and the sinusoidal-
linear trend of the hole discharge coefcient, Equation 10, was considered; when the needle-
seat relative displacement approaches its relative maximum value
r
M
, the discharge coef-
cient increases in time, which means that the efux through the nozzle holes is moving to
the stationary conditions. In order to describe this behavior, a transition phase between the
Fuel njection 104
(a) Steady ow rate (b) Stationary discharge coefcient
Fig. 7. Stationary efux through the nozzle
unsteady and the stationary values of the hole discharge coefcient at this needle lift was
considered. This phase was modeled as a temporal exponential curve, namely,
h
=
d
h
(
r
M
) + [
s
h
(
r
M
)
d
h
(
r
M
)] [1 exp (
t t
0
)] (13)
where t
0
is the instant in time at which the needle-seat relative displacement approaches its
maximum value
r
M
,
d
h
_
r
M
_
and
s
h
_
r
M
_
are the unsteady and the stationary hole discharge
coefcients evaluated at this needle-seat relative displacement, and is the time constant of
this phenomenon, which have to be dened during the model tuning phase.
Figure 8 shows the computed nozzle hole discharge coefcient,
h
, dependence upon needle-
seat relative displacement, , in accordance to the proposed model, in a wide range of op-
erating conditions (which are showed by rail pressure p
r0
and energisation time ET
0
in the
legend).
Examining the discharge coefcient,
h
, trends for the three main injections (ET
0
= 780 s, 700
s and 670 s) during the opening phase, it is interesting to note that for a given value of the
needle lift, lower discharge coefcients are to be expected at higher operating pressures. This
can be explained considering that the ow takes longer to develop if the pressure differential,
and thus the steady state velocity to reach is higher.
The main injection trends also show the transition from the sinusoidal to the linear depen-
dence of the transient discharge coefcient on needle lift.
The phase in which the needle has reached the maximum value and the discharge coefcient
increases in time from unsteady to stationary values is not very evident in main injections,
because the former increases enough during the opening phase to approach the latter. This
happens because the needle reaches sufciently high lifts as to have reduced effect on the ow
in the nozzle holes, and the longer injection allows time for complete ow development.
Conversely, during pilot injections (ET
0
=300 s), the needle reaches lower maximum lifts,
hence lower values of the unsteady discharge coefcient, so that the phase of transition to
the stationary value lasts longer. The beginning of this transition can be easily identied by
analyzing the curves marked with dots and crosses in Figure 8. The point at which they
depart from their main injection counterpart (same line style but without markers) marks the
beginning of the exponential evolution in time to stationary value of discharge coefcient.
For both pilot and main injections, the nozzle hole discharge coefcient remains constant, and
equal to the stationary value, during the injector closing phase, as shown by the horizontal
prole of the trends in Figure 8.
The needle-seat discharge coefcient
s
has to be modeled too. It is assumed as needle lift
dependent according to (Xu et al., 1992) where this coefcient was experimentally evaluated
after removing the nozzle tip. A three segment trend is considered, as shown in Fig. 8, but it
is worth to point out that it plays a marginal role in the injection system simulation because
its values are higher than 0.8 for most needle lift values.
Fig. 8. Needle-seat and holes discharge coefcient
2.1.4 One-dimensional model: pipe ow model
A one-dimensional modelling approach is followed in order to model the uid ow in the
pipe connecting injector and rail and in the injector internal duct that carries the uid from
the inlet to the delivery chamber. This is necessary to correctly take into account pressure
wave propagation that occurs in those elements. The pipe ow conservation equations are
written for a single-phase uid because in the common-rail injection system cavitation does
not appear in the connecting pipe. An isothermal ow is assumed and only the momentum
and mass conservation equations need to be solved
w
t
+ A
w
x
= b (14)
where w =
_
u
p
_
, A =
_
u 1/
c
2
u
_
, b =
_
4/d
0
_
and is the wall shear stress that is evaluated under the assumption of steady-state friction
(Streeter et al., 1998).
The eigenvalues of the hyperbolic system of partial differential Equations 14 are = u c,
real and distinct. The celerity c of the wave propagation can be evaluated as
c =
_
c
l
_
1 + K
p
E
l
E
p
d
p
t
p
_ (15)
Fuel njection 106
where the second term within brackets takes into account the effect of the pipe elasticity; K
p
is the pipe constraint factor, depending on pipe support layout, E
p
the Youngs modulus of
elasticity of the pipe material, d
p
the pipe diameter and t
p
the pipe wall thickness (Streeter
et al., 1998). Being the pipe ends rigidly constrained, the pipe constrain factor K
p
can be
evaluated as
K
p
= 1
2
p
(16)
where
p
is the Poissons modulus of the pipe material.
Pipe junctions are treated as minor losses and only the continuity equation is locally written.
As mentioned before, this simple pipe ow model is not suitable when cavitation occurs.
This is not a limitation when common-rail injection system are modelled because of the high
pressure level at which these systems always work. In order to model conventional injection
systems, as pump-pipe-nozzle systems, it is necessary to employ a pipe ow model able to
simulate the cavitation occurrence. For this purpose the authors developed an appropriate
second order model (Dongiovanni et al., 2003).
2.1.5 Fluid properties
Thermodynamic properties of oil are affected by temperature and pressure that remarkably
vary in the common rail injection system operation eld. Density, wave propagation speed
and kinematic viscosity of the ISO4113 air-free test oil had been evaluated as function of pres-
sure and temperature (Dongiovanni, 1997). These oil properties were approximated with an-
alytic functions of the exponential type in the range of pressures from 0.1 to 200 MPa and
temperatures from 10

C to 120

C. These analytic relations were derived from the actual
property values supplied by the oil maker, by using the least-square method for non-linear
approximation functions with two independent variables. The adopted formulae are:
l
(p, T) = K
1
+
_
1 exp
_
p
K
2
__
K
3
p
K
4
(17)
E
l
(p, T) = K
E1
+
_
1 exp
_
p
K
E2
__
K
E3
p
K
E4
(18)
l
(p, T) = K
1
+ K
2
p
K
3
(19)
The K
Ei
, K
i
and K
i
are polynomial functions of temperature T
K
i
=
l
i
j=0
K
i,j
T
j
i = 1, 2, 3, 4 (20)
and the numerical coefcients that appear in them are reported in Table 2 according with
SI units: pressure [p] = bar, temperature [T] =
C, density [
l
] = kg/m
3
, bulk modulus
[E
l
] = MPa and kinematic viscosity [
l
] = mm
2
/s
Finally, the celerity of the air free oil is evaluate in accordance with c
l
=
_
E
l
/
l
.
By using these approximation functions, the maximum deviation between experimental and
analytical values in the examined range of pressure and temperature has been estimated as
being lower than 0.2% for density, 1.2% for bulk modulus, 0.6% for celerity and 18%
for kinematic viscosity.
K
j= 0 j= 1 j= 2
K
1,j
8.3636e2 -6.7753e-1 -
K
2,j
1.5063e2 -2.4202e-1 -
K
3,j
1.7784e-1 1.4640e-3 1.5402e-5
K
4,j
7.8109e-1 -8.1893e-4 -
K
E
j= 0 j= 1 j= 2
K
E1,j
1.7356e3 -1.0908e1 2.2976e-2
K
E2,j
7.5540e1 - -
K
E3,j
1.5050 -3.7603e-3 -
K
E4,j
9.4448e-1 3.9441e-4 -
K
j=0 j=1 j=2 j=3

K
1,j
6.4862 -1.5847e-1 1.6342e-3 -6.0334e-6
K
2,j
4.0435e-4 -2.3118e-6 - -
K
3,j
1.4346 -6.2288e-3 3.3500e-5 -
Table 2. Polynomial coefcients for ISO4113 oil
2.2 Electromagnetic model
A model of the electromechanical actuator that drives the control valve must be realized in
order to work out the net mechanical force applied by the solenoid on its armature, for a given
current owing in the solenoid. The magnetic force applied by the solenoid on the armature
F
Ea
can be obtained by applying the principle of energy conservation to the armature-coil
system (Chai, 1998; Nasar, 1995). In the general form it can be written as follows:
V I dt = F
Ea
dx
a
+ dW
m
(21)
where V I dt represents the electric energy input to the system, F
Ea
dx
a
is the mechanical work
done on the armature and dW
m
is the change in the magnetic energy.
From Faradays law, voltage V may be expressed in terms of ux linkage (N
d
dt
) and Equation
21 becomes
N I d = F
Ea
dx
a
+ dW
m
(22)
as shown in (Chai, 1998; Nasar, 1995); by considering and x
a
as independent variables,
Equation 22 can be reduced to
F
Ea
=
W
m
x
a
(23)
The magnetic circuit geometry of the control valve needs to be thoroughly analyzed in order
to evaluate the magnetic energy stored in the gap. Fig. 9a shows the path of the signicant
magnetic uxes, having neglected secondary leakage uxes and ux fringing.
Exploiting the analogy between Ohms and Hopkinsons law, it is possible to obtain the mag-
netic equivalent circuit of Fig. 9b where NI is the ampere-turns in the exciting coil and
j
(j = 1, .., 5) are the magnetic reluctances. When the magnetic ux ows across a cross-
section area A
a
constant along the path length l, the value of the j-th reluctance can be ob-
tained by:
Fuel njection 108
(a) Magnetic path (b) Magnetic equivalent circuit
Fig. 9. Magnetic model sketch
j
=
l
j
0
A
aj
(j = 1, 2) (24)
When the ux ows across a radial path, the reluctance can be evaluated as
j
=
1
2
0
t
j
ln
_
d
e
d
i
_
j
(j = 3, 4, 5) (25)
being t the radial thickness, d
e
and d
i
the external and internal diameter of the gap volume.
Reluctance of the ferromagnetic components was neglected because it is several order of mag-
nitude lower than the corresponding gap reluctance.
Circuit of Fig. 9b is solved using Thevenins theorem, and the equivalent circuit reluctance
connected to the magnetomotive force generator is determined as
=
1
+

2
4
+
2
5
(
3
+
4
)
4
+ (
2
+
5
) (
3
+
4
)
(26)
The magnetic energy W
m
is stored in the volume of the electromechanical actuator, but only
the portion of energy stored in the gap between control-valve body and magnetic core de-
pends on the armature lift x
a
. Consequently, being the magnetization curve of non-ferromagnetic
materials (oil in the gaps) linear, Equation 23 can be written as
F
Ea
=
1
2
2
d
dx
a
=
1
2
_
NI
_
2
d
dx
a
(27)
To complete the model, it was necessary to take into account the saturation phenomenon that
occurs to every ferromagnetic material. That is, a magnetic ux cannot increase indenitely, as
the material presents a maximum magnetic ux density after which the curve B H is almost
at. In this model we assumed a simplied magnetization curve, given by :
B =
_
H H < H
+
0
(H H
) H H
(28)
thus neglecting material hysteresis and non-linearity.
As a result of the saturation phenomenon, the maximum force of attraction is limited because
the maximum magnetic ux which can be obtained in the j-th branch of the circuit is approx-
imately
Mj
H
j
A
j
(29)
being
0
negligible with respect .
The most important parameters in the electromagnetic model are set as reported in Table 3.
N B
= H
[T] t
3
[mm] t
4
[mm] t
5
[mm]
32 2.5 0.65 1.5 0.05
Table 3. Most important electromagnetic model parameters
The model was employed to evaluate the inductance of the solenoid when mounted on the
injector body. In this case, with the valve actuator in the closed position, an inductance of 134
H was evaluated. Employing a sinusoidal wave generator at a frequency of 5 kHz, which
is high enough to make negligible the mechanical system movements, an inductance of 137
H was measured. The accordance between experimental and theoretical inductance value
indirectly validates the electromagnetic model and the parameters value.
Fig. 10a shows the theoretical (solid line) driving actuator force when the actual energizing
current (dashed line) is used to feeding the injector solenoid.
(a) Magnetic force and feeding current (b) Inductance and armature lift
Fig. 10. Magnetic model results
Furthermore, we point out that the measure of the injector coil inductance L = N/
2
could
be used to indirectly evaluate the control valve lift, due to the dependence of reluctance upon
armature distance from the solenoid (Equation 24 ).
Bearing in mind that, by applying Ohms law to the solenoid coil, the inductance L could be
evaluated as:
L =
_
(V RI) dt
I
(30)
hence only the measurement of solenoid current I and voltage V would be required to calcu-
late L.
Fuel njection 110
Fig. 10b draws the theoretical inductance L, which was calculated according to Equation 24
and opportunely scaled, compared to the experimental valve lift x
c
, showing a good agree-
ment between the two trends, and hence the potential of this non-invasive measurement tech-
nique. However, Equation 30 is only applicable when electric current is owing in the solenoid
coil so, for example, it is not possible to use this method to record the the control valve closure
trend because, as Fig. 10b shows, this usually begins when the solenoid current is null.
A possible way to solve this problem would be to inject an additional, small amplitude, high
frequency (around 1 MHz) current into the coil, but this technique has not yet been tested by
the authors.
2.3 Mechanical model
All mechanical devices that can move during injector functioning (i.e. needle, control piston
and control valve) are modelled using the conventional mass-spring-damper scheme, gov-
erned by a mechanical equilibrium equation, in which the dynamic parameters are function
of element position.
m
j
d
2
x
j
dt
2
+
j
dx
j
dt
+ k
j
x
j
+ F
0j
= F
j
(31)
where m
j
is the displacing mass,
j
the damping coefcient, k
j
the spring stiffness and F
0
j
the spring preload; the bar above the symbols indicates that these coefcients are evaluated
according to the relative position of the moving elements.
2.3.1 Control piston, needle and nozzle model
The high working pressures in the common-rail injection system stress its components and
cause appreciable deformation of them. In order to take into account the effects of the axial
deformation of nozzle and injector body, the nozzle is modelled by means of a conventional
mass-spring-damper scheme as well as the needle and the control piston, while the injector
body is modelled by means of a simple spring having appropriate stiffness.
Injector needle, control piston and nozzle forma three degrees of freedommechanical system,
which can be modelled as shown in Figure 11a. Three equilibrium equations are needed to
describe the system motion, one for each element. With reference to Fig. 11a, the dynamic
equilibriumEquation 31 is written using the following denition for the control piston (j = P),
the needle (j = n) and the nozzle (j = N):
external force F
j
:
F
P
= F
c
+ p
T
S
P
+ F
R(Pn)
F
R(Pb)
F
n
= p
T
S
n
F
R(Pn)
+ F
S
+ F
R(nN)
F
N
= F
S
+ p
e
S
n
F
R(nN)
(32)
where F
R(ij)
is the force that i-th and j-th element apply to each other when they are in contact,
F
c
= p
uA
S
P
+ p
dZ
(S
P
S
P
) (33)
and
F
S
= p
S
S
S
+ p
D
S
D
+ [ p
S
+ (1 ) p
D
)] (S
n
S
D
S
S
) (34)
where = 0 indicates that the nozzle is closed while = 1 indicates open nozzle conditions.
Damping coefcient
j
, stiffness k
j
and preload F
0j
are evaluated as follows for:
control piston
(a) Control piston and needle (b) Control valve
Fig. 11. Dynamic models
x
P
< X
MP
l
P

P
=
P
k
P
= 0 F
0P
= 0
X
MP
l
P
x
P

P
=
b
+
P
k
P
= k
b
F
0P
= k
b
(X
MP
l
P
)
(35)
needle
x
n
x
N
< 0
n
=
b
+
n
k
n
= k
b
+ k
n
F
0n
= F
0n
0 x
n
x
N
< X
Mn
l
n

n
=
n
k
n
= k
n
F
0n
= F
0n
X
Mn
l
n
x
n
x
N

n
=
b
+
n
k
n
= k
b
+ k
n
F
0n
= F
0n
k
b
X
Mn
(36)
nozzle
x
n
x
N
< 0
N
=
b
+
N
k
N
= k
b
+ k
N
F
0N
= 0
0 x
n
x
N

N
=
N
k
N
= k
N
F
0N
= 0
(37)
2.3.2 Control valve model
The control valve contains two mobile parts: the pin element having mass m
c
and the armature
element of mass m
a
; they can be modelled with the two degrees of freedom scheme shown in
Fig. 11b. The two dynamic equilibrium equations are written in a similar fashion as Equation
31 where j = a indicates the armature and j = c the control-pin. The external forces F
j
can be
evaluated as
F
a
= F
Ea
F
R(ca)
F
c
= (p
dA
p
T
)S
c
+ F
R(ca)
+ F
R(cb)
(38)
where F
Ea
represents the electromagnetic action that the current generates when it ows in
the solenoid coil determined as shown in section 2.2.
Fuel njection 112
Damping coefcient
j
, stiffness k
j
and preload F
0j
are evaluated as follows:
pin element
x
c
< 0
c
=
b
+
c
k
c
= k
b
+ k
c
F
0c
= F
0c
0 x
c
< X
Mc
l
c

c
=
c
k
c
= k
c
F
0c
= F
0c
X
Mc
l
c
x
c

c
=
b
+
c
k
c
= k
b
+ k
c
F
0c
= F
0c
k
b
(X
Mc
l
c
)
(39)
armature
l
Mc
X
Mc
+ x
c
x
a

a
=
a
k
a
= k
a
F
0a
= F
0a
x
a
> l
Mc
X
Mc
+ x
c

a
=
b
+
a
k
a
= k
b
+ k
a
F
0a
= F
0a
k
b
(l
Mc
X
Mc
+ x
c
)
(40)
2.3.3 Mechanical components deformation
The axial deformation of needle, nozzle and control piston have to be taken into account.
These elements are considered only axially stressed, while the effects of the radial stress are
neglected. For the sake of simplicity, the axial length of control piston (l
P
), needle (l
n
), and
nozzle (l
N
) can be evaluated as function of the axial compressive load (F
C
) in each element.
Therefore, the deformed length l of these elements, which are considered formed by m parts
having cross section A
j
and initial length l
0
j
, is evaluated as follows
l =
m
j
l
0
j
_
1
F
C
j
EA
j
_
(41)
where E is Youngs modulus of the considered material.
The axial deformation of the injector body is taken into account by introducing in the model
the elastic elements indicated as k
B
and k
Bc
in Figure 11.
The injector body deformation cannot be theoretically calculated very easily, because one
should need to take into account the effect and the deformation of the constraints that x
the injector on the test rig. For this reason, in order to evaluate the elasticity coefcient of k
B
and k
Bc
, an empirical approach is followed, which consists in obtaining a relation between
the axial length of these elements and the uid pressure inside the injector body. As direct
consequence, the maximum stroke of the needle-control piston (
M
) and of the control-valve
(X
Mc
) can be expressed as a function of the injector structural stress.
(a) Needle (b) Control valve
Fig. 12. Effect of pressure on the maximum moving element lift
Figure 12 reports the actual maximum needle-control piston lift (circular symbols) as a func-
tion of rail pressure. At the rail pressure of 30 MPa the maximumneedle-control piston lift was
not reached, so no value is reported at this rail pressure. The continuous line represents the
least-square t interpolating the experimental data and the dashed line shows the maximum
needle-control piston lift calculated by considering only nozzle, needle and control-piston ax-
ial deformation. The difference between the two lines represents the effect of the injector body
deformation on the maximum needle-control piston lift. This can be expressed as a function
of rail pressure and, for the considered injector, can be estimated in 0.41 m/MPa. By means
of the linear t (continuous line) reported in Figure 12 it is possible to evaluate the parameters
K
1
= 1.59 m/MPa and K
2
= 364 m that appear in Eq. 11.
In order to evaluate the elasticity coefcient k
Bc
, an analogous procedure can be followed
by analyzing the maximum control-valve lift dependence upon fuel pressure, as shown in
Figure 12. It was found that the effect of injector body deformation was that of reducing the
maximum control valve stroke of 0.06 m/MPa.
(a) p
r0
=140 MPa, ET
0
= 1230 s (b) p
r0
=80 MPa, ET
0
= 1230 s
Fig. 13. Deformation effects on needle lift
The relevance of the deformation effects on the injector predicted performances is shown in
Fig. 13. The left graph shows the control piston lift at a rail pressure of 140 MPa generated with
an energizing time ET
0
of 1230s, while the right graph shows the same trend at a rail pressure
of 80 MPa, and generated with the same value of ET
0
. The experimental results are drawn by
circular symbols, while lines refer to theoretical results. The dashed lines (Model a) show the
theoretical control piston lift evaluated by only taking in to account the axial deformation of
the moving elements and nozzle, while the continuous lines (Model b) show the theoretical
results evaluated by taking into account the injector body deformation too. The difference
between the two models is signicant, and so is the underestimation of the volume of uid
injected per stroke (4.3% with p
r0
=140MPa and ET
0
of 1230s, 3.6% with p
r0
=80MPa, ET
0
of
1230s). This highlights the necessity of accounting for deformation of the entire injector body,
if accurate predictions are sought.
Indeed, the maximum needle lift evaluation plays an important role in the simulation of the
injector behaviour in its whole operation eld because it inuences both the calculation of the
injected ow rate (as the discharge coefcients of needle-seat and nozzle holes depend also
on needle lift) and of the injector closing time, thus strongly affecting the predicted volume of
fuel injected per cycle.
The deformation of the injector body also affects the maximum control valve stroke, and a
similar analysis can be performed to evaluate its effects on injector performance. Our study
showed that this parameter does not play as important a role as the maximum needle stroke,
because the effective ow area of the A hole is smaller than the one generated by the displace-
ment of the control valve pin, and thus it is the A hole that controls the efux from the control
volume to the tank.
Fuel njection 114
2.3.4 Masses, spring stiffness and damping factors
Components mass and springs stiffness k
j
can be easily estimated. Whenever a spring is in
contact to a moving element, the moving mass m
j
value used in the model is the sum of the
element mass and a third of the spring mass. In this way it is possible to correctly account for
the effect of spring inertia too.
The evaluation of the damping factors
j
in Equation 31 is considerably more difcult. Con-
sidering the element moving in its liner, like needle and control piston, the damping factor
takes into account the damping effects due to the oil that moves in the clearance and the fric-
tion between moving element and liner. The oil ow effect can be modelled as a combined
Couette-Poiseuille ow (White, 1991) and the wall shear stress on the moving element surface
can be theoretically evaluated. Experimental evidences show that friction effects are more rel-
evant than the uid-dynamics effects previously mentioned. Unfortunately, these can not be
theoretically evaluated because their intensity is linked to manufacturing tolerances (both ge-
ometrical and dimensional). Therefore, damping factors must be estimated during the model
tuning phase.
(a) Main injection: ET
0
=780s, p
r0
=135 MPa (b) Pilot injection: ET
0
=300s, p
r0
=80 MPa
Fig. 14. Comparison between numerical and theoretical results
3. Model tuning and results
Any mathematical model requires to be validated by comparing its results with the experi-
mental ones. During the validation phase some model parameters, which cannot be experi-
mentally or theoretically evaluated, have to be carefully adjusted.
The model here presented was tested comparing numerical and experimental control valve
lift x
c
, control piston lift x
P
, injected ow rate Q and injector inlet pressure p
in
in several
operating conditions. Figure 15 shows two of these validation tests and the good accordance
between experimental and numerical results is evident.
Table 4 shows the value of the parameters that were adjusted during the tuning phase. These
values can be used as starting points for the development of new injector models, but their
exact value will have to be dened during model tuning for the reasons explained above.
After the tuning phase the model can be used to reproduce the injection system performance
in its whole operation eld. By way of example, Fig. 15 shows the experimental and numerical
volume injected per stroke V
f
and the percentage error of the numerical estimation.
(a) Injected uid volume per stroke (b) Model error
Fig. 15. Model validation
Eq. 10 Eq. 12 Eq. 13 Eq. 31
d
h
(
0
)
d
h
(
M
) K
3
K
4

n

N

P

c

a
0.75 0.85 0.28 m/MPa 63 m 25 s 6.1 6310 6.5 28 5.1 [kg/s]
Table 4. Tuning dened parameters
Fuel njection 116
4. Nomenclature
Symbol Denition Unit
A Geometrical area m
2
C Uniform pressure chamber
c Wave propagation speed m/s
d Hole || Pipe diameter m
e Eccentricity m
E Youngs modulus Pa
ET Injector solenoid energisation time s
F Force N
f Friction factor
I Electric current A
K Coefcient
k Spring stiffness N/m
l Length m
m Mass kg
N Number of coil turns
p Pressure Pa
Q Flow rate m
3
/s
r Rail || Fillet radius m
R Hydraulic resistance
Re Reynolds number
S Surface area m
2
t Time s
u Average cross-sectional velocity of the uid m/s
V Valve || Volume m
3
W Energy J
X Distance m
x Displacement || Axial coordinate m
Damping factor kg/s
switch (0=nozzle closed,1=nozzle open)
Increment || Drop
Magnetic ux Wb
Needle-seat relative displacement m
Contraction || Discharge coefcient
Density kg/m
3
Wall shear stress || Time constant Pa || s
Reluctance H
1
Subscript Denition
A Control-volume discharge hole
a Armature
B Injector body
b Seat
C Compression
c Control valve
D Delivery
Symbol Denition Unit
d Downstream
E Electromechanical
e Injection environment External
f Fuel
h Hole
l Inlet loss Liquid phase
in Injector inlet
M Maximum value
m Magnetic
N Nozzle
n Needle
P Piston
R Reaction Force
r Rail
S Sac
s Needleseat
T Tank
u Upstream
v Vapour
vc Vena contracta
Z Control-volume feeding hole
0 Reference value
Superscripts Denition
d Dynamic
r Relative
s Steady-state
5. References
Amoia, V., Ficarella, A., Laforgia, D., De Matthaeis, S. & Genco, C. (1997). A theoretical code
to simulate the behavior of an electro-injector for diesel engines and parametric anal-
ysis, SAE Transactions 970349.
Badami, M., Mallamo, F., Millo, F. & Rossi, E. E. (2002). Inuence of multiple injection strate-
gies on emissions, combustion noise and bsfc of a di common rail diesel engines, SAE
paper 2002-01-0503.
Beatrice, C., Belardini, P., Bertoli, C., Del Giacomo, N. & Migliaccio, M. (2003). Downsizing
of common rail d.i. engines: Inuence of different injection strategies on combustion
evolution, SAE paper 2003-01-1784.
Bianchi, G. M., Pelloni, P. & Corcione, E. (2000). Numerical analysis of passenger car hsdi
diesel engines with the 2nd generation of common rail injection systems: The effect
of multiple injections on emissions, SAE paper 2001-01-1068.
Boehner, W. &Kumel, K. (1997). Common rail injection systemfor commercial diesel vehicles,
SAE Transactions 970345.
Brusca, S., Giuffrida, A., Lanzafame, R. &Corcione, G. E. (2002). Theoretical and experimental
analysis of diesel sprays behavior frommultiple injections common rail systems, SAE
paper 2002-01-2777.
Fuel njection 118
Canakci, M. & Reitz, R. D. (2004). Effect of optimization criteria on direct-injection homo-
geneous charge compression ignition gasoline engine performance and emissions
using fully automated experiments and microgenetic algorithms, J. of Engineering for
Gas Turbines and Power 126: 167177.
Catalano, L. A., Tondolo, V. A. & Dadone, A. (2002). Dynamic rise of pressure in the common-
rail fuel injection system, SAE paper 2002-01-0210.
Catania, A., Dongiovanni, C., Mittica, A., Badami, M. &Lovisolo, F. (1994). Numerical analysis
vs. experimental investigation of a distribution type diesel fuel injection system, J. of
Engineering for Gas Turbines and Power 116: 814830.
Catania, A. E., Dongiovanni, C., Mittica, A., Negri, C. & Spessa, E. (1997). Experimental eval-
uation of injector-nozzle-hole unsteady ow-coefcients in light duty diesel injection
systems, Proceedings of the Ninth Internal Pacic Conference on Automotive Engineering,
Bali, Indonesia.
Chai, H. (1998). Electromechanical Motion Devices, Pearson Professional Education.
Coppo, M. & Dongiovanni, C. (2007). Experimental validation of a common-rail injec-
tor model in the whole operation eld, J. of Engineering for Gas Turbines and Power
129(2): 596608.
Dongiovanni, C. (1997). Inuence of oil thermodynamic properties on the simulation of a
high pressure injection system by means of a rened second order accurate implicit
algorithm, ATA Automotive Engineering pp. 530541.
Dongiovanni, C., Negri, C. & Roberto, R. (2003). A uid model for simulation of diesel in-
jection systems in cavitating and non-cavitating conditions, Proceedings of the ASME
ICED Spring Technical Conference, Salzburg, Austria.
Ficarella, A., Laforgia, D. & Landriscina, V. (1999). Evaluation of instability phenomena in a
common rail injection system for high speed diesel engines, SAE paper 1999-01-0192.
Ganser, M. A. (2000). Common rail injectors for 2000 bar and beyond, SAE paper 2000-01-0706.
Henelin, N. A., Lai, M.-C., Singh, I. P., Zhong, L. & Han, J. (2002). Characteristics of a common
rail diesel injection system under pilot and post injection modes, SAE paper 2002-
010218.
Lefebvre, A. (1989). Atomization and Sprays, Hemisphere Publishing Company.
Munson, B. R., Young, D. F. & Okiishi, T. H. (1990). Fundamentals of Fluid Mechanics, Wiley.
Nasar, S. (1995). Electric machines and power systems : Vol. 1. Electric Machines, McGraw-Hill.
Park, C., Kook, S. & Bae, C. (2004). Effects of multiple injections in a hsdi diesel engine
equipped with common rail injection system, SAE paper 2004-01-0127.
Payri, R., Climent, H., Salvador, F. J. & Favennec, A. G. (2004). Diesel injection system mod-
elling. methodology and application for a rst-generation common rail system, Pro-
ceedings of the Institution of Mechanical Engineering Vol. 218 Part D.
Schmid, M., Leipertz, A. & Fettes, C. (2002). Inuence of nozzle hole geometry, rail pres-
sure and pre-injection on injection, vaporization and combustion in a single-cylinder
transparent passenger car common rail engine, SAE paper 2002-01-2665.
Schommers, J., Duvinage, F., Stotz, M., Peters, A., Ellwanger, S., Koyanagi, K. & Gildein, H.
(2000). Potential of common rail injection system passenger car di diesel engines,
SAE paper 2000-01-0944.
Streeter, V. L., White, E. B. & Bedford, K. W. (1998). Fluid Mechanics, McGraw-Hill.
Stumpp, G. &Ricco, M. (1996). Common rail - an attractive fuel injection systemfor passenger
car di diesel engines, SAE Transactions 960870.
Von Kuensberg Sarre, C., Kong, S.-C. & Reitz, R. D. (1999). Modeling the effects of injector
nozzle geometry on diesel sprays, SAE paper 1999-01-0912.
White, F. M. (1991). Viscous Fluid Flow, McGraw-Hill.
Xu, M., Nishida, K. & Hiroyasu, H. (1992). A practical calculation method for injection pres-
sure and spray penetration in diesel engines, SAE Transactions 920624.
Yamane, K. & Shimamoto, Y. (2002). Combustion and emission characteristics of direct-
injection compression ignition engines by means of two-stage split and early fuel
injection, J. of Engineering for Gas Turbines and Power 124: 660667.
Fuel njection 120
The investigation of the mixture formation upon fuel injection into high-temperature gas ows 121
X

The investigation of the mixture formation upon
fueI injection into high-temperature gas fIows

Anna Maiorova, AIeksandr Sviridenkov and VaIenlin Trelyakov
Ccn|ra| |ns|i|u|c cf Atia|icn Mc|crs namc af|cr P.|. 8aranct
Russia

1. Introduction
Conluslion of a fueI in lhe conluslion chanlers of a gas-lurline engine and a gas-lurline
pIanl is cIoseIy connecled vilh lhe processes of nixing (Lefelvre, 1985). Invesligalions of
lhese processes carried oul ly lolh experinenlaI and conpulalionaI nelhods have recenlIy
lecone especiaIIy cruciaI lecause of lhe necessily of soIving ecoIogicaI prolIens.
One of lhe nosl pressing prolIens al presenl is accounl for lhe infIuence of dropIels on an
air fIov. In sone of lhe regines of chanler operalion lhis nay Iead lo a sulslanliaI, aInosl
lvofoId, change in lhe Iong range of a fueI spray and, consequenlIy, lo corresponding
changes in lhe dislrilulions of lhe concenlralions of fueI phases.
In lhis chapler physicaI nodeIs of lhe processes of inlerphase heal and nass lransfer and
conpulalionaI lechniques lased on lhen are suggesled. The presenl vork is a conlinualion
of research ly Maiorova & Trelyakov, 2OO8. We sel oul lo caIcuIale lhe fieIds of air veIocily
and lenperalure as veII as of lhe dislrilulion of a Iiquid fueI in noduIe conluslion
chanlers vilh accounl for lhe processes of healing and evaporalion of dropIels in lhose
regines lypicaI of conluslion chanlers in vhich lhere is a sulslanliaI inlerphase exchange.
Il is cIear lhal vhen a "coId" fueI is suppIied inlo a "hol" air fIov, lhe dropIels are healed and
lhe air surrounding lhen is cooIed. Il is evidenl lhal al snaII fIov rales of lhe fueI lhis
cooIing can le negIecled. The ain of lhis vork is lo ansver lvo queslions: hov nuch lhe air
fIov is cooIed ly fueI in lhe range of paranelers lypicaI of reaI conluslion chanlers, and
hov far lhe region of fIov cooIing exlends. Moreover, lhe dependence of lhe fIov
characlerislics on lhe neans of fueI spraying (pressure alonizer, jelly or pneunalic) and
aIso on lhe spraying air lenperalure is invesligaled.

2. Statement of the ProbIem
Schenes of caIcuIaled areas are presenled on fig. 1. CaIcuIalions vere carried oul for lhe
veIocily and lenperalure of lhe nain air fIov U
O
= 2O n s and T
O
= 9OO K, fueI veIocily V
f
=
8 n/s, fueI lenperalure T
f
= 3OO K. The gas pressure al lhe channeI inIel vas equaI lo 1OO
kIa.
The firsl nodeI seIecled for invesligalion (fig. 1-a) is a slraighl channeI of reclanguIar cross
seclion 15O nn Iong inlo vhich air is suppIied al a veIocily U
O
and lenperalure T
O
. Il vas
7
Fuel njection 122
assuned lhal lhe slaIIing air fIov al lhe inIel had a deveIoped lurluIenl profiIe and lhal lhe
spraying air had a uniforn profiIe. Injeclion of a fueI vilh a lenperalure T
f
inlo lhe channeI
al a veIocily V
f
is nade lhrough a hoIe in lhe upper vaII of lhe channeI vilh lhe aid of an
injeclor inslaIIed aIong lhe nornaI lo lhe IongiludinaI axis of lhe channeI haIfvay lelveen
lhe side vaIIs. In nodeIing lhe pneunalic injeclor il is considered lhal, coaxiaIIy vilh lhe
fueI suppIy, lhe spraying air is fed al a veIocily U
1
and lenperalure T
1
inlo lhe channeI
lhrough a reclanguIar hoIe of size 4.5 3.75 nn. In nodeIing a jelly injeclor, ve assune lhal
lhe spraying air is alsenl.

(a)
(b)
Iig. 1. Schenes of caIcuIaled areas
The varialIe paranelers of lhe caIcuIalion vere lhe veIocily and lenperalure of lhe
spraying air: U
1
= O-2O ns and T
1
= 3OO-9OO K, as veII as lhe sunned coefficienl of air
excess lhrough lhe noduIe = 1.35-5.4.
The vaIues of lhe regine paranelers are presenled in TalIe 1. Regine 1 corresponds lo jel
spraying of a fueI, regine 2 ~ lo pneunalic spraying of a fueI ly a coId air jel, and regine
3 ~ lo pneunalic spraying ly a hol air jel in lhe Iiniling case of equaIily lelveen lhe
lenperalures of lhe spraying air and nain fIov.

Varianl Regine 1 Regine 2 Regine 3
U
1
, n/s U
1
, n/s T
1
, K U
1
, n/s T
1
, K
1 5.4 O 2O 3OO 2O 9OO
2 2.7 O 2O 3OO 2O 9OO
3 1.35 O 2O 3OO 2O 9OO
TalIe 1. Operaling Iaranelers for lhe fIov in a slraighl channeI.

The second nodeI (fig. 1-l) is lhe fIov lehind lvo coaxiaI lules in radius of 5 and 4O nn,
lule Ienglh is 24O nn. Heal-nass lransfer of drop-forning fueI vilh lhe co-svirIing lvo-
phase lurluIenl gas fIovs is caIcuIaled. In lhis case injeclion of a fueI is nade lhrough a
pressure or pneunalic alonizer aIong lhe IongiludinaI axis. Regine paranelers
corresponds regines 2 and 3 fron lalIe 1 and = 3.3. InIel condilions vere conslanl axiaI
veIocily, lurluIenl inlensily and Ienglh. AxiaI svirIers are sel in inIel seclions. The
langenliaI veIocily sel conslanl in lhe ouler channeI. The fIov in lhe cenlraI lule exil seclion
corresponded lo soIid lody rolalion Iav. The vane angIes in inner and ouler channeIs (
1
and
O
) varied fron O lo 65
o
.

3. CaIcuIation Technique
CaIcuIalions of lhe fIov of a gas phase are lased on nunericaI inlegralion of lhe fuII syslen
of slalionary ReynoIds equalions and lolaI enlhaIpy conservalion equalions vrillen in LuIer
varialIes. The lechnique of aIIoving for lhe infIuence of dropIels on a gas fIov is lased on
lhe assunplion lhal such an aIIovance can le nade ly inlroducing addilionaI sunnands
inlo lhe source lerns of lhe nass, nonenlun, and energy conservalion equalions. The
lransfer equalions vere vrillen in lhe foIIoving conservalive forn:

div
giau S
(1)

Here S
is lhe inlerphase source lern lhal descriles lhe infIuence of dropIels on lhe
corresponding characlerislics of fIov. The densily and pressure are ensenlIe-averaged
(according lo ReynoIds) and aII lhe renaining dependenl varialIes ~ according lo Iavre,
i.e., vilh lhe use of densily as a veighl coefficienl.
Wrillen in lhe forn of Lq. (1), lhe syslen of equalions of conlinuily (u 1,
u
O, S
u
O),
nolion (u= U
gi
, i = 1, 2, 3), and of lolaI enlhaIpy conservalion h (S
h
O) is soIved ly lhe
SinpIe finile-difference ileralion nelhod (Ialankar, 198O). The vaIIs vere considered
Fuel njection 124
lhernaIIy insuIaled. To find lhe coefficienls of lurluIenl diffusion, use is nade of lhe
oussinesq hypolhesis on lhe Iinear dependence of lhe conponenls of lhe lensor of
lurluIenl slresses on lhe conponenls of lhe lensor of defornalion rales of average nolion
and lvo equalions of lransfer of lurluIence characlerislics (k-) in lhe nodificalion lhal
lakes inlo accounl lhe infIuence of fIov lurluIence ReynoIds nunlers on lhe lurluIenl
characlerislics of fIov (Chien, 1982). Here, lhe loundary condilions of zero veIocily are
inposed on lhe soIid vaIIs. Ior svirI fIovs caIcuIalions lhe nodeI vas nodernized lo lake
inlo accounl lhe svirI effecl in lurluIence slruclure (KoosinIin al aI., 1974).
In lhe alsence of chenicaI reaclions lhe gas nixlure is considered lo consisl of lvo
conponenls: kerosene vapors (vilh a noIecuIar veighl of O.168 kg noIe) and air (vilh a
convenlionaI noIecuIar veighl of O.O29 kg noIe). Ior lhe nass fraclion of kerosene vapors
n
f
lhe equalion of lransfer of lhe lype (1) is soIved, and lhe nass fraclion of air is
delernined fron lhe condilion under vhich lhe suns of lhe nass fraclions of aII lhe
conponenls are equaI lo unily.
The caIcuIalions of lhe dislrilulion of fueI are lased on lhe soIulion of a syslen of equalions
of nolion, healing, and evaporalion of individuaI dropIels vrillen in lhe Lagrange
varialIes. The infIuence of lurluIenl puIsalions onlo lhe nolion of dropIels and on lhe
change in lheir shape in lhe process of lheir nolion is considered lo le negIigilIy snaII.
Then lhe equalions lhal descrile lhe processes of nolion, healing, and evaporalion have lhe
foIIoving forn:

m
uv
ut
R

(2)
uT
ut

Re
Pi

(3)
uB
ut

ln
c
L
T
Re
(4)

We consider lhal lhe Iav of lhe resislance of dropIels is lhe sane as lhal of lhe resislance of
soIid sphericaI parlicIes of dianeler D
d

R
=O.5C
R
S
g
WW
, C
R
= 24Re
1
+ 4.4Re
O.5
+O.32 , S = D
d
2
4 (5)

In nodeIing a fueI spray il vas assuned lhal il had a poIydisperse slruclure vilh lhe size
dislrilulion of dropIels oleying lhe Rosin-RannIer Iav (Dilyakin al aI., 1977) vilh
exponenl 3 and nean-nedian dianeler 5O n. The range of lhe sizes of dropIels vas
divided inlo 14 inlervaIs. The angIe dislrilulion of dropIels vas laken lo le uniforn, and
lhe vorking fueI vas TS-1 kerosene.
The inlerphase source lerns are caIcuIaled logelher vilh lhe dislrilulion of lhe Iiquid fueI
fron lhe condilions of lhe fuIfiIInenl of lhe Iavs of conservalion of nonenlun, nass, and
heal of lhe gas-dropIel syslen. Il is considered lhal lhe corresponding lerns in lhe
equalions for lhe lurluIence characlerislics can le negIecled.
Since physicaIIy lhe source lern S

in lhe conlinuily equalion, jusl as lhe source lern in lhe
equalion of lransfer of n
f
, S
, is lhe increase in lhe concenlralion of lhe fueI vapor per unil

line equaI lo lhe rale of Iiquid evaporalion, lhen

S
v
=
(6)

vhere C
v
is lhe rale of change of C
v
due lo lhe inlerphase exchange.
The inlerphase source lerns in lhe equalions of conservalion of nonenlun conponenls are
lhe conponenls of lhe veclor of lhe rale of change in lhe gas nonenlun due lo lhe
exchange vilh dropIels in a unil voIune S
. These quanlilies are delernined fron lhe

equalion of conservalion of nonenlun for lhe gas-dropIel syslen:

A(n
d
v
d
)+A(n
g
0
g
)= O
(7)

vhere n
g
is lhe nass of lhe isoIaled eIenenl of lhe gas voIune Av. Here and leIov, il is
assuned lhal lhe voIune of fueI dropIels is negIigilIy snaII as conpared lo lhe voIune
occupied ly lhe gas.
Assuning Al
d
(lhe residence line of a dropIel in lhe voIune eIenenl Av) lo le snaII enough,
ve nay repIace lhe second lern in (7) ly S
AvAl
d
. This gives us an approxinale expression
lo delernine S
:

S
(8)

vhere AV
d
is a change in lhe dropIel veIocily during ils residence in lhe eIenenlary voIune,
i neans individuaI dropIel.
The Iasl lern in reIalion (8) descriles lhe gas nonenlun increnenl al lhe expense of lhe
vapor fueI phase nonenlun reIaled lo lhe eIenenlary voIune Av and lhe line of dropIel
evaporalion in lhis voIune, since AC
v
= AC
f
. Il is assuned lhal lhe fueI vapor and air in lhe
voIune Av nix inslanlaneousIy. When Av O, Al
d
O, ve ollain an exacl expression for
S
in a differenliaI forn:

S
(9)

Here C
f,i
denoles a fraclion of lhe ilh dropIel in lhe voIunelric concenlralion of Iiquid.
The sunned vaIue of lhe rale of change in lhe nonenlun of a unil voIune of gas is equaI
lo

S

(1O)

vhere sunnalion is carried oul over aII lhe dropIels.
The inlerphase source lern in lhe lransfer equalion of lhe varialIe U
r, S
is delernined
fron lhe equalion of conservalion of anguIar nonenlun for lhe gas-dropIel syslen. This
lean has lhe foIoving forn:

S
i C

(11)
Fuel njection 126
The inlerphase source lern in lhe equalion for enlhaIpy S
lhal descriles heal exchange

lelveen dropIels and lhe gas fIov is delernined fron lhe equalion of conservalion of lhe
lolaI enlhaIpy of lhe gas-dropIel syslen, vhich has lhe forn

A(m
+A(m
)= -LA n
g

(12)

The expression on lhe righl-hand side of equaIily (12) delernines lhe energy spenl on lhe
lransilion of lhe dropIel Iiquid of nass An
d
= An
g
inlo lhe gaseous slale, and Ah
d
and An
d
are changes in lhe enlhaIpy and nass of lhe dropIel during ils residence in lhe voIune Av.
Assuning lhe line Al
d
lo le snaII enough, ve repIace lhe second lern in expression (12) ly
S
AvAl
d
. Then lhe approxinale expression for delernining S
+L

(13)

Using lhe definilion of lhe enlhaIpy h
d
= c
f
T, ve viII revrile (13) in lhe forn

S
(14)

When Av O, Al
d
O, ve ollain an expression for S
in a differenliaI forn:

S
(15)

The sunned vaIue of S
(infIov of heal fron lhe Iiquid phase lo lhe unil voIune

of gas) is equaI lo

S

(16)

vhere sunnalion is carried oul over aII lhe dropIels.
The vaIues Av
d,i
Al
d,i
, AT
d,i
Al
d,i
and AC
f,i
Al
di
or dv
d,i
dl,
,
dT
d,i
dl and dC
f,i
dl are laken fron
lhe soIulion of lhe equalion of nolion and healing of an individuaI dropIel.
The lechnique of caIcuIalion of a lvo-phase fIov is lased on lhe soIulion of a conjugale
prolIen of fIov of lhe gas and Iiquid nedia and heal exchange lelveen lhen. Iirsl lhe
prolIen of lhe nolion of a gas is soIved vilhoul accounl for lhe infIuence of lhe nolion of
dropIels on lhe fIov and lhen, lased on lhe veIocily and lenperalure fieIds ollained, lhe
dislrilulion of lhe Iiquid fueI is caIcuIaled as veII as lhe inlerphase source lerns. Al lhe
second slage, lhe gasdynanic and lenperalure fieIds are recaIcuIaled vilh accounl for lhe
inlerphase sources (lhe resuIls of lhe firsl slage are used as lhe iniliaI condilions). When
needed, lhe process is repealed severaI lines. The convergence crileria of lhe ileralion
process are considered lo le lhe alsence of changes in lhe veIocily and lenperalure fieIds
fron ileralion lo ileralion for lhe gas fIov and slaliIizalion over lhe ileralions of lhe
coordinale of lhe naxinun vaIue of lhe concenlralion of dropIels al lhe oulIel of lhe nodeI
vilhin lhe Iinils of one nesh of lhe finile-difference grid.
4. Testing of the CaIcuIation Technique
The firsl nodeI (fig. 1-a) vas originaIIy used lo lesl lhe caIcuIalion nelhod. As a resuIl of
nelhodicaI caIcuIalions a finile-difference grid uniforn in lhe x and z direclions vas
seIecled. The grid aIong lhe y axis vas nade finer lovard lhe channeI vaIIs according lo lhe
exponenliaI Iav vilh exponenl O.91. The lolaI nunler of nodes in lhe grid vas 11111141 =
5O5,161.
In experiences of aulhors il vas spenl Iaser visuaIizalion of a slrean and poslprocessing of
pholos ly a nelhod of lhe gradienl anaIysis. The conparison on Iig.2 shovs lhal lhe
conpulalionaI lechnique descriles veII lhe experinenlaI dala on lhe configuralion of lhe
fueI spray.

Iig. 2. IsoIines of voIunelric concenlralions of fueI dropIels lehind lhe pressure alonizer in
lhe cenlraI IongiludinaI seclion of lhe reclanguIar nixer, gray Iines - caIcuIalion, coIor Iines -
experinenlaI dala (gradienl anaIysis), T
O
= 3OO K

5.ResuIts of CaIcuIations
The resuIls of caIcuIalion for lhe slraighl channeI of reclanguIar cross seclion are presenled
in fig. 3 - 11. The veIocily fieId in lhe vicinily of lhe pIace of fueI injeclion for lhe jelly (U
1
=
O) and pneunalic (U
1
= 2O n s) sprayings are given in Iigs. 3 and 4, respecliveIy. Here and
leIov, lhe resuIls vere nade nondinensionaI lhrough division ly lhe characlerislic
dinension H = 5O nn, vhich is lhe heighl of lhe channeI (R
O
for lhe axisynnelric nixer),
and ly lhe characlerislic veIocily U
O
= 2O n s.
Fuel njection 128

Iig. 3. CaIcuIaled veclor veIocily fieId in lhe cross seclion of lhe reclanguIar nixer x = O.28
vilh jelly suppIy of fueI (regine 1, U
1
= O), = 1.35

In lhe alsence of fueI suppIy al U
1
= O lhe fIov is honogeneous and isolhernaI. In lhe case
of lhe jelly spraying, as a resuIl of lhe inleraclion of dropIels vilh lhe nain air fIov, on lolh
sides of lhe cenler of lhe injeclion hoIe, zones of reverse fIov inilialed ly dropIels are
olserved (Iig. 3), vhich increase vilh lhe fueI fIov rale. Al lhe sane line, lhe very vaIues of
lhe secondary fIov veIocilies are aInosl an order of nagnilude snaIIer lhan lhe charac-
lerislic fIov veIocily. In lhe IongiludinaI seclion lhe shape of lhe veIocily profiIes preserves
ils inIel configuralion.
In suppIying spraying air (Iig. 4) lhe nain roIe in lhe fornalion of lhe gas veIocily fieIds is
pIayed ly lhe inleraclion of air slreans of lhe nain and spraying air. Thus, lehind lhe
injecled jel a secondary fIov is forned in lhe forn of a lhree-dinensionaI zone of reverse
fIovs. The infIuence of lhe process of inleraclion of dropIels vilh air on lhe fIov slruclure is
praclicaIIy unnolicealIe for lhe cases considered (lhe pallerns of fIov for aII lhe regines are
praclicaIIy idenlicaI). Moreover, lhe deplh of penelralion of fueI-air jels inlo lhe slaIIing air
fIov decreases vilh increase in lhe lenperalure of lhe spraying air due lo lhe decrease in
lhe injecled gas nonenlun.
z

(a) (l)
Iig. 4. CaIcuIaled veclor veIocily fieId in lhe cenlraI IongiludinaI seclion of lhe reclanguIar
nixer vilh pneunalic suppIy of a fueI, a) spraying ly a coId air jel (regine 2, U
1
= 2O ns,
T
1
= 3OO K), l) spraying ly a hol air jel (regine 3, U
1
= 2O ns, T
1
= 9OO K)
Iigures 5 - 8 presenl lhe dislrilulions of dinensionIess voIunelric concenlralions of a Iiquid
fueI c
f
. The resuIls vere nade nondinensionaI lhrough division ly lhe vaIue of lhe nain air
fIov densily al lhe inIel.

(a) (l)
Iig. 5. IsoIines of voIunelric concenlralions of fueI dropIels in lhe cenlraI IongiludinaI
seclion of lhe reclanguIar nixer vilh jelly suppIy of fueI (regine 1, U
1
= O), a) = 5.4, l) =
1.35

A conparalive anaIysis of lhe concenlralion fieIds in jelly spraying for various vaIues of
(fig. 5-6) shovs lhal for lhe higher vaIues of lhe fueI fIov rale lhere corresponds a vider fueI
spray. This spray is nore exlended, ils inner region is characlerized ly higher vaIues of
concenlralions, and il occupies a grealer voIune. Moreover, lhe pallerns of lhe dislrilulions
of concenlralions of dropIels are idenlicaI, indicaling lhe insignificanl infIuence of dropIels
on lhe gas fIov veIocily fieIds.

(a) (l)
Iig. 6. IsoIines of voIunelric concenlralions of fueI dropIels in lhe lransverse seclion x = O.28
of lhe reclanguIar nixer vilh jelly suppIy of fueI (regine 1, U
1
= O), a) = 5.4, l) = 1.35

Iigure 7 presenls lhe caIcuIaled dislrilulions of dinensionIess voIunelric concenlralions of
a Iiquid fueI c
f
for coId spraying in lhe characlerislic seclions of a nixer: in lhe IongiludinaI
Fuel njection 130
seclion lhal passes lhrough lhe cenler of lhe injeclion hoIe (z = O) and in lhe lransverse
seclion innedialeIy lehind lhe injeclion hoIe al lhe dislance x = O.28 fron lhe inIel seclion.
A conparison vilh Iig. 5-6 shovs lhal in lhe case of pneunalic spraying lhe pallerns of
fueI dislrilulion change apprecialIy. Hovever, in lhis case loo lhe infIuence of exchange ly
nonenlun lelveen lhe air and dropIels on lhe dislrilulion of concenlralions is hardIy
nolicealIe.

(a) (l)
Iig. 7. IsoIines of voIunelric concenlralions of fueI dropIels in lhe cenlraI IongiludinaI (a)
and lransverse x = O.28 (l) seclions of lhe reclanguIar nixer vilh pneunalic suppIy of fueI,
spraying ly a coId air jel (regine 2, U
1
= 2O n s, T
1
= 3OO K), = 1.35

Iron lhe graphs of lhe dislrilulions of lhe voIunelric concenlralions of fueI dropIels il is
seen lhal on lhe vhoIe lhe Ialler foIIov lhe air fIov. The spIilling of lhe fueI jel in lhe
lransverse direclion in pneunalic spraying is associaled vilh lhe appearance of inlense
circuIalion fIovs in lhe vake of lhe spraying air jel. The alsence of such spIilling in jelly
spraying indicales lhal lhe secondary fIovs induced ly dropIels are insufficienlIy inlense.
We nole lhal vhen a high-lenperalure air jel is injecled inlo a slaIIing fIov, lhe deplh of fueI
penelralion inlo a nixer is snaIIer lhan in lhe case of spraying ly a coId jel (see Iig. 8). This
effecl is due, firsl of aII, lo lhe Iessening of lhe penelraling pover of an air jel (injeclion of a
gas of a snaIIer densily) and, second, lo lhe enhancenenl of lhe processes of healing and
evaporalion of dropIels in a high-lenperalure air fIov of lhe injecled jel.

0.02
0.02
0.02
0.22
0.22
1.02
2.62
0.42
0.82
0.02
0.62
0.62
0.02
z
y
-0.2 -0.1 0 0.1
0.5
0.6
0.7
0.8
0.9
4.82
4.62
5.02
1.02
2.42
1.22
0.62
0.22
0.02
0.02
0.02
x
y
0 0.2 0.4 0.6 0.8
0.3
0.5
0.7
0.9

Iig. 8. IsoIines of voIunelric concenlralions of fueI dropIels in lhe cenlraI IongiludinaI (a)
and lransverse x = O.28 (l) seclions of lhe reclanguIar nixer vilh pneunalic suppIy of fueI,
spraying ly a hol air jel (regine 3, U
1
= 2O n s, T
1
= 9OO K), = 1.35

Thus, in lolh jelly and pneunalic spraying of a fueI for lhe regines considered il is possilIe
lo negIecl lhe exchange of nonenlun lelveen lhe gas and dropIels and judge lhe
inleraclion of dropIels vilh lhe air fIov fron lenperalure fieIds. QuanlilaliveIy lhe inlensily
of heal lransfer is characlerized, firslIy, ly lhe dinensions of lhe region in vhich lhe gas
lenperalure is snaIIer lhan lhal of lhe surrounding fIov (in lhis case lhe loundary of lhis
region is T = 9OO K) and, secondIy, ly lhe nininun gas lenperalure in lhe conpulalionaI
donain. The forner quanlily indicales lhe parl of lhe space vhere lhe air lenperalure
undervenl a change and lhe Ialler ~ lhe quanlily of heal laken ly dropIels fron lhe gas.
The vaIues of lhe nininun gas lenperalures are given in TalIe 2 for aII lhe operaling
condilions considered.

Regines Varianl 1 Varianl 2 Varianl 3
1 638 539 447
2 3OO 3OO 3OO
3 724 612 5O2
TalIe 2. Mininun Cas Tenperalure, K, in lhe reclanguIar nixer

5.92
4.92
4.82
2.72
3.22
1.92
0.92
0.22
0.22 0.12
0.02
0.02
0.02
x
y
0 0.2 0.4 0.6 0.8
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.32
0.82
0.62
0.62
0.22
0.02
0.02
0.42
0.02
x
y
-0.2 -0.1 0 0.1
0,5
0,6
0,7
0,8
0,9
Fuel njection 132

Iig. 9. IsoIines of air lenperalures in lhe cenlraI IongiludinaI (a), lransverse x = O.28 (l) and
cross y = O.95 (c) seclions of lhe reclanguIar nixer of lhe reclanguIar nixer vilh jelly suppIy
of fueI (regine 1, U
1
= O), = 1.35

The caIcuIalions have shovn lhal even in lhe alsence of suppIy of lhe spraying air lhe gas
lenperalure depends sulslanliaIIy on lhe vaIues of operaling condilions. The dislrilulions
of air lenperalures in lhe alsence and in lhe presence of a spraying air are presenled in Iigs.
9 and 1O - 11 respecliveIy. Iigure 9 characlerizes lhe direcl infIuence of heal exchange
519
579
629
679
729
769
809
839
849
859
879
869
889
899
889
639
689
739
799
879
899
869
799
819
779
759
739 719
699
769
789
y
x 0 0,2 0,4 1.5 0,8 1 1,2 1,4
0,4
0,5
0,6
0.7
0,8
0,9

539
569
689
719
769
739
869
889
899
899
879
869
899
y
z

-0.2 --0,1 0 0.1
0,6
0,7
0.7
0,9

489
589 669 709 739
759
769
779
789
819
859
879
899
889
849
839
799
749
679
849
739
739 729
779
799
829
849
869
889
899
899
829
849
x
z
0 0.2 0.4 0.6 0.8 1 1.2 1.4
-0.3
-0.2
-0.1
0
0.1
0.2

l
a
c
lelveen lhe gas and dropIels on lenperalure fieIds, since in lhe alsence of lhis exchange air
has lhe sane iniliaI lenperalure over lhe enlire region of fIov. Iron lhe dislrilulions of
lenperalures in lhe IongiludinaI seclions of lhe nodeI il is seen lhal al = 1.35 lhe region of
heal lransfer al x = 1.6 exlends in lhe direclion of lhe y axis lo lhe dislance Ay = O.55. As
caIcuIalions shoved, al = 5.4 lhis dislance is equaI lo Ay = O.42. The nininun
lenperalures lhal correspond lo lhese varianls are equaI lo 447 and 683 K (TalIe 2). Ior lhe
varianl = 2.7 lhis quanlily is equaI lo 539 K. Thus, on increase in lhe fueI fIov rale lhrough
a jel injeclor lhe infIuence of dropIels on lenperalure fieIds lecones nore and nore
apprecialIe.

Iig. 1O. IsoIines of air lenperalures in lhe cenlraI IongiludinaI seclion of lhe reclanguIar
nixer vilh pneunalic suppIy of fueI, spraying ly a coId air jel (regine 2, U
1
= 2O n /s, T
1
=
3OO K), a) = 5.4, l) = 1.35

As caIcuIalions shov, on injeclion of a coId spraying air (Iig. 1O), vhen heal lransfer is
nainIy delernined ly lhe inleraclion of lhe nain and spraying fIovs, lhis effecl is virluaIIy
unnolicealIe. When a hol spraying air is injecled (T
1
= 9OO K), heal lransfer viII again le
749
779
759
739
709
689
309
529
589
639
659
699
719
729
759
709
659
619
529
839
849
769
779
769
839
869
869
869
849
879
879
889
889 899
899
899
x
y
0 0.2 0.4 0.6 0.8 1 1.2 1.4
0.4
0.5
0.6
0.7
0.8
0.9
309
429
529
559
589
609
649
689
719
739 689
679
749
789
779
419
539
569
619
679
699
679
709
739
789
759
679
599
839
849
789
879
889
899
899
899
889
819
679
x
y
0 0,2 0,4 0,6 0,8 1 1,2 1,4
0,3
0,4
0,5
0,6
0,7
0,8
0,9
a
b
Fuel njection 134
delernined ly lhe inleraclion of air fIovs vilh dropIels, and lherefore lhe infIuence of lhe
fueI fIov rale on lhe fornalion of lenperalure fieIds lecones apprecialIe (TalIe 2). The
corresponding graphs are presenled in Iig. 11. Il is seen lhal in lhese cases lhe infIuence of
dropIels nanifesls ilseIf virluaIIy in lhe enlire fIov region.

Iig. 11. IsoIines of air lenperalures in lhe cenlraI IongiludinaI seclion of lhe reclanguIar
nixer vilh pneunalic suppIy of fueI, spraying ly a hol air jel (regine 3, U
1
= 2O n /s,
T
1
= 9OO K), a) = 5.4, l) = 1.35

Considering lhe nodeI of heal lransfer suggesled in lhe presenl vork, lvo nonenls nusl le
noled. The firsl is lhal lhe change in lhe gas lenperalure occurs oving lo lhe lransfer of heal
fron lhe gas lo dropIels and is spenl lo heal and evaporale lhen. As caIcuIalions shov, lolh
Ialler processes are essenliaI despile lhe facl lhal lhe lasic fraclion of dropIels (D
d
< 1OO n)
evaporales ralher rapidIy in lhe high-lenperalure air fIov (T
1
= 9OO K). The second nonenl is
lhal healing and evaporalion are lhe nechanisns lhal underIie heal lransfer in lhe very gas
phase and lhey are aIso lvo in nunler. The firsl is lhe convenlionaI diffusion lransfer of heal
and lhe second ~ ils conveclive lransfer due lo secondary fIovs vhich are eilher inilialed ly
dropIels or resuIl fron lhe fIov of lhe slaIIing slrean around lhe spraying air jels.
In lhe case of jelly suppIy of fueI lhe incipienl secondary fIovs are of Iov inlensily, and
dropIels are veakIy enlrained ly such fIovs. This is expressed as lhe alsence of individuaI
899
899
899
889
889
879
879
879
869
869
869
879
879
859
849
839
839
819
799
799
739
869
x
y
0 0,2 0,4 0,6 0,8 1 1,2 1,4
0,4
0,5
0,6
0,7
0,8
0,9

529
679
739
659
849
889
899
899
889
879
869
859
819
749
789
809
819
799
789 769
709
829
839
849
859
829
x
y
0 0.2 0.4 0.6 0.8 1 1.2 1.4
0,4
0,5
0,6
0.7
0,8
0,9

a
b
vorlex slruclures in lhe dislrilulions of lolh concenlralions and lenperalures in lhe
lransverse seclions of lhe noduIe. The Iovering of lhe gas lenperalure occurs excIusiveIy al
lhe expense of inlerphase exchange. Vorlex slruclures are cIearIy seen in lransverse seclions
vilh pneunalic spraying on lhe graphs of lhe dislrilulion of fueI concenlralions. A
conparison lelveen lhe dislrilulions of lenperalures and concenlralions in lhese cases
shovs lhal lhe concenlralion profiIes are nuch narrover lhan lhe corresponding
lenperalure profiIes in lolh IongiludinaI and lransverse direclions. This is associaled vilh
lhe inlense diffusion heal fIuxes, vilh lhe dropIels nainIy foIIoving lhe air fIov. Allenlion
is aIso dravn lo lhe facl lhal lhe penelraling aliIily of a "coId" fueI-air jel is higher lhan lhal
of a "hol" one due lo lhe foIIoving lvo reasons: lhe greal energy of lhe "coId" jel and lhe
nore inlense process of healing and evaporalion of dropIels in lhe "hol" jel.
A conparison of gas cooIing in spraying of a fueI ly a hol air jel and in jelly spraying shovs
lhal aIlhough lhe fueI is injecled inlo fIovs vilh idenlicaI lenperalures, in lhe second case
lhe Iovering of lhe gas lenperalure is nore apprecialIe. This seens lo le due lo lhe facl
lhal on injeclion of dropIels inlo a slaIIing air fIov lhe veIocily of dropIels reIalive lo lhe gas
is higher lhan in lhe case of injeclion inlo a cocurrenl fIov. The rale of lhe evaporalion of
dropIels is aIso higher and, consequenlIy, lhe conpIele evaporalion of dropIels occurs over
snaIIer dislances and in snaIIer voIunes, lhus Ieading lo lhe effecl noled. The lolaI quanlily
of heal lransferred fron air lo dropIels is lhe sane in lolh cases, lul lhe differences
olserved aIIov one lo nake differenl fueI-air nixlures ly suppIying a fueI eilher inlo a
cocurrenl air fIov or inlo a slaIIing one.

(a) (l)
Iig. 12. CaIcuIaled veclor veIocily fieId in lhe IongiludinaI seclion of lhe axisynnelric
nixer, a) -
1
=
O
= 3O, l)
1
=
O
= 6O

The resuIls of caIcuIalion for lhe axisynnelric nixer (fig. 1-l ) are presenled in fig. 12 - 18. The
alove-slaled concIusions are appIicalIe and lo a fIov leyond lhe coaxiaI lules. Hovever in
lhe case of lhe svirI lhe region of fIov cooIing significanlIy depends on lhe operaling
condilions. This effecl is connecled vilh lhe alsence or presence of paraxiaI reverse zone. The
veIocily fieId in lhe vicinily of lhe pIace of fueI injeclion are given in fig. 12. As caIcuIalions
have shovn, lhe lasic roIe in fornalion of veIocily fieIds is pIayed ly a svirI. In svirIing
fIovs vilh

1
> 45 lhere occurs fIov separaled zone. IIov pallerns al nixlure of slreans vilh
idenlicaI (T
1
= T
O
=9OO K ) and various (T
1
= 3OO K, T
O
=9OO K ) lenperalure are aInosl lhe
Fuel njection 136
sane. The infIuence of lhe nean of spraying and lhe process of inleraclion of dropIels vilh air
on lhe fIov slruclure is praclicaIIy unnolicealIe for lhe cases considered.
In fig. 13 - 14 piclures of lrajeclories of lhe dropIels projecled on IongiludinaI seclion of lhe
nixer are resuIled.

Iig. 13. Trajeclories of lhe dropIels in lhe axisynnelric nixer upon fueI injeclion inlo
isolhernaI svirIing fIovs (spraying ly pneunalic alonizer vilh spray angIe 4O),
T
O
= T
1
= 9OO K, a)
1
=
O
= 3O, l)
1
=
O
= 6O

Iig. 14. Trajeclories of lhe dropIels in lhe axisynnelric nixer upon fueI injeclion inlo
nonisolhernaI svirIing fIovs (spraying ly pneunalic alonizer vilh spray angIe 4O), T
O
=
9OO K, T
1
= 3OO K, a)
1
=
O
= 3O, l)
1
=
O
= 6O

To various coIors in draving lhere correspond lrajeclories vilh various iniliaI dianelers of
dropIels. Iron conparison of lhe presenled piclures of lrajeclories il is visilIe, lhal
dislinclions in inleraclion of a fueI spray vilh an air fIov Iead lo significanl differences in
dislrilulions of drops in a vorking voIune. In lhe case of reverse zone (fig. 13 l and 14 l)
dropIels are shifled lo lhe vaII. The lenperalure node aIso pIays lhe inporlanl roIe in
fornalion of a fueI spray. Il is visilIe, lhal al T
1
= T
O
= 9OO K, oving lo evaporalion of drops,
lheir lrajeclories appear nore shorlIy, lhan al nolion in a fIov vilh T
1
= 3OO K. As
caIcuIalions have shovn lhe infIuence of inlerphase exchange on lrajeclories and lhe
dislrilulion of concenlralions is insignificanl.

(a) (l)
Iig. 15. IsoIines of air lenperalures in lhe IongiludinaI seclion of lhe axisynnelric nixer
upon fueI injeclion inlo isolhernaI svirIing fIovs (spraying ly pneunalic alonizer),
T
O
= T
1
= 9OO K, - a) -
1
=
O
= 3O, l)
1
=
O
= 6O
899
891
8 03
795
859
835
x
r
0 0. 2 0.4 0.6 0. 8
0
0. 1
0. 2
0. 3
r
803
811
819
811
899
899
891
891
883
875
867
859
851
843
5
827
x 0 0,1 0,2 0,3
0
0,1
0,2
0,3
a
l
So jusl as in lhe case of reclanguIar nixer il is possilIe lo negIecl lhe exchange of
nonenlun lelveen lhe gas and dropIels and lo judge lhe inleraclion of dropIels vilh an
air fIov fron lenperalure fieIds. Ils cIear lhal lhe grealesl cooIing of a gas fIov ly dropIels
occurs on lhe naxinun gas lenperalure. The dislrilulions of air lenperalures on injeclion
of a hol spraying air are given in Iig. 15. Thal lenperalure fieIds lo lhe fuII are delernined
ly lhe inleraclion of air fIovs vilh dropIels. Iron conparison of dravings in fig 15 a) and
l) il is visilIe, lhal areas of infIuence of dropIels on a gas fIov are various aIso lhey are
delernined in lhe core ly fIov hydrodynanics. In a case
1
=
O
= 3O, lhe fIov is no
separaled and lhe area of cooIing of gas is slrelched aIong an axis. In a case
1
=
O
= 6O
lhere exisls lhe paraxiaI reverse zone. As resuIl lhe Iasl dropIels are shifled lo lhe vaII
logelher vilh cooIed gas. AnaIogous isolhernaIs of gas al fueI spraying fron one source
(suppIy ly pressure alonizer) are resuIled in fig. 16 a) and `16 l).

(a) (l)
upon fueI injeclion inlo isolhernaI svirIing fIovs (spraying ly pressure alonizer ), T
O
= T
1

= 9OO K, - a)
1
=
O
= 3O, l)
1
=
O
= 6O

a) (l)
upon fueI injeclion inlo nonisolhernaI svirIing fIovs (spraying ly pneunalic alonizer),
T
O
= 9OO K, T
1
= 3OO K,
1
=
O
= 3O, a) - vilhoul an inlerphase exchange, l) - laking inlo
accounl an inlerphase exchange
284
308
300
588
652
852
884
892
x
y
0 0.2 0.4 0. 6 0.8
0
0.1
0.2
0.3
0.4
0.5
300
308
404 596
692
892
860
348
x
y
0 0.2 0.4 0.6 0.8
0
0.1
0.2
0.3
0.4
0.5
Fuel njection 138

During injeclion of a coId spraying air lhe heal lransfer is delernined lolh lhe inleraclion of
lhe nain and spraying fIovs and lhe inleraclion of air fIovs vilh dropIels. Cas isolherns in
lhis case are resuIled on fig. 17 and 18, accordingIy for
1
=
O
= 3O and
1
=
O
= 6O.

(a) (l)
upon fueI injeclion inlo nonisolhernaI svirIing fIovs (spraying ly pneunalic alonizer),
T
O
= 9OO K, T
1
= 3OO K,
1
=
O
=6O, a) - vilhoul an inlerphase exchange, l) - laking inlo
accounl an inlerphase exchange

Il is cIear lhal in lhe considered cases heal exchange in lhe core is delernined ly inleraclion
of gas fIovs. The inlerphase exchange changes fieIds of lenperalures onIy near lo a fueI
suppIy pIace, i.e. in order area in lhe size O.2 R
O
.

6. ConcIusions
In aII neans of spraying, for lhe regines considered il is possilIe lo negIecl lhe exchange of
nonenlun lelveen lhe gas and dropIels and lo judge lhe inleraclion of dropIels vilh an
air fIov fron lenperalure fieIds.
Injeclion of a fueI ly a jel injeclor nay cause a sulslanliaI change in lhe gas lenperalure. In
lhe given case il occurs due lo heal lransfer fron lhe gas lo dropIels and is spenl on lheir
healing and evaporalion. In lhe case of pneunalic spraying of a fueI ly a coId air jel lhe
infIuence of inlerphase exchange is insignificanl. Heal lransfer is predoninanlIy delernined
ly lhe inleraclion of lhe nain and spraying fIovs. During injeclion of a hol spraying air,
vhen heal lransfer inside lhe gas fIov is Iess inlense, lhe infIuence of lhe injeclion of a fueI
on lhe fornalion of lenperalure fieIds again lecones apprecialIe. Hovever, in lhis case lhe
gas is cooIed Iess lhan in jelly spraying. This effecl is due lo lhe facl lhal vhen dropIels are
injecled inlo a slaIIing air fIov, lhe rale of lheir evaporalion is higher lhan during injeclion
inlo a cocurrenl fIov.
In lhe case of lhe svirI lhe region of fIov cooIing significanlIy depends on lhe operaling
condilions. This effecl is connecled vilh lhe alsence or presence of paraxiaI reverse zone.
The concIusions dravn confirn lhe necessily of laking inlo accounl lhe processes of
inlerphase heal and nass exchange vhen invesligaling lhe nixlure fornalion.

284
300
396
564
716
884
892
x
y
0 0. 2 0. 4 0.6 0.8
0
0.1
0.2
0.3
0.4
0.5
812
668
540
444
452
308
324
300
0 0.2 0. 4 0. 6 0.8 1
0
0.1
0.2
0.3
0.4
0.5
0.6
7. The further deveIopment of a caIcuIation method
The furlher deveIopnenl of a conpulalionaI lechnique shouId acluale lhe accounl of
coaguIalion and lreakage of dropIels. The caIcuIalions resuIled leIov iIIuslrale lhe
inporlance of lurluIenl coaguIalion of dropIels of lhe spraying fueI lehind injeclors in
conluslion chanlers.
The nain assunplions of physicaI characler inposed on syslen coaguIalion of parlicIes,
consisl in lhe foIIoving. The nunler of parlicIes is greal enough, lhal il vas possilIe lo
appIy funclion of dislrilulion of parlicIes on veighls and in co-ordinale space. OnIy linary
coIIisions are considered, lhe coIIisions conserve lhe nass and voIune, and lhe aerosoI
parlicIes coaguIale each line lhey coIIide. Wilhin lhe SnoIuchovskys lheorelicaI
franevork (see IriedIander al aI., 2OOO), al any line, each aerosoI parlicIe couId le forned
ly an inleger nunler of lase parlicIes ( or nononers), vhich vouId le lhe snaIIesl, sinpIe
and slalIe parlicIes in lhe aerosoI, and lhe densily of lhe nunler of parlicIes vilh k
nononers, n
k
, as a funclion of line, vouId le lhe soIulion of lhe foIIoving laIance
equalion:

un
ut

(17)

Non-negalive funclion K
ij
is caIIed as a coaguIalion kerneI, il descriles parlicuIar inleraclion
lelveen parlicIes vilh voIunes i and j. The firsl lern al lhe righl hand side of Lq. (17) is lhe
produclion of lhe parlicIes vilh k nononers due lo coIIisions of parlicIes vilh i and j
nononers such lhal i + j = k, and lhe second lern is lhe consunplion of parlicIes vilh k
nononers due lo coIIisions vilh olher aerosoI parlicIes.
The najorily of aclivilies on coaguIalion research concern lo alnospheric aerosoIs in vhich
lhis process lasicaIIy is caIIed ly rovn diffusion. SliII in sprays lehind injeclors lhe nain
aclion caIIing increase of lhe sizes of drops, is lurluIenl coaguIalion. Ior such environnenls
lhe coaguIalion kerneI can le recorded in lhe forn of (Kruis & Kuslers, 1997)

K

(18)

Here a
1
and a
2
- radiuses of parlicIes i and j, W
s
- reIalive parlicIe veIocily due lo inerliaI
lurluIenl effecls and W
a
- reIalive parlicIe veIocily due lo shear lurluIenl effecls.
The syslen of equalions (17-18) vas soIved ly lhe finile-difference nelhod (Maiharju, 2OO5).
As a resuIl of lhe soIulion of lhe equalions of lurluIenl coaguIalion il is invesligaled lhe
infIuence of anlienl nediun properlies on grovlh rale of dropIels lehind lhe fronl
noduIe. In parlicuIar lhe infIuence of speed of a dissipalion of lurluIenl energy, lhe iniliaI
size of dropIels and anlienl pressure on dislrilulion of dropIels in lhe sizes on various
dislances lehind an injeclor vas invesligaled. The varialion of lhe nean- nedian dianeler
of dropIels on line (dislance fron an injeclor) for dropIels of lhe iniliaI size 5 and 1O
nicrons and nornaI anlienl pressure is shovn in fig. 19. The researches carried oul have
shovn lhal coaguIalion process can consideralIy change lhe sizes of dropIels. The iniliaI
dianeler of dropIels essenliaIIy infIuences coaguIalion process. So, al increase in lhe iniliaI
Fuel njection 140
size of drops vilh 5 n lo 1On, lhe reIalive nean nedian dianeler of dropIels in O.O1
seconds is increased al 1.2 line (see fig. 19).

Iig. 19. The dependence of reIalive size of dropIels in spray lehind injeclor on coaguIalion
line, lIue Iine - D
nO
= 5n, read Iine - D
nO
= 1O n

Iig. 2O. The dependence of reIalive size of dropIels in spray lehind injeclor on conluslion-
chanler pressure.
0 0.002 0.004 0.006 0.008 0.01

1
1.05
1.1
1.15
1.2
1.25
time [sj
Dm/Dmo
1 3 5 7 9 11 13 15 17 19 21
1
1.05
1.1
1.15
1.2
1.25
1.3
P, bar
Dm/Dmo

Iig. 21. The dislrilulion of voIunelric concenlralion on lhe sizes of dropIels,
lIue Iines - iniliaI dislrilulion, red Iines - dislrilulion in O.O1 seconds, a) - c = 1n
2
/s
3
,
l) c = 1OOn
2
/s
3

In fig. 2O dala aloul infIuence of anlienl pressure on coaguIalion of dropIels of lhe
kerosene spray are resuIled. CaIcuIalions are execuled al vaIue of c 1On
2
/s
3
and iniliaI
D
n
= 5n. Il's evidenlIy fron lhe pIol al pressure varialion fron 1 lo 25 lar lhe nean size
of dropIels as a resuIl of coaguIalion for O.O1 seconds is increased approxinaleIy al 3O .
Rale of a dissipalion of lurluIenl energy is lhe essenliaI paraneler delernining a kerneI of
lurluIenl coaguIalion K (x, y). Lslinalions shov, lhal lehind fronl devices of conluslion
chanlers lhe vaIue of rale of a lurluIenl energy dissipalion varies fron 1 lo 1OO n
2
/s
3
. In
dravings 21- a) and l) dislrilulions of voIunelric concenlralion C
f
for lvo vaIues of a rale
of dissipalion of lurluIenl energy are presenled. The increase in dissipalion Ieads lo
dispIacenenl of dislrilulion of voIunelric concenlralion in area of lhe lig sizes.O So lhe
nain fraclion of drops of spraying Iiquid viII faII lo drops vilh sizes, 1O lines nagniludes
surpassing iniliaI drops.
Thus, anlienl pressure, rale of dissipalion of lurluIence energy and lhe iniliaI size of lhe
dropIels Ieaving an injeclor nake essenliaI inpacl on coaguIalion of dropIels.
Il is necessary lo nole, lhal in disperse syslens, excepl process of coaguIalion vhich
conducls lo inlegralion of parlicIes, lhere are cases vhen lhe inlegraled parlicIe lreaks up
on snaII sponlaneousIy or under lhe infIuence of exlernaI forces. Therefore coaguIalion
process viII le acconpanied ly alonizalion of drops as a resuIl of aerodynanic effecl of air.
Thus as coaguIalion as lreaking of dropIels are desiralIe lo lake inlo accounl vhen
caIcuIaling lhe nixlure fornalion.

8. AcknowIedgement
This vork vas supporled ly lhe Russian Ioundalion for asic Research, projecl No. O8-O8-
OO428.

10
-6
10
-4
10
-2
0
1
2
3
4
x 10
-3
Dm/2 [mj
C
f
a
10
-6
10
-4
10
-2
0
1
2
3
4
x 10
-3
Dm/2 [mj
C
f
l
Fuel njection 142
9. Notation
C
f
, voIunelric concenlralion of a Iiquid fueI, kg n
3
, c
f
, coefficienl of specific heal of Iiquid, }
(kgK), c
pg
, coefficienl of specific heal of gas al conslanl pressure, } (kgK), C
R
, coefficienl of
dropIel resislance, C
v
, concenlralion of fueI vapor per unil voIune, kg n
3
, D
d
, dropIel
dianeler, n, D
n
, dropIel nean nedian dianeler, n, H, channeI heighl, n, h, specific lolaI
enlhaIpy, } kg, k, energy of lurluIence per unil nass, n
2
s
2
, L, Ialenl heal of evaporalion, }
kg, n
d
, nass of a dropIel, kg, n
f
, nass fraclion of kerosene vapors, n
k
, densily of lhe
nunler of parlicIes vilh k nononers, Ir =
g
c
pg
g
, IrandlI nunler, R
, force of
aerodynanic resislance, Re =
g
D
d
W
g
, ReynoIds nunler of a dropIel, S
u
, inlernaI source
lern in lhe equalion of lransfer of lhe varialIe u, T, lenperalure, K, l, line, s, 0
g, veclor of
averaged gas veIocily, U
gi
(i = 1, 2, 3), conponenls of lhe veclor of averaged gas veIocily, n
/s, v
d
, veclor of dropIel veIocily, W
= v
d
-0
g, veclor of dropIel veIocily reIalive lo gas, x, y, z,

Carlesian coordinales, x, r, , cyIindricaI coordinales, , sunned coefficienl of air excess,
u
, coefficienl of diffusion lransfer of varialIe u, Al
d
, line of dropIel residence in lhe
voIune eIenenl, s, Av, eIenenlary voIune, n
3
, , rale of dissipalion of lurluIence energy,
n
2
s
3
,
g
, lhernaI conduclivily of gas, W (nK),
g
, coefficienl of dynanic viscosily of gas,
kg (ns), , densily, kg n
3
, u, dependenl varialIe,
1,

O
, vane angIes of svirIers in inner
and ouler channeIs, . Sulscripls and superscripls: O, nain fIov, 1, spraying air, g, gas, f,
Iiquid fueI, d, dropIel, inl, inlerphase, v, vapor-Iike fueI, i, individuaI dropIel.

10. References
Chien K.}. (1982). Irediclions of channeI and loundary-Iayer fIovs vilh Iov-ReynoIds-
nunler lurluIence nodeI. A|AA ]., VoI. 2O, 33-38.
Dilyakin Yu. I., KIyachko L. A., Novikov . V. and V. I. Yagodkin. (1977). Spraqing cf |iquis
(in Russian), Mashinoslroenie, Moscov.
IriedIander, S. K. (2OOO). Smc|c, Dus| an Hazc. Oxford Univ. Iress, Oxford.
Lefelvre A.H. (1985). Gas Tur|inc Ccm|us|icn, Henisphere IulIishing corporalion,
Washinglon, Nev York, London.
KoosinIin M.L., Launder .L., Sharna .}. (1974). Irediclion of nonenlun, heal and nass
lransfer in svirIing lurluIenl loundary Iayers. Trans. ASM|, Scr. C., VoI. 96, No 2,
2O4.
Kruis I. L & Kuslers K.A. (1997) The CoIIision Rale of IarlicIes in TurluIenl IIov, Cncm.
|ng. Ccmm. VoI. 158, 2O1-23O.
Maiharju S.A (2OO5). Acrcsc| qnamics in a |ur|u|cn| jc|, A Tncsis |nc Dcgrcc Mas|cr cf Scicncc in
|nc Graua|c Scncc| cf Tnc Onic S|a|c Unitcrsi|q.
Maiorova A.I. & Trelyakov V.V. (2OO8). Characlerislic fealures of lhe process of nixlure
fornalion upon fueI injeclion inlo a high-lenperalure air fIov. ]curna| cf
|nginccring Pnqsics an Tncrmcpnqsics, VoI. 81, No. 2, 264-273. ISSN: 1O62-O125.
Ialankar S. (198O). Numcrica| Hca| Transfcr an ||ui ||cu, Henisphere IulIishing, Nev
York.

ntegrated numerical procedures for the design, analysis and optimization of diesel engines 143
X

Integrated numericaI procedures for the design,
anaIysis and optimization of dieseI engines

DanieIa Siano
1
, Ialio ozza
2
and MicheIa Cosla
1

1
|s|i|u|c Mc|cri - CNR
2
D|M| - Unitcrsi|a i Napc|i
|TA|Y

1. Introduction
olh lhe design and anaIysis of a dieseI engine requires lhe inlegralion of accurale
lheorelicaI nelhods, resorling lo 1D - 3D CID nodeIIing and vilro-acouslic engine anaIysis.
In lhis chapler, lhe alove nunericaI approaches viII le deepIy presenled and inlegraled lo
perforn a dieseI engine design and/or anaIysis. As knovn in facl, lhe possiliIily lo sinuIale
lhe physicaI and chenicaI processes characlerising lhe operalion of inlernaI conluslion
engines ly using appropriale codes and high perfornance conpulers is conlinuousIy
spreading. These sinuIalions can predicl, as an exanpIe, fueI consunplion, loxic enissions
and noise radialion. y varying lhe design and/or conlroI paranelers, differenl engine
configuralions or vorking condilions can le lesled and lheir perfornances conpared.
Oplinizalion lechniques (IapaIanlros el aI. 2OOO, Slephenson, 2OO8, Cosla el aI., 2OO9),
properIy nalched vilh lhe various sinuIalion procedures, are hence lhe nosl suilalIe looI
lo idenlify oplinaI soIulions alIe lo gain prescriled oljeclives on engine efficiency, pover
oulpul, noise, gas enissions, elc.. The choice of lhe oplinizalion goaI, noreover, slriclIy
depends on lhe appIicalion lype and lhe definilion of a conpronise soIulion anong lhe
confIicling needs is in nany cases required.
Concerning lhe design of a conluslion engine, a conpIicaled and nuIli-oljeclive lask is lo
le afforded, since il generaIIy requires lhe fuIfiInenl of various oljeclives and conslrainls,
as high efficiency and pover oulpul, Iov noise and gas enissions, Iov cosl, high reIialiIily,
elc. A looI for nuIli-oljeclive oplinizalion, lherefore, can le considered as fundanenlaI al
lhe engine design slage, in order lo gain insighl inlo lhe conpIicaled reIalionships lelveen
lhe physicaI enlilies invoIved in lhe design and design-dependenl paranelers. UIlinaleIy,
oplinizalion can grealIy reduce lhe line-lo-narkel of nev engine prololypes.
Oplinizalion lechniques can successfuIIy le appIied lo anaIyze lhe operaling condilions of
exisling engines, loo. In lhis case, lhe oplinizalion process can le focused on lhe seIeclion of
lhe conlroI paranelers in order lo ollain an oplinaI engine lehaviour. Il is veII knovn, in
facl, lhal conluslion deveIopnenl and enission produclion depend on a conpIex
inleraclion anong differenl paranelers, naneIy injeclion noduIalion and phasing (Slolz el
aI. 2OOO), loosl pressure, LCR fraclion, svirI ralio, fueI properlies, and so on. The oplinaI
choice of a so Iarge nunler of paranelers depends on speed and Ioad condilions, and il is
8
Fuel njection 144

reIaled lo lhe fuIfiInenl of a nunler of conlrasling oljeclives, Iike reduced NOx, Sool, HC,
CO, fueI consunplion and noise enissions.
In lhe presenl chapler lhe ciled approach lo lhe design and anaIysis of a dieseI engine viII
le expIained. The discussion viII le organized in lhe foIIoving paragraphs, each regarding
a differenl case sludy. In parlicuIar, lhe firsl paragraph is focused on lhe descriplion of
singIe nelhodoIogies and lo lheir inlegralion:
- A 1D sinuIalion of lhe vhoIe propuIsion syslen is reaIized ly neans of a proprielary
code. Il aIIovs lo delernine engine-lurlocharger nalching condilions and is alIe lo
conpule pressure, lenperalure and gas conposilion al lhe inlake vaIve cIosure. The
Ialler dala represenl iniliaI condilions for lhe successive 3D anaIysis.
- A 3D sinuIalion of lhe engine cyIinder is deveIoped ly expIoiling geonelricaI
infornalion derived ly lhe engine CADs. The in-cyIinder pressure cycIe during lhe
cIosed vaIve period, is predicled, slarling fron lhe iniliaI condilions provided ly lhe
1D code.
- 1D or 3D conpuled pressure cycIes are lhen uliIized vilhin vilro-acouslic anaIyses
aining lo eslinale lhe conluslion radialed noise. Depending on lhe appIicalion,
differenl approaches are foIIoved:
- ILM-LM approach: ILM anaIysis is appIied lo delernine lhe vilralion of lhe
engine skin surface, Direcl oundary LIenenl Melhod (DLM) soIves lhe
exlerior acouslic radialion according lo lhe ISO direclives, lo predicl lhe
radialed overaII noise IeveI. This nelhod is uliIised during lhe engine design
phase.
- SinpIified approach: An anaIylicaI nodeI lased on lhe deconposilion of in-
cyIinder pressure cycIes is deveIoped lo eslinale lhe radialed noise IeveI. Sone
coefficienls incIuded in lhe alove correIalion are properIy luned lo gel a good
agreenenl vilh lhe acouslic experinenlaI dala. This nelhod is appIied during
lhe engine anaIysis phase.
- The nunericaI nodeIs are in differenl vays coupIed lo lhe oplinizalion code, lo
idenlify lhe oplinaI design paranelers or lhe injeclion slralegies, lo lhe ain of
reaIizing lhe naxinizalion of lhe engine perfornance, lhe reduclion of lhe NOx and
sool enissions, and lhe reduclion of lhe radialed noise, al a conslanl Ioad and
rolalionaI speed.

The second paragraph iIIuslrales lhe design and oplinizalion of a nev lvo-slroke dieseI
engine suilalIe for aeronaulicaI appIicalions. The engine, equipped vilh a Connon RaiI
fueI injeclion syslen, is conceived in a lvo-slroke unifIov configuralion, ained al achieving
a veighl lo pover ralio equaI lo one kg/kW. olh CID 1D and 3D anaIyses are carried oul
lo supporl lhe design phase and lo address sone parlicuIar aspecls of lhe engine operalion,
Iike lhe scavenging process, lhe engine-lurlocharger nalching, lhe fueI injeclion and lhe
conluslion process. The exchange of infornalion lelveen lhe lvo codes aIIovs lo inprove
lhe accuracy of lhe resuIls. Conpuled pressure cycIes are aIso uliIized lo nunericaIIy
predicl lhe conluslion noise, lasing on lhe inlegralion of ILM and LM codes. The
ollained resuIls are suilalIe lo le used as driving paranelers for successive engine
oplinizalion. In order lo inprove lhe engine perfornance and vilro-acouslic lehaviour, lhe

1D nodeI, luned vilh infornalion derived fron lhe 3D code, is Iinked lo lhe oplinizalion
code. A conslrained nuIli-oljeclive oplinizalion is perforned lo conlenporary nininize
lhe fueI consunplion and lhe naxinun in-cyIinder lenperalure and pressure gradienl
direclIy reIaled lo lhe noise enission. In lhis vay a leller seIeclion of a nunler of engine
paranelers is carried oul (exhausl vaIve opening, cIosing and Iifl, inlake porls heighls, slarl
of injeclion, elc).

The lhird paragraph, indeed, descriles an environnenlaI and energelic oplinizalion of a
naluraIIy aspiraled, Iighl-duly direcl injeclion (DI) dieseI engine, equipped vilh a Connon
RaiI injeclion syslen. An experinenlaI canpaign is iniliaIIy carried oul lo gain infornalion
on perfornance and noise IeveIs on lhe engine and lo acquire lhe dala required lo vaIidale
lhe 1D, lhe 3D nodeI and lhe conluslion noise procedure. As in lhe previous case, a
preIininary nunericaI sinuIalion is carried oul. Then, an oplinizalion process is sellIed in
order lo idenlify lhe conlroI paranelers of a lhree puIses injeclion profiIe, for a conslanl
overaII nass of injecled fueI. These paranelers are assuned as independenl varialIes of lhe
nuIli-oljeclive oplinizalion looI and are seIecled vilh lhe ain of sinuIlaneousIy
nininizing fueI consunplion, poIIulanl enissions and radialed noise.

2. NumericaI Procedures
1D 5Imu!atInn: lhe 1D sinuIalion of lhe vhoIe propuIsion syslen is reaIized ly neans of
lhe 1Dine soflvare deveIoped al lhe MechanicaI Lngineering Deparlnenl of lhe Universily
of NapIes Iederico II (Siano el aI., 2OO8, Cosla el aI., 2OO9) and ly expIoiling geonelricaI
infornalion of lhe inlake and exhausl syslen derived ly lhe engine CAD. The vhoIe engine
is firslIy schenalized as a nelvork of pipe and pIenuns, lhen, lhe 1D fIov equalions are
soIved in each pipe consliluling lhe inlake and lhe exhausl syslen. The gas inside lhe
cyIinder is indeed lrealed as a zero-dinensionaI lhernodynanic syslen.
The code soIves lhe 1D fIov equalions in lhe inlake and exhausl pipes:

O
O
O
O
O
O
(
(
|
|
.
|
\
|
+ +
O
O
O
O
=
+
=
=
G[
G
X[
G[
G
X[
'
T
G[
G
X+
X
X
/
&
'
I
G[
G
X
G[
G
X
X[
X[
X+
S X
X
[
[
(
X
I
U
S
I
U
I
U
1
1
4
1
2
2
1
1
2
2
S F U

(1)

vhere , u, p, L=c
v
T+u
2
/2, H=c
p
T+u
2
/2 respecliveIy represenl lhe densily, lhe veIocily, lhe
pressure, and lhe lolaI energy and enlhaIpy per unil nass. The source lern 5 lakes inlo
accounl lhe ducl area varialion aIong lhe fIov direclion, dD/dx, lhe vaII heal exchange, q,
Fuel njection 146

and lhe friclion Iosses. The Iasl lvo equalions descrile lhe scaIar lransporl of chenicaI
species, x
r
and x
f
leing lhe residuaI gases and fueI nass fraclion, respecliveIy. These
equalions aIIov lo conpule lhe conposilion of lhe gases fIoving in lhe inlake and exhausl
syslens and lherefore lo eslinale aIso lhe in-cyIinder charge conposilion.
In lhe case of lurlocharged engine, lhe perfornance naps of lhe lurlocharger group are
enpIoyed lo conpule lhe engine-lurlocharger nalching.
Concerning lhe nodeIing of lhe conluslion process, a cIassicaI Wiele equalion is uliIized lo
conpule lhe heal reIease rale in lhe engine. Iroper vaIues of lhe conluslion process
duralion during lolh prenixed and diffusive phases are specified. This preIininary
approach is sulsliluled in lhe foIIoving ly lhe nore delaiIed 3D anaIysis Ialer descriled.

-360 -270 -180 -90 0 90 180 270 360
Crank AngIe, deg
0
10
20
30
40
50
60
70
P
r
e
s
s
u
r
e
,

b
a
r
Exp.
1D ResuIts
1500 rpm, BMEP=1.5 bar
C1 C2 C3
C4
C5
C6

-360 -270 -180 -90 0 90 180 270 360
Crank AngIe, deg
0
10
20
30
40
50
60
70
P
r
e
s
s
u
r
e
,

b
a
r
Exp.
1D ResuIts
2500 rpm, BMEP=2.5 bar
C1
C2 C3
C4
C5
C6

Iig. 1. 1D conpuled pressure cycIes in differenl operaling condilions

As an exanpIe, Iigure 1 dispIays lhe conparison lelveen conpuled and experinenlaI
pressure cycIes in a lurlocharged six cyIinders engine, for lvo differenl operaling
condilions. The agreenenl aIong lhe conpression slroke indicales lhal a good engine-
lurlocharger nalching can le reached. Slarling fron lhe conpuled iniliaI condilions al IVC,
lhe 3D nodeI is expecled lo furlher inprove conluslion phase anaIysis.

3D 5Imu!atInn: The AVL IIRL 3D code is enpIoyed. Il represenls a nuIlipurpose looI,
speciaIIy conceived for engine appIicalions. The firsl slep of lhe anaIysis concerns lhe
generalion of lhe 3D donain represenling lhe conpulalionaI grid. This is effecled ly neans
of lhe seni-aulonalic procedure of lhe code, naned Iane Advanced Hylrid, vhich aIIovs
lo reach a good conpronise lelveen accuracy and reduced nunler of ceIIs. The
conpulalionaI period is suldivided inlo inlervaIs, each perlinenl lo a grid of a chosen size,
lhal is deforned as lhe pislon noves unliI predefined crank angIes, slarling fron vhich
nev grids vilh differenl sizes are used. This occurs lhrough a procedure of re-napping of
lhe conpuled varialIes, lerned rezone, avoiding an excessive ceIIs defornalion. Wilhin
each ceII, lhe ReynoIds Averaged Navier-Slokes equalions are nunericaIIy soIved lo
conpule lhe 3D fIov fieId and lhe lhernodynanic condilions inside lhe cyIinder. The CID
anaIysis accounls for lhe fueI spray dynanics and for lhe sulsequenl chenicaI reaclions,
Ieading lo lhe prediclion of lhe rale of heal reIease (CoIin and enkenida, 2OO4) , poIIulanls
fornalion and in-cyIinder pressure cycIe.
The fueI spray spalio-lenporaI dynanics is sinuIaled according lo a Discrele DropIels
ModeI (DDM) (Liu and Reilz, 1993, ORourke. 1989, Dukovicz, 198O), vhere lhe LuIerian
descriplion of lhe gaseous phase is coupIed vilh a Lagrangian approach lo lhe sludy of lhe
Iiquid fIov. ModeIIing of spray accounls for prinary and secondary alonizalion,
evaporalion, coaIescence, lurluIence effecls and possilIe cavilalion vilhin lhe nozzIe.
LxanpIes of spray, conluslion and enission caIcuIalions viII le presenled in paragraphs 2
and 3, vilh reference lo seIecled case sludies.

Despile lhe nodeIIing of lhe conluslion and noxious enission, lhe 3D code can le aIso
enpIoyed lo inprove lhe accuracy of lhe previousIy descriled 1D nodeI lhrough lhe
lheorelicaI evaIualion of lhe discharge coefficienls lhrough vaIves or porls. The alove vaIues
are usuaIIy derived fron Iileralure infornalion in 1D nodeIIing and, of course, lhe reIaled
accuracy is Iiniled. AIlernaliveIy, lhe Ialler can le evaIualed ly nunericaIIy sinuIaling lhe
3D air fIov vilhin lhe engine inlake syslen. The conpulalion is perforned under lhe
hypolhesis of sleady condilions, in such a vay lo reproduce a possilIe experinenl reaIisalIe
over a fIov rale lesl lench. An exanpIe grid used for lhis kind of anaIysis is shovn in
Iigure 2. Il refer lo a lvo-slroke engine nore deepIy descriled in paragraph 2 and cIearIy
exhilils lhe geonelricaI characlerislics of lhe air adnission voIune, vilh one infIov ducl
and lhe lhree cyIinders of a lank pIaced vilh lheir axes in lhe direclion orlhogonaI lo lhal of
lhe air infIov. OnIy lhe cenlraI cyIinder is considered as opened. The voIune corresponding
lo lhe inlake porls of lhe cenlraI cyIinder is neshed, lhose of lhe IaleraI cyIinders are nol
incIuded in lhe conpulalionaI donain for lhe sake of sinpIicily. The lolaI nunler of ceIIs is
518346, 33O746 of vhich are hexahedricaI, lhus assuring a cerlain grid reguIarily. The
fourleen porls of lhe cenlraI cyIinder forn one lIock vilh an exlernaI cyIindricaI area,
vhose design foIIovs lhe geonelric characlerislics of lhe cyIinder jackel. The grid in lhis
zone, quile veII visilIe in Iigure 2, is nade parlicuIarIy lhick, since il conprehends 148415
ceIIs. Conpulalion is perforned for fixed vaIues of lhe enlering air nass fIov rale and
aIvays selling lhe slalic pressure and lenperalure al lhe oulIel seclion equaI lo alnospheric
condilions. The porls are opened lo lhe 1OO, 75, 5O and 25 for a nass fIov rale
ranging fron O.O3 lo O.15 kg/s.
Iigure 3 shovs a viev of lhe veIocily nagnilude dislrilulion ollained as a resuIl of lhe
caIcuIalion for inlake porls conpIeleIy opened. The veIocily veclor nagnilude is considered
Fuel njection 148

over a pIane orlhogonaI lo lhe cyIinders axes, culling lhe inlake porls exil seclion in lhe
niddIe. Air veIocily dislrilules non-unifornIy over lhe fourleen porls surfaces: lhe seven
porls vhich are cIoser lo lhe air infIov ducl are leller served and nainIy conlrilule lo fiII
lhe cyIinder vilh lhe fresh charge. The veIocily in correspondence of lhe olher seven porls
nainlains Iover. The shape of lhe porls, vilh lheir surfaces posilioned in langenliaI
direclion vilh respecl lo lhe cyIinder exlernaI surface, aIIovs a svirI nolion of lhe air
enlering lhe cyIinder.

Iig. 2. Crid enpIoyed for lhe evaIualion of lhe inlake porls discharge coefficienl.

Iig. 3. VeIocily fieId on a pIane orlhogonaI lo lhe cyIinder axis


The resuIls of lhe conpulalions aIIov lhe delerninalion of lhe discharge coefficienl as lhe
ralio lelveen lhe effeclive nass fIov rale,
HII
P , and lhe lheorelicaI one,
WK
P :

WK
HII
'
P
P
&
= (2)

The lheorelicaI nass fIov rale is evaIualed as a funclion of inIel pressure and lenperalure
and fIov area, corresponding lo lhe inposed effeclive nass fIov rale.
Iigure 4 sunnarises lhe resuIls of lhe caIcuIalion for various surface percenlages in lhe
opening of lhe inlake porls. The discharge coefficienl nainlains aInosl conslanl vilh lhe
enlering nass fIov rale and ranges lelveen aloul O.6 and O.98, due lo lhe reduclion of lhe
exil seclion area used for lhe evaIualion of lhe lheorelicaI nass fIov rale. These resuIls viII
le direclIy enpIoyed in lhe 1D nodeI, for accuracy inprovenenl.

0 0.04 0.08 0.12 0.16
mass flow rate (kg/s)
0
0.2
0.4
0.6
0.8
1
n
t
a
k
e

p
o
r
t
s

d
i
s
c
h
a
r
g
e

c
o
e
f
f
i
c
i
e
n
t
Port opening 100%
Port opening 75%
Port opening 50%
Port opening 25%

Iig. 4. Discharge coefficienls evaIualed for various opening of lhe inlake porls.

VIbrn-acnustIc ana!yscs: a ILM anaIysis is conducled in order lo evaIuale lhe engine
surface vilralions induced ly lhe conluslion process evoIulion. The previousIy predicled
pressure cycIe is enpIoyed in lhis phase lo conpule lhe forces acling on lhe inlernaI
slruclure. The ollained vilralionaI oulpul dala represenl lhe loundary condilions lo le
appIied lo lhe LM code for lhe finaI evaIualion of lhe radialed sound pover. The seIecled
codes lo evaIuale lhe sound radialion noise fron lhe engine lIock surface are lhe
MSC/Naslran
TM
and LMS/Sysnoise
TM
. olh approaches require lhe deveIopnenl of a
delaiIed 3D nesh and are very line-consuning. Ior lhis reason, lhis delaiIed nelhodoIogy
cannol le direclIy appIied vilhin lhe oplinizalion Ioop. AddilionaI insighls of lhe alove
sunnarized approach viII le given in paragraph 2.

AIlernaliveIy, a sinpIified and recenlIy proposed nelhodoIogy (Torregrosa el aI., 2OO7,
Iayri el aI., 2OO5) can le uliIized for lhe prediclion of lhe overaII conluslion noise, vhich
Fuel njection 150

incIudes in lhe correIalion a slricl dependency on lhe engine operaling condilions and
injeclion slralegy. The nain idea lehind lhis lechnique is lhe deconposilion of lhe lolaI in-
cyIinder pressure signaI according lo lhree nain conlrilulions: conpression-expansion,
conluslion and resonance pressures:

UHV FRPE PRW WRW
S S S S + + = (3)

The firsl conlrilulion (aIso referred as pseudo-nolored signaI) is onIy reIaled lo voIune
varialion, and is used as a reference signaI. Il is delernined ly a direcl in-cyIinder pressure
acquisilion during a fueI svilch-off operalion. The lhird lern (resonance pressure) is indeed
reIaled lo high-frequency pressure fIuclualions, induced ly lhe vave refIeclions in lhe
conluslion chanler, nainIy occurring al auloignilion line. Il is conpuled lhrough a high
pass-land fiIler (alove 45OO Hz) of lhe lolaI pressure signaI. y difference, lhe conluslion
pressure (second lern in eq. 3) can le easiIy delernined, loo. The lhree lerns in eq. 3 are
reporled for conparison in figure 5. Despile lhe presence of lhe previousIy discussed high-
frequency anpIiludes, lhe resonanl pressure is significanlIy Iover lhan olher conlrilulions.
NeverlheIess, il nay sliII exerl a non-negIigilIe effecl on lhe overaII noise.

-90 -60 -30 0 30 60 90
Crank AngIe, [deg]
0
25
50
75
100
P
r
e
s
s
u
r
e

[
b
a
r
]
TotaI Pressure
Motored
Combustion
Resonance
-20 -10 0 10 20
Crank AngIe [deg]
-0.4
-0.2
0
0.2
0.4
0.6
P
r
e
s
s
u
r
e

[
b
a
r
]
Resonance Pressure

Iig. 5. Deconposilion of lhe lolaI pressure in nolored, conluslion and resonance
conlrilulions.

The lhree deconposed pressures are uliIized lo conpule lvo characlerislic indices I
1
and I
2

defined as:

(
(
(
(
|
.
|
\
|
|
.
|
\
|
+ |
.
|
\
|
=
PRW
FRPE FRPE
LGOH
GW
GS
GW
GS
GW
GS
Q
Q
,
max
2 max 1 max
1
(4)


(
(
|
|
.
|
\
|
=
)
)
GW S
GW S
,
PRW
UHV
2
2
6
10 2
10 log (5)

The I
1
index is a funclion of lhe naxinun pressure gradienl of lhe conluslion conlrilulion,
occurring afler lhe piIol (dp
nax1
/dl)c
onl
and lhe nain injeclion (dp
nax2
/dl)
conl
. The I
1
index
is aIso non-dinensionaIized over lhe naxinun pressure gradienl of lhe pseudo-nolored
pressure (dp
nax
/dl)
nol
. In lhe case of a singIe-shol injeclion, a unique lern is of course
presenl in lhe eq. (4) nuneralor. The I
2
index lakes inlo accounl lhe acouslic energies (}p
2
dl)
associaled vilh resonance and nolored pressure signaIs. An addilionaI index In is finaIIy
defined accounling for nechanicaI noise conlrilulion, reIaled, as slaled, lo lhe soIe engine
speed:

n
i|c
n
|
n
1O
Iog
| |
=
|
\ .

(6)

n leing lhe engine speed and n
idIe
lhe idIe rolalionaI speed.
asing on lhe alove definilions, lhe OveraII Noise (ON) can le finaIIy conpuled as:

2 2 1 1 0
, & , & , & & 21
Q Q
+ + + = (7)

C
i
leing proper luning conslanls, depending on lhe engine archileclure and size.
IoIIoving lhe reIalions (4-7), a MalIal rouline vas deveIoped lo properIy process lhe in-
cyIinder pressure cycIe and conpule lhe various noise indices and lhe overaII noise.
The nelhod is vaIidaled on acouslic neasurenenls laken on a connerciaI engine, as
reporled in figure 6. The agreenenl ollained is salisfaclory al each engine speed. A
naxinun alsoIule error of aloul 1.3 d is found al a nediun engine regine.

1200 1600 2000 2400 2800 3200
Engine Speed [rpm]
90
95
100
105
110
115
O
v
e
r
a
I
I

N
o
i
s
e

[
d
B
]
Exp
CaIc

Iig. 6. Conparisons on lhe overaII noise
Fuel njection 152

The nelhod can le appIied lo lolh experinenlaI and nunericaI pressure cycIes and slriclIy
depends on lhe engine operaling condilions and injeclion slralegy. Once vaIidaled, lhis
sinpIified approach is direclIy incIuded in lhe oplinizalion Ioop lo predicl lhe overaII noise.

OptImIzatInn prnccss: lhe lriefIy descriled 1D, 3D and acouslic looIs are coupIed logelher
vilhin an oplinizalion Ioop searching for design or conlroI paranelers nininizing fueI
consunplion, gaseous enissions and radialed noise. The IogicaI deveIopnenl of lhe
oplinizalion prolIen is deveIoped vilhin lhe ModeIRONTILR
TM
environnenl. Ior each
sel of lhe design or conlroI paranelers, lhe 1D, 3D and acouslic looIs are aulonalicaIIy
slarled. 1D resuIls aIIov lo run lhe 3D code fron reIialIe condilions al lhe inlake vaIve
cIosure. Then, lhe 3D conpuled pressure cycIe is aulonalicaIIy given in inpul lo a MalIal
TM

rouline conpuling lhe overaII conluslion noise. SinuIlaneousIy, lhe Indicaled Mean
effeclive pressure (IMLI), is relurned lack lo lhe oplinizer, logelher vilh lhe NO and sool
IeveIs al lhe end of lhe 3D run. A nuIli-oljeclive oplinizalion is so defined lo
conlenporariIy search lhe naxinun IMLI, lhe nininun sool, lhe nininun NO and lhe
nininun overaII noise. To soIve lhe alove prolIen, genelic aIgorilhns (Sasaki, 2OO5) are
usuaIIy uliIized, enpIoying a range adaplalion lechnique lo overcone line-consuning
evaIualions. As usuaI in nuIli-oljeclive oplinizalion prolIens, a nuIlipIicily of soIulions is
expecled, leIonging lo lhe so-caIIed Iarelo fronliers. In order lo seIecl a singIe oplinaI
soIulion anong lhe Iarelo-fronlier ones, lhe MuIli Crileria Decision Making looI (MCDM)
provided in nodeIRONTILR
TM
is enpIoyed. This aIIovs lhe definilion of preferences
expressed ly lhe user lhrough direcl specificalion of allrilules of inporlance (veighls)
anong lhe various oljeclives. Depending on lhese reIalions, lhe MCDM looI is alIe lo
cIassify aII lhe soIulions vilh a rank vaIue. The highesl rank soIulion is lhe one lhal leller
salisfies lhe preference sel.

In lhe foIIoving paragraphs, lvo exanpIes are presenled vhere lhe descriled nelhodoIogy
is appIied lo perforn lhe design of a Tvo-Slroke Lngine for aeronaulicaI appIicalion and lo
seIecl an oplinaI fueI injeclion slralegy for a Iighl-duly aulonolive engine.

3. OptimaI Design of a Two-Stroke Engine for aeronauticaI appIication
In lhis paragraph, sone aspecls concerning lhe deveIopnenl of a prololype of a dieseI
engine suilalIe for aeronaulicaI appIicalions are discussed (Siano el aI., 2OO8). The engine
ained al achieving a veighl lo pover ralio equaI lo one kg/kW (22O kg for 22O kW) is
conceived in a lvo slroke UnifIov configuralion and is consliluled ly six cyIinders
dislriluled on lvo paraIIeI lanks. asing on a firsl choice of sone geonelricaI and
operalionaI dala, a preIininary fIuid-dynanic and acouslic anaIysis is carried oul al lhe sea
IeveI. This incIudes lhe engine-lurlocharger nalching, lhe eslinalion of lhe scavenging
process efficiency, and lhe sinuIalion of lhe spray and conluslion process, arising fron a
Connon RaiI injeclion syslen. olh 1D and 3D CID nodeIs are enpIoyed.
A CAD of lhe engine under invesligalion is shovn in figure 7. Six cyIinders are dislriluled
on lvo paraIIeI lanks vilh separale air adnission. The supercharging syslen consisls of a
dynanicaI lurlocharger coupIed lo a nechanicaI one (of lhe rools lype), serving lhe engine
slarl-up, as veII. An aulonolive derived rools conpressor is chosen vilh a lransnission-
ralio equaI lo 5. As a firsl slep, a preIininary 1D sinuIalion of lhe enlire propuIsion syslen

is reaIized ly neans of lhe previousIy descriled 1D soflvare, and ly expIoiling geonelricaI
infornalion derived ly lhe engine CAD. Iigure 8 reporls lhe engine 1D schenalizalion
incIuding lhe lhree cyIinders, lhe lurlocharger group (C-T), lhe inlercooIer (IC) and lhe
nechanicaI supercharger (C), coupIed lo lhe engine shafl. A vasle-gale vaIve (Y) is aIso
considered upslrean of lhe lurline. HaIf engine is schenalized, due lo lhe synnelry
properly of lhe lvo engine lanks. Lach of lhe lhree cyIinders is connecled lo lhe inlake
pIenun lhrough fourleen inIel porls and lo lhe exhausl pIenun lhrough lvo exhausl vaIves.
In lhe 1D conpulalion, lhe 3D conpuled discharge coefficienls are enpIoyed. Scavenging is
indeed considered as in lhe niddIe lelveen lhe lvo opposile Iinils so-caIIed of perfecl
dispIacenenl and perfecl nixing. In olher vords, a paraneler, vnix, represenling a reIalive
veighl faclor lelveen lhe occurrence of a perfecl nixing and a perfecl dispIacenenl
process, is assuned equaI lo lhe vaIue of O.67. The alove paraneler resuIls, once again, fron
accurale anaIyses carried oul on lhe engine cyIinder ly neans of lhe 3D code (figure 9).

Iig. 7. 3D cad viev of lhe 6 cyIinder, lvo-slroke DieseI engine.

A1
D1 E1 E3 E5
S0
G1 G2 G3 G4 G5
D3 D5
PA
C1
I1
C3
I3
C5
I5
ST
00 ambiente
Condotti I1, I3, I5 (14 per ogni cilindro)
Waste Gate
IC Intercooler
C
C A0
T
BY
00
00
TV
00
CM
IC
RM
A1
D1 E1 E3 E5
S0
G1 G2 G3 G4 G5
D3 D5
PA
C1
I1
C1
I1
C1
I1
C3
I3
C3
I3
C3
I3
C5
I5
C5
I5
C5
I5
ST
00 ambiente
Waste Gate
IC Intercooler
00 ambiente
Waste Gate
IC Intercooler
C
C A0
T
BY
00 00
00 00
TV
00 00
CM
IC
RM

Iig. 8. 1D schenalizalion of lhe AVIO3 engine

Fuel njection 154

Iig. 9. 3D anaIysis for lhe caIcuIalion of lhe scavenging efficiency
The 3D anaIysis aIso provides lhe definilion of a proper heal reIease Iav, lo le incIuded in
lhe 1D nodeI, assuning injeclion in one shol. Iigure 1O shovs lhe 3D conpuled pressure
cycIe in conparison vilh lhe resuIls of lhe 1D nodeI. An idea of hov lhe injeclion slralegy
affecls conluslion is aIso given. Il is evidenl lhal a loo advanced injeclion nakes for a loo
nuch high pressure peak, vhich nay le dangerous in lerns of nechanicaI slresses, vhereas
a Iale injeclion nakes for a Iov cycIe area, hence a Iov pover oulpul and high fueI
consunplion.

260 280 300 320 340 360 380 400 420 440 460
Crank angle ()
0
20
40
60
80
100
120
P
r
e
s
s
u
r
e

(
b
a
r
)
3D - SO 20BTDC
3D - SO 25 BTDC
3D - SO 15 BTDC
1D

Iig. 1O. Conparison of lhe in-cyIinder pressure as ollained ly lhe 1D and lhe 3D codes for
SOI al 2O TDC. The 3D sinuIalions are aIso reIevanl lo SOI al 15 and 25 TDC.

In paraIIeI lo lhe 1D and 3D anaIyses, an acouslic sludy is aIso carried oul lo predicl lhe
conluslion noise
radialion foIIoving lhe ILM/LM approach.


Iig. 11. Mesh nodeIs of lhe engine lIock and cyIinder Iiners.

In parlicuIar, lhe IL nodeI is deveIoped suldividing lhe engine inlo singIe groups, each
nanuaIIy neshed and finaIIy assenlIed. Tvo parls are nainIy considered, as shovn in
Iigure 11: lhe engine lIock and lhe cyIinder Iiners. A non aulonalic neshing process is
required lo handIe lhe greal conpIexily of lhe cyIinder geonelry especiaIIy concerning lhe
presence of lhe fourleen inIel porls. Wilh lhe purpose of gelling infornalion aloul lhe skin
surface vilralions, a frequency response anaIysis is conducled using as excilalion lhe 1D
conpuled pressure forces acling inside lhe cyIinders during lhe conluslion process al lhe
24OO rpn engine speed.

Iig. 12. HenisphericaI surface vilh fieId poinls and Sound Iover nap according lo lhe ISO
3746 direclive.

eside lhe caIcuIalion of lhe surface veIocily, a loundary eIenenl nesh is reaIised vilh a
reduced nunler of nodes and eIenenls. The ollained vilralionaI oulpul dala represenl lhe
loundary condilions lo le appIied lo lhe LM for lhe finaI evaIualion of lhe radialed sound
pover. The approach used is lhe ATV nelhodoIogy (Acouslic Transfer Veclors). This
lechnique, lhrough lhe preIininary evaIualion of lhe lransfer funclions of surface-receivers
(nicrophones), aIIovs lo evaIuale lhe ansver lo differenl loundary condilions, as lhe
Fuel njection 156

appIicalion of fine-Ioads or nuIli-frequency excilalions (engine noise). The acouslic
radialion can le so evaIualed fron lhe caIcuIalion of lhe sound pressure on a virluaI
neasurenenl surface lhal conpIeleIy conlains lhe radianl surface.
Ior lhe neasurenenl of lhe radialed pover, an henisphericaI surface is crealed around lhe
engine nodeI, according lo nornalive ISO 3746.
Iigure 12 shovs lhe alove surface, posilioned al lhe dislance of one neler fron lhe engine,
vhich incIudes lhe nineleen fieId poinls (virluaI nicrophones) used lo gel infornalion
aloul lhe noise radialion. In lhe sane figure lhe resuIling sound pressure nap is aIso
pIolled. In parlicuIar, il is possilIe lo nole lhal lhe najor conlrilulion lo lhe overaII noise
cones fron a IaleraI parl, corresponding lo lhe carler, vhich presenls a snaIIer lhickness
vilh respecl lo lhe olher engine parls. A non negIigilIe conlrilulion aIso cones fron lhe
engine lop, exciled ly lhe sulsequenl conluslion process evenls.
Iigure 13 dispIays lhe frequency speclrun of lhe average sound pover radialion on lhe
surface. Il is inporlanl lo renark lhe presence of lvo lonaI peaks al lhe frequencies of 44O
Hz (118 d) and 1O6O Hz corresponding lo a resonance phenonenon vilh lhe fundanenlaI
firing frequency al aloul 4O Hz (24OO rpn).
In concIusion, il can le slaled lhal a greal noise radialion is reveaIed in correspondence of
resonance condilions.
This kind of inlegralion of differenl nunericaI procedures aIIovs lo predicl, vilh a good
accuracy, lhe engine radialed noise and can le used in a pre-design phase in order lo
characlerize lhe acouslic lehaviour of lhe engine slruclure.

Iig. 13. Average Sound Iover radialion on lhe henisphericaI surface.

The ileralive exchange of infornalion lelveen lhe 1D and 3D codes aIIovs lo define lhe
nain perfornance oulpuls of lhe engine under deveIopnenl. AIlhough lhe nunericaI
anaIysis confirns lhe possiliIily lo reach lhe prescriled pover oulpul vilh lhe inposed
Iinilalion on lhe naxinun pressure (126 lar), il aIso puls inlo evidence lhe occurrence of a
high vaIue of lhe rake Specific IueI Consunplion (SIC = 258 g/kWh). The acouslic
anaIysis aIso eslinales lhe presence of a high conluslion noise IeveI, vilh a sound pover

peak of aloul 118 d, slriclIy reIaled, as knovn, lo lhe naxinun in-cyIinder pressure
gradienl reached during lhe conluslion process.

In order lo inprove lhe overaII perfornance characlerislics of lhe engine, an oplinizalion
procedure is carried oul lo lhe ain of finding a leller seIeclion of sone geonelricaI and
operaling paranelers. In parlicuIar, a differenl phasing of lolh exhausl vaIves and inlake
porls is considered, logelher vilh a differenl phasing of lhe injeclion Iav. The alove
paranelers acluaIIy affecl aIso lhe supercharging IeveI, and, for lhis reason, lhe 1D code
nusl le nandalory uliIized in lhe oplinizalion procedure. The 1D anaIysis, hovever,
incIudes lhe delaiIs of lhe previous 3D sludy in lerns of lolh scavenging efficiency,
discharge coefficienls and heal reIease rale.
Iigure 14 dispIays lhe Iogic charl of lhe oplinizalion procedure, deveIoped in lhe
ModeIronlier graphicaI environnenl. The independenl varialIes considered are:
- LVO: Lxhausl VaIve Opening, deg
- LVD: Lxhausl VaIve Duralion, deg
- LVL: Lxhausl VaIve Lifl, nn
- IIO: Inlake Iorl Opening, deg
- IIL: Inlake Iorl vidlh, nn
- TH}: Slarl of injeclion, deg

Iig. 14. Logic charl of lhe oplinizalion procedure.

Al each ileralion, lhe vaIues of lhe alove varialIes are aulonalicaIIy vrillen in lhe inpul fiIe
of lhe 1D code. ModeIronlier lhen runs lhe 1D code and exlracls lhe required oulpul
resuIls. Afler lhal, lhe independenl varialIes are ileraliveIy changed vilhin prescriled
inlervaIs lo lhe ain of finding lhe nininun fueI consunplion. AddilionaI oljeclives are
aIso specified concerning lhe nininizalion of lhe pressure gradienl and lhe nininizalion of
lhe naxinun average lenperalure inside lhe cyIinder. In lhis vay lolh noise and NOx
enissions are expecled lo le reduced. Of course, each sel of lhe independenl varialIes nusl
aIso guaranlee lhe possiliIily lo reach lhe prescriled pover oulpul (11O kW per lank) vilh
a naxinun pressure Iiniled lo 126 lar. These lvo addilionaI requirenenls are fuIfiIIed
lhrough lhe definilion of lvo conslrainl varialIes in lhe Iogic schene of figure 14.
Fuel njection 158

Sunnarizing, a nuIli-oljeclive conslrained oplinizalion prolIen is sel-up, as foIIovs:
Oljeclive 1: nin (SIC)
Oljeclive 2: nin (dp/dlhela
nax
)
Oljeclive 3: nin (T
nax
)
Conslrain 1: p
nax
< 126 lar
Conslrain 2: Iover > 1O8 kW
To soIve lhe alove prolIen, lhe ARMOCA aIgorilhn is uliIized. The Ialler leIongs lo lhe
calegory of genelic aIgorilhns and enpIoys a range adaplalion lechnique lo carry oul line-
consuning evaIualions.
The specificalion of 3 oljeclives delernines lhe exislence of a lvo-dinensionaI Iarelo
fronlier (Iarelo surface) incIuding aII lhe soIulions of lhe oplinizalion prolIen.
Differenl seclions of lhe Iarelo surface are represenled in figure 15 lhal highIighls lhe
presence of a cIear lrade-off lelveen lhe lhree specified oljeclives. Due lo lhe slrong
correIalion lelveen lhe naxinun pressure and naxinun lenperalure, a siniIar lrade-off
lehaviour is found lelveen lhe fueI consunplion and lhe naxinun pressure.
AII lhe dispIayed poinls, hovever, respecl lhe specified Conslrain 1. The iniliaI design poinl
ollained in lhe previousIy discussed preIininary sinuIalion, is Iocaled far avay fron lhe
Iarelo fronliers, as highIighled in lhe Iigure 15. A reIevanl inprovenenl of aII lhe lhree
oljeclives, hence, is sureIy reaIized.

230 240 250 260 270 280
BSFC, g/kWh
2
3
4
5
6
M
a
x
i
m
u
m

P
r
e
s
s
u
r
e

G
r
a
d
i
e
n
t
,

b
a
r
/
d
e
g
InitiaI Design
OptimaI
SoIution
230 240 250 260 270 280
BSFC, g/kWh
2040
2080
2120
2160
M
a
x
i
m
u
m

T
e
m
p
e
r
a
t
u
r
e
,

K
InitiaI Design
OptimaI
SoIution

Iig. 15. Oplinizalion resuIls. Trade-off anaIysis.

In order lo seIecl a singIe soIulion anong lhe ones Iocaled on lhe Iarelo fronliers, lhe MuIli
Crileria Decision Making looI (MCDM) provided in nodeIRONTILR
TM
is enpIoyed. Il
aIIovs lhe definilion of preferences expressed ly lhe user lhrough direcl specificalion of
allrilules of inporlance (veighls). SIC and pressure gradienl vere considered as lhe nosl
reIevanl paranelers. Depending on lhe alove reIalions, lhe MCDM looI is alIe lo cIassify aII
lhe soIulion vilh a rank vaIue. The soIulion vhich ollains lhe highesl rank, lherefore, can
le idenlified. asing on lhe descriled nelhodoIogy, lhe soIulion vilh lhe highesl rank vaIue

is lhe one characlerized ly lhe idenlificalion nunler (ID) 238. The Ialler is aIso depicled
aIong lhe Iarelo fronliers in Iigure 15.

DcsIgn ID 0 238 238-0
Input VarIab!cs Va!uc Va!uc Dc!ta
LVO, deg ATDC 8O.OO 94.17 14.17
LVD, deg 135.OO 155.96 2O.96
LVL, nn 12.OO 14.66 2.66
IIO, deg ATDC 111.5O 119.74 8.24
TH}, deg ATDC 347.49 351.69 4.2
IIL, nn 9.52O 12.126 2.6O6
Transfcr VarIab!cs Va!uc Va!uc Dc!ta
IID, deg 137.OO 12O.53 -16.47
LVC, deg ATDC 215.OO 25O.13 35.13
IIC, deg ATDC 248.5O 24O.27 -8.23
IIH, nn 26.98 2O.99 -5.99
ObjcctIvcs Va!uc Va!uc % Var
Min(SIC), g/kWh 257.98 233.O1 -9.68
Min(dI/dlh), lar/deg 5.665 3.676 -35.11
Min(Tnax), K 2136.3 2O73.5 -2.94
CnnstraInts Va!uc Va!uc Dc!ta
Inax < 126 lar 125.83 97.43 -28.4
Iover Oulpul > 1O8 kW 11O.O3 11O.OO ----
TalIe 1. Conparison lelveen iniliaI soIulion (ID=O) and gIolaI oplinun (ID=238)

The posilion of lhe oplinaI soIulion aIso puls inlo evidence lhal lhe MCDM procedure
effecliveIy reaIizes a conpronise lelveen lhe confIicling needs, quanlified ly lhe allrilules
of inporlance descriled. In addilion, lhis procedure defines a slandardized nelhod for lhe
seIeclion of lhe gIolaI oplinun.
TalIe 1 reporls a conparison lelveen lhe iniliaI and oplinaI soIulions in lerns of lolh
independenl (or inpul) varialIes, oljeclives paranelers and conslrainls. Sone olher
lransfer varialIes, direclIy derived fron lhe inpul dala, are aIso Iisled.
The lalIe puls inlo evidence lhal a SIC inprovenenl higher lhan 9 can le reached,
logelher vilh a reIevanl reduclion of lolh pressure gradienl, naxinun lenperalure and
naxinun pressure. This neans lhal lolh a Iover noise and NOx enission are expecled,
logelher vilh veII Iover lhernaI and nechanicaI slresses on lhe engine.
The alove resuIls are ollained lhanks lo a deIayed opening of lhe exhausl vaIve and lo an
increased duralion of exhausl phase. ConlenporariIy, a Iover heighl and a grealer vidlh of
lhe 14 inlake porls are aIso seIecled ly lhe oplinizalion procedure.

Fuel njection 160

-200 -100 0 100 200
Crank AngIe, deg
0
40
80
120
160
P
r
e
s
s
u
r
e
,

b
a
r
InitiaI Condition
OptimaI SoIution
-200 -100 0 100 200
Crank AngIe, deg
400
800
1200
1600
2000
2400
T
e
m
p
e
r
a
t
u
r
e
,

K
InitiaI Condition
OptimaI SoIution

Iig. 16. IniliaI and oplinaI pressure and lenperalure cycIes.

The deIayed opening of lhe exhausl vaIve aIso produces an increased expansion vork, as
cIearIy olservalIe in lhe in-cyIinder pressure cycIe pIolled in Iigure 16. The sane figure
highIighls lhal a very Iover pressure peak is ollained as a consequence of a Iover
supercharging IeveI and a deIayed injeclion slarl (see TH} varialIe in TalIe 1). SiniIar
consideralions can le drav Iooking al lhe average in-cyIinder lenperalure profiIe.
Despile lhe Iover loosl pressure, lhe nel shafl pover renains lhe sane, as requesled ly lhe
Conslrain 2, nainIy due lo a Iover nechanicaI energy alsorled ly lhe rools conpressor.

Il is vorlh pulling inlo evidence lhal each nodificalion lo lhe engine geonelry aIso
delernines a change in lhe operaling condilions in lerns of lhe super-charging IeveI. This,
logelher vilh a differenl pover alsorplion of lhe rools, requires a conlroI of lhe vasle-gale
opening in order lo reach lhe prescriled pover oulpul al lhe engine shafl. In lhis sense, lhe
oplinizalion design regards lhe vhoIe propuIsion syslen, since il keeps inlo accounl lhe
conpIex inleraclion lelveen lhe various engine conponenls.

4. OptimaI seIection of fueI injection strategies
for a Iight-duty automotive engine
In lhis paragraph, a 3D nodeIing and an oplinizalion procedure is appIied lo a naluraIIy
aspiraled Iighl-duly dieseI engine (5O5 cn3 dispIacenenl). The engine is equipped vilh a
nechanicaI IueI Injeclion Syslen (IIS) and is originaIIy designed for non-road appIicalions.
Slarling fron lhe alove lase engine, a nev prololype, equipped vilh a Connon RaiI (CR)
IIS, is deveIoped for leing inslaIIed on snaII cily-cars. The lehavior of lhe CR injeclion
syslen is firslIy experinenlaIIy anaIyzed, in order lo define lhe spray slruclure and injeclion
rale reaIized under differenl operaling condilions. As an exanpIe, in figure 17, lhe injeclion
rales reIaled lo lhree differenl Ioad condilions are conpared. They are neasured ly an AVL
Injeclion Cauge Rale Syslen vorking on lhe osch lule principIe. In addilion, experinenlaI

dala on lhe spray lip penelralion are avaiIalIe fron lhe anaIysis of lhe Iiquid fueI spray
inages, carried oul ly inage processing procedures (AIfuso el aI., 1999, di Slasio el aI.,
1999). These dala are enpIoyed lo vaIidale lhe spray nodeI in lhe 3D CID anaIysis (AIIocca
el aI. 2OO4).

0 500 1000 1500 2000 2500 3000 3500
Time, s
0
0.04
0.08
0.12
0.16
I
n
j
e
c
t
i
o
n

R
a
t
e
,

m
g
/
s
Low Load (Mf=3.40 mg, Pinj=28 MPa)
Medium Load (Mf=11.87 mg, Pinj=71 MPa)
High Load (Mf=26.35 mg, Pinj=140 MPa)

Iig. 17. LxperinenlaI injeclion rale of lhe CR-IIS.

Iigure 18 sunnarizes lhe resuIls of lhe preIininary nunericaI luning of lhe spray lreak-up
nodeI, ly conparing lhe experinenlaIIy neasured penelralion Ienglh and lhe nunericaI
resuIls. The Huh-Cosnan and lhe Wave nodeI are lolh lesled and luned ly a change in lhe
conslanls delernining lhe aerodynanic lreak-up line, C2 and C1. Lven vilh a vaIue of 4O
for lhe C2 conslanl, lhe Huh-Cosnan nodeI undereslinales lhe spray penelralion Ienglh,
vhereas quile reIialIe resuIls are achieved ly aclivaling lhe Wave nodeI vilh C1=6O.

0 100 200 300 400 500 600 700 800 900 1000
Time, s
0
10
20
30
40
50
60
T
i
p

P
e
n
e
t
r
a
t
i
o
n
,

m
m
ExperimentaI
NumericaI (Wave C
1
=60)
NumericaI (Wave C
1
=30)
NumericaI (Huh-Gosman C
2
=40)

Iig. 18. NunericaI and experinenlaI spray penelralion Ienglh.
Fuel njection 162

The luned spray nodeI is parl of a nore conpIele 3D CID anaIysis. Iigure 19 shovs a lop
viev of lhe unslruclured grids enpIoyed in lhe caIcuIalions.

Iig. 19. A lop viev of lhe grid al lhe DC, and a lollon viev of lhe grid al lhe TDC

During lhe 3D anaIysis, a lhree puIses injeclion slralegy is specified as shovn in Iigure 2O,
conpared lo lhe acluaI experinenlaI profiIe. Iive degrees of freedon - naneIy lhe slarl of
piIol injeclion (soip), lhe dveII line lelveen lhe firsl and second puIse (dveII_1), lhe dveII
line lelveen lhe second and lhird puIse (dveII_2), and lhe percenlages of fueI nass
injecled during lhe firsl lvo puIses - conpIeleIy define lhe overaII injeclion profiIe.

0 500 1000 1500 2000 2500
Time, s
0
0.02
0.04
0.06
0.08
I
n
j
e
c
t
i
o
n

R
a
t
e
,

m
g
/
s
ExperimentaI ProfiIe
Parameterized ProfiIe
1
2
3 4
8
10
9
piIot%_1
soip
dweII_1
piIot%_2
5
6 7
dweII_2

Iig. 2O. Iaranelric Injeclion slralegy al nediun Ioad

In lhis vay, ly varying lhe alove 5 paranelers, differenl conluslion deveIopnenls and
noxious enissions arises. Lach predicled pressure cycIe is aIso processed lo eslinale lhe
conluslion-radialed noise, vilh lhe sinpIified approach previousIy descriled.
The oplinizalion prolIen is sellIed in order lo idenlify lhe 5 conlroI paranelers vilh lhe
ain of sinuIlaneousIy nininizing fueI consunplion, poIIulanl enissions and radialed
noise. The IogicaI deveIopnenl of lhe oplinizalion prolIen vilhin lhe ModeIRONTILR
TM

environnenl is expIained in figure 21.

Iigure 22 dispIays lhe scaller charls of lhe 44O poinls conpuled aIong lhe oplinizalion
process, highIighling lhe conpIex inleraclions anong lhe various oljeclives. A cIear lrend
exisls lelveen lhe IMLI and lhe OveraII Noise. A grealer dispersion of lhe resuIls is found
Iooking al lhe lrade-off lelveen NO and sool nass fraclions.

Iig. 21. Logic schene of lhe oplinizalion process vilhin ModeIRONTILR

The MuIli Crileria Decision Making looI (MCDM) provided in nodeIRONTILR
TM
is
finaIIy enpIoyed lo seIecl singIe soIulions anong lhe ones reporled in figure 22.

0 1 2 3 4 5
HP-IMEP, bar
96
100
104
108
112
O
v
e
r
a
I
I

N
o
i
s
e
,

d
B
428
115
297
10
-6
10
-5
10
-4
10
-3
NO Mass Fraction
10
-5
10
-4
10
-3
S
o
o
t

M
a
s
s

F
r
a
c
t
i
o
n
115
428
297

Iig. 22. Scaller charls of lhe oplinizalion process

Three differenl soIulions are idenlified, lhe firsl one seIecling lhe IMLI and sool as lhe nosl
inporlanl paranelers (soIulions #297). In lhe second and lhird one, lhe inporlance of NO
enission and OveraII Noise are nore and nore increased (soIulions #428 and #115,
respecliveIy).
Fuel njection 164

Iigure 23 conpares lhe reIaled oplinaI injeclion slralegies, vhiIe figure 24 finaIIy shovs
lhe pressure cycIes, lhe heal reIease rales, and lhe NO and sool produclion. High IMLI and
Iov sool are ollained vilh a very advanced slarl of lolh piIol and nain injeclions (soIulions
#297). This slralegy delernines lhe highesl pressure peak and IMLI, vhiIe al lhe sane line
producing lhe highesl heal reIease peak and pressure gradienl, responsilIe of increased NO
anounls and Noise IeveI. AIlhough inlroducing sone IMLI and sool penaIizalion, a
radicaIIy Iover Noise IeveI and a leller NO enission is found vilh a deIayed soip and
snaIIer dveII lines (soIulion #428).

-120 -100 -80 -60 -40 -20 0 20
Crank AngIe, deg
0
0.02
0.04
0.06
0.08
I
n
j
e
c
t
i
o
n

R
a
t
e
,

m
g
/
s
# 297 (High IMEP & Low Soot)
# 428 (Intermediate Case)
# 115 (Reduced NO & Noise)

Iig. 23. OplinaI injeclion slralegies

-60 -40 -20 0 20 40 60
Crank AngIe, deg
0
20
40
60
80
100
P
r
e
s
s
u
r
e
,

b
a
r
0
40
80
120
R
a
t
e

o
f

H
e
a
t

R
e
I
e
a
s
e
,

J
/
d
e
g
# 297
# 428
# 115
Noise, HP-IMEP = 106.4 dB, 5.16 bar
101.1 dB, 4.89 bar
98.7 dB, 4.0 bar
Pressure
ROHR
-20 0 20 40 60
Crank AngIe, deg
10
-7
10
-6
10
-5
10
-4
10
-3
N
O

M
a
s
s

F
r
a
c
t
i
o
n
10
-7
10
-6
10
-5
10
-4
10
-3
10
-2
10
-1
S
o
o
t

M
a
s
s

F
r
a
c
t
i
o
n
# 297
# 428
# 115
NO
Soot

Iig. 24. Conparisons of lhe pressure cycIes, heal reIease, NO and sool enission

In lhis case a very snoolher heal reIease is found (fig. 24). The nain palh for a sulslanliaI
noise reduclion is indeed lhe specificalion of a nuch deIayed piIol and a sliII nore deIayed
nain injeclion (soIulion #115). olh lhese effecls, in facl, conlrilule in reducing lhe heal
reIease rale and lhe pressure gradienl during lhe conluslion process, even if a non-
negIigilIe nechanicaI oulpul Ioss has lo le expecled. This Iasl slralegy prolalIy represenls
lhe lesl conpronise soIulion in lerns of poIIulanl species produclion and radialed noise.

Hovever, leing lhe presenl resuIls ollained al a conslanl overaII injecled fueI nass, lhe
Iover pover oulpul ollained in lhis case direclIy inpacls on lhe specific fueI consunplion
and on lhe CO2 enission. This denonslrales lhe difficuIlies in idenlifying an oplinaI
injeclion slralegy, alIe lo conpIying vilh so nany confIicling needs, even vhen a highIy
noduIaled injeclion process is considered.
As a concIusive renark, lhe oplinizalion procedure is alIe lo caplure lhe expecled effecls of
lhe injeclion paranelers on lhe overaII perfornance and radialed noise and provides a
nelhod for lhe choice lhe lesl conpronise soIulion. The nelhodoIogy can le easiIy
exlended lo nuIlipIe operaling condilions and can incIude addilionaI varialIes (injeclion
pressure, svirI ralio, loosl pressure, LCR rale) and oljeclives (CO and HC produclion). In
lhis vay a varialIe conluslion node, Iike slandard prenixed-diffusive, HCCI (ression el
aI, 2OO8, ZavaIa el aI. 2OO1), ILTC (Shinazaki el aI. 2OO3, KaIghalgi, 2OOO9), elc. can le
reaIized, depending on operaling condilions and seIecled oljeclives.

5. ConcIusion
The presenl chapler presenled lvo differenl exanpIes for lhe design and anaIysis of nodern
CR equipped DieseI engines. A looI for nuIli-oljeclive oplinizalion vas found very usefuI
al lhe engine design slage lo reduce lhe line-lo-narkel of nev engine prololypes, lhanks lo
lhe possiliIily lo effecliveIy evaIuale lhe infIuence of geonelricaI and operaling engine
paranelers. In addilion, a siniIar lechnique vas appIied lo find lhe oplinaI seIeclion of lhe
conlroI paranelers (naneIy lhe injeclion slralegy) in order lo ollain an leller engine
lehaviour. To reach lhe alove oljeclives, differenl sinuIalion lechniques vere enpIoyed,
resorling lo 1D, 3D and acouslic anaIyses. More accurale or sinpIified approaches vere
presenled and differenlIy used aIong lhe differenl phases of lhe engine deveIopnenl
process. In each case, a conlinuous exchange of infornalion anong lhe various nelhods
aIIoved lo inprove lhe overaII sinuIalion accuracy and resuIls reIialiIily. The vhoIe
procedure hence represenls a very usefuI looI lo reduce lhe huge experinenlaI aclivily
usuaIIy required lo design and deveIop a nodern CR dieseI engine.

6. References
AIfuso, S., AIIocca, L., Corcione, I.L., di Slasio S. (1999) Inage Diagnoslics of Connon
RaiI DieseI Sprays LvoIving in Nylrogen Anlienl al Differenl Densilies, ICL99
InlernaI Conluslion Lngines: Lxperinenls and ModeIing.
AIfuso, S., AIIocca, L., Aurienna, M., Capulo, C., Corcione, I.L., Monlanaro, A., VaIenlino,
C. (2OO5) AnaIysis of a High Iressure DieseI Spray al High Iressure and
Tenperalure Lnvironnenl Condilions, SAL Iaper 2OO5-O1-1239
AIIocca, L., Corcione, I.L., Cosla, M. (2OO4) NunericaI and LxperinenlaI AnaIysis of
MuIlipIe Injeclion DieseI Sprays, SAL Iaper 2OO4-O1-1879
ression, C., SoIeri, D., Dehoux, S., AzouIay, D., Hanouda, H., Doradoux, L., Cuerrassi, N.
Lavrence, N.}. (2OO8) A Sludy of Melhods lo Lover HC and CO Lnissions in
DieseI HCCI, SAL Iaper 2OO8-O1-OO34
CoIin, O., and enkenida, A. (2OO4) "The 3-Zones Lxlended Coherenl IIane ModeI
(LCIM3Z) for Conpuling Irenixed/Diffusion Conluslion", OiI & Cas Science
and TechnoIogy, III, VoI. 59, N. 6, pp. 593-6O9
Fuel njection 166

Cosla, M., Siano, D., VaIenlino, C., Corcione, I.L., ozza, I. (2OO9) Irediclion and
Oplinizalion of lhe Ierfornances, Noxious Lnissions and Radialed Noise of a
Lighl Duly Connon-RaiI DieseI Lngine, proceedings of 9lh InlernalionaI
Conference on Lngines and VehicIes (ICL2OO9)
di Slasio, S., AIfuso, S., AIIocca, L., Corcione, I.L. (1999) LxperinenlaI Sludy on lhe
Alonizalion Mechanisn for IueI Sprays LvoIving in Alnospheres of Differenl
Nalure and Densily - InechL Seninar IulIicalion 1, 17, pp.241-255
Dukovicz, }.K. (198O) "A IarlicIe-IIuid NunericaI ModeI for Liquid Sprays", }. Conp.
Ihysics, 35, 229-253
KaIghalgi, C. (2OO9) Is CasoIine lhe esl IueI for Advanced DieseI Lngines` - IueI Lffecls
in Irenixed-Lnough conpression ignilion Lngines, Tovards CIean DieseI
Lngines, TCDL2OO9
Liu, A.. and Reilz, R.D. (1993) "ModeIing lhe Lffecls of Drop Drag and reak-up on IueI
Sprays", SAL 93OO72
O'Rourke, I.}. (1989) SlalislicaI Iroperlies and NunericaI InpIenenlalion of a ModeI for
DropIel Dispersion in TurluIenl Cas, }. Conpul. Ihysics 83
IapaIanlros, I.V., and WiIde, D.}.(2OOO) IrincipIes of OplinaI Design ModeIing and
Conpulalion, Canlrilde Universily Iress, Canlridge
Iayri, I., roalch, A., Tornos, ., Maranl, V., (2OO5) Nev nelhodoIogy for in-cyIinder
pressure anaIysis in direcl injeclion dieseI engines~appIicalion lo conluslion
noise, Meas. Sci. TechnoI. 16 54O-547 doi:1O.1O88/O957-O233/16/2/O29
Sasaki, D. (2OO5) ARMOCA, An efficienl MuIli-Oljeclive Cenelic AIgorilhn, TechnicaI Reporl
Shinazaki, N., Tsurushina, T., Nishinura, T. (2OO3) DuaI Mode Conluslion Concepl vilh
Irenixed DieseI Conluslion ly Direcl Injeclion Near Top Dead Cenler, SAL
2OO3-O1-O742
Siano, D., ozza, I., Cosla, M. (2OO8) OplinaI Design of a Tvo-Slroke DieseI Lngine for
AeronaulicaI AppIicalions Concerning lolh ThernofIuidynanic and Acouslic
Issues, IMLCL2OO8-68713, Iroceedings of 2OO8 ASML InlernalionaI MechanicaI
Lngineering Congress and Lxposilion, oslon, Massachusells, USA.
Slephenson, I.W. (2OO8) MuIli-Oljeclive Oplinizalion of a Charge Air CooIer using
nodeIRONTILR and ConpulalionaI IIuid Dynanics, SAL Iaper 2OO8-O1-O886
Slolz, M., Schonners, }., Duvinage, I., Ielrs, A., LIIvanger, S., Koynagi, K., CiIdein, H.
(2OOO) IolenliaI of Connon-RaiI Injeclion Syslen for Iassenger Car Di DieseI
Lngines, SAL Iaper 2OOO-O1-O944
Torregrosa, A.}., roalch, A., Marlin }., MoneIIella, L. (2OO7) Conluslion noise IeveI
assessnenl in direcl injeclion DieseI engines ly neans of in-cyIinder pressure
conponenls, Meas. Sci. TechnoI., 18 2131-2142, doi:1O.1O88/O957-O233/18/7/O45,
ZavaIa, I.A.C., Iinlo, M.C, IavaneIIo, R., Vaqueiro }. (2OO1) Conprehensive Conluslion
Noise Oplinizalion, Sae Iaper 2OO1-O1-15O9

X

Hydrogen fueIIed scramjet combustor
- the impact of fueI injection

Wei Huang
12
, Zhen-guo Wang
1
, Mohaned Iourkashanian
2
,
Lin Ma
2
, Derek .Inghan
2
, Shi-lin Luo
1
and }un Liu
1
1
Cc||cgc cf Acrcspacc an Ma|cria|s |nginccring, Na|icna| Unitcrsi|q cf Dcfcnsc
Tccnnc|cgq, Cnangsna, Hunan, Pccp|cs Rcpu||ic cf Cnina, 410073
2
Ccn|rc fcr C|D, Scncc| cf Prcccss, |ntircnmcn|a| an Ma|cria|s |nginccring,
Unitcrsi|q cf |ccs, Uni|c Kingcms, |S2 9]T

1. Introduction
The scranjel engine is one of lhe nosl pronising propuIsive syslens for fulure hypersonic
vehicIes. Over lhe Iasl fifly years lhe scranjel engine lechnoIogy has leen inlensiveIy
invesligaled and severaI such engines have leen fIighl-lesled in recenl years (NeaI, MichaeI,
& AIIan, 2OO5, IauI, Vincenl, Lual, & }eryI, 2OO4). Research on supersonic conluslion
lechnoIogies is of greal significance for lhe design of lhe engine and nany researchers pay
significanl allenlion lo lhe hypersonic airlrealhing propuIsion. The nixing and diffusive
conluslion of fueI and air in convenlionaI scranjel engines lake pIace sinuIlaneousIy in lhe
conluslor (Huang, Qin, Luo, & Wang, 2O1O). Since lhe inconing supersonic fIov can slay in
lhe conluslor onIy for a very shorl period of line, i.e. of lhe order of niIIiseconds (Aso,
Inoue, Yanaguchi, & Tani, 2OO9, Huang c| a|., 2O1O, Hyungseok, Hui, }aevoo, & Yunghvan,
2OO9), and lhe vhoIe process of conluslion has lo le conpIeled vilhin lhis shorl duralion,
lhis is a significanl reslriclion lo lhe design of lhe scranjel engine. In order lo soIve lhis
prolIen, hydrogen, one of lhe nosl pronising fueIs for lhe airlrealhing engine vilh ~1O
lines fasler reaclion lhan hydrocarlons, is videIy used in lhe scranjel conluslor.
In recenl years, a cavily fIanehoIder, vhich is an inlegraled fueI injeclion/fIane-hoIding
approach, has leen proposed as a nev concepl for fIane hoIding and slaliIizalion in
supersonic conluslors (AIejandro, }oseph, & Visvanalh, 2O1O, Chadvick c| a|., 2OO5,
Chadvick, SuIalh, & }anes, 2OO7, DanieI & }anes, 2OO9, Cu, Chen, & Chang, 2OO9, }eong,
O'yrne, }eung, & Houvong, 2OO8, Kyung, Seung, & Cho, 2OO4, Sun, Ceng, Liang, & Wang,
2OO9, Vikranadilya & Kurian, 2OO9). The presence of a cavily on an aerodynanic surface
couId have a significanl inpacl on lhe fIov surrounding il. The fIov fieId inside a cavily
fIanehoIder is characlerized ly lhe recircuIalion fIov lhal increases lhe residence line of
lhe fIuid enlering lhe cavily, and lhe cavily fIane provides a source of heal and radicaIs lo
ignile and slaliIize lhe conluslion in lhe core fIov.
Hovever, so far, lhe fIov fieId in lhe scranjel conluslor vilh nuIlipIe cavily fIanehoIders
has leen rareIy discussed, and lhis is an inporlanl issue as il can provide sone usefuI
guidance for lhe furlher design of lhe scranjel conluslor. MuIli-cavily fIanehoIder can
9
Fuel njection 168

produce Iarger drag forces on lhe scranjel conluslor, as veII as inprove lhe conluslion
efficiency of lhe conluslor. A laIance lelveen lhese lvo aspecls viII le very inporlanl in
lhe fulure design of lhe propuIsion syslen in hypersonic vehicIes. Al lhe sane line, lhe
conluslor configuralion, i.e. lhe divergence angIe of each slage, nakes a Iarge difference lo
lhe perfornance of lhe conluslor. Researchers have shovn lhal (Huang, Li, Wu, & Wang,
2OO9) lhe effecl of lhe divergence angIes of lhe poslerior slages on lhe perfornance of lhe
scranjel conluslor is lhe nosl inporlanl, and lhe effecl of lhe divergence angIe on lhe firsl
slage is lhe Ieasl inporlanl. When lhe Iocalion of lhe fueI injeclion noves forvard, lhe effecl
of lhe divergence angIe of lhe forner slages lecones nore inporlanl.
In lhis chapler, lhe lvo-dinensionaI coupIed inpIicil ReynoIds Averaged Navier-Slokes
(RANS) equalions, lhe slandard |- lurluIence nodeI (Huang & Wang, 2OO9, Launder &
SpaIding, 1974) and lhe finile-rale/eddy-dissipalion reaclion nodeI (Nardo, CaIchelli,
MongieIIo, Ciannarlini, & RufoIoni, 2OO9) have leen enpIoyed lo invesligale lhe effecl of
lhe Iocalion of lhe fueI injeclion on lhe conluslion fIov fieId of a lypicaI hydrogen-fueIed
scranjel conluslor vilh nuIli-cavilies.

2. PhysicaI modeI and numericaI method
The engine invesligaled adopls lhe singIe-expanded conluslor and fraclionaI conluslion
node, and il consisls of an isoIalor and lhree slaged conluslors, see Iig. 1. There are four
cavily fIane hoIders Iocaled on lhe upper and Iover vaIIs of lhe firsl and lhe second slaged
conluslors, respecliveIy. Hydrogen is injecled fron lhe sIol, Iocaled al 5nn fron lhe
Ieading edge of lhe four cavily fIane hoIders on lolh lhe upper and Iover vaIIs of lhe firsl
and lhe second slaged conluslor. The vidlh of lhe sIol is 1nn.
Assuning lhal lhe heighl of lhe isoIalor H
i
is 1 unil, lhe dislance lelveen lhe upslrean
forvard face of lhe cavily fIanehoIder in lhe upper vaII and lhal in lhe Iover vaII of each
slaged conluslor is O.183 aIong lhe x axis. The dinensions of lhe conponenls of lhe
scranjel conluslor are shovn in TalIe.1, vhere |
i
, |
c1
, |
c2
and |
c3
are lhe Ienglhs of lhe
isoIalor, lhe firsl slaged conluslor, lhe second slaged conluslor and lhe lhird slaged
conluslor, respecliveIy. The divergence angIes of lhe firsl slaged conluslor,
1
, lhe second
slaged conluslor,
2
and lhe lhird slaged conluslor,
3
are 2.O degree, 3.5 degree and 4.O
degree, respecliveIy.

Iig. 1. A schenalic of a lypicaI scranjel conluslor lhal has leen invesligaled.

Hi |i |c1 |c2 |c3 1/() 2/() 3/()
1.O 7.O 8.8 12.8 5.8 2.O 3.5 4.O
TalIe 1. CeonelricaI dinensions of lhe scranjel conluslor.

The prinary geonelry paranelers of lhe cavily fIanehoIder: lhe Ienglh of lhe cavily
fIanehoIder |=1.376, lhe heighl of lhe Ieading edge D
u
=O.275, lhe ralio of Ienglh-lo-heighl

|/D
u
=5.O, lhe svepl angIe =45 and lhe heighl of lhe lraiIing edge D
d
=O.275. A schenalic
diagran of a lypicaI cavily fIanehoIder lhal has leen invesligaled is shovn in Iig. 2.

Iig. 2. A schenalic of a lypicaI cavily fIanehoIder lhal has leen invesligaled.

TalIe.2 shovs lhe loundary condilions enpIoyed in lhe conpulalionaI fIuid dynanics
(CID) nodeIs. The ralio of lhe oxygen gas noI fraclion lo lhe nilrogen gas noI fraclion al
lhe enlrance of lhe conluslor is 23:77, vilh lhe Mach nunler leing 3.2, lhe lolaI pressure
2.9MIa and lhe lolaI lenperalure 15O5.OK. The hydrogen is injecled inlo lhe core fIov vilh
sonic veIocily, as shovn in TalIe.2. The slalic pressure and lenperalure of lhe injeclion are
1O6OKIa and 25OK, respecliveIy.

Ma Pe/KIa Te/K YN2 YO2 YH2
The enlrance of lhe conluslor 3.2 58.66 493.77 O.77 O.23 O
The exil of lhe injeclion 1.O 1O6O 25O O.O O.O 1.O
TalIe 2. oundary condilions for lhe nunericaI nodeI.

In lhe CID nodeI, lhe slandard |- lurluIence nodeI is seIecled. This is lecause of ils
roluslness and ils aliIily lo fil lhe iniliaI ileralion, design Ieclolype and paranelric
invesligalion. Iurlher, lecause of lhe inlense lurluIenl conluslion effecls, lhe finile-
rale/eddy-dissipalion reaclion nodeI is adopled. The finile-rale/eddy dissipalion nodeI is
lased on lhe hypolhesis of infinileIy fasl reaclions and lhe reaclion rale is conlroIIed ly lhe
lurluIenl nixing. olh lhe Arrhenius rale and lhe nixing rale are caIcuIaled and lhe snaIIer
of lhe lvo rales is used for lhe lurluIenl conluslion (ILULNT, 2OO6). WhiIe a no-sIip
condilion is appIied aIong lhe vaII surface, al lhe oulfIov aII lhe physicaI varialIes are
exlrapoIaled fron lhe inlernaI ceIIs due lo lhe fIov leing supersonic.

3. ModeI vaIidation
In order lo vaIidale lhe presenl nunericaI nelhod for conpuling lhese conpIex fIuid fIovs
in lhe scranjel conluslor vilh nuIli-cavilies, lhree conpulalionaI cases are invesligaled,
naneIy, lhe prolIens of an injeclion fIov, a cavily fIov and a fueI-rich conluslion fIov.
The grids for lhe geonelries are slruclured and generaled ly lhe connerciaI soflvare
Canlil, and lhe grids are dislriluled nore denseIy near lhe vaIIs and in lhe vicinily of lhe
shock vave generalion in order lo resoIve lhe loundary Iayers.

3.1 Injection fIow
In lhis firsl case, lhe physicaI nodeI lhal vas experinenlaIIy invesligaled ly Weidner el
aI.(Weidner & Drunnond, 1981) is enpIoyed since lhe nodeI has a good lvo-dinensionaI
slruclure and il can le used lo vaIidale lhe correclness of lhe injeclion phenonenon in lhe
scranjel conluslor.
Fuel njection 170

The experinenlaI lesl invesligales lhe phenonenon of lhe lraverse injeclion of heIiun inlo
paraIIeI air fIov, naneIy =9O, and lhe selup of lhe experinenl is schenalicaIIy shovn in
Iig. 3. The air slrean is inlroduced fron lhe Iefl hand side of a reclanguIar channeI vhich is
25.4cn Iong and 7.62cn high. The slalic pressure of lhe air slrean is P=O.O663MIa, lhe slalic
lenperalure is T=1O8.OK and lhe March nunler is M=2.9. The heIiun is injecled al sonic
condilion fron a O.O559cn sIol inlo an air slrean fron lhe lollon surface of lhe reclanguIar
channeI al a Iocalion vhich is 17.8cn dovnslrean fron lhe enlrance of lhe channeI. The
fIov condilions for lhe heIiun al lhe sIol exil are P=1.24MIa, T=217.OK and M=1.O.

Iig. 3. Schenalic of lhe physicaI nodeI invesligaled for injeclion fIov.

Iig. 4. Slalic pressure dislrilulion aIong lhe lollon vaII of lhe channeI for lhe differenl grid
syslens.

In order lo invesligale grid independency of lhe nunericaI sinuIalions, lhree sels of nesh
vilh differenl nunlers of ceIIs have leen enpIoyed, naneIy approxinaleIy 19,2OO, 38,O8O
and 76,23O ceIIs, respecliveIy. Iig. 4 shovs lhe slalic pressure dislrilulion aIong lhe lollon
vaII of lhe channeI for lhe lhree differenl grids. Il is olserved lhal lhe shock vave can le
caplured accuraleIy for aII lhree differenl grid scaIes, and lhe pressure dislrilulions aIong
lhe lollon vaII of lhe channeI in lhe dovnslrean region of lhe injeclion sIol are aInosl lhe
sane for lhe lhree grids enpIoyed. Wilh differenl grid scaIes, lhe Iocalion of lhe
disappearance of lhe reallachnenl region and lhe Iocalion of lhe generaled shock vave can

le predicled reasonalIy accuraleIy vhen conpared vilh lhe experinenlaI dala, see Iig. 5.
This neans lhal lhe difference in lhe lhree grid syslens enpIoyed in lhe sinuIalions nakes
onIy a snaII difference lo lhe nunericaI prediclions for lhe inleraclion lelveen lhe air
slrean and lhe injeclion.

Iig. 5 shovs a conparison lelveen lhe experinenlaI dala and lhe conpulalionaI
prediclions for lhe pressure aIong lhe lollon vaII. The reference pressure P
rcf
is O.O663MIa.
Il is olserved lhal lhe conpulalionaI resuIls ollained in lhis invesligalion shov good
quaIilalive agreenenl vilh lhe experinenlaI dala for lolh lhe upslrean and dovnslrean
regions of lhe injeclion.

Iig. 5. Conparison lelveen lhe experinenlaI dala of Weidner el aI. (Weidner &
Drunnond, 1981) and lhe predicled conpulalionaI pressures aIong lhe lollon vaII.

Iig. 6 shovs a conparison lelveen lhe experinenlaI dala and lhe predicled conpulalionaI
pressures al a dislance of 3.81cn dovnslrean of lhe injeclion sIol vhen lhe reference
pressure is O.21MIa and lhe reference heighl is 7.62nn. Il is olserved lhal lhere is a rapid
pressure drop al a dislance of aloul 1.524cn (i.e. y/h=O.2) fron lhe lollon vaII, and lhis is
lhe Iocalion vhere lhe separaled region disappears dovnslrean of lhe injeclion sIol. This
rapid pressure drop is foIIoved ly a pressure rise in lhe cenlraI region of lhe channeI, and
lhis is lhe inlerseclion poinl lelveen lhe shock vave and lhe lransverse Iine al lhis Iocalion.
Al lhe sane line, ve olserve lhal lhere are aIso sone discrepancies lelveen lhe
experinenlaI dala and lhe caIcuIaled resuIls lecause of lhe conpIex fIov fieId in lhe vicinily
of lhe injeclion exil and lhe inaccuracy of lhe k- lurluIenl nodeI lo sinuIale lhe separalion
region generaled jusl upslrean and dovnslrean of lhe injeclor.

Fuel njection 172

Iig. 6. Conparison lelveen lhe experinenlaI dala of Weidner el aI. (Weidner & Drunnond,
1981) and lhe conpulalionaI pressures al a dislance of 3.81cn dovnslrean of lhe injeclor.

The heIiun nass fraclion dislrilulion al a dislance of 3.81cn dovnslrean of lhe injeclor, as
ollained fron lhe conpulalionaI nodeI, agrees reasonalIy veII vilh lhe experinenlaI dala,
see Iig. 7, aIlhough lhere is a sIighl underprediclion ly lhe nunericaI sinuIalion. Il shouId
le noled lhal lhe heighl is nondinensionaIized ly lhe heighl of lhe channeI, naneIy

Iig. 7. Conparison lelveen lhe experinenlaI dala of Weidner el aI. (Weidner &
Drunnond, 1981) and lhe conpuled vaIue for lhe heIiun nass fraclion al a dislance of
3.81cn dovnslrean of lhe injeclor.

n=7.62cn.
Iron lhe resuIls presenled in Iigs. 5, 6 and 7, il is found lhal lhe nalhenalicaI and
conpulalionaI nodeI can reasonalIy accuraleIy sinuIale lhe inleraclion lelveen lhe air
slrean and lhe injeclion. In parlicuIar, lhe nodeI can caplure lhe shock vave and predicl
lhe paranelric dislrilulion. Therefore ve concIude lhal lhe nalhenalicaI and
conpulalionaI nodeI can le used vilh confidence lo invesligale lhe fIov fieId of lhe
scranjel conluslor.


3.2 Cavity fIow

Iig. 8. WaII slalic pressure dislrilulions for: (a) |/D=3 and no svepl angIe, (l) |/D=5 and
no svepl angIe, and (c) |/D=3 vilh lhe svepl angIe 3O.
Fuel njection 174

The second nodeI considered foIIovs lhe experinenlaI vork of Cruler el aI. (Cruler,
aurIe, Malhur, & Hsu, 2OO1) vho sludied severaI cavily configuralions for an unhealed
fIov al Mach 3. Cavilies vilh a deplh of 8.9nn vere used in lhe experinenlaI vork and for
lhe condilions of |/D=3, |/D=5 vilhoul a svepl angIe, and |/D=3 vilh lhe svepl angIe ()
of 3O, see Iig. 2. In addilion, lhe slagnalion lenperalure (T
O
) and slagnalion pressure (P
O
) of
lhe free slrean are 3OOK and 69OkIa, respecliveIy. This physicaI nodeI is used lo vaIidale
lhe correclness of lhe predicling fIov pasl lhe cavily fIanehoIder in lhe scranjel conluslor.
Iig. 8 shovs lhe vaII pressure dislrilulions for |/D=3, |/D=5 vilhoul a svepl angIe, and
|/D=3 vilh lhe svepl angIe 3O. Tvo sels of nesh, vilh differenl nunler of ceIIs, have
leen enpIoyed in order lo invesligale lhe grid independency of lhe nunericaI sinuIalions,
naneIy approxinaleIy 36,4OO and 147,2OO ceIIs have leen enpIoyed.
In Iig. 8, lhe effeclive dislance conprises of lhe cavily upslrean Ieading edge fron lhe
separalion corner, lhe cavily fIoor and lhe cavily lraiIing edge (Kyung c| a|., 2OO4). A good
agreenenl is olserved lelveen lhe conpuled and experinenlaI resuIls, and lhe difference
in lhe lvo nunlers of grids enpIoyed in lhe sinuIalions produces prediclion lhal nakes
aInosl no difference for lhe unhealed cavily fIov. We olserve lhal lhe nunericaI nelhod
enpIoyed in lhis invesligalion can le used vilh confidence lo sinuIale lhe fIov fieId of lhe
scranjel conluslor vilh nuIli-cavilies, and invesligale lhe effecl of lhe fueI injeclion
Iocalion on lhe perfornance of lhe scranjel conluslor.

3.3 FueI-rich combustion fIow fieId
The lhird nodeI considered foIIovs lhe experinenlaI configuralion and fIov condilions for
lhe case invesligaled ly Wang Chun el aI. (Wang, Silu, Ma, & Yang, 2OOO), and lhis nodeI is
used lo vaIidale lhe correclness of lhe conluslion nodeI enpIoyed in lhis invesligalion.
The geonelry consisls of a slraighl channeI vilh a Ienglh of 37Onn foIIoved ly a divergenl
channeI vilh a divergenl angIe of 3.6. There is a cIaploard lelveen lhe enlrance of lhe air
and lhe enlrance of hol gas, see Iig. 9, and lhe Ienglh of lhe cIaploard is 6nn. AII lhe
dinensions used in lhe CID nodeI are exaclIy lhe sane as in lhe experinenlaI
configuralion. The air and hol gas fIov condilions are presenled in TalIe.3.

Iig. 9. The geonelry of lhe conluslor invesligaled (Unil: nn)(Wang el aI., 2OOO).

IIov P
s
/MIa T
s
/K Ma Mass fraclion
C
2
H
4
O
2
CO
2
H
2
O N
2

Air O.O977 491.9 2.O9 - O.233O - O.O52O O.715O
Hol gas O.1731 1771.9 1.25 O.1O59 O.O1O3 O.12O5 O.1566 O.6O67
TalIe 3. Iaranelers al lhe enlrance of lhe supersonic conluslor(Wang c| a|., 2OOO).

ConpulalionaI sinuIalions have leen perforned vilh a coarse and a fine conpulalionaI
nesh consisling of 8,7OO (CID1) and 16,9OO ceIIs (CID2), respecliveIy. Iig. 1O shovs lhe
conparisons of lhe vaII pressure dislrilulions ollained fron lhe presenl CID caIcuIalions

and lhe experinenlaI dala of Wang Chun el aI. (Wang el aI., 2OOO). The soIid Iine represenls
lhe nunericaI resuIls fron lhe coarse nesh, CID1, and lhe dashed Iine is for CID2. Il can le
olserved lhal lhe slalic pressure dislrilulions on lhe lop and lollon vaIIs ollained ly lhe
CID resuIls shov good quaIilalive agreenenl vilh lhe experinenlaI resuIls. The CID
nodeI caplures lhe shock vave reasonalIy veII in lerns of lolh lhe Iocalion and slrenglh of
lhe vave syslen. The pressure dislurlance on lhe lop and lollon vaIIs is due lo lhe
conpression and expansion of lhe fIov lhal occurs aIlernaleIy in lhe nixing and expansion
seclions of lhe conluslor caused ly lhe shock vave syslen. Al lhe enlrance lo lhe nixing
seclion of lhe conluslor, due lo lhe differences in lhe fIov paranelers in lhe lvo supersonic
fIovs of air and hol slreans, and lhe effecl of lhe cIaploard, lhe expansion vave appears
during fIov expansions. When lhe lvo fIovs inlersecl, lhe fIov direclion changes, and lhe
lvo fIovs lecone conpressed (Silu, Wang, Niu, Wang, & Lu, 1999). Il is concIuded lhal lhe
CID approach used in lhis invesligalion can reasonalIy accuraleIy sinuIale lhese physicaI
phenonena in lhe scranjel conluslor.

Iig. 1O. WaII pressure conparisons of lhe CID caIcuIalions and lhe experinenlaI resuIls of
Wang Chun el aI. (Wang el aI., 2OOO): (a) lop vaII, and (l) lollon vaII.
Fuel njection 176

4. ResuIts and discussion
In order lo discuss lhe infIuence of lhe fueI injeclion Iocalion on lhe fIov fieId of lhe scranjel
conluslor vilh nuIlipIe cavily fIanehoIders, lhree sels of lhe fueI injeclion Iocalion are
enpIoyed in lhis invesligalion, naneIy, T
2
, T
4
and lolh T
2
& T
4
, in Iig. 1. The olher fueI
injeclion Iocalions are nol considered here, i.e. T
1
or T
3
, lecause pIacing lhe fueI injeclion
Iocalion cIoser lo lhe enlrance of lhe conluslor and nore concenlraled in a cerlain dislance
can le of nuch assislance in lhe oplinizalion of lhe perfornance of lhe conluslor, lul lhe
fueI injeclion Iocalion leing excessiveIy cIose lo lhe enlrance of lhe conluslor can cause lhe
inleraclion lelveen lhe isoIalor and lhe conluslor lo occur nore easiIy and push lhe shock
vave forvard, and lhis viII cause lhe inIel unslarl (Wu, Li, Ding, Liu, & Wang, 2OO7).
Iigs. 11-13 shov lhe paranelric conlours of lhe cases vilh lhe hydrogen injecled fron T
2
, T
4

and lolh T
2
& T
4
, respecliveIy. When lhe hydrogen is injecled fron lolh T
2
and T
4
, lhe shock
vave in lhe conluslor is pushed forvards inlo lhe isoIalor ly lhe inlense conluslion and a
high slalic pressure region forned lelveen lhe firsl upper cavily fIanehoIder and lhe
second upper cavily fIanehoIder, see Iig. 13 (a). Then if lhe fueI injeclion Iocalion noves
forvard, i.e. T
1
or T
3
, lhe shock vave is pushed oul of lhe isoIalor inlo lhe inIel and lhis
causes lhe inIel unslarl.
There exils a conpIex shock vave syslen in lhe conluslor. When lhe hydrogen is injecled
fron T
2
, lhe shock vaves generaled fron lhe Ieading edges of lhe firsl upper and Iover
cavily fIanehoIders inleracl and forn a high pressure region, see Iig. 11 (a). Al lhe sane
line, ve olserve lhal lhe high pressure region exisls nainIy in lhe vicinily of lhe injeclion
due lo lhe fueI conluslion. There is a Iov Mach nunler region generaled on lhe upper vaII
of lhe conluslor due lo lhe fueI injeclion, see Iig. 11 (l). MeanvhiIe, due lo lhe inleraclion
lelveen lhe shock vave and lhe loundary Iayer, lhere exisls a separalion region on lhe
Iover vaII of lhe conluslor, see Iig. 14 (a). The fueI injeclion nakes lhe vorlices in lhe
cavily fIanehoIder lecone Iarger and il defIecls inlo lhe core fIov. The shear Iayer forned
on lhe Ieading edge of lhe second upper cavily fIanehoIder inpinges on ils lraiIing edge,
and lhere are aInosl no vorlices in lhe firsl upper and Iover cavily fIanehoIders. The region
in lhe cavily fIanehoIders acls as a pooI lo provide lhe energy lo ignile lhe fueI and proIong
lhe residence line of lhe fIov in lhe conluslor. The Mach nunler in lhe cavily
fIanehoIders is nuch Iover lhan lhal in any olher pIace of lhe conluslor, excepl in lhe
separalion regions, see Iig. 11 (l), and lhe slalic lenperalure in lhe cavily fIanehoIders is
sIighlIy higher lhan lhal in lhe core fIov, see Iig. 11 (c). If ve change lhe geonelry of lhe
cavily fIanehoIder, il can acl as an ignilor in lhe scranjel conluslor, lul ve shouId


Iig. 11. Iaranelric conlours of lhe case vilh hydrogen injecled fron T
2
: (a) slalic pressure,
(l) Mach nunler, (c) slalic lenperalure, (d) H
2
nass fraclion, and (e) H
2
O nass fraclion.

Iig. 12. Iaranelric conlours of lhe case vilh hydrogen injecled fron T
4
: (a) slalic pressure,
(l) Mach nunler, (c) slalic lenperalure, (d) H
2
2
O nass fraclion.
Fuel njection 178

Iig. 13. Iaranelric conlours of lhe case vilh hydrogen injecled fron lolh T
2
and T
4
: (a) slalic
pressure, (l) Mach nunler, (c) slalic lenperalure, (d) H
2
2
O nass
fraclion.

consider lhe naleriaI of lhe cavily vhen operaling al such high lenperalures. Iurlher, lhe
conluslion of lhe hydrogen lakes pIace near lhe upper vaII of lhe conluslor, see Iig. 11 (d),
and lhe conluslion producl, naneIy, H
2
O nainIy dislrilules aIong lhe upper vaII. There is
aIso a snaII conluslion produclion in lhe firsl upper and Iover cavily fIanehoIders, see Iig.
11 (e), and il is lroughl forvard ly lhe recircuIalion zone.
When lhe hydrogen is injecled inlo lhe core fIov fron T
4
, lhe shock vave generaled fron
lhe Ieading edge of lhe firsl upper cavily fIanehoIder is nuch veaker lhan lhal generaled
fron lhe Ieading edge of lhe firsl Iover cavily fIanehoIder, and lhis nakes lhe shock vave,
afler lhe inleraclion, defIecl inlo lhe upper vaII of lhe conluslor. Iurlher, ve can olserve a
high pressure region generaled in lhe vicinily of lhe upper vaII, see Iig. 12 (a), and lhis is
differenl fron lhe case vilh lhe hydrogen injecled fron T
2
. The reason nay Iie in lhe
differences in lhe fueI injeclion Iocalions. Al lhe sane line, ve olserve lvo Iov Mach
nunler regions on lhe Iover vaII of lhe scranjel conluslor and lhis has leen caused ly lhe
recircuIalion zones, see Iig. 12 (l) and Iig. 14 (l), and lecause of lhe inleraclion of lhe shock
vave and lhe loundary Iayer, lhere aIso exisls a separalion area in lhe vicinily of lhe upper
vaII of lhe conluslor.
ecause of lhe varialion in lhe fueI injeclion Iocalion and lhe effecl of lhe shock vave, snaII
eddies are forned in lolh lhe upper and Iover cavilies of lhe firsl fIanehoIders, and il Iies
on lhe rear edge of lhe cavily, see Iig. 14 (l). The vorlices can acl as a recircuIalion zone for
lhe nixlure. Al lhis condilion, if lhe fueI is injecled fron lhe firsl slaged conluslor
sinuIlaneousIy, lhe perfornance of lhe conluslor viII le inproved since lhe residence line
is Ionger lhan in lhe case vhen lhe hydrogen is injecled fron T
2
. MeanvhiIe, lhe

dislrilulions of lhe fueI and lhe conluslion produclion are opposile lo lhe case vhen lhe
hydrogen is injecled fron T
2
, and lhey nainIy dislrilule aIong lhe Iover vaII of lhe scranjel
conluslor lecause of lhe fueI injeclion Iocalion, see Iig. 12(d) and (e). Due lo lhe fueI
injeclion leing lefore lhe cavily fIanehoIder, lhe eddy generaled in lhe second Iover cavily
fIanehoIder lecone Iarger lhan lefore, see Iig. 14 (l), naneIy lhe case vilhoul fueI injeclion
lefore lhe cavily fIanehoIder. The eddy is defIecled inlo lhe core fIov, and lhe shear Iayer
generaled al lhe Ieading edge of lhe second Iover cavily fIanehoIder inpinges on ils
lraiIing edge.

Iig. 14. SlreanIine dislrilulions in lhe scranjel conluslor vilh hydrogen injecled fron
differenl Iocalions: (a) T
2
, (l) T
4
, and (c) T
2
and T
4
.

When lhe hydrogen is injecled fron lolh T
2
and T
4
, lhe fIov fieId is lhe nosl conpIex in lhe
conluslor, see Iig. 13. Al lhis condilion, lhe shock vave is pushed oul of lhe conluslor
lecause of lhe inlense conluslion, and a Iarger Iov Mach nunler region is generaled on
lhe Iover vaII of lhe conluslor lecause of lhe slronger inleraclion lelveen lhe shock vave
and lhe loundary-Iayer, see Iig. 13 (l), and il spreads forvard lo lhe Iover vaII of lhe
isoIalor. A higher slalic pressure is ollained in lhe region lelveen lhe firsl and lhe second
cavily fIanehoIder, see Iig. 13 (a), and lhis is lhe nain cause for lhe spreading forvard of
lhe shock vave. Due lo lhe hydrogen injecled fron lolh T
2
and T
4
, lhe fueI and lhe
conluslion producl dislrilule lolh on lhe upper and Iover vaIIs of lhe conluslor, see Iig.
13 (d) and (e), and lhe conluslion occurs nainIy in lhe vicinily of lhe vaIIs. This iIIuslrales
lhal lhe injeclion pressure is nol high enough lo nake lhe fueI penelrale deeper. The
recircuIalion zone generaled al lhis condilion is nuch Iarger lhan lhal forned in lhe olher
lvo cases, and lhus lhe fIov can slay in lhe conluslor nuch Ionger, see Iig. 14(c). WhiIe
lraveIIing over lhe cavily, lhe injecled hydrogen inleracls vilh lhe slrong lraiIing edge shock
vave, vhich pIays an inporlanl roIe in lhe conluslion. The lraiIing edge shock vave can
inprove lhe slalic pressure and lhe slalic lenperalure of lhe fIov in lhe vicinily of lhe
lraiIing edge of lhe cavily fIanehoIder, and lhis can aIso lenefil lhe conluslion.

Fuel njection 180

5. ConcIusion
In lhis chapler, lhe lvo-dinensionaI coupIed inpIicil RANS equalions, lhe slandard |-
lurluIence nodeI and lhe finile-rale/eddy-dissipalion reaclion nodeI are inlroduced lo
sinuIale lhe conluslion fIov fieId of lhe scranjel conluslor vilh nuIlipIe cavily
fIanehoIders. The effecl of lhe fueI injeclion Iocalion on lhe fIov fieId of lhe conluslor has
leen invesligaled. We olserve lhe foIIoving:
z The nunericaI nelhods enpIoyed in lhis chapler can le used lo accuraleIy sinuIale
lhe conluslion fIov fieId of lhe scranjel conluslor, and predicl lhe deveIopnenl
slalus of lhe shock vave.
z The fueI injeclion Iocalion nakes a Iarge difference lo lhe conluslion fIov fieId of
lhe scranjel conluslor vilh nuIlipIe cavily fIanehoIders. The fIov fieId for lhe
case vilh hydrogen injecled fron lolh T
2
and T
4
is lhe nosl conpIex, and in lhis
silualion lhe shock vave has leen pushed forvard inlo lhe isoIalor. This causes lhe
loundary Iayer lo separale, generales a Iarge recircuIalion zone and reduces lhe
enlrance region of lhe infIov. If lhe fueI injeclion Iocalion noves sIighlIy forvard,
lhe shock vave nay le pushed oul of lhe isoIalor, and inlo lhe inIel. This viII do
danage lo lhe inIel slarl.
z The fueI injeclion Iocalion changes lhe generalion process of lhe vorlices in lhe cavily
fIanehoIders lo sone exlenl. When lhe hydrogen is injecled fron T
2
, lhere is no
vorlex fornalion in lolh lhe upper and Iover cavily of lhe firsl fIanehoIder. When
lhe hydrogen is injecled fron T
4
, snaII eddies are generaled in lhe firsl upper and
Iover cavily fIanehoIders. Iurlher, if lhe hydrogen is injecled fron lolh T
2
and T
4
,
lhe eddies in lhe firsl upper and Iover cavily fIanehoIders lecone Iarger, and lhis is
due lo lhe spread of lhe shock vave pushed ly lhe higher slalic pressure lecause of
lhe nore inlense conluslion.
z The fueI injeclion varies lhe dinension of lhe eddy generaled in lhe nearly cavily
fIanehoIder. Due lo lhe fueI injeclion, lhe eddy generaled in lhe nearly cavily
fIanehoIder lecones Iarger, over lhe cavily and defIecls inlo lhe core fIov. This
nakes a Iarger recircuIalion zone lhan lhe case vilhoul fueI injeclion.
z The cavily is a good choice lo slaliIize lhe fIane in lhe hypersonic fIov, and il
generales a recircuIalion zone in lhe scranjel conluslor. Iurlher, if ils geonelry can
le designed properIy, il can acl as an ignilor for lhe fueI conluslion, lul lhe
naleriaI of lhe cavily fIanehoIder shouId le considered for operaling al lhose high
lenperalures.

6. AcknowIedgement
The firsl aulhor, W Huang vouId Iike lo express his sincere lhanks for lhe supporl fron lhe
LxceIIenl Craduale Sludenl Innovalive Irojecl of lhe NalionaI Universily of Defense
TechnoIogy (No.O7O1O1) and lhe Hunan IrovinciaI Innovalion Ioundalion for
Ioslgraduale (No.32O6). AIso he vouId Iike lo lhank lhe Chinese SchoIarship CounciI (CSC)
for lheir financiaI supporl (No. 2OO9611O36).


7. References
AIejandro, M. ., }oseph, Z., & Visvanalh, R. K. (2O1O). IIane slaliIizalion in snaII cavilies.
A|AA jcurna|, 48(1), 224-235.
Aso, S., Inoue, K., Yanaguchi, K., & Tani, Y. (2OO9). A sludy on supersonic nixing ly
circuIar nozzIe vilh various injeclion angIes for air lrealhing engine. Ac|a
As|rcnau|ica, 65, 687-695.
Chadvick, C. R., }anes, I. D., Kuang-Yu, H., }effrey, M. D., Mark, R. C., & CanpleII, D. C.
(2OO5). SlaliIily Iinils of cavily-slaliIized fIanes in supersonic fIov. Prccccings cf
|nc Ccm|us|icn |ns|i|u|c, 30, 2825-2833.
Chadvick, C. R., SuIalh, K. D., & }anes, I. D. (2OO7). VisuaIizalion of fIanehoIding
nechanisns in a supersonic conluslor using ILII. Prccccings cf |nc Ccm|us|icn
|ns|i|u|c, 31, 25O5-2512.
DanieI, }. M., & }anes, I. D. (2OO9). Conluslion characlerislics of a duaI-node scranjel
conluslor vilh cavily fIanehoIder. Prccccings cf |nc Ccm|us|icn |ns|i|u|c, 32, 2397-
24O4.
ILULNT, I. (2OO6). ||U|NT 6.3 Uscr's Guic. Lelanon, NH: IIuenl Inc.
Cruler, M. R., aurIe, R. A., Malhur, T., & Hsu, K. Y. (2OO1). IundanenlaI sludies of cavily-
lased fIanehoIder concepls for supersonic conluslors. ]curna| cf Prcpu|sicn an
Pcucr, 17(1), 146-153.
Cu, H.-l., Chen, L.-h., & Chang, X.-y. (2OO9). LxperinenlaI invesligalion on lhe cavily-lased
scranjel nodeI. Cnincsc Scicncc 8u||c|in, 54(16), 2794-2799.
Huang, W., Li, X.-s., Wu, X.-y., & Wang, Z.-g. (2OO9). Configuralion effecl anaIysis of
scranjel conluslor lased on lhe inlegraI laIancealIe nelhod. ]curna| cf
As|rcnau|ics, 30(1), 282-286.
Huang, W., Qin, H., Luo, S.-l., & Wang, Z.-g. (2O1O). Research slalus of key lechniques for
shock-induced conluslion ranjel (shcranjel) engine. SC||NC| CH|NA
Tccnnc|cgica| Scicnccs, 53(1), 22O-226.
Huang, W., & Wang, Z.-g. (2OO9). NunericaI sludy of allack angIe characlerislics for
inlegraled hypersonic vehicIe. App|ic Ma|ncma|ics an Mccnanics(|ng|isn |i|icn),
30(6), 779-786.
Hyungseok, S., Hui, }., }aevoo, L., & Yunghvan, . (2OO9). A sludy of lhe nixing
characlerislics for cavily sizes in scranjel engine conluslor. ]curna| cf |nc Kcrcan
Sccic|q, 55(5), 218O-2186.
}eong, L. }., O'yrne, S., }eung, I. S., & Houvong, A. I. I. (2OO8). Invesligalion of supersonic
conluslion vilh angIed injeclion in a cavily-lased conluslor. ]curna| cf Prcpu|sicn
an Pcucr, 24(6), 1258-1268.
Kyung, M. K., Seung, W. ., & Cho, Y. H. (2OO4). NunericaI sludy on supersonic conluslion
vilh cavily-lased fueI injeclion. |n|crna|icna| ]curna| cf Hca| an Mass Transfcr, 47,
271-286.
Launder, . L., & SpaIding, D. . (1974). The nunericaI conpulalion of lurluIenl fIovs.
Ccmpu|cr Mc|ncs in App|ic Mccnanics an |nginccring, 3(2), 269-289.
Nardo, A. D., CaIchelli, C., MongieIIo, C., Ciannarlini, S., & RufoIoni, M. (2OO9). C|D
mcc|ing cf an cxpcrimcn|a| sca|c mcc| cf a |rappc tcr|cx ccm|us|cr. Iaper presenled
al lhe LCM 2OO9 Iourlh Luropean conluslion neeling, Vienna, Auslria.
Fuel njection 182

NeaI, L. H., MichaeI, K. S., & AIIan, I. (2OO5). ||ign| a|a ana|qsis cf HqSnc| 2. Iaper presenled
al lhe 13lh AIAA/CIRA InlernalionaI Space IIanes and Hypersonic Syslens and
TechnoIogies Conference, USA.
IauI, L. M., Vincenl, L. R., Lual, T. N., & }eryI, R. H. (2OO4). NASA hypersonic fIighl
denonslralors-overviev, slalus, and fulure pIans. Ac|a As|rcnau|ica, 55, 619-63O.
Silu, M., Wang, Z.-c., Niu, Y.-l., Wang, C., & Lu, H.-p. (1999). Invesligalion of supersonic
conluslion of hydrocarlon fueI-riched hol gas. ]curna| cf Prcpu|sicn Tccnnc|cgq,
20(6), 75-79.
Sun, M.-l., Ceng, H., Liang, }.-h., & Wang, Z.-g. (2OO9). Mixing characlerislics in a
supersonic conluslor vilh gaseous fueI injeclion upslrean of a cavily fIanehoIder.
||cu Tur|u|cncc Ccm|us|, 82, 271-286.
Vikranadilya, N. S., & Kurian, }. (2OO9). Iressure osciIIalions fron cavilies vilh ranp. A|AA
jcurna|, 47(12), 2974-2984.
Wang, C., Silu, M., Ma, }.-h., & Yang, M.-I. (2OOO). NunericaI sinuIalion on supersonic
conluslion of fueI-rich hol gas. ]curna| cf Prcpu|sicn Tccnnc|cgq, 21(2), 6O-63.
Weidner, L. H., & Drunnond, }. I. A. (1981). Paramc|ric s|uq cf s|agc fuc| injcc|cr
ccnfigura|icns fcr scramjc| app|ica|icns. Iaper presenled al lhe 17lh
AIAA/SAL/ASML }oinl IropuIsion Conference, Uniled Slales.
Wu, X.-y., Li, X.-s., Ding, M., Liu, W.-d., & Wang, Z.-g. (2OO7). LxperinenlaI sludy on effecls
of fueI injeclion on scranjel conluslor perfornance. Cnincsc ]curna| cf Acrcnau|ics,
20(6), 488-494.

Plasma ame sustained by microwave and burning hydrocarbon fuel: ts applications 183
X

PIasma fIame sustained by microwave and
burning hydrocarbon fueI: Its appIications

YongcheoI Hong
1
and Han Sup Uhn
2

1
Na|icna| |usicn Rcscarcn |ns|i|u|c
2
Kuanguccn Unitcrsi|q
1,2
Kcrca

1. Introduction
ThernaI pIasna lorches have leen deveIoped for various induslriaI appIicalions. Induslries
require lhen lo le high pover, conlaninanl-free, Iov-nainlenance, Iov-cosl, and Iarge-
voIune. IrincipaIIy, lhe pIasna lorch is a device lo produce an arc pIasna coIunn lelveen
lvo eIeclrodes. There are severaI kinds of pIasna lorches, incIuding dc arc lorch, induclion
lorch, and high-frequency capacilive lorch. The dc arc lorch is operaled ly lhe dc eIeclric
fieId lelveen lvo eIeclrodes al a severe environnenl of high arc currenl in lhe range fron
severaI lens lo lhousands of anperes. Therefore, lheir eIeclrodes are repIaced oflen due lo
lheir Iiniled Iifeline, in parlicuIar an oxidalive environnenl. AInosl aII radio frequency
lorches are inducliveIy coupIed discharges. Their lypicaI lhernaI efficiencies ( of pover
effecliveIy dissipaled in lhe pIasna forning gas) are in lhe range of 4O-5O (Iauchais &
VardeIIe, 1997). These convenlionaI lorches aIso have a snaII voIune of pIasna, high
operalionaI cosl and require nany expensive addilionaI syslens for operalion. AIlhough lhe
convenlionaI pIasna lorches are used in nany induslriaI appIicalions, lhe vide acceplance
of lhese processes is Iiniled ly econonic, conpelilive, reIialiIily, and olher concerns. Iron
lhe reason alove-nenlioned, lhey nay nol le usefuI in environnenlaI appIicalions.
In order lo overcone prolIens reIaled vilh lhe convenlionaI pIasna lorch, an eIeclrodeIess
nicrovave pIasna lorch al alnospheric pressure vas deveIoped (Hong el aI., 2OO3, Kin el
aI., 2OO3). Microvave pIasnas operaled al lhe alnospheric pressure, especiaIIy vaveguide-
lased pIasna, have leen suljecl of increased allenlion during lhe Iasl decade (Margol, 2OO1,
Moisan el aI., 2OO1). Such an inleresl cones fron lheir polenliaI and acluaI use in various
appIicalions, incIuding excilalion sources for eIenenlaI anaIysis, Iighling, and purificalion
or renedialion of gas effIuenls delrinenlaI lo lhe environnenl (Harlz el aI., 1998, Woskov &
Haddi, 1999). The nicrovave pIasna lorch can easiIy le nade ly nodifying lypicaI
househoId nicrovave ovens as inexpensive nelhod (Kin el aI., 2OO3). Therefore, lhe
nicrovave pIasna lorch is sinpIe, conpacl and econonicaI. Iurlhernore, in previous
vorks, lhe nicrovave pIasna lorch has leen invesligaled in various appIicalions such as
lhe alalenenl of CI
4
, NI
3
and SI
6
, lhe eIininalion of chenicaI and lioIogicaI varfare
agenls, and synlhesis of carlon nanolule, lilaniun dioxide, lilaniun nilride, and zinc oxide
(Hong el aI., 2OO4, Kin el aI., 2OO7). AIlhough lhe nicrovave pIasna lorch in air discharge
10
Fuel njection 184

provides high pIasna densily and high gas lenperalure in inexpensive vays (Creen, 2OO1),
lhe pIasna voIune and lenperalure of lhe nicrovave pIasna lorch decrease draslicaIIy
oulside lhe discharge lule, lherely Iiniling ils capaliIily of luIk lrealnenl of vasle. Ior
exanpIe, lhe gas fIov rales in lhe lrealnenl experinenls of CI
4
and phosgene vere onIy
lens of Iilers per ninule (Ipn), aIlhough high deslruclion and renovaI efficiency nore lhan
96 had leen acconpIished (Hong el aI., 2OO3, Hong el aI., 2OO5). In lhis conlexl, pIasna
fIanes nade of a nicrovave pIasna and a fueI lurning fIane have leen deveIoped for
producing an enIarged high-lenperalure pIasna fIane ly injecling a hydrocarlon fueI inlo
lhe nicrovave pIasna lorch. The conpIele conluslion of a hydrocarlon fueI in lhe
nicrovave pIasna lorch causes lhe increase of pIasna fIane voIune, high lenperalure
zone, and residence line of largel naleriaIs in high lenperalure zone.
There are nany experinenlaI invesligalions reIaled lo lhe fueI conluslion using arc pIasna
lorch and nicrovave. Since lhe experinenlaI research in lhe 197Os (Weinlerg el aI., 1978),
sludies on pIasna lorch ignilers have leen exlensiveIy conducled. The advanlages of a
pIasna lorch in conluslion are ils roIe as a source of a radicaI pooI and high lenperalure.
RecenlIy, Takila el aI. shoved lhal a consideralIe increase in lurning veIocily ly addilion of
radicaIs occurred onIy in lhe case vhen lhe nixlure lenperalure vas high and lhe nixlure
incIuded a Iarge nunler of radicaIs (Takila, el aI., 2OO1). AIso, Masuya c| a|. invesligaled lhe
ignilion of H
2
, H
2
/N
2
, H
2
/air, and O
2
/N
2
in high-lenperalure airfIov ly an arc lorch
(Masuya el aI., 2OO2). The effecls of nicrovave radialion on conluslion vere
experinenlaIIy invesligaled ly nany researchers. Ior inslance, Ogava el aI. invesligaled lhe
infIuence of nicrovaves on CH
4
/air Ianinar fIanes in a cavily resonalor exciled al 2.45
CHz ly neasuring lhe lurned-gas lenperalure, lrighlness, and eIeclron lenperalure
(Ogava el aI., 1998). IinaIIy, il is concIuded lhal lhe conluslion enhancenenl ly
nicrovave is due lo lhe nicrovave healing of lhe luIk gases in lhe fIane zone, and lhus
yieIding a grealer fIane lenperalure. Therefore, lhe nicrovave pIasna lurning syslen can
provide a near perfecl conluslion of a hydrocarlon fueI gas vilh air or oxygen gas due lo
lhe high-lenperalure pIasna and a Iarge quanlily of radicaIs in lhe nicrovave pIasna. Ior
inslance, in deslrucling fIuorinaled conpound gases (ICs) using a nicrovave pIasna lorch
(Hong el aI., 2OO3), lhe inpacl process of eIeclrons dissociales or ionizes IC noIecuIes, and
olher pIasna consliluenls converl inlo lenign or nore lrealalIe producls ly injecling inlo
lhe cenler parl of lhe nicrovave pIasna lorch fIane suslained fron any gas nixlure.
Hovever, lhe nicrovave lorch alaled lhe conlaninanls onIy in 1O-2O Ipn nilrogen gas
conlaninaled ly CI
4
, SI
6
, and NI
3
gases. The reason for lhe alalenenl Iinilalion of lhe
nicrovave pIasna lorch is due lo lhe snaII voIune of lhe lorch pIasna and lhe shorl
residence line of conlaninanls in lhe reaclor. AcluaIIy, lhe voIune of lhe pIasna fIane
suslained ly lhe nicrovave pIasna and lurning hydrocarlon fueI is nuch Iarger lhan lhal
of lhe nicrovave pIasna lorch. Therefore, lhe Iarge-voIune, high lenperalure pIasna
fIane is expecled lo overcone lhe alalenenl Iinilalion al high fIov rale.
This chapler conlrilules lhe conluslion enhancenenl of a hydrocarlon fueI augnenled ly
lhe nicrovave pIasna lorch. AIso, lhis chapler descriles lhe configuralion of pIasna fIane
generalor nade of a nicrovave pIasna and a fueI-lurning fIane, invesligales physicaI
properlies conlaining ils lenperalure and oplicaI enission, and IaslIy shovs lhe
experinenlaI resuIls in alaling fIuorinaled conpound gases, deconlaninaling chenicaI and
lioIogicaI varfare agenls, and eIininaling odorous chenicaI agenls.


2. Microwave pIasma torch
Iigure 1 dispIays a schenalic configuralion of lhe alnospheric pressure nicrovave pIasna
syslen. The design and operalion of lhe alnospheric nicrovave pIasna lorch are lriefIy
sunnarized here for conpIeleness, aIlhough lhey have leen reporled in delaiI in previous
arlicIes (Hong el aI., 2OO3, Kin el aI., 2OO3). Il is conprised of lhe 2.45 CHz nicrovave
generalor, WR-34O vaveguide conponenls, incIuding an isoIalor, a direclionaI coupIer, a 3-
slul luner, and a nicrovave pIasna lorch as a fieId appIicalor. The WR-34O vaveguide (86
nn 43 nn) used in lhe nicrovave pIasna lorch vas lapered lo a reduced cross-seclion
of 86 nn 2O nn. In lhe unIoaded vaveguide vilhoul pIasna, lhe reduclion of lhe
vaveguide heighl provides an increase in lhe eIeclric fieId slrenglh even vilh lhe sane
nicrovave pover. Due lo lhe Ioynling lheoren, ve gel aloul a
2
lines higher eIeclric
fieId slrenglh in lhe pIasna lorch region. The discharge lule vas inserled verlicaIIy,
perpendicuIar lo lhe vide vaII of lhe vaveguide. The discharge lule vas Iocaled al a 1/4
vaveIenglh avay fron lhe shorled end of lhe vaveguide. The highesl-inlensily eIeclric fieId
occurs al lhis Iocalion confirned ly a high frequency slruclure sinuIalor code (Kin el aI.,
2OO3). The nicrovave radialion generaled fron nagnelron passes lhrough lhe 3-slul luner,
is guided lhrough lhe lapered vaveguide, and enlers lhe discharge lule nade of fused
siIica. The cenler axis of lhe discharge lule vilh an ouler dianeler of 3O nn is Iocaled one-
quarler vaveIenglh fron lhe shorled end of lhe vaveguide. The lule penelrales lhrough lhe
vide vaveguide vaIIs, as shovn in Iig. 1. The ignilers nol shovn in Iig. 1 vilh lheir
lerninaI eIeclrodes inside lhe discharge lule iniliale lhe pIasna. The pIasna generaled
inside lhe discharge lule is slaliIized ly a svirI gas, vhich enlers lhe discharge lule
sidevays, crealing a vorlex fIov (CulsoI el aI., 1998) in lhe lule. The inpedance of lhe
pIasna and lhe fieId appIicalor lo lhe characlerislic inpedance of lhe WR-34O vaveguide
vere nalched ly luning lhe 3-slul luner. The refIecled pover adjusled ly lhe 3-slul luner
is aInosl zero. Lven vilh aII lhe luning sluls conpIeleIy vilhdravn, refIecled pover is
lypicaIIy Iess lhan 1O of lhe forvard pover (Hong el aI., 2OO3). The forvard and refIecled
povers are nonilored ly lhe direclionaI coupIer.

Iig. 1. Schenalic presenlalion of lhe 2.45 CHz nicrovave syslen conponenls and lhe
nicrovave pIasna lorch.
Fuel njection 186

A nunler of experinenlaI resuIls in alnospheric nicrovave pIasnas have leen reporled
(Creen el aI., 2OO1, Moon el aI., 2OO2). Ior exanpIe, Creen el aI.

neasured lhe lorch fIane
lenperalure inside a discharge lule ly naking use of lhe Ie I enission Iines in lhe 37O-377
nn range (Creen el aI., 2OO1). The nicrovave pIasna lorch is siniIar or aInosl sane vilh
ours. The lenperalure profiIes are aInosl fIal oul lo lhe Iargesl neasuralIe pIasna radius of
1O nn vilh a naxinun vaIue of 655O 35O K on axis al an air of 28 Ipn and 1.4 kW pover.
The fIane lenperalure al lhe 1O nn radius is sliII 8O of ils vaIue on axis. CeneraIIy, lhe
pIasna coIunn Ienglh in lhe nicrovave pIasna lorch depends on lhe anounl of svirI gas.
In previous Iileralure (Kin el aI., 2OO3), lhe pIasna coIunn Ienglh vas aloul 2O-3O cn for 1
kW nicrovave pover, for a discharge lule vilh 27 nn inner dianeler and for 2O Ipn of air
svirI gas. The pIasna coIunn Ienglh vas reduced lo 1O cn vhen lhe svirI gas increased
fron 2O lo 8O Ipn. The nicrovave pIasna lorch can le operaled in various gases. In lhis
conlexl, Iigs. 2(a)-(f) reveaI lhe nicrovave discharge pIasnas in 1O Ipn argon, 1 Ipn argon,
1O Ipn heIiun, 1O Ipn nilrogen, 1O Ipn air, and nixlure of 5 Ipn nilrogen and 1O Ipn
heIiun, respecliveIy.

Iig. 2. Various alnospheric pressure nicrovave pIasnas al (a) 1O Ipn argon, (l) 1 Ipn
argon, (c) 1O Ipn heIiun, (d) 1O Ipn nilrogen, (e) 1O Ipn air, and (f) nixlure of 5 Ipn
nilrogen and 1O Ipn heIiun. Then, lhe appIied nicrovave pover is approxinaleIy 1 kW.

3. PIasma fIame generator
3.1 Arrangement of pIasma fIame generator
Iigure 3(a) shovs lhe schenalic viev for lhe pIasna fIane generalor nade of lhe
nicrovave pIasna and a fueI-lurning fIane. The nain parls of experinenlaI configuralion
for lhe pIasna fIane generalor, as shovn schenalicaIIy in Iig. 3(a), consisl of lhe nicrovave

pIasna lorch, a fueI-injeclor, and a pIasna fIane exil. Air, oxygen or a nixlure of air and
oxygen can le used as a svirI gas. Therefore, lhe svirI gas provides alonic oxygen and
noIecuIar singIel oxygen of high-densily (Lai el aI., 2OO5) for near perfecl conluslion of
hydrocarlon fueIs, vhich is sprayed fron lhe fueI injeclor in Iig. 3(a). The fueI injeclor,
vhich is a lypicaI fueI nozzIe used in hone-loiIers, is equipped vilh lhe slainIess sleeI lule
and provides fueI for pIasna. The injeclor is inslaIIed conliguousIy lo lhe upper side of lhe
vaveguide, as shovn in Iig. 3(a). The inner dianeler of slainIess sleeI lule has lhe sane
inner size as lhe discharge lule and is inslaIIed on lhe lapered vaveguide lo suslain a
sleady vorlex fIov of lhe svirI gas. The hydrocarlon fueI injecled inlo pIasna nixes vilh
lhe svirI gas (air or oxygen) and exlends lhe pIasna fIane lo lhe open air, evaporaling
inslanlaneousIy and lreaking dovn lhe noIecuIar slruclure ly energelic eIeclrons and high
lenperalure. The lenperalures al differenl posilions of lhe pIasna fIane vere neasured ly
a lhernocoupIe of R lype. The narks L
O
~ L
15
in Iig. 3(a) indicale lhe lenperalure-
neasurenenl poinls. Ior exanpIe, lhe nark L
6
represenls a neasurenenl poinl 6 cn avay
fron lhe pIasna fIane exil. The nark L
-1.5
is lhe posilion corresponding lo a neasurenenl
poinl 1.5 cn avay lhe direclion of lhe vide-vaII vaveguide. Iig. 3(l) is lhe cross-seclionaI
viev of svirI generalor narked vilh dolled Iine in Iig. 3(a). The svirI gases are injecled
lovards langenliaI direclion lhrough four langenliaI hoIes. The hoIes for lhe svirI gases are
incIined lovards axiaI direclion ly 3O .

Iig. 3. (a) Schenalic viev shovs a pIasna fIane generalor vilh lhe nicrovave pIasna lorch.
The fueI injeclor vas inslaIIed adjacenl lo lhe upper side of lhe vaveguide. (l) Cross-seclionaI
viev dispIays a svirI gas generalor vilh four langenliaI hoIes for lhe dolled Iine in (a)

3.2 Temperature profiIe of pIasma fIames

3.2.1 PIasma fIame from kerosene
CeneraIIy, fIanes aIready conlain a veakIy ionized pIasna vilh lypicaI densily grealer lhan
1O
1O
ions/cn
3
. Ior exanpIe, lhe oxidalion of nelhanoI ly alonic oxygen is 1O niIIion lines
Fuel njection 188

fasler lhan lhal ly oxygen noIecuIes al lhe gas lenperalure of 13OO K (Uhn, 1999). If so, as
nenlioned earIier, lecause lhe nicrovave pIasna lorch has high pIasna densily of
~1O
13
/cn
3
in air discharge and high lenperalure of aloul 65OO K al lhe cenler axis (Creen el
aI., 2OO1), ve expecl lhal lhe nicrovave pIasna lorch can acconpIish near perfecl
conluslion of fueI. In lhis regard, Iigs. 4(a) and (l) shov lhe nicrovave pIasna-lurner
fIanes lefore and afler lhe fueI injeclion al lhe appIied nicrovave pover of 1.5 kW,
respecliveIy. In Iig. 4(a) and (l), a nixlure of 5O Ipn air and 1O Ipn oxygen as a svirI gas
vas injecled inlo lhe nicrovave pIasna lorch, vhiIe 5O Ipn air as a svirI gas and 1O Ipn
oxygen vilh a fueI lhrough lhe fueI injeclor vere injecled in Iig. 4(c). Iigure 4(a) is a piclure
of lhe pIasna lorch fIane vilhoul fueI injeclion. The fIane vas nol expanded lo lhe exil of
lhe slainIess sleeI lule of 1O cn in Ienglh. Hovever, as shovn in Iig. 4(l) and (c), lhe lurner
fIane shol oul lhrough lhe exil of lhe slainIess sleeI lule vhen O.O25 Ipn kerosene vas
injecled as a fueI inlo lhe nicrovave pIasna lorch. The lurner fIane dianeler and Ienglh
fron lhe fIane exil vere aloul 8 cn and 4O cn, respecliveIy. As a naller of facl, lhe fueI
injeclor in lhis vork vas inslaIIed jusl alove lhe vaveguide and vas 2 cn avay fron lhe
vaveguide excilalion region, as shovn in Iig. 3(a). When 1O Ipn oxygen gas vas added lo
lhe nicrovave pIasna lorch, nol shovn in Iig. 4, il vas olserved lhal lhe pIasna fIane
coIor changed fron a yeIIovish vhile lo a lIuish vhile, inpIying lhe phenonenon of near
perfecl conluslion.

Iig. 4. Microvave pIasna-lurner fIanes lefore (a) and afler (l and c) a fueI injeclion al lhe
appIied nicrovave pover of 1.5 kW. A nixlure of 5O Ipn air and 1O Ipn oxygen as a svirI
gas vas injecled inlo lhe nicrovave pIasna lorch in (a) and (l), vhiIe 5O Ipn air as a svirI
gas and 1O Ipn oxygen vilh a fueI lhrough lhe fueI injeclor vere injecled in (c) (Hong el aI.,
2OO6).

The lenperalure of lhe nicrovave pIasna lorch fIane al lhe cenler of lhe fIane exil (nark
L
O
in Iig. 3) neasured ly a lhernocoupIe vas onIy 55O K, vhen 6O Ipn air as a svirI gas
vas injecled. Hovever, lhe lenperalure of lhe lurner fIane vilh lhe addilion of O.O25 Ipn
kerosene draslicaIIy increased lo aloul 138O K. Moreover, lhe lenperalure of lhe lurner
fIane vilh lhe addilion of O.O25 Ipn kerosene and 1O Ipn oxygen gas draslicaIIy increased
lo aloul 17OO K. Tenperalure dislrilulions aIong lhe radiaI and axiaI direclions al differenl
kerosene fIov rales vere neasured vilh lhe addilion of oxygen. Iigure 5(a) shovs lhe
radiaI lenperalure profiIes al narks L
O
, L
3
, and L
6
vhen a nixlure of 4O Ipn air and 2O Ipn
oxygen as a svirI gas and O.O25 Ipn kerosene vere injecled. The Ienglh of lhe fIane vas
aloul 3O cn. Lach Iine narked ly lhe reclangIes, circIes, and lriangIes indicales lhe radiaI
lenperalure dislrilulion al O, 3, and 6 cn fron lhe slainIess sleeI lule. In conparison vilh
lhe lenperalure dislrilulions al L
3
and L
6
, lhe dislrilulion al L
O
decreases nore draslicaIIy

oving lo lhe snaII inner dianeler of lhe exil. CeneraIIy speaking, lhe voIune of lhe lurner
fIane decreases as oxygen fIov rale increases, vhiIe lhe fIane lenperalure al lhe cenler axis
increases. As shovn in Iig. 5(a), lenperalures al L
O
, L
3
, and L
6
dislrilule fron 175O lo 185O K,
reveaIing approxinaleIy uniforn lenperalure-dislrilulion in lhe axiaI direclion. Iigure 5(l)
presenls lhe axiaI lenperalure profiIes al differenl kerosene fIov rales, vhen a nixlure of 5O
Ipn air and 1O Ipn oxygen as a svirI gas vas injecled. Lach Iine narked ly lhe reclangIes,
circIes, and lriangIes indicales lhe axiaI lenperalure profiIes al kerosene fIov rales of O.O31,
O.O25, O.O19 Ipn, slarling fron lhe slainIess sleeI lule. The fueI injeclors of O.O31 and O.O25
Ipn spray lhe fueI in a shape of a conic sheII vilh 8O
o
spraying angIe. The injeclor of O.O19
Ipn sprays lhe fueI in lhe shape of a soIid cone vilh 6O
o
spraying angIe. As shovn in Iig.
5(l), il is olserved lhal lhe lrends of lhe lenperalure profiIes al axiaI posilions O and 3 cn
depend on lhe lype and angIe of fueI injeclion. The difference of lhe nixlure of svirI gas and
fueI in lhe slainIess sleeI lule causes differenl lrend in lenperalure profiIes. CeneraIIy
speaking, fIane lenperalure and voIune increase as lhe kerosene fIov rale increases.

Iig. 5. RadiaI lenperalure profiIe (a) vhen a nixlure of 4O Ipn air and 2O Ipn oxygen as a
svirI gas and O.O25 Ipn kerosene vere injecled inlo lhe nicrovave pIasna lorch and axiaI
lenperalure profiIe (l) al kerosene fIov rales of O.O31, O.O25, and O.O19 Ipn vhen a nixlure
of 4O Ipn air and 2O Ipn oxygen as a svirI gas vas injecled (Hong el aI., 2OO6).

3.2.2 PIasma fIame from dieseI

Iig. 6. DieseI nicrovave pIasna-lurner fIanes (a) lefore and (l and c) afler a fueI injeclion
al lhe appIied nicrovave pover of 1.2 kW. 5O Ipn air as a svirI gas and 1O Ipn oxygen vilh
O.O19 Ipn dieseI lhrough lhe fueI injeclor vere injecled, vhere (c) is a fronl viev of piclure
(l) (Hong & Uhn, 2OO6).
Fuel njection 190

In Iig. 6, lhe pIasna fIanes lefore and afler lhe injeclion of dieseI fueI vere conpared.
SiniIar lo lhe kerosene nicrovave pIasna lurner reporled in lhe previous vork (Hong, el
aI., 2OO6), lhe dieseI nicrovave pIasna fIane aIso shovs lhal lhe voIune is nore lhan 5O
lines lhal of lhe lorch pIasna, lurning dieseI fueI inslanlaneousIy. Iigure 6(a) is a piclure of
lhe nicrovave pIasna lorch fIane operaled al 1.2 kW nicrovave pover, 5O Ipn air as a
svirI gas, and 1O Ipn oxygen lhrough lhe fueI injeclor vilhoul dieseI. Iigure 6(l) is a piclure
of lhe pIasna fIane generaled ly 5O Ipn air as a svirI gas and 1O Ipn oxygen vilh O.O19
Ipn dieseI injeclion lhrough lhe fueI injeclor. Iigure 6(c) is a fronl viev of Iig. 3(l). WhiIe
lhe lorch fIane in Iig. 6(a) is veakIy expanded lo lhe exil of lhe slainIess sleeI lule of aloul
1O cn in Ienglh, lhe lurner fIanes shol oul lhrough lhe exil of lhe slainIess sleeI lule vhen
O.O19 Ipn dieseI vas injecled as a fueI inlo lhe nicrovave pIasna lorch, as shovn in Iigs.
6(l) and (c). The pIasna fIane dianeler and Ienglh fron lhe fIane exil vere aloul 8 cn and
3O cn, respecliveIy. The Iiquid dieseI can evaporale inslanlaneousIy, lreaking dovn lhe
noIecuIar slruclure ly lhe nicrovave pIasna coIunn vilh high gas lenperalure, and lurn
innedialeIy vilh oxygen. The fueI injeclor in lhis vork vas inslaIIed jusl alove lhe
vaveguide and vas approxinaleIy 2 cn avay fron lhe vaveguide excilalion region, as
shovn in Iig. 3(a). When 2O Ipn oxygen gas vilh O.O19 Ipn dieseI vas addilionaIIy injecled
inlo lhe nicrovave pIasna lorch, nol shovn in Iig. 6, il vas olserved lhal lhe pIasna fIane
coIor changed fron a yeIIovish vhile lo a lIuish vhile, indicaling a high lenperalure al
oxygen injeclion. In lhis conlexl, ve neasured lhe gas lenperalure of lhe dieseI nicrovave
pIasna fIane in lerns of oxygen conlenl.

Iig. 7. AxiaI lenperalure profiIes of lhe pIasna fIane in lerns of O
2
conlenl () al O.O19 Ipn
dieseI. TolaI fIov rale of nixlure of air and O
2
vas 7O Ipn and lhe appIied nicrovave
pover vas 1.2 kW. The nixlure vas injecled as a svirI gas (Hong & Uhn, 2OO6).

Iigure 7 exhilils lhe gas lenperalure profiIes of lhe dieseI nicrovave pIasna fIanes
neasured axiaIIy ly a lhernocoupIe in lerns of oxygen conlenl () al L
-O.5
, L
O
, L
3
, and L
6
in
Iig. 3(a). The lolaI fIov rale of lhe nixlure conposed of air and oxygen vas fixed al 7O Ipn
and lhe nixlure vas injecled inlo lhe nicrovave pIasna lorch as a svirI gas. Ior exanpIe,
O
2
conlenl of 54 represenls lhe nixlure conposed of 4O Ipn air and 3O Ipn O
2
. In lhe case

of 2O O
2
, vhiIe lhe gas lenperalures of lhe nicrovave pIasna lorch fIane al four
neasurenenl posilions dislrilule in lhe range of 4OO-85O K, lhe gas lenperalures of lhe
pIasna fIane vilh O.O19 Ipn dieseI are in lhe range of 14OO-16OO K. Moreover, vhen O
2

conlenl al L
-1.5
posilion increases fron 2O lo 54, lhe gas lenperalure significanlIy
increases fron 158O K lo 221O K. In generaI, lhe dieseI pIasna fIane shovs siniIar
properlies vilh lhe kerosene pIasna fIane such as Iig. 4.

3.2.3 PIasma fIame from methane
Iigure 8(a) shovs lhe nicrovave pIasna lorch fIane al appIied pover of 1.2 kW vhen 6O
Ipn air as a svirI gas vas injecled. As shovn in Iig. 8(l), hovever, once CH
4
(1O Ipn in lhis
lesl) as a hydrocarlon fueI is injecled inlo lhe nicrovave pIasna lorch lhrough lhe fueI
injeclor in Iig. 3(a), voIune of lhe CH
4
pIasna lurner fIane significanlIy increases, enilling
niIky vhile Iighls. Iig. 8(c) is lhe CH
4
fIane piclure vilhoul lhe nicrovave pIasna. UnIike
Iig. 8(l), Iig. 8(c) shovs lhe fIane vilh Iighl lIue coIor. The fIane lenperalure of Iig. 8(l)
and (c) al lhe poinl of L
O
in Iig. 3(a) is ~159O K and 125O K, and lhe visuaI fIane Ienglh of Iig.
2(l) and (c) is 18 cn and 11 cn, respecliveIy. Iron anolher experinenl resuIl of CH
4

nicrovave pIasna lurner nol shovn in lhis chapler, il is idenlified lhal CH
4
injeclion rale
up lo 3O Ipn (near O.36 g/s) is reasonalIe al 1.2 kW nicrovave pIasna lorch and
sloichionelric fueI/air nixlure. Ior praclicaI appIicalion such as pover pIanl, a nicrovave
pIasna lorch vilh 915 MHz or 896 MHz nicrovave syslen or nuIlipIe nicrovave pIasna
lorches nay le suilalIe for ollaining high pover.

Iig. 8. IIasna-lurner vilh (a) nicrovave pIasna onIy, (l) nicrovave pIasna + CH
4

injecled al 1O Ipn and (c) CH
4
fIane onIy (ang el aI., 2OO6).

In lhis regard, ve neasured lenperalures of lhe CH
4
nicrovave pIasna lurner fIane al
differenl posilions presenled in Iig. 3(a). The reclanguIar narks in Iig. 9(a) represenl lhe
axiaI lenperalure profiIe of lhe pIasna lurner fIane al narks L
O
-L
15
vhen 6O Ipn air as a
svirI gas, and a nixlure of 1O Ipn CH
4
and 4O Ipn air lhrough lhe fueI injeclor vere injecled
lhe nicrovave pIasna lorch. Therefore, lhe vaIue of nixlure ralio of air/CH
4
is 1O : 1. The
lenperalure of lhe nicrovave pIasna lorch fIane al lhe cenler of lhe fIane exil (nark L
O
in
Iig. 1) vas onIy 6OO K vilhoul CH
4
. As shovn in Iig. 9 (a), lhe lenperalure of lhe pIasna
lurner fIane increased lo aloul 189O K vhen 1O Ipn CH
4
vas injecled. And lhen lhe visuaI
Ienglh of lhe lurner fIane vas aloul 24 cn. Il is veII-knovn lhal lhe adialalic fIane
lenperalure of a CH
4
/air fIane is aloul 2222 K. 189O K in lhis lesl is fIane lenperalure
neasured al lhe poinl avay 5 cn fron lhe fueI injeclor in Iig. 1. Therefore, lhe lurner fIane
lenperalure near a region of fueI injeclion nay le as high as lhal of CH
4
/air fIane in
Fuel njection 192

adialalic condilion. The circuIar narks in Iig. 9(l) indicale lhe axiaI lenperalure profiIe of
lhe lurner fIane vhen 4O Ipn air as a svirI gas, and a nixlure of 1O Ipn CH
4
and 6O Ipn air
vas injecled lhrough lhe fueI injeclor. Wilh 1O Ipn CH
4
, lenperalure of lhe lurner fIane
increased fron 6OO K lo 168O K. And lhen lhe visuaI Ienglh of lhe lurner fIane in Iig. 9(l)
vas approxinaleIy 3O cn. Conpared Iig. 9(a) vilh (l), lhe lenperalure profiIe in Iig. 9(a)
faIIs rapidIy al axiaI posilion of 6 cn, vhereas lhe lenperalures in Iig. 9(l) reduce genlIy
aIong vilh axiaI direclion. In lhis conlexl, Iig. 9 inpIies lhal lhe lenperalure and Ienglh of
lhe lurner fIane can le conlroIIed ly injeclion vay or nixing rale of air and fueI. In generaI,
il is recognized lhal lhe use of a lhernocoupIe for neasurenenl of fIane lenperalures nay
encounler sone prolIens. AIso, fIanes aIready conlain a veakIy ionized pIasna vilh
lypicaI densily grealer lhan 1O
1O
ions/cn
3
(Uhn, 1999). Hovever, lhe lhernocoupIe used in
lhis lesl is perfeclIy covered vilh aIunina (AI
2
O
3
). So pIasna inpacls in lenperalure
neasurenenls nay le negIecled.

Iig. 9. AxiaI lenperalure profiIes of lhe CH
4
augnenled nicrovave pIasna lurner
neasured al posilions L
O
-L
15
as denoled in Iig. 3(a). (a) 6O Ipn svirI air + nixlure of 1O Ipn
CH
4
and 4O Ipn air. (l) 4O Ipn svirI air + nixlure of 1O Ipn CH
4
and 6O Ipn air (ang, el aI.,
2OO6)

Iigure 1O shovs lhe radiaI lenperalure profiIe of lhe CH
4
augnenled nicrovave pIasna
lurner al narks L
O
in Iig. 1. The reclanguIar narks in Iig. 1O(a) indicale lhe radiaI
lenperalure profiIe of lhe lurner fIane vhen 6O Ipn air as a svirI gas, and a nixlure of 1O
Ipn CH
4
and 4O Ipn air lhrough lhe fueI injeclor vere injecled lhe nicrovave pIasna lorch.
As shovn in Iig. 1O(a), lhe lenperalure of lhe lurner fIane decreased lo aloul 118O K,
rapidIy. The circuIar narks in Iig. 1O (l) indicale lhe radiaI lenperalure profiIe of lhe lurner
fIane vhen 4O Ipn air as a svirI gas, and a nixlure of 1O Ipn CH
4
and 6O Ipn air lhrough
lhe fueI injeclor vere injecled lhe lorch. The lenperalure of lhe lurner fIane decreased lo
aloul 137O K, sIovIy. Iigs. 9 and 1O shoved lhe axiaI and radiaI lenperalure profiIes in CH
4

augnenled nicrovave pIasna lurner fIane, respecliveIy. The perfornance of CH
4

nicrovave pIasna lurner significanlIy depends on lhe physicaI and chenicaI properlies of
nicrovave pIasna lorch. The lheorelicaI descriplion of lhe nicrovave pIasna lorch is
leyond lhe scope of lhe presenl sludy. Hovever, one can refer lhe previous arlicIes (Kin el

aI., 2OO3, Margol, 2OO1, Moon el aI., 2OO2) descriling lhe alnospheric pressure nicrovave
pIasna lorch.

Iig. 1O. RadiaI lenperalure profiIes of lhe CH
4
augnenled nicrovave pIasna lurner
neasured al posilion L
O
in Iig. 3(a). (a) 6O Ipn svirI air + nixlure of 1O Ipn CH
4
and 4O Ipn
air. (l) 4O Ipn svirI air + nixlure of 1O Ipn CH
4
and 6O Ipn air (ang el aI., 2OO6).

The lenperalure profiIes in Iigs. 9 and 1O vere changed vilh addilion of lhe sane CH
4

quanlily al differenl svirI air fIov rales and air fIov rales lhrough lhe injeclor. When a svirI
air fIov rale is nore lhan lhal lhrough lhe injeclor, lhe vorlex fIovs inside lhe slainIess sleeI
lule in Iig. 3(a) can survive againsl air fIov lhrough lhe injeclor, increasing lhe conluslion
line of CH
4
, confining lhe CH
4
fIanes axiaIIy, and lhus increasing lhe lenperalure al L
O

poinl. On lhe olher hand, vhen a svirI air fIov rale is Iess lhan lhal lhrough lhe injeclor, an
air fIov lhrough lhe injeclor can suppress vorlex fIov ly svirI air injeclion and increase
axiaI fIov veIocily and fIane Ienglh. Therefore, even lhough lhe sane CH
4
fIov rale is
injecled, each lenperalure profiIe in Iigs. 9 and 1O can le changed due lo differenl gas
injeclion nelhods.

3.3 SimpIe description of atomic oxygen density in pIasma fIames
IrincipaIIy, a discharge pIasna and a high lenperalure environnenl generale nany
chenicaIIy aclive radicaIs. Ior exanpIe, oxygen alons can le generaled ly lhe pIasna and
lhernaI dissocialions of oxygen noIecuIes, i.e., O
2
O + O. IIasna dissocialion incIudes
dissocialive reconlinalion of noIecuIar oxygen ions, eIeclron inpacl dissocialion of oxygen
noIecuIes, and dissocialive allachnenl of oxygen negalive ions (Uhn, 1999). ThernaI
dissocialion of oxygen noIecuIes has reaclion conslanl (Hong & Uhn, 2OO6) | = 2.7 1O
11

(T
R
/T
I
)
2
exp(-59429/T
I
) s
-1
, vhere T
R
and T
I
represenl roon and pIasna fIane lenperalure,
respecliveIy, in unils of KeIvin. The oxygen alons reconline vilh reconlinalion coefficienl
o = 2.3 1O
-14
(T
R
/T
I
)
2
cn
3
s
-1
, forning oxygen noIecuIes (Hong & Uhn, 2OO6). The oxygen
alon nay aIso forn ozone vilh oxygen noIecuIe lul ozone dissociales rapidIy due lo high
pIasna lenperalure. Therefore, ozone fron lhe nicrovave pIasna lorch is nol produced.
The rale equalion of oxygen alon densily n
O
is given ly
Fuel njection 194

2
2
2 ,
2
2 2
GQ
NQ Q
GW
o =
(1)
vilh lhe soIulion
2
2
( ) tanh( ),
2
2
NQ W
Q W
o q
=
(2)
vhere
2
2 1/
2
Q N =
, n
O2
is lhe noIecuIar oxygen densily and lhe faclor 2 in fronl of |
represenls 2 alons fron one noIecuIar dissocialion. Ior inslance, lhe pIasna fIane
lenperalures fron lhe fueI injeclion poinl lo L
9
poinl in Iig. 3(a) range in T
I
= 4OOO-15OO K
(Hong, el aI., 2OO6). The oxygen alon fornalion ly lhe pIasna nay le significanl, lul il is
difficuIl lo find lhe pIasna effecls in lhe pIasna fIane. NegIecling lhe pIasna effecls and
assuning T
I
= 2OOO K as an average vaIue, ve find q = 1.5 s

and n
O
(| = ) = 1.3 1O
15
/cn
3

for n
O2
= 6 1O
17
/cn
3
. Assuning lhe residence line | = O.O6 s

in lhe slainIess sleeI lule in Iig.
3(a) as a lypicaI vaIue, lhe oxygen alon densily is caIcuIaled lo le n
O
= 5.7 1O
13
/cn
3
fron
Lq. (2), vhich effecliveIy conlusls hydrocarlon fueIs. The oxygen alon densily increases
draslicaIIy vilh lhe high pIasna-fIane lenperalure originaled fron lhe pIasna lorch.

3.4 InfIuence of microwave pIasma in pIasma fIame

Iig. 11. Conparison of OH noIecuIar inlensilies for CH
4
fIane-onIy (gray) and pIasna
fIane vilh CH
4
(lIack) (Hong & Uhn, 2OO6).

As alove-nenlioned, ve conpared CH
4
fIane-onIy and lhe pIasna fIane vilh CH
4
ly
providing visuaI changes of fIane coIor, fIane Ienglhs, and fIane lenperalure ly a
lhernocoupIe. These nay le under lhe infIuence of lhe nicrovave pIasna on lhe
conluslion fIane. Here, ve presenl lhe infIuence on lhe nicrovave pIasna ly olserving
hydroxyI (OH) noIecuIes in an enission speclrun as a supporling dala. OH noIecuIar
speclrun is necessariIy olserved in nany kinds of fIanes and hol gases conlaining oxygen
and hydrogen. In lhis sense, ve conpared lhe enission inlensilies of OH radicaI for CH
4

fIane-onIy and CH
4
fIane conlined vilh lhe nicrovave pIasna al L
O
poinl in Iig. 3(a).
LxperinenlaI paranelers correspond lo lhe curve in Iig. 1O(a). Iigure 11 conpares lhe

reIalive enission inlensily of OH for CH
4
fIane-onIy (gray Iine) and CH
4
fIane conlined
vilh lhe nicrovave pIasna (lIack Iine). Tvo OH enission inlensilies vere nornaIized ly
lhe inlensily of CH
4
fIane conlined vilh lhe nicrovave pIasna. In generaI, perfecl
conluslion of hydrocarlon fueIs produces gaseous valer and carlon dioxide as finaI
resuIlanl producls. OH species are essenliaI inlernediales during lhe process of valer
produclion. OH enission inlensily slrongIy depends on lhe densily of alonic oxygen.
According lo lhe sinpIe descriplion for alonic oxygen nenlioned earIier, lhe alonic oxygen
densily vas eslinaled lo le n
O
= 5.7 1O
13
/cn
3
fron Lq. (2), vhich effecliveIy conlusls
hydrocarlon fueIs. UIlinaleIy, oxygen alons produced ly lhe nicrovave pIasna in air are
very heIpfuI for CH
4
conluslion, lherely exhililing slrong OH inlensily as shovn in Iig. 11.
On lhe conlrary, OH inlensily of CH
4
fIane-onIy is very veak in lhe conparison vilh lhe
pIasna fIane of CH
4
conluslion. This difference of OH inlensily refIecls one of rolalionaI
lenperalure (T
rol
) of OH noIecuIes lo le equaI lo lhe gas lenperalure of lhe fIanes (de
Izarra, 2OOO).

Iig. 12. Conparison of lhe neasured dala vilh lhe sinuIaled OH speclrun for CH
4
fIane-
onIy and pIasna fIane vilh CH
4
yieIding lhe rolalionaI lenperalures (T
rol
) of 13OO K and
195O K al L
O
poinl in Iig. 3(a) (Hong & Uhn, 2OO6).

In lhis conlexl, Iig. 12 shovs lhe unresoIved OH noIecuIar Iines (A
2
+
,
=O
X
2
,
=O
)
olserved in lhe vaveIenglh of 3O6-31O nn vilh a speclraI resoIulion of O.35 nn. T
rol
vas
delernined in lhis sludy ly conparing lhe sinuIaled OH speclrun vilh lhe neasured
speclrun ollained al a reIaliveIy Iov speclraI resoIulion. The nelhod for ollaining
sinuIaled OH speclrun al a given lenperalure vas provided in previous arlicIes (de Izarra,
2OOO). The gas lenperalures for CH
4
fIane-onIy and lhe pIasna fIane vilh CH
4
in Iig. 12
Fuel njection 196

are delernined lo le approxinaleIy 13OO K and 195O K, respecliveIy, shoving lhe infIuence
of lhe nicrovave pIasna on CH
4
conluslion.
A Iarge, high-lenperalure pIasna fIane nay le suilalIe for a luIk naleriaI lrealnenl, in
parlicuIar environnenlaI appIicalion. As nenlioned earIier, lhe nicrovave pIasna lorch in air
discharge has snaII voIune and ils lenperalure decreases draslicaIIy in lhe axiaI direclion. Ior
exanpIe, CI
4
, NI
3
, and SI
6
alalenenl experinenls conducled in our research group shoved
deslruclion efficiency nore lhan 9O onIy in conlaninanl fIov of 2O Ipn (Hong el aI, 2OO3). In
order lo overcone lrealnenl Iinilalion of lhe nicrovave pIasna lorch, a looI for an enIarged
high-lenperalure pIasna fIanes vas designed. As shovn in Iig. 9, lhe lenperalure difference
lelveen lhe CH
4
fIane-onIy and lhe CH
4
nicrovave pIasna lurner fIane is approxinaleIy
64O K. Therefore, lhe pIasna consliluenls, such as alonic oxygen and noIecuIar singIel oxygen,
produced in lhe nicrovave pIasna lorch can le very heIpfuI for hydrocarlon fueI conluslion
and nay le usefuI in lhe lhernaI lrealnenl processes.

4. AppIications using pIasma fIames
4.1 Mass purification of contaminated air with chemicaI and bioIogicaI warfare agents
The eIininalion experinenl of any chenicaI varfare agenl is aInosl inpossilIe in an ordinary
Ialoralory due lo safely issues. Considering lhus, lhe experinenlaIisls cuslonariIy carry oul a
sinuIaled experinenl ly naking use of loIuene gas. Ior lhis sane reason, lhe lioIogicaI
varfare agenls are nol used in an ordinary Ialoralory (Hong, el aI., 2OO4). The airlorne
lioIogicaI varfare agenls Iike nicroles or lacleria are allached lo organic or inorganic aerosoIs
and are spread vhen aerosoI parlicIes fIoal around. ConsequenlIy, lhe eIininalion of sool fron
a dieseI engine as lhe sinuIaled carrier aerosoI of lioIogicaI agenls vas carried oul.
The reaclion chanler for nass lrealnenl of conlaninaled air vas designed speciaIIy, providing
lhe necessary residence line for lhe lesl deconlaninalion effecls. The delaiI expIanalion of lhe
reaclion chanler has leen reporled in previous arlicIe (Uhn el aI., 2OO6). The loxic varfare
agenls conlaninaling air enlers lhe inner conparlnenl lhrough sIils fron lhe ouler conparlnenl
and are eIininaled nainIy ly oxidalion process exposed lo lhe high-lenperalure pIasna fIane
vilh alundanl oxygen alons in lhe inner conparlnenl. The deslruclion nodeI of lhe chenicaI
and lioIogicaI varfare agenls can le expressed as (Hong el aI., 2OO4)

) exp(
0
|
(
;
;
=
, (3)
vhere X represenls lhe Ieflover concenlralion of lhe varfare agenls afler lhe pIasna fIane
lrealnenl and X
O
is lhe iniliaI concenlralion lefore lhe lrealnenl, | denoles lhe energy
densily (in unils of jouIes per Iiler) deposiled on lhe conlaninaled air ly lhe pIasna fIane
during lhe lrealnenl and | represenls lhe energy densily required for lringing dovn lhe
concenlralion lo 1/e of ils iniliaI concenlralion, i.e. lhe energy densily needed for 63
deconposilion. Designaling R as lhe fIov rale of lhe conlaninaled air, ve nole R| =
conslanl for specified physicaI paranelers of lhe deconlaninalion syslen. In olher vords,
lhe energy densily | deposiled ly lhe pIasna fIane during lhe lrealnenl is inverseIy
proporlionaI lo lhe airfIov rale R. Assuning lhal X
1
and X
2
correspond lo lhe Ieflover
concenlralions for lhe fIov rales R
1
and R
2
, respecliveIy, ve find lhe reIalionship

,
) / ln(
) / ln(
1 0
2 0
2
1
; ;
; ;
5
5
=
(4)

vhich reIales lhe Ieflover concenlralion X lo lhe airfIov rale R. We can find lhe Ieflover
concenlralion X
2
in lerns of R
2
if ve knov lhe concenlralion X
1
in lerns of R
1
.
As an exanpIe, ve used loIuene (C
7
H
8
) as a sinuIaled chenicaI varfare agenl, and kerosene
and nelhane vere used as lhe hydrocarlon fueIs. A Iiquid fueI is leller lhan a gaseous fueI
vhen perlaining lo lhe inslance of conpaclness and noliIily. A suclion fan suppIied lhe
conlaninaled air vilh evaporaled loIuene lo lhe reaclion chanler. The airfIov rale vas R =
5,OOO Ipn. 4O Ipn of lhe conpressed air vas suppIied lo lhe svirI gas. The injeclion rales of
lhe kerosene in lhis experinenl vere 1.15 kg/hr (~ O.3 gaI/hr), 1.46 kg/hr and 1.87 kg/hr.
The O.3 gaI/hr nozzIe is lhe snaIIesl fueI nozzIe ever found. The nelhane fIov rales vere 5
Ipn, 1O Ipn, 15 Ipn, 2O Ipn and 3O Ipn. The energy conlained in kerosene and in nelhane
are 1O
7
caI/kg and 9.52 1O
6
caI/n
3
, respecliveIy. The fueI injeclion rale can le lransIaled
inlo valls. Ior exanpIe, a 1.15 kg/hr injeclion rale of kerosene is 13.3 kW and a 2O Ipn
injeclion rale of nelhane is aIso 13.3 kW. The energy densily | in Lq. (3) can le caIcuIaled ly
naking use of lhe fueI pover and airfIov rale. The nicrovave pover vas 1.4 kW and lhe
iniliaI loIuene concenlralion vas X
0
= 17O parlicuIales per niIIion (ppn). The kerosene
injeclion rales 1.15 kg/hr, 1.46 kg/hr and 1.87 kg/hr vilh lhe nicrovave pover of 1.4 kW
correspond lo lhe energy densily | = 176.4 }/L, 219.4 }/L, and 276.3 }/L, respecliveIy, for R
= 5,OOO Ipn. The size of lhe reaclion chanler used in lhe experinenl vas 22 cn dianeler
and 3O cn Iong. The conpaclness and Iighlveighl of lhe deconlaninalion syslen are lhe
key issues for a quick and easy appIicalion in Iife-lhrealening silualions. Therefore, lhe
reaclion chanler nusl le as snaII as possilIe for a specified airfIov rale. The reaclion
chanler of 22 cn dianeler and 3O cn Ienglh is good for lhe airfIov rale of 5,OOO Ipn. The
Ieflover concenlralion X of lhe loIuene had leen neasured ly naking use of deleclor lules
fron lhe CASTLCH Conpany. The gas chronalography (CC) or lhe Iourier lransforn
infrared (ITIR) can le used for nore accurale dala. In spile of lhis, lhose diagnoslic looIs
nay give conpIeleIy vrong neasurenenl vaIues, lecause loIuene is in Iiquid forn al lhe
roon lenperalure of one alnospheric pressure. A sanpIe Ieading lo lhe diagnoslic looIs can
easiIy le spoiIed ly loIuene condensalion. The neasurenenl ly deleclor lules can le done
al lhe fIane exil of lhe reaclion chanler vilhoul any deIay or any inlerference. Therefore,
lhe deleclor lule nay reIialIy neasure lhe Ieflover loIuene, aIlhough lhe dala nay have a
Iarge error lar. Iigure 13 shovs lhe Ieflover loIuene-concenlralion rale in lerns of energy
densily for kerosene (cIosed square dols) and nelhane (open square dols) fueI injeclions.
Lach dala poinl in Iig. 13 represenls lhe average of 8 repealed neasurenenls. The
reclanguIar dols al | = 16.7 }/L represenl loIuene deconposilion onIy ly lhe nicrovave
lorch pIasna vilh 1.4 kW. The lypicaI error in lhe neasurenenl as shovn in lhe open
square dol al | = 16.7 }/L of Iig. 13 is aloul 5 associaled vilh lhe deleclor lule. The error
lars of nosl olher dala are snaIIer lhan lhe dol size in Iig. 13. The loIuene curve in Iig. 13
for kerosene vas ollained fron Lq. (3) vilh lhe |-vaIue lhal vas Ieasl-squared filled lo lhe
dala poinls (cIosed square dols). The |-vaIue of lhe loIuene deconposilion ly lhe pIasna
fIane is | = 84.76 }/L for kerosene, vhich is nuch Iess lhan | = 393 }/L ly lhe puIse corona
(Ienelranle el aI., 1997) and | = 173 }/L ly lhe nicrovave pIasna lorch (Hong el aI., 2OO4).
The loIuene curve in Iig. 13 for nelhane vas ollained fron Lq. (3) vilh lhe |-vaIue lhal vas
Ieasl-squared filled lo lhe dala poinls (open square dols). The |-vaIue of lhe loIuene
deconposilion ly lhe pIasna fIane is | = 62.74 }/L for nelhane. CIearIy, lhe loIuene
deconposilion ly lhe high-lenperalure pIasna fIane is far nore efficienl lhan lhal ly lhe
puIse corona or ly lhe nicrovave lorch. Iurlhernore, lhe presenl deconposilion syslen is
Fuel njection 198

very conpacl and Iighl lo le handy for various appIicalions. The lenperalure of lhe reaclion
chanler vaII and lhe exil gas is nol hol due lo a Iarge anounl of airfIov. In facl, lhe ouler
vaII of lhe reaclion chanler onIy feeIs varn.

Iig. 13. Leflover loIuene and sool concenlralions in lerns of lhe energy densily |. The
cIosed and open square dols represenl neasurenenl dala for kerosene and nelhane
injeclions, respecliveIy, and lhe cIosed circuIar dols are lhe sool concenlralion dala for
nelhane injeclion (Uhn el aI., 2OO6).

An eIininalion experinenl of lhe airlorne lioIogicaI varfare agenls is very difficuIl lecause
of lhe conpIexily of delecling lhe agenls lefore and afler lhe pIasna fIane lrealnenl.
Spores of lhe lioIogicaI varfare agenls are usuaIIy allached lo aerosoI parlicIes. The
eIininalion of aerosoI parlicIes nay indireclIy shov eIininalion of lhe airlorne lioIogicaI
varfare agenls. LIininalion of sool fron lhe dieseI engine, vhich can le seen as airlorne
aerosoI parlicIes, vas olserved in lhe experinenl. The lurning kerosene nay generale ils
ovn sool, vhich nay inlerfere vilh lhe olservalion of lhe dieseI engine sool, hence lhe
gaseous fueI of nelhane vas used in lhe experinenl. The nelhane injeclion rale vas 15 Ipn,
25 Ipn and 3O Ipn in lhe sool eIininalion. The discharge gas fron a 1O,OOOcc lus dieseI
engine al 8OO rpn vas used as lhe conlaninaled air vilh sool. The airfIov rale al lhe engine
exil vas 8,OOO Ipn, vhich is eslinaled lo le 3,5OO Ipn al lhe end of lhe laiI pipe due lo lhe
cooIing of lhe anlienl air. The energy densily lherefore vas caIcuIaled ly lhe nelhane
injeclion inlo lhe airfIov of 3,5OO Ipn. While fiIlers caplured sool fron lhe discharge gas. A
snoke neler fron OSCH, vhich delernines opacily, neasured lhe caplured sool-anounl
in lhe fiIler. The renaining sool (cIosed circuIar dols) in reIalive lo lhe unlrealed case is
pIolled in Iig. 13 in lerns of lhe energy densily for nelhane injecled inlo pIasna. The sool
vas aInosl conpIeleIy eIininaled al | = 34O }/L corresponding lo lhe 3O Ipn nelhane
injeclion. The |-vaIue of lhe sool eIininalion vas delernined ly lhe Ieasl-squared-filled lo
lhe experinenlaI dala (cIosed circuIar dols) in Iig. 13 and is given ly | = 138.O2 }/L. The
pIasna fIane is an effeclive nean lo eIininale lhe sool fron lhe dieseI engine. This neans
lhal lhe pIasna fIane nay effecliveIy eIininale airlorne aerosoI parlicIes. Mosl of lhe
aerosoIs are nade of hydrocarlon naleriaIs, vhich can easiIy le oxidized al a high-
lenperalure pIasna fIane vilh lhe lenperalure higher lhan 1OOO degrees CeIsius. The
lioIogicaI agenls consisling of lacleria and virus nay nol survive as lhey go lhrough lhe
high-lenperalure pIasna fIane. Therefore, lhe pIasna fIane nay effecliveIy eIininale lhe

airlorne lioIogicaI varfare agenls. A differenl experinenlaI olservalion confirned lhal lhe
pIasna fIane of lhe kerosene or dieseI injecled inlo lhe lorch pIasna does nol produce ils
ovn sool. In lhis conlexl, lhe pIasna fIane can aIso le usefuI for lhe eIininalion of sool
fron dieseI engines in lrucks, in luses, in lrains and in ships.
Il is noled fron Lq. (4) lhal lhe airfIov rale can increase ly reslricling lhe deconposilion
rale. Ior exanpIe, lhe Ieflover concenlralion of loIuene al lhe kerosene fueI rale of 1.87
kg/hr corresponding lo | = 276.3 }/L in Iig. 2 is X
1
/X
O
= O.O2 for R
1
= 5,OOO Ipn. Sulsliluling
lhese nunlers inlo Lq. (4), ve find lhal R
2
= 19,56O Ipn for X
2
/X
O
= 1/e. Aloul 2O,OOO Ipn
of lhe conlaninaled air vilh loIuene can le lrealed if lhe lrealnenl is al a 63 percenl
eIininalion requirenenl.
As nenlioned earIier, lhe conpaclness and Iighlveighl of lhe deconlaninalion syslen are
crilicaI issues for rapid noliIily and quick inslaIIalion in Iife lhrealening silualions. The
reaclion chanler size used in lhe exanpIes presenled earIier is 22 cn dianeler and 3O cn
Iong, vhich Iinils lhe airfIov rale. The Iinear dinension of lhe vaveguide and discharge
lule in lhe pIasna lorch syslen is proporlionaI lo lhe vaveIenglh of nicrovaves. Therefore,
lhe lorch pIasna voIune is inverseIy proporlionaI lo lhe square of lhe nicrovave
frequency. Ior exanpIe, lhe lorch pIasna voIune increases 7 lines ly changing lhe
nicrovave frequency fron 2.45 CHz lo 915 MHz vilh an addilionaI pover. The Iarger
voIune of lhe pIasna fIane in an increased reaclion chanler vilh Iov-frequency
nicrovaves and addilionaI fueI neans lhe nore lrealnenl of lhe airfIov rale. The lrealnenl
voIune can easiIy le enhanced ly increasing lhe size of lhe pIasna fIane in an enIarged
reaclion chanler. Therefore, lhere viII le no scienlific prolIen lo exlend lhe lrealnenl
voIune lo 1OO,OOO Ipn, aIlhough lhe syslen size nay increase accordingIy.

4.2 EIimination of air contaminated with odorous chemicaI agents

Iig. 14. LxperinenlaI sel-up for eIininaling NH
3
and H
2
S as odor-causing chenicaI
naleriaIs ly naking use of a nicrovave pIasna lurner. The insel is lhe piclure of lhe
kerosene pIasna fIane (Hong el aI., 2OO7).
Fuel njection 200

The insel in Iig. 14 shovs lhe piclure of lhe pIasna fIane produced fron lhe nicrovave
pIasna lurner al 1.4 k}/s pIasna pover vilh no refIecled pover and 1.15 kg/hr kerosene.
In Iig. 14, lhe lIover fan connecled lo lhe reaclion chanler ly four slainIess sleeI leIIovs
sucks up air conlaninaled odorous gases, and lransfers lhe conlaninanls inlo lhe reaclion
chanler. Il can suck and lIov airfIov nore lhan 5 OOO Iillers per ninule (Ipn) al Ieasl. The
reaclion chanler consisls of inner and ouler conparlnenl, providing a space lelveen lhen.
The conlaninaled air vas injecled inlo lhe reaclion chanler via four injeclion porls in
langenliaI direclion inslaIIed on lhe ouler conparlnenl, lherely rolaling in lhe space. In
lurn, lhe rolaling airfIovs enler lhe inner conparlnenl vilh langenliaI sIils, vhich are aIso
in langenliaI direclions aIong lhe inner surface of lhe inner conparlnenl vaII, nixing vilh
lhe pIasna fIane nade of alnospheric nicrovave pIasna and a fueI-lurning fIane. The
dinensions of lhe reaclion chanler used in lhe experinenl vere 22 cn dianeler and 3O cn
Iong. The pIasna fIane and lhe conlaninaled air in lhe inner conparlnenl rolale in lhe
sane direclion, providing lhe necessary residence line for lhe lesl eIininalion effecls. These
sequenliaI processes lhen eIininale lhe odorous chenicaI agenls in lhe passing air.
Aqua annonia (NH
4
OH) vas used lo ollain NH
3
gas in lhe sinuIaled experinenl for
eIininaling NH
3
and vas nainlained al 6O
C ly a vaporizalion device in Iig. 14. On lhe

olher hand, in case of lhe sinuIaled experinenl of eIininaling H
2
S, gas-phase H
2
S vas
direclIy injecled inlo lhe lIover fan and vas nixed vilh air. The lIover fan suck up air
conlaninaled vilh NH
3
and H
2
S gas, and lransferred lhe conlaninanls inlo lhe reaclion
chanler. And lhen lhe lolaI air-fIov rale vas approxinaleIy 5 OOO Ipn. 4O Ipn of lhe
conpressed air as a svirI gas vas injecled inlo lhe nicrovave pIasna lorch. The injeclion
rales of lhe kerosene vere 1.15 kg/hr, 1.46 kg/hr and 1.87 kg/hr. The 1.15 kg/hr nozzIe is
lhe snaIIesl fueI nozzIe ever found. The nelhane fIov rales vere 5 Ipn, 1O Ipn, 15 Ipn, 2O
Ipn and 3O Ipn. The energy conlained in kerosene and in nelhane are 1O
7
caI/kg and 9.52
1O
6
caI/n
3
, respecliveIy. The fueI-fIov rales injecled can le lransIaled inlo jouIes per second.
The pover of 1.15 kg/hr kerosene energy corresponds lo 13.3 k}/s and lhal of 2O Ipn
nelhane energy is aIso 13.3 k}/s. The delaiIed sinuIaled experinenls for eIininaling NH
3

and H
2
S vere carried oul in lerns of lhe inpul energy densily of lhe nicrovave pIasna
lurner. Ior inslance, lhe kerosene injeclion rales of 1.15 kg/hr, 1.46 kg/hr, and 1.87 kg/hr
vilh lhe 1.4 k}/s pIasna pover correspond lo lhe inpul energy densilies 176.4 }/L, 219.4 }/L,
and 276.3 }/L, respecliveIy, for lhe lolaI air-fIov rale of 5 OOO Ipn.
In lhis vork, lhe experinenlaI resuIls vere presenled ly naking use of a sinpIe firsl order
decay nodeI for eIininaling largel chenicaIs. The deslruclion nodeI (Hong el aI., 2OO4) of
lhe odorous chenicaIs can le expressed as X/X
0
= exp(-|/), vhere X represenls lhe Ieflover
concenlralion of lhe odorous chenicaIs afler lhe pIasna fIane lrealnenl and X
O
is lhe iniliaI
concenlralion lefore lhe lrealnenl, | denoles lhe inpul energy densily (in unils of jouIes per
Iiler) deposiled on lhe conlaninaled air ly lhe pIasna fIane during lhe lrealnenl and |
represenls lhe energy densily required for lringing dovn lhe concenlralion lo 1/e of ils
iniliaI concenlralion, i.e. lhe energy densily needed for 63 deslruclion. The Ieflover
concenlralions of NH
3
and H
2
S vere neasured ly enpIoying deleclor lules fron lhe
CASTLCH Conpany in }apan. The neasurenenl ly deleclor lules vas done al lhe fIane
exil of lhe reaclion chanler vilhoul any deIay or any inlerference. Therefore, lhe deleclor
lule nay reIialIy neasure lhe Ieflover NH
3
and H
2
S, aIlhough lhe dala nay have a Iarge
error lar. The dala poinls in Iig. 15 indicale lhe average Ieflover NH
3
(open circIe dols) and
H
2
S (open square dols) concenlralions ollained fron lhe repealed neasurenenls in lerns of

lhe inpul energy densilies ly neans of lhe nelhane pIasna lurner. The cIosed square dols
are lhe Ieflover H
2
S concenlralions ly neans of lhe kerosene pIasna lurner. The iniliaI
concenlralions of NH
3
and H
2
S vas X
0
= 159 ppn and 12O ppn, respecliveIy. The curves in
Iig. 15 represenl lhe Ieasl squared fils lo lhe experinenlaI dala poinls for lhe nicrovave
pIasna lurner. LvenluaIIy, lhe |-vaIues of lhe NH
3
and H
2
S eIininalion ly lhe nelhane
pIasna lurner are 39.69 }/L and 56.45 }/L, respecliveIy. On lhe olher hand, lhe |-vaIue of
H
2
S eIininalion ly lhe kerosene pIasna lurner is 46.52 }/L. The |-vaIues of NH
3
and H
2
S
eIininalion ly nelhane pIasna lurner are consideralIy Iess lhan 62.74 }/L for loIuene and
138.O2 }/L sool eIininalion, vhich vere reporled in lhe previous docunenl (Uhn el aI.,
2OO6). In lhe recenl arlicIe for deconposilion of H
2
S and NH
3
using a pIale-lo-vire puIse
corona reaclor (Huang el aI., 2OO1), lhe |-vaIues of H
2
S (X
0
= 148 ppn) and NH
3
(X
0
= 58
ppn) deconposilion vere 65 }/L and 6O }/L, respecliveIy. CIiding arc discharges (DaIaine
el aI., 1998, Czernichovski A. 1994) have leen used as olher exanpIe of H
2
S depoIIulion.
Czernichovski

(Czernichovski, 1994) reporled lhal 7 Nn
3
/h of air conlaninaled vilh O.7
H
2
S vas conpIeleIy purified al lhe energy consunplion of O.14 kWh per Nn
3
vilhoul any
prehealing. The energy in lringing dovn lhe concenlralion of ils iniliaI concenlralion lo
zero vas eslinaled lo le 54O }/L. In Iig. 15, lhe energy is approxinaleIy 3OO }/L. Lven
lhough lhe iniliaI concenlralions are differenl for H
2
S eIininalion, lhis vork reveaIs lhal lhe
kerosene nicrovave pIasna lurner nay le nore effeclive lhan lhe puIse corona reaclor

(Shi
el aI., 2OO5) and lhe gIiding arc discharge

(Czernichovski, 1994) in a slandpoinl of energy
consunplion. Iron lhe sinpIe descriplion for alonic oxygen produced in lhe nicrovave
pIasna lurner (Hong & Uhn, 2OO6), lhe alonic oxygen densily n
c
vas caIcuIaled lo le n
c
=
5.7 1O
13
/cn
3
, vhich effecliveIy conlusls hydrocarlon fueIs. Il is aIso enphasized lhal a
Iarge voIune of air can le lrealed ly a conpacl apparalus in lhis sludy.

Iig. 15. IIols of Ieflover H
2
S and NH
3
concenlralion in lerns of lhe inpul energy densily |.
The cIosed and open square dols represenl lhe dala poinls of H
2
S concenlralions for
kerosene and nelhane injeclion, respecliveIy, and lhe open circIe dols are NH
3

concenlralion dala for nelhane injeclion. Lach dala poinl indicales lhe average vaIue of
eighl repealed neasurenenls (Hong el aI., 2OO7).
Fuel njection 202

4.3 Destruction of fIuorinated compound gases
In NI
3
alalenenl, NI
3
, N
2
, O
2
and CH
4
vere prenixed in lhe gas-nixing vesseI and injecled
fron lhe side of lhe nicrovave pIasna lorch lhrough IC and CH
4
gas injeclor. NI
3
can le
direclIy ionized, allached, or dissocialed lo NI
x
(x=O, 1, 2) radicaIs ly eIeclron inpacl
processes, and lhe reaclion is expressed as

NI
3
NI
x
+ I
q
(| = 6.6 1O
-8
c
(-2416O/T)
cn
3
/s), (5)

vhere q is 1 or 2, and T is gas lenperalure. The nicrovave pIasna lurner produces high-
lenperalure, Iarge-voIune pIasna fIane (Hong el aI., 2OO6). NI
3
gas is easiIy deconposed
in high-lenperalure environnenl. Ior exanpIe, lhe reaclion rale in Lq. (5) is 4.2 1O
-12

cn
3
/s al 25OO K. In facl, average lenperalure of lhe nelhane pIasna lurner fron lhe CH
4

injeclor lo 2O cn avay is approxinaleIy 25OO K (ang el aI., 2OO6). Therefore, alalenenl of
IC gases using lhe nelhane nicrovave pIasna lurner is acconpIished ly lolh pIasna and
lhernaI deconposilion. The chenicaI reaclions descriled leIov are considered in a
slandpoinl of lhe addilive gas used for effeclive alalenenl, aIlhough lhere are nany olher
possilIe reaclions. When O
2
as an addilive gas is used lo alale NI
3
, lhe desired reaclion
palhvay of O
2
is lo oxidize lhe nilrogen in NI
3
lo N
x
O
q
. Whenever dialonic oxygen
noIecuIes neel eIeclrons, lhey undergo dissocialive allachnenl, vhich produces an O
radicaI and O
-
ion (Hong el aI., 2OO3) for lhe eIeclron lenperalure in lhe range of lhe presenl
experinenl. These oxygen alons reacl vilh lhe NI
x
radicaIs. The chenicaI reaclion
equalions are
O + NI
2
NI + OI (| = 1O
-12
cn
3
/s), (6)
O + NI NO + I (| = 1O
-12
cn
3
/s), (7)
O + OI O
2
+ I (| = 5 1O
-11
cn
3
/s). (8)

ased on Lqs. (6)-(8), lhe finaI lyproducls are nilrogen nonoxide and fIuorine al lhe
dovnslrean of lhe reaclor. AIso CH
4
eIeclron inpacl dissocialion produces H radicaIs. H
radicaIs are precursors for IC renedialion. As an exanpIe, lhe chenicaI reaclions of NI
x
ly
H radicaIs (Chang el aI., 2OOO) are presenled:

H + NI N + HI (| = 1.5 1O
-13
cn
3
/s), (9)
H + I HI (| = 1.6 1O
-9
cn
3
/s). (1O)

As shovn in Lqs. (9) and (1O), lhe slalIe lyproducl HI is forned ly CH
4
. Il is veII knovn lhal
HI is valer soIulIe and is easiIy caplured ly passing lhrough a connerciaI vel scruller.
AIlhough lhe pIasna lenperalure decreases vilh lhe radius and lhe Ienglh of lhe pIasna lorch
fIane, NI
3
is easiIy deconposed lo NI
x
radicaIs vilh a high reaclion rale al given
lenperalures, as shovn Lq. (5). Therefore, ve expecl lhal aII reaclions presenled in Lqs. (5) (1O)
occur in lhe core of lhe pIasna lurner fIane and HI conlenls increases in lhe aflergIov.
In addilion lo NI
3
alalenenl, oxygen alons aIso reacl vilh SI
x
(x=1-5) radicaIs produced
ly eIeclron inpacl processes, crealing addilionaI SO
2
or SO noIecuIes and forning SOI
2

and SO
2
I
2
noIecuIes ly I
2
reaclions dovnslrean of lhe pIasna. The chenicaI reaclion
equalions (IIunl & Ryan, 1988)

are

O + SI
5
SOI
4
+ I (| = 2 1O
-11
cn
3
/s), (7)
O + SI
4
SOI
4
(| = 2 1O
-14
cn
3
/s), (8)

O + SI
2
SOI
2
(| = 1.O8 1O
-1O
cn
3
/s), (9)
O + SI
4
SOI + I (| = 7.63 1O
-11
cn
3
/s), (1O)
O + SI SOI (| = 1.7 1O
-1O
cn
3
/s), (11)
O + SOI Iroducls (| = 7.9 1O
-11
cn
3
/s). (12)

In lhe CI
4
alalenenl, lhe desired reaclion palhvay of O
2
is lo oxidize lhe carlon in CI
4
lo
CO
2
. When dialonic oxygen noIecuIes neel eIeclrons, lhey undergo dissocialeve
allachnenl lhal producing O radicaI and O
-
ion al lhe eIeclron lenperalure in lhis range of
presenled experinenl. These oxygen alons reacl vilh lhe CI
x
radicaIs. The chenicaI
reaclion equalions are (Hong el aI., 2OO3)

O + CI
3
COI
2
+ I (| = 3.1 1O
-11
cn
3
/s), (13)
O + CI
2
COI + I (| = 1.4 1O
-11
cn
3
/s), (14)
O + CI
2
CO + 2I (| = 4.O 1O
-11
cn
3
/s), (15)
O + CI CO + I (| = 2.4 1O
-11
cn
3
/s). (16)

The hydrogen radicaIs produced fron lhe deconposilion of CH
4
reacl vilh fIuorine species
and forn sinpIe, slalIe lyproducl HI, as shovn in Lq. (1O).
As previousIy nenlioned, ITIR vas enpIoyed lo idenlify lhe concenlralion changes of NI
3
,
SI
6
, CI
4
and lhe pIasna lyproducl lefore and afler lhe pIasna lurner lrealnenl. The
perfornance of lhe nicrovave pIasna alalenenl device vas descriled in lerns of DRL.
The DRL represenls lhe percenlage of IC gas lhal has leen deslroyed. In olher vords, lhe
definilion of DRL is
DRL () = (S
leforc
- S
afler
) / S
leforc
1OO, (17)

vhere S
leforc
and S
afler
are lhe nain peak area of lhe IC gases lefore and afler lhe pIasna
lurner lrealnenl, respecliveIy.

Iig. 16. ITIR speclra (a) lefore and (l) afler lhe nicrovave pIasna lurner alalenenl of NI
3

vilh lhe conponenls of 4O Ipn conpressed air as svirI gas, 25O Ipn N
2
, 3O Ipn O
2
, 15 Ipn
CH
4
and O.6 Ipn NI
3
al lhe appIied pIasna pover of 1.2 kW (Hong el aI., 2O1O).

Fuel njection 204

Iigures 16(a) and (l) shov infrared lransnilled speclra, denonslraling lefore and afler lhe
appIicalion of lhe nicrovave pIasna lurner lo lhe conponenls of 4O Ipn svirI air, a
nixlure of 3O Ipn O
2
, 25O Ipn N
2
, 15 Ipn CH
4
and O.6 Ipn NI
3
al lhe appIied nicrovave
pover of 1.2 kW, respecliveIy. The energy conlained in lhe CH
4
is aloul 9.52 1O
6
caI/n
3
.
The CH
4
injeclion rale can le lransIaled inlo vall. Ior exanpIe, 15 Ipn injeclion rale of
nelhane is 9.98 kW. Therefore, lolaI appIied pover is approxinaleIy 11.2 kW. As nenlioned
earIier, lhe nicrovave pIasna lorch shoved a lrealnenl Iinilalion al high fIov rales vilh
Iov renovaI efficiency due lo lhe draslic decrease of pIasna voIune and lhe shorl residence
line of conlaninanls. The vavenunlers 91O and 1O32 cn
-1
on lhe horizonlaI Iine represenl
lhe signalure of lhe NI
3
species. Afler lhe nicrovave pIasna lurner vas lurned on, lhe gas
slrean vas anaIyzed. As expecled, lhe CH
4
pIasna lurner deslroyed aInosl aII of lhe NI
3

conlaninanls vilh CH
4
deconposilion rale cIose lo 1OO. The NO
2
curve in Iig. 16(l)
represenls lhe increases of lhe NO
2
noIecuIes during lhe nicrovave discharge, lecause
nilrogen in NI
3
is oxidized ly O
2
. Moreover, in air pIasna, N
x
O
q
neulraI noIecuIes, N
x
O
y
-

and N
x
O
y
+
ions, viII le forned, vhere NO
2
and N
2
O, N
2
O
2
-
and N
3
O
+
are expecled lo le
doninanl neulraI species, negalive and posilive ions, respecliveIy. Therefore, NO
2
is forned.
CeneraIIy, NO in lhe IITR speclrun is idenlified al 195O-181O cn
-1
. As shovn in Iig. 16(l),
lhe NO peak nay le screened ly slrong H
2
O peaks or lhe fornalion of NO nay le nininaI
if lhere is any. HI can le forned lhrough lhe reaclion of I
2
vilh H
2
O vhich nay le in
dovnslrean of lhe pIasna lurner.

Iig. 17. ITIR speclra (a) lefore and (l) afler lhe nicrovave pIasna lurner alalenenl of SI
6

vilh lhe conponenls of 4O Ipn conpressed air as svirI gas, 12O Ipn N
2
, 3O Ipn O
2
, 15 Ipn
CH
4
and O.1 Ipn SI
6
al lhe nicrovave pIasna pover of 1.4 kW (Hong el aI., 2O1O).

Iigure 17 dispIays ITIR speclra lefore and afler lhe appIicalion of lhe nicrovave pIasna
lurner lo lhe conponenls of 4O Ipn svirI air, a nixlure of 3O Ipn O
2
, 12O Ipn N
2
, 15 Ipn
CH
4
and O.1 Ipn SI
6
al lhe appIied nicrovave pover of 1.4 kW. The lransIaled lolaI pover
for 15 Ipn CH
4
is approxinaleIy 11.2 kW. The vavenunler 948 cn
-1
on lhe horizonlaI Iine
represenls lhe signalure of lhe SI
6
concenlralion, as shovn Iig. 17(a). Afler lhe pIasna
lurner lrealnenl in Iig. 17(l), SI
6
concenlralion vas significanlIy reduced, shoving high
DRL of 99.92. SO
2
noIecuIes al 1361 cn
-1
as an inporlanl lyproducl for deconposilion of
SI
6
vere delecled. The nicrovave pIasna lurner sinuIlaneousIy deslrucled and lurned

oul lhe SI
6
conlaninanls ly exolhernic reaclions vilh O
2
in conpressed air injecled as a
svirI gas, forning HI noIecuIes.
Iigure 18 shovs infrared lransnilled speclra, denonslraling lhe nicrovave pIasna lurner
alalenenl for CI
4
. In lhe sane experinenlaI condilion vilh lhe case of SI
6
, a nixlure of 4O
Ipn N
2
and O.O5 Ipn CI
4
as conlaninanl vas injecled. In Iig. 18(a), lhe vavenunler 1283
cn
-1
on lhe horizonlaI Iine represenls lhe signalure of lhe CI
4
concenlralion. In addilion lo
NI
3
and SI
6
, CI
4
vas aIso alaled in lhe experinenlaI condilion, shoving 98.1 DRL in Iig.
18(l). AIlhough lhe lrace of CO and CO
2
as resuIlanl producls of CI
4
oxidalion vas nol
idenlified due lo kerosene conluslion, nain lyproducls are CO and CO
2
.

Iig. 18. ITIR speclra (a) lefore and (l) afler lhe nicrovave pIasna lurner alalenenl of CI
4
(Hong el aI., 2O1O).

Iigure 19 shovs lhe pIol of lhe DRLs of SI
6
and CI
4
versus nilrogen fIov rales,
corresponding lo Iigs. 17 and 18, respecliveIy. The conponenls of 4O Ipn svirI air, a
nixlure of IC gas vhich is O.1 Ipn SI
6
or O.O5 Ipn CI
4
, 3O Ipn O
2
, 15 Ipn CH
4
and various
N
2
gas fIov rales al lhe appIied nicrovave pover of 1.4 kW. DRLs vere neasured in lerns
of N
2
fIov rales. The N
2
fIov rale of SI
6
and CI
4
alalenenl vere fron 4O Ipn lo 16O Ipn
and 2O Ipn lo 12O Ipn, respecliveIy. In case of SI
6
alalenenl, DRLs vere nore lhan 99.9
unliI 12O Ipn N
2
, hovever, DRLs of over 14O Ipn N
2
vere decreased draslicaIIy. Al lhe N
2

fIov rale over 14O Ipn, pIasna lurner fIane lenperalure, voIune and residence line vere
decreased draslicaIIy. In lhis conlexl, DRL of CI
4
vas decreased draslicaIIy in lerns of over
6O Ipn N
2
fIov. Il is apparenl lhal lhe alalenenl of CI
4
is nuch nore difficuIl lhan lhal of
SI
6
. Ior inslance, lhe DRL of SI
6
is nore lhan 99.99 al 12O Ipn N
2
, vhiIe lhal of lhe CI
4
gas
is onIy 46.2. In parlicuIar, CI
4
has proven lo le difficuIl lo deslroy and renove lecause of
ils chenicaI lhernaI slaliIily due lo lhe slrong covaIenl nalure of ils londing. Dissocialion
energies in SI
6
and CI
4
are represenled ly e + SI
6
SI
5
+ I + e (4.O1 eV) and e + CI
4
CI
3

+ I + e (12.5 eV), respecliveIy. The CI
4
dissocialion energy is aloul lhree lines of lhe SI
6

dissocialion energy, vhich expIains lhe olservalions in Iig. 19.

Fuel njection 206

Iig. 19. DRLs of SI
6
and CI
4
pIolled in lerns of N
2
fIov rales (Hong el aI., 2O1O).

Creenhouse gases lhal are suljecl lo reduclion under lhe Kyolo IrolocoI incIude HICs,
IICs, SI
6
, N
2
), CH
4
, and CO
2
. Since lhe gIolaI varning polenliaI (CWI) of lhe lhree
fIuorinaled gas sulslilules is severaI hundred lo severaI lens of lhousand lines of lhal of
CO
2
, ulnosl efforls lovards reguIaling lheir enission are required. In lhe presenl vork, CO
2

enission fron lhe conluslion of 15 Ipn CH
4
is aloul 29 grans per ninule. And NI
3

enission fron O.5 Ipn is aloul 1.5 grans per ninule. CWI of NI
3
is very high as nany as
972O lines in respecl lo CO
2
. Therefore, ve expecl lhal CO
2
enission fron lhe presenl
pIasna syslen is ignoralIe (Hong el aI., 2O1O).

5. ConcIusion
In order lo significanlIy increase lhe voIune and lenperalure of lhe nicrovave pIasna
lorch, ve have deveIoped a nicrovave pIasna-lurner ly injecling hydrocarlon fueI in
Iiquid or gaseous slale inlo lhe nicrovave pIasna lorch generaled ly air and a nixlure of
air and oxygen. The nicrovave pIasna-lurner inpIies lhal a pIasna fIane of high-pover
IeveI can le ollained ly onIy injecling snaII quanlily of a hydrocarlon fueI. Ior exanpIe,
lhe energy conlained in dieseI and in nelhane are approxinaleIy 1O
7
caI/kg and 9.52 1O
6

caI/n
3
, respecliveIy. The fueI injeclion rale can le lransIaled inlo valls. LvenluaIIy, a O.O19
Ipn (~ 1.15 kg/hr) injeclion rale of dieseI is 13.3 kW and a 1O Ipn CH
4
corresponds lo
approxinaleIy 6.6 kW. A significanl lenperalure increase of lhe CH
4
nicrovave pIasna
lurner vas olserved in reIalive lo lhose of lhe nicrovave pIasna lorch and CH
4
fueI-onIy
fIane, shoving lhe infIuence of lhe nicrovave pIasna on lhe conluslion fIane.
In deconlaninalion lesl of chenicaI agenls, lhe pIasna fIane couId le a very effeclive nean
for deconlaninaling lhe airlorne chenicaI and lioIogicaI varfare agenls. The pIasna fIane
aIso provided a unique opporlunily of a sinuIlaneous eIininalion and lurnoul of chenicaI
and lioIogicaI varfare agenls diIuled in air. Hovever, il is essenliaI for lhe nalions and lheir
niIilary, vorIdvide, lo carry oul lhe deconlaninalion experinenl of lhe reaI chenicaI and
lioIogicaI varfare agenls for lhe experinenlaI dala, vilh vhich lhe pIasna fIane syslen
nakes lhe varfare agenls olsoIele. We aIso reporled lhe experinenlaI resuIls for renoving
IC gases ly using lhe nelhane nicrovave pIasna lurner and descriled lhe sinpIe nodeI
for effeclive alalenenl of IC gases. In NI
3
alalenenl, ve achieved a DRL of nore lhan
99.9999 in lhe nixlure of O.6 Ipn NI
3
, 4OO Ipn N
2
, 1OO Ipn O
2
and 15 Ipn CH
4
ly appIying

lhe nicrovave pover of 1.2 kW. In case of SI
6
, ve achieved a DRL of nore lhan 99.9 in
lhe nixlure of O.1 Ipn SI
6
, 12O Ipn N
2
, 3O Ipn O
2
and 15 Ipn CH
4
ly al lhe appIied
nicrovave pover of 1.4 kW. AIso, DRL of CI
4
is 94.72 in lhe nixlure of O.O5 Ipn CI
4
, 6O
Ipn N
2
, 3O Ipn O
2
and 15 Ipn CH
4
al lhe nicrovave pover of 1.4 kW. N
2
gas is required in
posl-punp syslens. The aforenenlioned characlerislics nake lhe nicrovave pIasna lurner
allraclive for lhe deslruclion and renovaI of chenicaIIy slalIe gases enilled ly lhe
seniconduclor induslry.

6. References
ang, C.U., Hong, Y.C., Cho, S.C., Uhn, H.S. & Yi, W.}. (2OO6): Melhane-Augnenled
Microvave IIasna urner, |||| Trans. P|asma Sci. VoI. 34, No. 5, pp 1751-1756,
ISSN OO93-3813.
Chang, }.-S., Urashina, K.C., Yananolo, K., Okayasu, T., Kalo, Y., Ivaizuni, T. &
Yoshinura, K. (2OOO): RenovaI of NI
3
fron seniconduclor-process fIue gases ly
landen packed-led pIasna and adsorlenl hylrid syslens, |||| Trans. |n. App|.
VoI. 36, No. 5, pp 1251-1259, ISSN OO93-9994.
Czernichovski, A. (1994): CIiding arc: appIicalions lo engineering and environnenl conlroI,
Purc App|. Cncm. VoI. 66, No. 6, pp 13O1-131O, ISSN 1365-3O75.
DeIaine, V., Cornier, }.M, IeIIerin, S. and Lefaucheux, I. (1998): H
2
S deslruclion in 5O hz
and 25 kHz gIiding arc reaclor, ]. App. Pnqs. VoI. 84, No. 3, pp 1215-1221, ISSN .
de Izarra, C. (2OOO): Conpuler sinuIalion of lhe UV OH land speclrun, |n|. ]. Mc. Pnqs. C
VoI. 11, No. 5, pp 987-998, ISSN 1793-6586.
Harlz, C.L., evan, }.W., }ackson, M.W. & Wofford, .A. (1998): Innovalive surface vave
pIasna reaclor lechnique for IIC alalenenl, |ntircn. Sci. Tccnnc|. VoI. 32, No. 5,
pp 682-687, ISSN OO13-936X.
Iauchais, I. & VardeIIe, A. (1997): ThernaI pIasnas, |||| Trans. P|asma Sci. VoI. 25, No. 6,
pp 1258-128O, ISSN OO93-3813.
Creen, K.M., orass, M.C., Woskov, I.I., IIores, C.}., Haddi, K. & Thonas, I. (2OO1):
LIeclronic excilalion lenperalure profiIes in an air nicrovave pIasna lorch, ||||
Trans. P|asma Sci. VoI. 29, No. 2, pp 399-4O6, ISSN OO93-3813.
CulsoI, A. & akken, }.A. (1998): A nev vorlex nelhod of pIasna insuIalion and expIanalion
of lhe Ranque effecl, ]. Pnqs. D. App|. Pnqs. VoI. 31, No. 6, pp 7O4-7O8, ISSN 1361-
6463.
Hong, Y.C., Kin, }.H. & Uhn, H.S. (2OO3): Alalenenl of CI
4
ly alnospheric pressure
nicrovave pIasna lorch, Pnqs. P|asmas VoI. 1O, No. 8, pp 341O-3414, ISSN 1O89-
7646.
Hong, Y.C., Kin, }.H. & Uhn, H.S. (2OO4): SinuIaled experinenl for eIininalion of chenicaI
and lioIogicaI varfare agenls ly naking use of nicrovave pIasna lorch, Pnqs.
P|asmas VoI. 11, No. 2, pp 83O-835, ISSN 1O89-7646.
Hong, Y.C., Uhn, H.S., Kin, H.S., Kin, M.}., Han, S.H., Ko, S.C. & Iark, S.K. (2OO5):
Deconposilion of phosgene ly nicrovave pIasna-lorch generaled al alnospheric
pressure, |||| Trans. P|asma Sci. VoI. 33, No. 2, pp 958-963, ISSN OO93-3813.
Hong, Y.C., Cho, S.C., ang, C.U., Shin, D.H., Kin, }.H., Uhn, H.S. & Yi, W.}. (2OO6):
Microvave pIasna lurner and lenperalure neasurenenls in ils fIane, App|. Pnqs.
|c||. VoI. 88, No. 2O, pp 2O15O2-2O15O4, ISSN OOO3-6951.
Fuel njection 208

Hong, Y.C., Shin, D.H. & Uhn, H.S. (2OO7): SinuIaled experinenl for eIininalion of air
conlaninaled vilh odorous chenicaI agenls ly nicrovave pIasna lurner, App|.
Pnqs. |c||. VoI. 91, No. 16, pp 1615O2-1615O4, ISSN OOO3-6951.
Hong, Y.C., Lho, T., Shin, D.H. & Uhn, H.S. (2O1O): RenovaI of fIuorinaled conpound gases
ly an enhanced nelhane nicrovave pIasna lurner, ]pn. ]. App|. Pnqs. VoI. 49, No.
1, pp O171O1-O171O6, ISSN 1347-4O65.
Huang, L., Nakajo, K., Ozava, S. & Malsuda, H. (2OO1): Deconposilion of dichIoronelhane
in a vire-in-lule puIsed corona reaclor, |ntircn. Sci. Tccnnc|. VoI. 35, No. 6, pp
1276-1281, ISSN OO13-936X.
Kin, }.H., Hong, Y.C., Kin, H.S. & Uhn, H.S. (2OO3): SinpIe nicrovave pIasna source al
alnospheric pressure, ]. Kcrcan Pnqs. Scc. VoI. 42, No. 92, pp S876-S879, ISSN 1976-
8524.
Kin, }.H., Hong, Y.C. & Uhn, H.S. (2OO7): inary oxide naleriaI nade fron a nixlure of Zn
and Cd in a nicrovave pIasna, Cncm. Pnqs. |c||. VoI. 443, No. 1, pp 122-126, ISSN
OOO9-2614.
Lai, W., Lai, H., Kuo, S.I., Tarasenko, O. & Levon, K. (2OO5): Deconlaninalion of lioIogicaI
varfare agenls ly a nicrovave pIasna lorch, Pnqs. P|asmas VoI. 12, No. 5, pp
O235O1-O235O6, ISSN 1O89-7646.
Margol, }. (2OO1): Sludies of enission speclra in heIiun pIasnas al alnospheric pressure and
IocaI lhernodynanicaI equiIilriun, Pnqs. P|asmas VoI. 8, No. 5, pp 2525-2532, ISSN
1O89-7646.
Masuya, C., Takila, K., Takahashi, K., Takalori, I. & Ohzeki, H. (2OO2): Lffecls of airslrean
nach nunler on H/N pIasna igniler, ]. Prcpu|. Pcucr VoI. 18, No. 3, pp 679-685,
ISSN O748-4658.
Moisan, M., Zakrzevski, Z. & Roslaing, }.C. (2OO1): Waveguide-lased singIe and nuIlipIe
nozzIe pIasna lorches: lhe TIACO concepl, P|asma Scurccs Sci. Tccnnc|. VoI. 1O, No.
3, pp 387-394, ISSN 1361-6595.
Moon, S.Y., Choe, W., Uhn, H.S., Hvang, Y.S. & Choi, }.}. (2OO2): Characlerislics of an
alnospheric nicrovave-induced pIasna generaled in anlienl air ly an argon
discharge exciled in an open-ended dieIeclric discharge lule, Pnqs. P|asmas VoI. 9,
No. 9, pp 4O45-4O51, ISSN 1O89-7646.
Ogava, S., Sakai, Y., Salo, K. & Sega, S. (1998): InfIuence of nicrovave on nelhane-air
Ianinar fIanes, ]pn. ]. App|. Pnqs. VoI. 37, No. 1, pp 179-185, ISSN 1347-4O65.
Ienelranle, .M., Hsiao, M.C., Merrill, .T., VoglIin, C.L., Kulhi, A., urkharl, C.I. &
ayIess, }.R. (1997): Idenlificalion of nechanisns for deconposilion of air
poIIulanls ly non-lhernaI pIasna processing, P|asma Scurccs Sci. Tccnnc|. VoI. 6,
No. 3, pp 251-2598, ISSN 1361-6595.
IIunl, I.C. & Ryan, K.R. (1998): Cas-phase reaclions in pIasnas of SI
6
vilh O
2
: Reaclions of
I vilh SOI
2
and SO
2
and reclions of O vilh SOI
2
, P|asma Cncm. P|asma Prcccss. VoI.
9, No. 3, pp 4O9-42O, ISSN O272-4324.
Shi, Y., Ruan, }., Wang, X., Li, W. & Tan, T. (2OO5): Deconposilion of nixed naIdoranls in a
vire-pIale puIse corona reaclor, |ntircn. Sci. Tccnnc|. VoI. 39, No. 17, pp 6786-6791,
ISSN OO13-936X.
Takila, K., Masuya C., Salo T. & }u, Y. (2OO1): Lffecl of addilion of radicaIs on lurning
veIocily, A|AA ]. VoI. 39, No. 4, pp 742-744, ISSN OOO1-1452.

Uhn, H.S. (1999): Iroperlies of pIasna fIanes generaled ly eIeclricaI lreakdovn in fIanes,
Pnqs. P|asmas VoI. 6, No. 11, pp 4366-4674, ISSN 1O89-7646.
Uhn, H.S., Hong, Y.C. & Shin, D.H. (2OO6): IIasna fIane for nass purificalion of
conlaninaled air vilh chenicaI and lioIogicaI varfare agenls, App|. Pnqs. |c||. VoI.
89, No. 12, pp 1215O4-1215O6, ISSN OOO3-6951.
Weinlerg, I.}., Hon, K., Oppenhein, A.K. & Teichnan, K. (1978): Ignilion ly pIasna jel,
Na|urc VoI. 272, No. 5651, pp 341-343, ISSN OO28-O836.
Woskov, I.I. & Haddi, K. (2OO2): Large eIeclrodeIess pIasnas al alnospheric pressure
suslained ly a nicrovave guide, |||| Trans P|asma Sci. VoI. 3O, No. 1, pp 156-157,
ISSN OO93-3813.
Fuel njection 210
X

The bIast furnace trazabiIity by heIium

RafaeI area^, Rann Marln D*, I. Ruiz uslinza*
and }avier Mochn*
` Nc|rija Unitcrsia, Pirinccs, 55, 28040 Mari Spain
Ccn|rc Nacicna| c |ntcs|igacicncs Mc|a|urgicas (C|N|M cf CS|C)
Ata. Grcgcric c| Amc, 8, 28040-Mari. Spain

1. Introduction
Iasl furnaces (I) nay le considered lo le anong lhe oIdesl lype of induslriaI equipnenl
lhal is sliII used reguIarIy loday. Iasl furnaces exisled in China fron aloul lhe 5lh cenlury
C, and in lhe Wesl fron lhe High MiddIe Ages. They spread fron lhe region around
Nanur in WaIIonia (eIgiun) in lhe Iale 15lh cenlury, leing inlroduced lo LngIand in 1491.
Wilh a fev varialions, lhey have leen used for over 6OO years.
The sleeI induslry uses lIasl furnaces lo chenicaIIy reduce iron ore (nainIy oxides),
renoving oxygen and enriching lhe ore inlo nelaIIic iron vilh a high degree of
nelaIIizalion (lolaI iron conlenl higher lhan 95-96), fron vhich is ollained pig iron
(Aranguren, & MaIIoI, 1962) vhich is lhe rav naleriaI used in lhe inlegraled sleeI induslry
lo nanufaclure sleeI (UNLSID, 1998).
A lIasl furnace is a verlicaI reaclor conprising an assenlIy of cyIindricaI or conicaI
eIenenls (Iigure 1). Anong ils nosl inporlanl fealures are lhe inlernaI voIune, vhich is
delernined ly lhe dianeler of lhe crucilIe, lhe furnace lollon, defined ly lhe desired
resuIls, vhich coIIecls lolh noIlen nelaI and sIag. Thus a furnace crucilIe vhich is 12
nelers in dianeler can produce nore lhan lhree niIIion lons of pig iron annuaIIy.
The lIasl furnace is a counlercurrenl reaclor in vhich lhe reducing gas is produced ly coke
gasificalion vilh lhe oxygen lIoving via luyeres. The reducing gas fIovs upvards reducing
lhe iron ores charged al lhe lop of lhe furnace. The lIasl furnace process is very conpIex
vilh nany infIuencing and correIaling faclors (SleiIer, 1998, Kundral, 1989).

11
Fuel njection 212

Iig. 1. A Iasl furnace diagran:(1) Hol lIasl fron Covper Slones (2). Hearlh (neIling zone).
(3). eIl (reduclion zone of ferrous oxide) (4). Slack (reduclion zone of ferric oxide (5). Throal
(pre-healing zone) (6). Ieed of ore, Iineslone and coke (7). Lxhausl gases (8) CoIunn of ore,
coke and Iineslone (9). RenovaI of sIag (1O). Tapping of noIlen pig iron (11). CoIIeclion of
vasle gases

2. The BIast Furnace
2.1. The main parts of bIast furnaces
The prinary syslens and nosl inporlanl conponenls are:
(a) Cas Lvacualion Syslen: Consisls of drainage pipes for lhe gas produced and cones
equipped vilh reIief vaIves and conlroIs for pressure, lenperalure, gas
conposilion, elc. (ilens 7 and 11 of lhe Iigure 1).
(l) Oven ody: The oven ilseIf, luiIl vilh refraclory naleriaIs vilh differenl
characlerislics, depending on lhe needs of each area and lhe furnace conslruclion.
Ils design provides for lhe operalions required, and lhe chenicaI exchange
lelveen soIids and gases, vhich coexisl in various proporlions in each area. The
design of lhe kiIn is varialIe, aIong vilh ils heighl, dislinguishing differenl areas,
vhich are descriled Ialer.
(c) CooIing Syslen: Varies according lo each zone in order lo provide lhe nosl
inlense, efficienl perfornance lo naxinize syslen reIialiIily, refraclory Iining
duraliIily, and, consequenlIy, oplinize lhe duralion of lhe furnace canpaign. Il
nornaIIy requires severaI cIosed cooIing circuils, in addilion lo conlroI syslens for
lenperalure, pressure, fIov, heal Ioss caIcuIalions, deleclion of Ieaks or faiIures of
refrigeraled unils, elc.
(d) ConlroI Syslen: To nanage lhe operalion of lhe furnace equipnenl, lhis syslen is
nade up of a Iarge nunler of sensors (lenperalure conlroIs, pressure, IeveI
sensors, gas anaIysis sensors, elc.) vhich, logelher vilh ils anciIIary faciIilies, forn
a gIolaI conlroI syslen.
Inside lhe lIasl furnace lhere are differenl chenicaI zones:
(a) Throal (see Iigure 2 and Ilen 5 in lhe Iigure 1): The lop of lhe oven vhere rav
naleriaIs such as iron ore, coke and fIux are charged. Iorned ly slraighl vaIIs,
Ioading shouId ensure a conlroIIed dislrilulion of each naleriaI inside lhe oven.
CurrenlIy lhe nosl connon syslen for conlroIIing Ioad dislrilulion, designed ly
IauI Wurlh, is a syslen using a gear lox, aIIoving peIIels lo le dovnIoaded inside
lhe oven in a honogeneous vay.
(l) Slack: (see Iigure 2 and Ilen 4 in lhe Iigure 1) Iocaled lelveen lhe end of lhe
hopper and lhe leIIy of lhe furnace, il conlains nosl of lhe Ioad and has a
lruncaled cone shape. This parl of lhe lIasl furnace has lhe jol of encouraging lhe
exchange of heal lelveen soIids and gases.
(c) eIl : (see Iigure 2 and Ilen 3 in lhe Iigure 1) This is lhe vider parl of lhe furnace,
and ils cyIindricaI in shape vilh slraighl vaIIs. As lhe charge descends and lhe
lenperalure ascends, nore space is needed as voIune increases. This area of lhe
kiIn is lhe Iink lelveen lhe slack and lhe hearlh.
(d) Hearlh : (see Iigure 2 and Ilen 2 in lhe Iigure 1):This is lhe furnace area vhere air
is injecled inlo lhe hol lIasl fron lhe sloves. The varn air reaches a circuIar lule
lhal is inserled inlo lhe furnace lhrough hoIes in nozzIes. Ils nission is lo enalIe
lhe conluslion of coaI. AuxiIiary fueIs are aIso injecled lhrough nozzIes. The
nunler of nozzIes, vhich are aIvays Iocaled equidislanlIy around lhe furnace,
varies fron one oven lo anolher depending on lhe oven size.
(e) CrucilIe: (see Iigure 2) This cyIindricaI shaped parl is lhe lollon of lhe furnace,
vhere lhe producls ollained in lhe process, pig iron and sIag, are coIIecled lhrough
a lrough. The crucilIe is lhe nosl crilicaI parl of lhe pIan for lhe seIeclion and
inslaIIalion of refraclories. The naleriaI used is carlon (2 n lhick in lhe vaIIs and
2-3 n for lhe lollon). In aII olher areas of lhe I, lhe refraclories are generaIIy of
an aIuninous lype and, in sone specific areas, nore duralIe naleriaIs (SiC or SiaI)
are used.

2.2. BIast furnace controI
ConlroI of lhe lIasl furnace process requires significanl neasures lo ensure adequale
progress and nanufacluring faciIilies reIaled lo lhe naking and sinlering of coke. This is lo
nainlain a slalIe oven operalion and lo conlroI and prolecl lhe funclioning of lhe
equipnenl inslaIIed. Wilh lhis in nind, sensors and neasuring equipnenl are used lo
conlroI lhe novenenl of naleriaIs and fIuids in and oul of lhe oven.
Specific conlroIs incIude:
(a) Inpul: rav naleriaIs, nain and auxiIiary fueIs, hol lIasl, oxygen injecled inlo lhe
lIasl, slean injecled vilh lhe hol lIasl (conlroI of hunidily, elc.).
Fuel njection 214
(l) Inside lhe furnace ilseIf: Screening and Ioading of rav naleriaIs, hopper and
syslen of dislriluling charge, pressures and lenperalures in differenl parls of lhe
oven, charge IeveIs, charge sanpIes in differenl areas, elc.
(c) Oulpul: iron and sIag, gases, dusl and sIudge recovered, and so on. Of parlicuIar
inporlance are conlroIs for lhe dislrilulion of lhe hopper charge and lhe
nelaIIurgicaI varialIes lhal define lhe process, draving on nelaIIurgicaI nodeIs of
laIancing heal and naller, in addilion lo slalislicaI quaIily conlroI, arlificiaI
inleIIigence, elc..
AII of lhese faclors conline lo achieve slalIe operalion of lhe I in order lo increase
produclivily, inprove producl quaIily (pig iron, sIag), increase lhe duralion of lIasl furnace
funclionaIily and reduce cosls (nininun consunplion of coke and auxiIiary fueI, elc.).

Iig. 2. The nain parls of lIasl furnaces
One of lhe key aspecls lhal infIuence produclivily (l/n3 usefuI voIune / day), produclion
(l / year) and duralion of a lIasl furnace canpaign, neasured in lolaI oulpul in reIalion lo
lhe size of lhe oven, is oplinaI conlroI of lhe process lhal resuIls in sleady oven produclion.

Olher aspecls lhal infIuence produclion, produclivily, and canpaign duralion are:
- The I design (profiIe, crucilIe dianeler, elc.)
- The slralegy chosen lo ollain a parlicuIar producl (iron quaIily, elc.)
- Coke quaIily and lhe use of auxiIiary fueIs such as coaI
- The anciIIary equipnenl vhich faciIilales peak charge dislrilulion (hopper), high
lenperalure lIover (lIasl), oplinaI oxygen injeclion, vork counler, IeveI and
reguIarily of lhe furnace funclionaIily, and so on
- The originaI design quaIily of lhe equipnenl, such as lhe linoniaI refraclory /
cooIing, equipnenl reIialiIily, appIied lechnoIogies, elc.
- The quaIily of operalion and nainlenance processes
- The quaIily and lype of rav naleriaIs used (percenlages and lypes of sinler,
peIIels, elc.).

2.3. BIast furnace functions
The funclions lo le perforned ly a I are quile conpIex, if lhe process is lo resuIl in lhe
required produclion voIune in an econonicaI fashion. These funclions incIude a nunler of
nechanicaI operalions, since lhe furnace needs lo conslanlIy and sleadiIy lring lhe charge of
soIid naleriaIs (coke, ore and fIux) leing pIaced in lhe hopper in rapid and evenIy
dislriluled conlacl vilh lhe ascending gas slrean.
Ior lhis lo happen on a reguIar lasis, lhe charge nusl le highIy porous and very veII
dislriluled so lhal ils inleraclion vilh lhe hol gas fIov is honogeneous. The grealer or
Iesser porosily of lhe charge depends on lhe shape and dinensions of lhe pieces leing used,
vilh grealer porosily leing associaled vilh spheres of equaI size, langenl and cenlers al lhe
verlices of a cule. In lhis case, lhe enply spaces Iefl lelveen lhe spheres represenl 52.3 of
voIune. If lhe spheres vere of equaI langenl, lul vilh cenlers al lhe verlices of a rhonloid,
lhe enply spaces vouId le 47.7 al nosl.
In praclice, lhe pieces of ore are nol sphericaI and do nol have lhe sane dinensions, so fine
pieces caughl in Iarger pieces can significanlIy Iover lhe porosily of lhe Ioad. Il can le seen,
lherefore, lhal preparing and cIassifying lhe charge is inporlanl, sinlering lhe fine pieces
vhiIe Ieaving lhe Iarger reIaliveIy honogenous ones unlrealed, as is lhe lrend in I
operalions. The lype of hopper, Wurlh or vilh leIIs, and lhe Ioading pallern are aIso
essenliaI for oplinaI dislrilulion.
In lhe case of using a hopper vilh leIIs, IHI or MOHR, lhe incIinalion of lhe cIosing leII, lhe
reIalive dianeler of lhe leII and lhe hopper and lhe Ienglh of lhe leII, aII infIuence lhe forn
of dislrilulion of soIid naleriaIs in lhe oven hopper. The specific veighl of lhe naleriaIs in
lhe charge and ils angIe of repose are aIso reIevanl.
In addilion, lhe I nusl dislrilule lhe Ioad so as lo pernil a reguIar descenl vilhoul
forning parlicuIar passagevays, vhich vouId give rise lo prolIens. Nor shouId lhe charge
preferenliaIIy descend ly lhe furnace vaIIs or lhrough lhe cenler of lhe hol gases, vhich
neans lhal lhe dislrilulion of lolh fine and Iarger pieces nusl le carefuIIy conlroIIed
lhroughoul lhe area of descenl.
esides aII lhis, as lhe descending charge undergoes changes in conposilion, lenperalure
and voIune, lhe design of lhe I nusl pronole lhese changes and lhe rapid, sleady and
consislenl descenl of lhe charge, sIag and iron fornalion and casling al lhe lollon, as veII
as lhe reIease of gases al lhe lop.

Fuel njection 216
2.4. The deadman
The parl of lhe furnace caIIed lhe deadnan, lelveen lhe leIl and lhe crucilIe, grealIy
inpacls lhe enlire process. Il pIays a key roIe in hol nelaI quaIily. The condilion of lhe
deadnan has a slrong infIuence on hol nelaI lenperalure and conposilion, and fIov
condilions. When hol nelaI can fIov freeIy lovards lhe lap hoIe and lhe deadnan is a
porous coke led, lhen condilions are good for lhe desuIphurizalion and carlonizalion of hol
nelaI. The deadnan aIso has a significanl effecl on Iining vear and canpaign Ienglh as il
conlroIs hol nelaI fIov in lhe hearlh.

2.5 Improving the Productivity
The lIasl furnace produclivily is lhe quolienl lelveen possilIe gas lhroughpul per unil of
line and required specific gas generalion for one lonne of hol nelaI ollained (Harling el aI.,
2OOO). ConsequenlIy, a produclivily increase requires an increase in lhe gas lhroughpul,
vhich inpIies, on lhe one hand, inprovenenl in furnace pernealiIily and, on lhe olher
hand, a reduclion in lhe specific gas requirenenls so lhen finaIIy neans a reduclion in lhe
specific consunplion of reducing agenls. IernealiIily is a neasure of lhe gas aliIily lo pass
lhrough lhe led of soIid naleriaIs, if lhe pernealiIily in lhe furnace is higher lhen lhe
furnace lurden novenenl and lhe reducing gas fIov lhrough lhe furnace are leller
(Iandey el aI., 1996).

The search for inprovenenl of overaII lIasl furnace pernealiIily resuIls lhus in furlher
inprovenenls in lhe foIIoving fieIds:
- lurden conposilion and quaIily,
- lehaviour of lurden during reduclion,
- lhe cohesive zone shape and posilion conlroI
- Iiquid evacualion fron lhe hearlh (sIag quanlily, coke size and Iiquid fIov condilions in
lhe hearlh and lapping praclices).

In aII lhese fieIds, lhe aclions aining al pernealiIily inprovenenls are generaIIy leneficiaI
for lhe consunplion of reducing agenls, naking il possilIe lo decrease lhe specific gas
consunplion, and lhus lo ollain high produclivily easier. eyond lhal, lhe oxygen
enrichnenl of lhe lIasl is lhe najor vay lo decrease lhe specific gas voIune. Hovever, lhis
lhing viII onIy le appIicalIe in conlinalion vilh luyere injeclion in order lo nainlain a
sufficienl gas quanlily lo overheal lhe lurden (lop gas lenperalure higher lhan 1OOC), in
addilion, lhe oxygen enrichnenl of lhe lIasl viII le adjusled lo nainlain lhe RAIT
(Racevay Adialalic IIane Tenperalure) in a sufficienlIy high vaIue lo ensure a good
gasificalion of lhe reducing agenl injecled, lul nol loo high lo avoid nechanicaI prolIens
on luyeres (Iornoso el aI. 1999). The oljeclive "high IeveIs of injeclion of puIverized coaI"
(and lhus Iov coke consunplion) is nol onIy conpalilIe vilh produclivily, lul aIso even
necessary lo increase lhe lIasl furnaces produclivily (alich el aI. 1996) (alich el aI. 1999).

IernealiIily of lhe ferrous lurden and coke coIunn for lhe gas fIov is Iinked logelher vilh
lhe increases of gas lhroughpul. A lurden coIunn (slruclure of charged naleriaIs, ores and
coke, vilh properlies lhal exlensiveIy assure lhe necessary void for an adequale
pernealiIily, in spile of nechanicaI, lhernaI and neIling area) are dovn lo lhe Iover
furnace. The nain funclion of coke is guaranleeing lhe pernealiIily for lhe gas in lhe dry
region alove lhe cohesive area, in lhe cohesive area ilseIf and in lhe hearlh. In lhe cohesive
zone lhe coke has an inporlanl roIe lecause lhe soflening and neIling iron-learing
naleriaIs can forn an inpernealIe Iayer (usly el aI.1994)(Serl el aI, 2OO4). Therefore,
nany quaIily crileria for ferrous lurden naleriaIs and coke have leen defined vilh ever
increasing denands. In aII cases, lhe generaI praclice has shovn lhal il is advisalIe lo
prepare an honogeneous nix of aII iron learing conponenls lefore charging lo lhe lIasl
furnace, lo achieve exceIIenl pernealiIily and suilalIe neIling lehaviour.

Then lhe gas dislrilulion conlroI is lhe resuIls of a conpronise anong differenl
requirenenls (SleiIer, 1998):
To achieve a conlroI gas fIov in order lo nainlain lhe shafl pernealiIily in spile of higher
IeveIs of injeclion of puIverized coaI. .
To adjusl lhe gas fIov aIong lhe vaII lelveen lvo Iinils, in order lo guaranlee Iov heal
Iosses and lhe alsence of scaffoId sinuIlaneousIy
To achieve efficienl gas soIid fIov in order lo pronole efficienl reducing condilions and Iov
reduclanl rale.

The lIasl furnace oplinizalion inpIies lhal lhe process noniloring has reached a high
slandard lo conlroI lhe process in ils inner parl. SpeciaI inporlance has lhe oplinisalion and
conlroI of conposilion and gas dislrilulion and lhe pressure Iosses inside lhe furnace.

SeveraI indirecl neasurenenls (lhernaI Iosses, gas anaIysis and lenperalure recorded ly
lurden proles) heIp lo caplure infornalion over lhe gas dislrilulion in lhe lIasl furnace
(Nikus & Saxen, 1996)(NicoIe el a. 2OOO). Wilh lhe oljeclive lo deveIop nev looIs lo inprove
lhe gas dislrilulion noniloring lhe heIiun lracing lechniques has leen lesled (HaveIange,
2OOO). This nelhod is enpIoyed lo nonilor lhe gas lransfer line fron luyere lo lhe lurden
prole. The anaIysis of lhis lransfer line fron lhe luyere lo lhe lop and ils reIalionship vilh
lhe process vouId le expIained.

3. ExperimentaI method
3.1. Theory
HeIiun lracing lechnique consisls in lo injecl heIiun in lhe lIasl furnace al lhe luyeres IeveI
and ils arrivaI al lhe lIasl furnace lop is delecled ly a nass speclroneler. The speclroneler
indicales lhe heIiun conlenl in oulloard gas and il is possilIe lo define lhe lransfer line as
lhe deIay lelveen lhe injeclion nonenl and lhe line vhen heIiun concenlralion reaches 1O
percenl of lhe naxinun delecled IeveI.

Inside lhe furnace lhe gas conposilion has an evoIulion fron lhe luyeres (vhere lhe gas is
nainIy conposed for O
2
and N
2
) lo rolaling chule (vhere lhe gas is nainIy conposed for
CO
2
, CO, H
2
and N
2
). The iniliaI oxygen, lhe oxygen produced in lhe deconposilions of
oxides and valer, parlicipales in lhe produclion of CO, CO
2
and H
2
O. The delecled He ppn
vouId le Iover vhen nore nass of gases (CO
2
, H
2
and CO) is produced. Using lhis dala,
ils possilIe lo ollain an indirecl neasure of lhe oxidanl reaclions lhal has leen produced
inside lhe furnace. The air is nore veighlalIe afler oxidanl reaclion lhen lhe neasured of
He ppn is Iover. As gas conposilion inlroduced in luyeres is siniIar and lhe He conlenl is
Fuel njection 218
lhe sane, lhe naxinun He ppn conlenl neasured in lop of lhe furnace can le reIaled vilh
lhe reaclivily inside lhe furnace.

In lhis vork lhe charge in lhe lIasl furnace viII le lrealed Iike ceranic foans. In aII cases
perpendicuIar seclion lo lhe fIov and lhe average veIocily nay le lased on lhe enlire cross
seclionaI area.

Using lhe Iake-Kozeny equalion or Kozeny-Carnan equalion, vhen lhere is a Ianinar fIov
in a porous nediun, and lhe urke-IIunner equalion, vhen lhe fIov is lurluIenl, lhen lhe
generaIised Lrgun`s equalion (Lrgun, 1957) is ollained for aII-Ianinar lo aII lurluIenl
fIov.

( ) ( )
2
2 2 3 3
150 1 1, 75 1
S S
3 X X
' '
| q |
| |
| |

| V = +
|
\ .
G G
(1)

vhere,
I = pressure (Ia)
u = veclor of veIocily (n/s)
= shape faclor (-)
q = viscosily of lhe fIuid (Ia s)
= fIuid densily (Kg/n
3
)
= porosily (-)
Dp = parlicIe dianeler of lhe granuIar nediun (n)

In olher vay Darcy`s Iav (Cisonni 2OO3)(Hager 1994) vas lased on experinenls vilh lhe
pressure drop p neasured over a finile Ienglh, L, in a sand pack of pernealiIily lensor, k,
and cross-seclionaI area, A. Ior Iinear fIov of an inconpressilIe fIuid of viscosily q lhrough
lhis sand pack, lhe fIov rale is reIaled lo olher faclor.

N 3
X
q
V
=
G
(2)

The Darcys equalion has leen used lo eslalIish a reIalionship lelveen lhe fIuid speed and
lhe pernealiIily as funclion of lhe difference of pressures and gas viscosily. The gas speed
is ollained dividing lhe dislance lraveIIed ly lhe lransfer line. Making lhe assunplion lhal
lhe inslanl pernealiIily is conslanl in lhe furnace (an averaged pernealiIily) lhen lhe
pernealiIily lensor is sinpIified lo a conslanl. In lhis vork lhe dislance lraveIIed ly lhe
gases and lhe gas pernealiIily viII le supposed conslanls for aII lhe neasures. The dislance
lraveIIed ly lhe gases and lhe pernealiIily of lhe furnace k, can le caIcuIaled fron lhe
equalion (3)


3 W
&
N
V
=
(3)

Where C is a conslanl and l is lhe lransfer line. Il is possilIe lo ollain lransfer lines in six
poinls using lhe in-lurden prole, so il is possilIe lo eslalIish a nean pernealiIily lurden
dislrilulion fron lhe luyeres up lo lhe lurden prole.
The sanpIing of gas for ils anaIysis lhrough lhe in-lurden prole vhal aIIovs conparing
direclIy lhe resuIls of lhe heIiun lracing lechnique vilh lhe gas conposilion and
lenperalure ollained ly neans of lhe prole.
Ior lhe lurden prole configuralion, lhe journey carried oul ly lhe gas fron lhe injeclion
unliI lhe deleclor, does nol vary vilh lhe neasurenenl poinl, lecause lhe prole is a lurden
lule in vhose end is lhe deleclor and fiIler. Ils fIov rale is neasured lo delecl cIogging of
lhe sanpIing Iine. The heIiun concenlralion is neasured ly a nass speclroneler.
To cIean lhe pipes lefore lhe injeclion of heIiun, a lhree vays eIeclrovaIve is allached lo lhe
syslen lo Iel lhe injeclion of N
2
. This viII ensure nol onIy lhe cIeaning of lhe pipes lul aIso
viII avoid any faIse neasurenenl of heIiun concenlralion and line lransfer due lo lhe
presence of residuaI heIiun in lhe pipes.

0 10 20 30 40 50 60 70 80
2
3
4
5
6
7
8
9
q
C
O

+

H
2
He ppm

Iig. 3. Il can see lhal higher naxinun He ppn are delecled, Iover gases area reacling inner
lhe furnace

Fuel njection 220
2.2. Measurements
The prole neasures in 6 poinls aIong a radius of lhe furnace fron nearesl area lo lhe vaII
up lo 4, 96 nelers of lhe vaII, near lhe furnace cenlre. When lhe lenperalure in lhe cenlre is
very high (holler lhan 95O C) lhe prole is nol inlroduced unliI lhe Iasl poinl, and in lhose
cases lhe Iasl neasures are Iosl.

l1
(s)
l2
(s)
l3
(s)
l4
(s)
l5
(s)
l6
(s)
Max1
(ppn)
Max2
(ppn)
Max3
(ppn)
Max4
(ppn)
Max5
(ppn)
Max6
(ppn)
99.3 2O4.7 26O.7 417.9 338.9 517.5 18.O 17.9 1O.8 14.6 11.4 1O.7
31.3 167.7 334.2 388.9 364.7 32 25.4 15.3 12.2 1O.9
29.2 1O3.9 1O8.7 117.5 198.O 433.5 34.1 11.6 1O 8.7 8.4 8.2
28.2 1O2.6 159.6 189.O 298.3 368.9 37.1 9.7 11.4 9.1 1O.7 8
62.2 121.2 135.O 26O.9 342.1 486.7 32 37.6 13.1 1O.4 1O.9 13
2O.5 122.3 179.2 258.O 397.2 48.1 19.2 41.5 25.2 19.2
123.O 181.7 299.3 288.8 3O3.2 337.2 31.3 37.9 62.7 42.6 31.1 34.3
99.O 117.7 123.9 186.4 283.2 25.7 73.3 43.2 25.6 1O.3
153.3 2O5.3 323.6 356.8 536.5 7O.3 2O.1 9.1 8.4 11.9
136.3 217.8 3OO.4 328.O 478.7 64 27 9.7 15.9 29.6
24O.8 249.1 247.O 182.1 454.5 75.2 69.5 2O.7 57.9 65.3
84.4 99.2 147.8 275.5 144.5 517.1 23.2 19.5 8.1 27.4 54.5 17.5
176.4 268.2 262.7 438.2 212.8 16 1O.4 25.8 2O.4 34.8
39.O 248.7 379.7 4O6.7 397.1 39.3 43.8 14.6 1O.7 1O.8
119.6 116.8 181.1 259.3 478.8 17.8 41.6 19.3 18 1O
112.6 128.5 127.4 194.9 317.O 49.6 26.3 57.2 69.2 16
135.6 233.2 369.9 433.8 496.1 29.5 73.5 1O 29 13
116.O 226.1 281.9 342.6 435.8 448.9 24.4 69.7 2O.8 4O.5 68.2 51.1
45.1 288.2 284.5 297.2 312.1 3O8.O 38.8 9.2 1O 38.6 49.8 17.1
96.1 145.3 217.2 373.9 435.5 548.5 49.3 13 8.6 12.1 9.5 8.6

TalIe 1. ResuIls of lhe neasurenenls reaIised vilh He lracing lechnique. On lhe Iefl of lhe
lalIe il can le olserved lhe nornaIised lransfer line in aII poinls of neasure fron vaII
(poinl 1) up lo cenlre of furnace (poinl 6). On lhe righl side of lhe lalIe ve can read lhe He
concenlralion al lhe lop or furnace. When lhe furnace vas holler lhe Iasl poinl (in lhe cenlre)
did nol neasure.

Il is necessary lo correcl lhe lolaI lransfer line previousIy ollained (li) lecause lhe gases are
picked up lo differenl lenperalures. When a furnace area is higher lenperalure, lhe gases
go up quicker and lhe lransfer line is reduced, lul lhal does nol indicale lhal lhe furnace
has a leller reaI pernealiIily. As lhe hollesl furnace is lhe cenlraI area if lhe infIuence of
lenperalure is nol correcled, seens as if a ligger pernealiIily in lhe cenlraI area and a
snaIIer fIovn in lhe vaII exisl. Then, il viII le used a nornaIised lransfer line (l
n
) (equalion)
lhal viII le independenl of lhe lenperalure (T
l
) inside lhe lIasl furnace. l
n
enalIes nore
reIialIe conparisons lhan l
i
vhen lhe gas fIovs fasler lecause lhe gas lenperalure is
higher aIlhough lhe pernealiIily inside lhe furnace is nol leller.

L
1
E
Q
W
7
7
W =
(4)

The heIiun concenlralion vas neasured ly a speclroneler and ils naxinun vas enpIoyed
lo predicl lhe gases concenlralion in lhe furnace. Ils concenlralion is a ralio lelveen heIiun
nass and air nass and aIlhough heIiun nass is conslanl, lecause is an inerl gas, lhe air
nass change vilh lhe reaclions produced in lhe furnace.

$LU
+H
PHDVXUHG +H
P
P
SSP =
(5)
3. Discussion
The quanlilies of CO and CO
2
al lhe lop gas are reIaled lo each olher in lhe paraneler elaCO
(
2
2
&2 &2
&2
&2
+
=
) caIcuIaled in each poinl of neasure, hovever lhe IineaI correIalion
anong naxinun He ppn in each poinl and neasured elaCO is aIvays negalive and snaII
(Iover lhan O.2) vhal vouId indicale lhal il is necessary lo consider nore infIuence faclors.
The delecled fraclion of H
2
in lhe lop of furnace is higher lhan in lhe air due lo lhe inner
reaclion in furnace. ConsequenlIy, lhe conpounds forn heavier gases, lherefore, lo nore H
2

Iess He ppn shouId le delecled. Whelher elaCO and H
2
are added, a negalive correIalion
vilh delecled He ppn of lhe O.3 is ollained (see Iig. 3). Il is possilIe lhal il exisl olher
varialIes lhal infIuence in lhis reIalion lul ve did nol sludy in lhis paper. Therefore, higher
naxinun He ppn are delecled, Iover gases are reacling inner lhe furnace.

On lhe olher hand, heIiun enpIoys nore line ascending vhen lhe lurden is Iess pernealIe
so lhal lhere is nore line for lhe inner reaclions lakes pIace. Il is possilIe lo have an
indicalor aloul hov lhe reaclion in lhe lurden evoIves if il is conpuled lhe ralio lelveen
lolh varialIes (line in seconds/naxinun He ppn). This ralio viII nane Reaclion and
vhen lhe line increases or lhe ppn dininishes, lhis paraneler (Reaclion) increases
indicaling lhal ligger reaclion lakes pIace in lhal area. The average neasure of lhis varialIe
in each poinl is shov in Iig. 4:

The oulIiers can le used lo delecl anonaIous silualion. Ior lhis reason aII neasures
evaIualed as nornaI slale vere seIecled lo caIcuIale lhe nean vaIue of lhe varialIe
Reaclion.

Fuel njection 222
Wall 2 3 4 Near Medium
0
5
10
15
20
25
M
e
a
n

(
s
/
p
p
m
)
Measure Point

Iig. 4. In lhis figure, ve can le see lhe nean vaIue of lhe ralio lelveen nornaIised lransfer
line and naxinun He ppn. Il can le olserved lhal lhis ralio (il viII nane as Reaclion)
increase in lhe cenlre of furnace and ils snaIIer in lhe vaII.

0
10
20
30
40
50
60
70
80
1 2 3 4 5 6
Measure Point
R
e
a
c
t
i
o
n

(
s
/
p
p
m
)

Iig. 5. In lhis figure has leen represenled lhe Reaclion vaIue ollained in aII neasures vhich
lechnicians evaIualed Iike nornaI slale of furnace. In lhe figure have leen incIuded lvo
gross Iines lhal represenl lhe nean vaIue of Reaclion pIus/Iess lhe doulIe of slandard
devialion for each poinl.
AddilionaIIy lo nornaI dala lvo caIcuIaled Iines are draving using lhe nean vaIue
pIus/Iess lhe doulIe of slandard devialion for each poinl. These lvo Iines viII le
considered as operalion Iinils for a nornaI furnace slale (see Iig. 5).
To represenl aII infornalion in an oplinun exposilion, a lernary diagran represenling lhe
nornaIised varialIe naned Reaclion has leen used. The six poinls of furnace define lhree
specific vorking areas in lhe lransversaI seclion of lhe shafl: vaII, nediun and cenlraI. The
vaII area corresponds vilh lhe lvo firsl poinls neasure, lhe cenlraI area corresponds vilh
lhe lvo Iasl poinls and lhe lvo inlernediale poinls correspond vilh nediun area.
Reaclions profiIes recorded during nornaI periods are shovn in Iig. 6. The ranges of
reaclion in vhich furnace is correclIy running are W (O-25), M (4O-6O) and C (25-
5O).

Iig. 6. In lhis figure il has leen represenled in a lernary diagran aII neasures evaIualed ly
lechnicians as nornaI condilions. Il can olserved lhal aII neasures area grouped

Il can le see lhal aII poinls seIecled as nornaI running are grouped in lhe sane area of lhe
diagran. The poinls oul lhis region represenl 'non oplinun slales of furnace, i.e, if lhe
poinl is dispIaced al righl lhen lhere are nore reaclion in lhe vaII lhan in lhe cenlre and lhe
furnace cooIer can le olserved.
The Iig. 7. represenls aII neasures vhere lhe evaIualion vas cerlificaled as alnornaI
slale. The poinls represenling lhe neasures are dispersed oul of lhe area defined
previousIy, excepl in lvo cases. Il can le olserved lhal aInosl aIvays vhere lhe lechnicians
indicaled lhal lhe furnace vas in alnornaI slale lhe neasure reaIised ly heIiun lrazing
Fuel njection 224
lechnique and evaIualed vilh lhe varialIe Reaclion shoved an alnornaI silualion in lhe
lernary diagran.
Wilh lhis resuIls, il is possilIe ollain onIy one paraneler lo characlerise lhe evoIulion of
lIasl furnace.

Iig. 7. In lhe figure il has leen represenled in a lernary diagran aII neasures evaIualed ly
lechnicians as alnornaI slale. Il can olserve lhal lhese neasures are dispersed ly lhe
lernary. The origin of an alnornaI slale in each neasure can le differenl

4. ConcIusions
CaIcuIaled varialIes fron lhe neasurenenls aIIov a concise characlerizalion of lhe lIasl
furnace slale. Cas lransfer neasurenenls can le considered as a nev looI lo evaIuale lhe
slale of a furnace in a specific nonenl lecause aIlhough lhe resuIls do nol have an excessive
repealaliIily lhey can le a heIp lo lhe operalor.
The nain advanlage viII le lhal enpIoyed onIy one neasure il viII le possilIe lo evaIuale
lhe furnace.

5. References
Aranguren, I. y MaIIoI, A.: (1962) Sicrurgia. LdiloriaI Dossal, S.A., 978-84-237-OO49-3, Spain
alich, A., Yarosevskii, S., Iornoso, A., Isidro, A., Ierreira, S., Cores, A. and Carcia, L.
(1996) |ncrcasc cf pu|tcrizc cca| usc cfficicncq in ||as| furnacc ISI} Inl., 36, 125O-1258,
ISSN:O915-1559
alich, A., Yarosevskii, S., Iornoso, A., Cores, A., Carcia, L., and Nozdrachev, V. (1999), :
Cc-injcc|icn cf Ncncc|ing Cca| an Na|ura| Gas in 8|as| |urnacc ISI} Inl., 39 229-238,
ISSN:O915-1559
usly, N.}., Iray, T.A.T., and CoIdring, D.C. (1994), Na|urc cf ccncsitc zcnc in ||as| furnacc
Ironnaking and SleeInaking, 21(3) 229-236
Lrgun, S. (1952), ||ui ||cu Tnrcugn Pac|c Cc|umns, Cncmica| |nginccring Prcgrcss, Chen.
Lng. Irogress., 48 (2), 89-94
Iornoso, A., alich, A., Cudenau, H.W:, Carcia, L. and Cores, A (1999) Hca| |xcnangc in |nc
Hcar|n cf a 8|as| |urnacc Opcra|ing ui|n Ccm|inc 8|as| Paramc|crs ISI} Inl., 39 (11),
1134-1139, ISSN: O915-1559
Cisonni, C. (2OO3). "Hcnrq Darcq an |nc pipc f|cu fcrmu|a", in Henry I. C. Darcy and Olher
Iioneers in HydrauIics: Conlrilulions in CeIelralion of lhe 2OOlh irlhday of
Henry IhiIilerl Caspard Darcy, C.O. rovn, }.D. Carlrechl, and W.H. Hager (eds),
ASCL, Reslon, VA. pg. 24-36.
Hager, W. (1994). "Dic nis|criscnc |n|uic||ung cr ||icssfcrmc|". Schviezer Ingenieur und
Archilekl. 9(24), 123-133
Harling, W., Langer, K., Lngen, H.. and Slricker, K.I, (2OOO) Prcccing cf |rcnma|ing Ccnf..,
VoI 59, Iillslurg, IennsyIvania, USA, : 9781886362437, 3
HaveIange, O., DanIoy, C., Iranssen, R., and onle, L.: (2OOO), Rev. MelaII.-CIT, 97 29
HaveIange, O., DanIoy, C., Iranssen, R.(2OO4), Tnc ca man, f|ca|ing cr nc|? Rev. MelaII.-
Iariz, 3, 195-2O1
Kundral, D.M., Ccnccp|ua| mcc| cf |nc ircn ||as| furnacc ccnsicring |ncrmcqnamic an |inc|ic
ccns|rain|s cn |nc prcccss MelaII. Trans. , 2O, 2O5-218, ISSN: 1O73-5615
NicoIe, R., SleiIer, }.M., and Manin, M. (2OOO), : Rev. MelaII.-CIT, 97, 61.
Nikus, M., and Saxen, H. (1996), : Ironnaking Conf. Iroc., VoI. 55, 189.
Iandey, .D., Sen, I.K. and Challrjee, A (1996).: Ironnaking and SleeInaking, 23, 31.
Serl, D., DanIoy, C., HaveIange, O., and Saiz de AyaIa, }. (2OO4) Ncu apprcacn fcr |nc
c|crmina|icn cf |nc ||as| furnacc ccncsitc zcnc, LCSC, ISN 92-894-7491-2.
Luxenlourg,
SleiIer, }.M. (1998), Prcccing cf |rcnma|ing Ccnf., VoI. 57, Toronlo, Onlario, Canada,, 161
UNLSID, (1998), |a fa|ricacicn c| Accrc, LdiloriaI Unin de Lnpresas Siderurgicas, 978-84-
6O5-729O-9, Spain
Fuel njection 226
Experimental investigations into the production behavior of methane
hydrate in porous sediment under ethylene glycol injection and hot brine stimulation 227
X

ExperimentaI Investigations into
the Production Behavior of Methane
Hydrate in Porous Sediment under EthyIene
GIycoI Injection and Hot Brine StimuIation

Xiao-Sen Li and Cang Li
Guangzncu |ns|i|u|c cf |ncrgq Ccntcrsicn, Cnincsc Acacmq cf Scicnccs
Pccp|cs Rcpu||ic cf Cnina

1. Introduction
NaluraI gas hydrales (NCH) are soIid, non-sloichionelric conpounds forned ly hosl valer
noIecuIes vilh snaII guesl noIecuIes, such as CH
4
, C
2
H
6
, C
3
H
8
, CO
2
, H
2
S, elc (SIoan & Koh,
2OO8). NaluraI gas hydrale deposils invoIve nainIy CH
4
, and occur in lhe pernafrosl and in
deep ocean sedinenl, vhere lhe necessary condilions of Iov lenperalure and high pressure
exisl for hydrale slaliIily. Lslinales of vorId hydrale reserves are very high, and vary fron
O.2 1O
15
- 12O 1O
15
n
3
of nelhane al STI (Slandard Tenperalure and Iressure) (SIoan &
Koh, 2OO8, MiIkov, 2OO4, KIauda & SandIer, 2OO5). Hovever, even vilh lhe nosl
conservalive eslinales, il is cIear lhal lhe energy in lhese hydrale deposils is IikeIy lo le
significanl conpared lo aII olher fossiI fueI deposils, and vas considered lo le a polenliaI
slralegic energy resource (Makogon el aI., 2OO7, CoIIell, 2OO4, Moridis el aI., 2OO9a).
Techniques for gas produclion fron hydrale reservoir are lased on lhree najor dissocialion
principIes, i.e.: 1) Depressurizalion (Yousif el aI., 1991a, Moridis el aI., 2OO4a, Moridis el aI.,
2OO7, Tang el aI., 2OO7), lo decrease lhe reservoir pressure leIov lhe hydrale dissocialion
pressure al a specified lenperalure, 2) ThernaI slinuIalion (Kanalh el aI., 1991, Li el aI.,
2OO6, 2OO8a, 2OO8l, Tang el aI., 2OO5a, Kavanura el aI., 2OO7), lo heal lhe reservoir alove
hydrale dissocialion lenperalure vilh hol valer, slean or hol lrine injeclion, 3) ChenicaI
inhililor slinuIalion (Sira el aI., 199O, Sung el aI., 2OO2, Li el aI., 2OO7a, 2OO7l, Kavanura el
aI., 2OO5a, 2OO5l), lo injecl chenicaIs, such as nelhanoI or LC lo shifl lhe hydrale pressure-
lenperalure equiIilriun condilions. Il is essenliaI for safeIy and efficienlIy producing
naluraI gas lo characlerize lhe hydrale dissocialion nechanisn and nuIliphase fIoving
nechanisn of dissocialed gas and valer in hydrale-exisling sedinenl.
LxperinenlaI invesligalions of hydrale dissocialion lehaviors under depressurizalion,
lhernaI slinuIalion and chenicaI inhililor slinuIalion in sedinenl have leen reporled.
Yousif el aI. (Yousif el aI., 199O, 1991a, 1991l) deveIoped a one-dinensionaI nodeI lo sludy
depressurizalion-induced hydrale dissocialion in lerea sandslone cores suggesled lhal a
noving loundary nodeI provide a salisfaclory fil lo hydrale dissocialion neasurenenls.
Kono el aI. (Kono el aI., 2OO2) neasured lhe dissocialion rale of nelhane gas hydrale ly
12
Fuel njection 228

depressurizing nelhod and derived lhe kinelic dissocialion rale equalion and lhe order of
reaclion. Tang el aI. (Tang el aI., 2OO5a, 2OO7) experinenlaIIy invesligaled lhe lenperalure
dislrilulion, gas and valer produclion rale, and lhe lhernaI efficiency during lhe hydrale
dissocialion process using depressurizalion nelhod and lhe hol valer injeclion using a one-
dinensionaI physicaI nodeI. Sung el aI. (Sung el aI., 2OO3) exanined lhe fIoving
characlerislics of lhe dissocialed gas and valer fron hydrale in porous rock ly lhe
depressurizalion and nelhanoI injeclion schenes, using eIeclric resislance lo dislinguish lhe
hydrale fornalion and dissocialion. Kanalh el aI. (Kanalh el aI., 1991)

invesligaled lhe
dissocialion characlerislics of lhe nelhane hydrales during lhe lrine injeclion vilh Iov saIl
concenlralions and high lenperalures, and neasured lhe effecl of lhe lenperalure, lhe
saIinily, and lhe injeclion rale of lhe lrine on lhe hydrale dissocialion rale. Kavanura el aI.
(Kavanura el aI., 2OO5a, 2OO5l, 2OO6) anaIyzed lhe dissocialion lehavior of an arlificiaI
hydrale core sanpIe in nelhanoI aqueous soIulion and lhe experinenl vas carried oul ly
varying lhe lenperalure and concenlralion of lhe nelhanoI aqueous soIulion. The
dissocialion kinelics of nixed gas hydrales lhal conlain propane as a guesl noIecuIe has
leen invesligaled as veII. Sira el aI. (Sira el aI., 199O) reporled lhe characlerislics of hydrale
dissocialion process during nelhanoI and LC injeclion. They concIuded lhal lhe rale of
hydrale dissocialion is a funclion of chenicaI concenlralion, injeclion rale, pressure,
lenperalure of chenicaI soIulion and hydrale-chenicaI inlerfaciaI area. Li el aI. (Li el aI.,
2OO7a, 2OO7l) experinenlaIIy invesligaled lhe gas produclion lehavior fron nelhane
hydrale in porous sedinenl ly injecling elhyIene gIycoI (LC) soIulion vilh differenl
concenlralions and differenl injeclion rales. The resuIls shoved lhal lhe produclion
efficiency is affecled ly lolh lhe LC concenlralion and lhe LC injeclion rale. The alove
experinenlaI vork focused on lhe characlerizalion of hydrale dissocialion during nelhanoI
and elhyIene gIycoI injeclion, vhiIe fev reporls are found aloul lhe invesligalion of lhe
dissocialion lehavior nelhane hydrale in lhe porous nedia under hol lrine injeclion.
Anong lhese nelhods, lhe lhernaI slinuIalion is capalIe of producing sulslanliaI anounls
of naluraI gas, and il viII le nore effeclive conlining vilh lhe depressurizalion or lhe
chenicaI injeclion nelhods. (Moridis el aI., 2OO3, 2OO4l, 2OO9l) So far, lhe inpacls of lhe
higher concenlralions (O~24 vl) of lhe injecled lrine soIulion al a vide lenperalure ranges
(-1~13O
o
C) are nol veII underslood yel, especiaIIy al lhe nediun lenperalure ranges. And
elhyIene gIycoI (LC) is videIy knovn as a lhernodynanic inhililor of gas hydrale, sludies
of naluraI gas dissocialion in lhe presence of LC are Iiniled, especiaIIy lhe inpacls of lhe LC
concenlralion and injeclion rale on hydrale dissocialion are nol veII underslood yel (Sira el
aI., 199O, Li el aI., 2OO7a, Kavanura el aI., 2OO5a).
In lhis vork, lhe produclion lehaviors of MH in unconsoIidaled sedinenl under hol lrine
and LC injeclion vere invesligaled in a deveIoped one-dinensionaI experinenlaI apparalus.
In lhe experinenls of lhe hol lrine injeclion, NaCI aqueous soIulion vilh lhe concenlralion
of O~24 vl and lhe lenperalure of -1~13O
o
C vas injecled inlo lhe vesseI vilh lhe injeclion
rale of 9.99nI/nin. The dissocialion kinelics of nelhane hydrale in porous sedinenl, lhe
produclion lehaviors of gas and valer, and lhe lhernaI and lhe energy efficiencies of lhe
hydrale recovery process under hol lrine injeclion vere invesligaled. The experinenls of
lhe LC injeclion vere carried oul ly varying lhe concenlralion and injeclion rale of lhe LC
soIulion. The reIalionship lelveen lhese lvo paranelers and lhe produclion lehavior vas
ollained. In addilion, lhe efficiency of lhe gas produclion process during MH dissocialion
vas delernined.

2. ExperimentaI Apparatus
The schenalic diagran for lhe one-dinensionaI experinenlaI apparalus used in lhis vork is
shovn in Iigure 1. The pressure vesseI vas innersed in an air lalh vilh lhe lenperalure
range fron -2O
o
C lo 8O
o
C, O.5
o
C lo nainlain a conslanl lenperalure. The vesseI is nade of
quarlz gIass and has an inlernaI dianeler of 3Onn and a Ienglh of 534nn, and il can le
operaled up lo 1OMIa. Iour resislance lhernonelers and lvo pressure sensors vilh lhree
differenliaI pressure lransducers vere unifornIy-spaced, shovn in Iigure 1, lo neasure lhe
lenperalure and pressure profiIe aIong lhe vesseI. The lhernonelers are Il1OO vilh lhe
range of -2O
o
C lo 2OO
o
C, O.1
o
C. The pressure lransducers are KLLLLR IA-21S 8O4OO O-
2OMIa, O.25. Tvo gas fIov nelers, vhich vere used lo neasure lhe gas injeclion or
produclion rale and lhe cunuIalive gas produced fron lhe vesseI, are lolh of DO7-11A/ZM,
O-1OOO nI/nin, 1 fron seven slar conpany. The pressure lransducers, lhernonelers,
gas fIov nelers, vere caIilraled using a pressure lesl gauge vilh lhe error of O.O5, a
nercury lhernoneler vilh lhe loIerance of O.O1
o
C, and a vel gas neler vilh lhe accuracy
of 1O nI/nin, respecliveIy. The lvo laIances, vhich vere are used lo neasure lhe rales of
Iiquid inpul and oulpul, respecliveIy, are Sarlorius S 22O2S, O-22OOg, O.O1g. In order lo
prolecl lhe nelering punp fron corrosion ly lhe hol lrine or chenicaI inhililors, lhe
niddIe conlainers have leen used for lhe soIulion injeclion. The dala acquisilion syslen
records pressure, differenliaI pressure, lenperalure, gas/valer injeclion rale and
produclion rale. The vesseI and lhe Ioading pipeIines vere covered fron a heal-insuIaling
naleriaI lo avoid lhe heal Ioss during lrine injeclion. The NaCI used vas anaIylicaIIy pure
suppIied ly Cuangzhou ChenicaI Reagenl Iaclory, China. The LC vas anaIylicaIIy pure,
and lhe nelhane gas vas produced ly Ioshan Kody Cas ChenicaI Induslry Co., Lld, China
vilh ils purily of 99.99.

Iig. 1. The schenalic pIol of lhe experinenlaI apparalus

Fuel njection 230

3. Hot brine stimuIation
3.1 ExperimentaI Procedures
During lhe experinenl, lhe rav dry quarlz sand vilh lhe size range of 3OO-45O n are
lighlIy packed in lhe vesseI, and lhen lhe vesseI vas evacualed lvice lo renove air in il vilh
a vacuun punp. NaCI aqueous soIulion vilh lhe concenlralion of 2 vl inslead of disliIIed
valer vas used for aII experinenls lecause of lhe higher fornalion rale of lrine soIulion
lhan lhal of lhe disliIIed valer in sedinenls. (Tang el aI., 2OO5l) Hence, in lhis vork lhe dry
quarlz sand in lhe vesseI vas velled al alnospheric pressure vilh 2 vl NaCI aqueous
soIulion, vhich vas injecled inlo lhe vesseI fron a lhe niddIe conlainers vilh a nelering
punp, vhen lhe anounl of lhe Iiquid driven fron lhe vesseI vas equaI lo lhose injecled.
Then lhe nelhane gas vas injecled inlo lhe vesseI up lo lhe pressure, vhich reached nuch
higher lhan lhe equiIilriun hydrale fornalion pressure al lhe vorking lenperalure. Afler
lhal, lhe vesseI vas cIosed as an isochoric syslen, and lhe lenperalure in lhe vesseI vas
graduaIIy decreased lo forn lhe hydrale ly changing lhe air lalh lenperalure. In lhis vork,
lhe lenperalure of lhe air lalh during lhe hdyrale fornalion is O
o
C. Il vas considered lhal
lhe fornalion had nol conpIeled unliI lhere vas no pressure decrease in lhe vesseI.

Runs
T lefore
fornalion
o
C
I lefore
fornalion MIa
I lefore
injeclion MIa
T of injecled
lrine
o
C
rine injecled
vl
H1 25.7 5.898 3.O41
13O
O
H2 23.2 5.635 3.O22 8
H3 2O.3 5.617 3.O95 16
H4 24.3 5.641 3.O97 24
H5 25.2 5.641 2.973
9O
O
H6 25.5 5.9O4 3.O52 8
H7 28.7 5.849 3.O95 16
H8 26.3 5.849 3.1O2 24
H9 24.O 5.922 2.888
5O
O
H1O 27.4 5.825 2.961 8
H11 28.8 5.934 2.955 16
H12 28.8 6.251 3.O1 24
H13 27.5 6.O13 2.967
-1
2
H14 27.8 6.OO7 3.O47 8
H15 26.4 5.916 3.113 16
H16 26.2 5.5O7 2.845 24
TalIe 1. LxperinenlaI condilions of MH dissocialion ly hol rine injeclion

Melhane hydrale dissocialion under hol lrine injeclion vas carried oul in lhe foIIoving
procedures. The lrine vilh cerlain concenlralion vas prepared in lhe niddIe conlainer. The
lack pressure reguIalor vas sel lo lhe cerlain required pressure vaIue, vhich is lhe syslen
pressure during lhe hydrale dissocialion process under hol lrine injeclion. The lrine vas
firslIy healed or cooIed ly lhe healer/cooIer in lhe niddIe conlainer. Then lhe lrine vas
injecled inlo lhe vesseI. Afler a fev ninules, gas and Iiquid legan lo reIease fron lhe vesseI
lhrough lhe oulIel vaIve. During lhe enlire fornalion and lhe dissocialion runs of nelhane
hydrale, lhe dala lhus as lhe lenperalure and lhe pressure in lhe vesseI, lhe gas produclion,
lhe anounl of injecled lrine, and lhe Iiquid produclion vere recorded al 6 second inlervaIs.
The lrine injeclion rale is 9.99 nI/nin. The delaiIed experinenlaI arrangenenl is shovn in
TalIe 1.

3.2 Hydrate Dissociation
In lhis vork, a lolaI of 16 experinenlaI runs vere carried oul lo invesligale inlo lhe
dissocialion lehaviors of nelhane hydrale in lhe porous nedia ly injecling lhe aqueous
soIulions of NaCI al -1, 5O, 9O and 13O
o
C, respecliveIy, and vilh lhe injeclion rale of 9.99
nI/nin, as shovn in TalIe 1.
In Run H1-Run H12, lrine soIulions vere vilh O, 8, 16 and 24 vl, respecliveIy. In Run H1,
Run H5 and Run H9, lhe nelhane hydrale vas dissocialed vilh lhe effecl of hol valer (O
vl lrine soIulion) al 5O, 9O and 13O
o
C, respecliveIy. In Run H13- Run H16, lhe nelhane
hydrale vas slinuIaled vilh lhe lrine soIulion vilh lhe concenlralions of 2, 8, 16 and 24
vl, respecliveIy, and al -1
o
C, Iover lhan lhal of lhe hydrale syslen in lhe vesseI (O
o
C).
Since lhe hydrale in lhe sedinenl in lhe vesseI fron 2 vl NaCI soIulion vas forned al O
o
C
vhen, as did in Run H13, lhe lrine vas injecled al 2 vl and -1
o
C inlo lhe vesseI, lhe
hydrale vas nol dissocialed fron lhe lrine soIulion. Hence, lhe purpose of Run H13 is lo
delernine lhe anounl of free gas in lhe vesseI. In addilion, in Run H14-Run H16, .lhe
hydrale vas dissocialed on accounl of lhe effecl of lhe lrine soIulion vilhoul lhernaI effecl.
In olher experinenlaI runs, lhe hydrale vas dissocialed fron lhe synergislic effecl of lhe
heal and saIl.
In lhis vork, lhe foIIoving processes vere invoIved, lrine repIacenenl of pore soIulion,
heal lransfer, and ion diffusion acconpanying lrine injeclion. Anong lhese processes, lhe
high saIinily or high lenperalure can cause hydrale dissociale ly shifling hydrale oulside
ils slaliIily regine. The high saIinily can resuIl eilher ly lhe direcl repIacenenl of lhe
originaI pore soIulion vilh high concenlralion of lrine or ly lhe diffusion of ions due lo a
concenlralion gradienl, and lhe high lenperalure can le caused eilher ly lhe direcl
repIacenenl of lhe originaI pore soIulion vilh high lenperalure lrine or ly lhernaI lransfer
due lo a lenperalure gradienl.

3.2.1 Hot water stimuIation
Iigure 2 shovs lhe curve of lhe lenperalure change vilh line al Iorl 2, Iorl 3 and Iorl 4
(lhe oulIel) in lhe vesseI vilh lhe injeclion of hol valer al 5O, 9O and 13O
o
C, respecliveIy, as
did in Run H9, Run H5 and Run H1. As shovn in Iigure 2, each lenperalure curve consisls
of lhe horizonlaI seclion and lhe ascending seclion. The horizonlaI seclion represenls lhe
hydrale non-dissocialion processes and lhe isolhernaIIy endolhernaI dissocialion processes
resuIling fron lhe direcl repIacenenl of lhe originaI pore soIulion and heal lransfer fron
Fuel njection 232

hol valer, lhe ascending seclion is lhe helerolhernaIIy endolhernaI process of lhe syslen in
lhe porous nedia in lhe effecl of hol valer afler lhe hydrale dissocialion has fuIIy conpIeled.
As shovn in Iigure 2, in lhe alove lhree runs, lhe dissocialion firslIy happened in lhe inIel
of lhe vesseI, and lhen in Iorl 2, Iorl 3 and Iorl 4 vilh line in lurn unliI lhe hydrale in lhe
vesseI vas conpIeleIy dissocialed. AccordingIy, il is considered lhal lhe dissocialion process
of lhe hydrale in lhe vesseI is lhe noving-forvard process of lhe hydrale dissocialion
loundary fron lhe inIel lo lhe oulIel. In olher vords, lhe fIoving of hol valer injecled in lhe
vesseI can le regarded as lhe noving of a pislon fron lhe inIel lo lhe oulIel.
0 20 40 60 80
-10
0
10
20
30
40
50
60
Port 3 Port 2
Port 1

T
e
m
p
e
r
a
t
u
r
e

/

o
C
Time / min
50
o
C
90
o
C
130
o
C

Iig. 1. The curve of lhe lenperalure change over line al Iorls 1-3 in lhe vesseI in Run H1,
Run H5 and Run H9 vilh lhe effecl of hol valer al 5O
o
C, 9O
o
C and 13O
o
C

3.2.2 Brine stimuIation
In Run H13-Run H16, lhe experinenls of lhe lrine slinuIalion vas carried oul. Iigure 3
shovs lhe curve of lhe lenperalure change over line al Iorl 4 in lhe vesseI vilh lhe
injeclion of lhe lrine soIulion al -1
o
C vilh 2, 8, 16 and 24 vl, respecliveIy. Ior Irol 1 and
Iorl 3, lhe characlerislics of lhe curve of lhe lenperalure change over line are siniIar. Since
in lhe alove experinenls, lhe lrine soIulion vas injecled inlo lhe vesseI al -1
o
C, Iover lhan
lhal of lhe hydrale syslen in lhe vesseI (O
o
C), lhe hydrale dissocialion can le onIy caused
fron lhe inhililors, nol fron lhe lhernaI effecl.
As shovn in Iigure 3 and discussed alove, under lhe injeclion of lhe lrine al 2 vl and -1

o
C inlo lhe vesseI, lhe hydrale vas nol dissocialed. Hovever, lhe hydrale dissocialion can le
caused ly lhe effecl of lhe lrine soIulion vilh higher concenlralions. As shovn in Iigure 3,
lhe process of lhe hydrale dissocialion is lhe process of lhe lenperalure decrease, vhich is
lhe resuIl of lhe presence of lhe lrine soIulion. Since lhe lenperalure drop vas caused ly
lhe heal laIance lelveen lhal needed for hydrale dissocialion and lhal suppIied fron

surrounding environnenl, lhe Iovesl poinl of lenperalure represenls lhe occasion vhen
hydrale dissocialed nosl inlenseIy. In addilion, il vas found lhal lhe line for lhe hydrale
dissocialion is shorlened and lhe degree of deplh (veII deplh) of lhe lenperalure drop
increases vilh lhe increase of lhe concenlralion of lhe lrine soIulion.
According lo lhe caIcuIalion, aloul 16 ninules has leen needed for lrine lo repIace lhe pore
valer around lhe lenperalure sensors of Iorl 4 in Run H13-Run H16 vilh lhe effecls of lhe
differenl NaCI concenlralions al -1
o
C. Hovever, lhe Iovesl poinls of lenperalure have
occurred afler Iapse of line vhen lhe repIacenenl had finished. This vas caused ly saIinily
change of pore valer due lo ion diffusion.
Iigure 16 gives lhe curve of lhe lenperalure change vilh line al Iorls 1-3 in lhe vesseI in
lhe presence of lrine soIulion vilh 24 vl and al -1
o
C. As shovn in Iigure 16, lhere is a
veII deplh of lhe lenperalure change in each lenperalure curve al Iorls 1-3, and lhe veIIs
appear vilh line in lurn and lhe deplhs of lhe veIIs fron Iorl 2 lo Iorl 4 graduaIIy increase.
In lhe process of lhe hydrale dissocialion, il nighl le caused ly lhe direcl repIacenenl of
pore valer vilh lrine al porls 1 and 2, resuIling in lhe lhernaI honogenizalion, vhiIe lhe
lenperalure change al Iorl 4 vas caused ly saIinily change of pore valer due lo ion
diffusion.
0 20 40 60 80 100 120 140 160
-5
-4
-3
-2
-1
0
1
2
3
16
T
e
m
p
e
r
a
t
u
r
e

o
I

P
o
r
t

3

/

o
C
Time / min
2%
8%
16%
24%
12min

Iig. 3. The curve of lhe lenperalure change over line al Iorl 4 in lhe vesseI in Run H13-Run
H16 vilh lhe effecls of lhe differenl lrine concenlralions al -1
o
C
Fuel njection 234

0 10 20 30 40 50 60
-4
-2
0
2

T
e
m
p
e
r
a
t
u
r
e

/

o
C
Time (min)
Port 1
Port 2
Port 3 12min

Iig. 4. The curve of lhe lenperalure change over line al Iorls 1-3 in lhe vesseI in Run H16
vilh lhe injeclion of 24 vl lrine soIulion al -1
o
C

3.2.3 Hot Brine stimuIation
Iigure 5 gives lhe lypicaI curve of lhe lenperalure change vilh line al Iorls 1-3 in lhe
vesseI in lhe presence of hol lrine soIulion vilh 24 vl and al 9O
o
C. Il is shovn fron lhe
figure lhal al Iorl 4, lhe curve can le divided inlo lhree seclions: lhe horizonlaI seclion, lhe
dovnvard seclion and lhe upvard seclion. The horizonlaI seclion represenls lhe non-
dissocialion and lhe isolhernaIIy endolhernaI dissocialion (phase lransfornalion)
processes of lhe hydrale sliII vilhoul lhe effecl of lhe inhililor. The dovnvard seclion is lhe
cooIing endolhernaI dissocialion process of lhe hydrale on lhe effecls of lhe hol valer and
lrine soIulion. In lhis seclion, vilh lhe increase of concenlralion of lrine soIulion vilh line,
vhich acls on lhe surface of lhe hydrale, lhe lenperalure of lhe hydrale graduaIIy decreases
and lhe hydrale graduaIIy dissociales unliI lhe dissocialion is conpIeled vhiIe lhe
concenlralion of lrine soIulion reaches lhe naxinun vaIue. The upvard seclion is onIy lhe
helerolhernaIIy endolhernaI process of lhe syslen in lhe porous nedia in lhe effecl of heal
afler lhe hydrale dissocialion has fuIIy conpIeled. In lhe seclion, lhere are no lhe phase
lransfornalion. As shovn in Iigure 5 lhal lhe characlerislics of lhe lenperalure changes
vilh Iorls 1 and 2 are siniIar vilh Iorl 4. Ior olher saIl concenlralions and olher
lenperalures of lhe injecled hol soIulions, lhe characlerislics of lhe lenperalure change are
aIso siniIar vilh lhe alove. In addilion, as shovn in lhe figure, lhe fIoving of hol lrine
valer injecled in lhe vesseI can le aIso regarded as lhe noving of a pislon fron inIel lo
oulIel, as anaIyzed in Iigure 2.
Tenperalure changes in Iorl 4 in Run H4, Run H8, Run H12 and Run H16 over line vilh
lhe injeclion of lhe lrine of 24 vl al -1, 5O, 9O, 13O
o
C, respecliveIy, have leen shovn in

Iigure 6. The experinenlaI resuIls iIIuslrale lhal vilh lhe lrine injecled al lhe sane
concenlralions lhe sane Iovesl vaIue of lenperalure decrease of lhe hydrale syslen al lhe
sane porl has leen produced and il is independenl of lhe iniliaI lenperalures of lhe injecled
soIulions. The lenperalure changes over line vilh lhe lrine injecled al lhe olher sane
concenlralions al -1, 5O, 9O, 13O
o
C shov lhe siniIar characlerislics.
Iigure 7 gives a lypicaI curve of lhe lenperalure change over line al Iorl 4 vilh Run H1-
Run H4 lhrough injecling lrine soIulion vilh lhe concenlralions of O, 8, 16, and 24 vl,
respecliveIy, al 13O
o
C. As shovn in Iigure 7, il is noled lhal lhe line for lhe hydrale
dissocialion shorlened and lhe degree of lhe deplh (veII deplh) of lhe lenperalure drop
increases vilh lhe increase of lhe concenlralion of lrine soIulion. Ior olher cerlain
lenperalures vilh lhe differenl injeclions of lrine soIulion of O, 8, 16 and 24 vl,
respecliveIy, lhe siniIar characlerislics can le ollained.
The dissocialion processes of hydrale have leen dispIayed lhrough lenperalure curves al
various porls changing over line. Hovever, for 2 vl and 8 vl saIinily curves in Iigure 3,
lenperalure shovs an increase aloul O.2- O.3
o
C during aloul 2 or 3 ninules earIy. This is
due lo heal lransfer fron lhe air lalh afler lhe air lalh had leen opened parliaIIy lo lurn on
inpul vaIve and oulpul vaIve on lhe purpose of lhe injeclion of Iiquid as shovn in Iigure 1.
Heal lransfer lo or fron lhe air lalh affecled aII lhe lenperalure neasurenenls during
aloul 2 or 3 ninules earIy. In spile of lhal, lhis increase or drop does nol denoIish lhe dala
expIain alove lecause il vas nuch Iover lhan lhe veII deplh of lhe lenperalure change in
lhe lenperalure curves occurring Ialer.
0 20 40 60 80 100 120 140
-5
0
5
10
15
20
25
30
35
40
45
50

T
e
m
p
e
r
a
t
u
r
e

/

o
C
Time (min)
Port 1
Port 2
Port 3

Iig. 5. The curve of lhe lenperalure change over line al Iorls 1-3 in lhe vesseI in Run H8
vilh lhe injeclion of 24 vl lrine soIulion al 9O
o
C
Fuel njection 236

0 20 40 60 80 100 120 140
-10
0
10
20
30
T
e
m
p
e
r
a
t
u
r
e

o
f

P
o
r
t

3

/

o
C
Time (min)
-1
o
C
50
o
C
90
o
C
130
o
C

Iig. 6. The curve of lhe lenperalure change over line al Iorl 4 in lhe vesseI in Run H4, Run
H8, Run H12 and Run H16 vilh lhe injeclion of 24 vl lrine soIulion al lhe differenl
lenperalures
0 10 20 30 40 50 60 70 80
-6
-4
-2
0
2
4
6
8
10
12
14
16
18
20

T
e
m
p
e
r
a
t
u
r
e

/

o
C
Time (min)
0
8
16
24

Iig. 7. The curve of lhe lenperalure change over line al Iorl 4 in lhe vesseI in Run H1-Run
H4 vilh lhe effecls of lhe differenl lrine concenlralions al 13O
o
C.


3.3 Gas production
A lypicaI curve of lhe accunuIalive gas produclion for lhe vhoIe gas produclion process in
Run H9 is given in Iigure 8. As shovn in Iigure 8, lhe gas produclion process vilh lhe hol
lrine or hol valer injeclion in lhe vesseI can le divided inlo lhree seclions. In Seclion I, lhe
free nelhane gas in lhe vesseI is reIeased, and inslanlaneousIy gas produclion rale increases
rapidIy. The gas produclion rale couId le expressed ly lhe sIope of lhe curve of lhe
accunuIalive gas produclion. Afler lhe free gas reIeased, lhe gas produclion rale decreases
renarkalIy. This seclion is lhe hydrale dissocialion and gas produclion process and
considered lo le Seclion II. Aflervards in Seclion III, lhe hydrale dissocialion process has
finished, and lhere is onIy lhe residuaI gas reIease fron lhe vesseI. (SIoan & Koh, 2OO8) As
shovn in Iigure 8, lhere are lvo infIexion poinls on lhe curve of lhe accunuIalive gas
produclion vilh line. The Iefl poinl indicales lhe end of free gas reIease process (Seclion I)
and lhe leginning of lhe hydrale dissocialion process (Seclion II). The righl one neans lhe
end of hydrale dissocialion process and lhe leginning of produclion process of lhe residuaI
gas (Seclion III).
Iigure 9 gives lhe accunuIalive gas produclion over line vilh lhe 2 vl lrine soIulion
injeclion al -1
o
C, vhich is a lypicaI case of lhe gas produclion vilhoul lhe effecls of lhernaI
and lrine. Il can le seen fron lhe figure lhal lhere is onIy lhe free gas produclion vilhoul
lhe dissocialed gas fron lhe hydrale in lhis case.
Iigure 1O shovs lhe accunuIalive gas produclion in Seclion II vilh lhe hol valer injeclion al
5O, 9O and 13O
o
C, respecliveIy, as did in Run H9, Run H5 and Run H1. The hydrale
dissocialion rale increases vilh lhe increase of lhe lenperalure of lhe injecled hol valer
during lhe hydrale dissocialion process (CoeI el aI., 2OO1).
Iigure 11 gives lhe accunuIalive gas produclion in Seclion II al 5O
o
C vilh lhe injeclions of
lhe lrine soIulion in lhe concenlralion range of O~24 vl. The hydrale inslanlaneous
dissocialion rale couId le increased ly injecling lrine soIulion olher lhan valer, and il is
reIaled lo lhe concenlralion of injecled lrine soIulion. When lhe lrine concenlralion is Iess
lhan 16 vl, lhe dissocialion rale increases vilh lhe lrine concenlralion. Il is noled lhal lhe
hydrale inslanlaneous dissocialion rale is approxinaleIy lhe sane vilh lhe injeclion of lrine
soIulion of 16 vl and 24 vl al 5O
o
C. In olher vords, if lhe lrine concenlralion conlinues
rising afler reaching cerlain vaIue, lhe concenlralion has IillIe effecl on lhe hydrale
inslanlaneous dissocialion rale. Hence, in lhe process of hydrale dissocialion vilh lhe
injeclion of hol lrine, il is nol necessary lo use lhe lrine soIulion vilh very high
concenlralions. The accunuIalive gas produclion and lhe hydrale inslanlaneous
dissocialion rale al olher cerlain lenperalure such as -1, 9O, and 13O
o
C, vilh lhe injeclions of
lhe lrine soIulion in lhe concenlralion range of O~24 vl shov lhe siniIar lehavior.
Fuel njection 238

0 10 20 30 40 50 60 70 80 90
0
2000
4000
6000
8000
10000
12000
14000
16000
0
200
400
600
800
1000
1200
STD: Standard State
Section
Section
Section

C
u
m
u
l
a
t
i
v
e

w
a
t
e
r

m
a
s
s

(
g
)

T
h
e

g
a
s

c
u
m
u
l
a
t
i
v
e

p
r
o
d
u
c
t
i
o
n

(
S
T
D

m
l
)
Time (min)
gas produced volume
water injected
water produced

Iig. 8. The accunuIalive gas produclion and lhe accunuIalive nass of valer injecled and
produced over line in Run H9 vilh lhe injeclion of hol valer al 5O
o
C
0 5 10 15 20 25 30
0
500
1000
1500
2000
2500
3000
3500
4000
C
u
m
u
l
a
t
i
v
e

l
i
q
u
i
d

m
a
s
s

(
g
)

T
h
e

g
a
s

c
u
m
u
l
a
t
i
v
e

p
r
o
d
u
c
t
i
o
n

(
S
T
D

m
l
)
Time (min)
gas produced volume
0
50
100
150
200
250
300
350
400
water injected
water produced

Iig. 9. The accunuIalive gas produclion and lhe accunuIalive nass of lrine injecled and
produced in Run H13 vilh lhe injeclion of 2 vl lrine soIulion al -1
o
C

0 10 20 30 40 50 60 70 80 90
0
1000
2000
3000
4000
5000
6000
7000

T
h
e

g
a
s

c
u
m
u
l
a
t
i
v
e

p
r
o
d
u
c
t
i
o
n

(
S
T
D

m
l
)
Time (min)
50
o
C
90
o
C
130
o
C

Iig. 1O. The accunuIalive gas produclion al seclion II in Run H1, Run H5 and Run H9 vilh
lhe effecls of hol valer al 5O
o
C, 9O
o
C and 13O
o
C
0 10 20 30 40 50 60 70
0
1000
2000
3000
4000
5000
6000
7000
8000

T
h
e

g
a
s

c
u
m
u
l
a
t
i
v
e

p
r
o
d
u
c
t
i
o
n

(
S
T
D

m
l
)
Time (min)
0 wt%
8 wt%
16 wt%
24 wt%

Iig. 11. The accunuIalive gas produclion al seclion II in Run H9-Run H12 vilh lhe effecls of
lhe differenl lrine concenlralions al 5O
o
C

Fuel njection 240

3.4 Liquid production
As shovn in Iigure 8, during free gas produclion, vilh hol valer or hol lrine injeclion lhere
is IillIe Iiquid produclion. This slage is one process lhal free gas in lhe vesseI is drived oul,
and in lhis slage, lhe injecled Iiquid soIulion slays in lhe vesseI. During lhe hydrale
dissocialion, lhe Iiquid produclion rale is sIighlIy higher lhan lhe soIulion injeclion rale, due
lo lhe valer produced fron lhe hydrale dissocialion. Afler lhe hydrale dissocialion process
finished, lhe Iiquid produclion rale is equaI lo lhe soIulion injeclion rale.

3.5 Production efficiency anaIysis
In lhis vork, lo delernine lhe efficiency of gas produclion fron lhe hydrale ly hol lrine
injeclion, lhe lhernaI efficiency and lhe energy ralio are invesligaled. The lhernaI efficiency
is defined as lhe ralio of lhe heal quanlily for hydrale dissocialion lo lhe lolaI heal inpul,
vhich is defined as lhe anounl of heal needed lo raise lhe lenperalure of lhe hydrale
syslen in lhe vesseI up lo lhe injeclion lenperalure. Thus, vhen lhe fIuid is injecled al O
o
C
or Iess lhan O
o
C, lhe lhernaI efficiency is zero, and lhere is no lhernaI effecl on lhe hydrale
syslen in lhe vesseI ly lhe fIuid injecled. The energy ralio is defined as lhe ralio of lhe
conluslion heal quanlily of produced gas lo lhe lolaI inpul heal quanlily (Li el aI., 2OO6,
2OO8l).
ThernaI efficiencies and energy ralios for lhe hydrale produclion in lhe alove various
experinenlaI runs under hol valer and hol lrine injeclions are shovn in Iigures 12 and 13,
respecliveIy. As shovn in Iigures 12 and 13, lhe lhernaI efficiency and lhe energy ralio
decrease vilh lhe increase of lhe lenperalure of injecled hol valer al lhe O vl saIinily. Ior
lhe case of lhe injeclion of hol lrine soIulion, lhe lhernaI efficiency and lhe energy ralio
increase vilh lhe increase of lhe concenlralion of injecled hol lrine vilh lhe cerlain
lenperalure. Ior hydrale dissocialion, nore poverfuI lenperalure-driving force cones
forlh resuIling fron increasing saIinily and lhus hydrale dissociales nore rapidIy resuIling
in snaIIer lhe lolaI heal inpul. Then, increasing lhernaI efficiency and energy ralio have
leen ollained.
Hovever, vilh lhe differences of lhe lenperalures of lhe injecled hol lrine, lhe degrees of
lhe increases of lhe lhernaI efficiency and lhe energy ralio are differenl. As shovn in
Iigures 12, 13, il is noled lhal al Iov lenperalure, 5O
o
C, lhe increase effecliveness of lhe
lhernaI efficiency and lhe energy ralio is apparenl vilh lhe increase of lhe concenlralion of
hol lrine. Whereas, al high lenperalure lhus as 13O
o
C, lhere are onIy a IillIe increase for
lhen. Hence, il is suggesled lhal in lhe gas hydrale produclion ly lhe hol lrine injeclion, lhe
appropriale lenperalure in conjunclion vilh lhe high concenlralion of lrine soIulion lrings
reIalive high recovery efficiency. The injeclion vilh loo high lenperalure resuIls in lhe
energy Ioss.

0 5 10 15 20 25
0.0
0.1
0.2

T
h
e
r
m
a
l

e
f
f
i
c
i
e
n
c
y
Salinity wt%
50
o
C
90
o
C
130
o
C

Iig. 12. ThernaI efficiencies of gas produclion vilh lhe saIinily al 5O
o
C, 9O
o
C and 13O
o
C
0 5 10 15 20 25
3
6
9
12
15

E
n
e
r
g
y

r
a
t
i
o

f
o
r

w
h
o
l
e

p
r
o
c
e
s
s
brine salinity wt%
50
o
C
90
o
C
130
o
C

Iig. 13. Lnergy ralios of gas produclion vilh lhe saIinily al 5O
o
C, 9O
o
C and 13O
o
C

Fuel njection 242

4. EG stimuIation
4.1 ExperimentaI Procedures
During lhe experinenl, lhe rav dry quarlz sand vilh lhe size range of 3OO-45O n are
lighlIy packed in lhe vesseI, and lhen lhe vesseI vas evacualed lvice lo renove air in il vilh
a vacuun punp. The quarlz sand in lhe vesseI vas velled lo saluralion vilh disliIIed valer
using a nelering punp. The sand sedinenl vas saluraled vhen lhe anounl of valer
produced fron lhe vesseI vas equaI lo lhe anounl of valer injecled. Il vas assuned lhal
lhe voIune of valer injecled in lhe vesseI vas lhe lolaI voIune avaiIalIe in lhe vesseI. Then
lhe nelhane gas vas injecled inlo lhe vesseI unliI lhe pressure in lhe vesseI reaches nuch
higher lhan lhe equiIilriun hydrale fornalion pressure al lhe vorking lenperalure. Afler
lhal, lhe vesseI vas cIosed as an isochoric syslen. The lenperalure vas graduaIIy decreased
lo forn lhe hydrale ly changing lhe air lalh lenperalure. The hydrale fornalion vas
considered lo le conpIeled unliI lhere vas no pressure decrease in lhe syslen. The hydrale
fornalion process in generaI Iasls for 2 lo 5 days.
The hydrale dissocialion ly LC injeclion vas carried oul in lhe foIIoving procedures.
IirslIy, lhe LC soIulion vilh lhe desired concenlralion vas prepared in lhe niddIe
conlainers. The lack pressure reguIalor vas sel lo 3.8MIa, vhich is lhe syslen pressure
during lhe hydrale dissocialion process under LC injeclion. Then lhe dissocialion run vas
slarled ly injecling lhe LC soIulion fron lhe niddIe conlainers inlo lhe vesseI. The LC
soIulion vas cooIed dovn lo lhe lenperalure in lhe air lalh lefore injecled inlo lhe vesseI.
Afler injecling lhe LC soIulion for approxinaleIy 5 nins, hydrale legan lo dissociale and
gas and valer soIulion vere olserved lo reIease fron lhe vesseI lhrough lhe oulIel vaIve.
The gas produclion process Iasled for 3O-1OO nin, depending on lhe LC concenlralions and
injeclion rales. When lhere vas no significanl gas reIeased, lhe LC injeclion vas finished
and lhe syslen pressure vas reIeased lo 1 aln. graduaIIy. During lhe enlire dissocialion
run, lhe lenperalure and pressure in lhe vesseI, lhe gas produclion, lhe anounl of LC
soIulion injecled and lhe valer produclion vere recorded al 2 seconds inlervaIs.

4.2 Hydrate Formation
TalIe 2 provides lhe hydrale fornalion condilions. The voIune of lhe valer and gas lefore
hydrale fornalion is equaI lo lhe lolaI voIune of valer, gas and hydrale afler hydrale
fornalion:
V
v1
+V
g1
= V
v2
+V
g2
+V
h2
(1)

Il vas assuned lhal lhere is 5.75 noI valer in 1noI nelhane hydrale, and lhe densily of
nelhane hydrale is O.94 g/cn
3
and valer in lhe vesseI is inconpressilIe. The voIune of lhe
gas in lhe vesseI afler hydrale fornalion vas caIcuIaled ly lhe pressure and lenperalure
condilions in lhe vesseI using lhe Ieng-Rolinson equalion. The inIel and oulIel pressures of
lhe vesseI change sinuIlaneousIy due lo lhe high porosily and pernealiIily of lhe sedinenl,
so lhe pressure in lhe vesseI in lhis vork lakes lhe average of lhe inIel and oulIel pressures.
Iigure 14 shovs a lypicaI experinenlaI resuIl of lhe pressure and lenperalure profiIes vilh line
during MH fornalion in lhe sedinenl. Il can le seen fron Iigure 14 lhal lhe pressure profiIe
during MH fornalion couId le divided inlo four seclions. In seclion I (O nin-175 nin), lhe
lenperalure decreased fron 17.O
o
C lo 2.O
o
C in isochoric condilion, and lhe pressure decreases
fron 5.4 MIa lo 5.1 MIa due lo lhe gas adsorplion on porous lhe quarlz sand and lhe gas
conlraclion in lhe vesseI. Afler seclion I, lhe cIosed syslen vas nainlained al a conslanl

lenperalure (2.O
o
C) unliI lhe end of lhe experinenl. In seclion II (175 nin-28O nin), lhe pressure
of lhe cIosed syslen vas alove 5.O MIa, vhich vas nuch higher lhan lhe pure hydrale
equiIilriun pressure of 3.5 MIa al 2.O
o
C. (SIoan & Koh, 2OO8) This seclion vas considered lo le
lhe hydrale nucIealion process, and in lhis period of line lhere vas no hydrale forned in lhe
vesseI. (Ian el aI., 2OO6) The seclion III is lhe hydrale fornalion process. In lhis seclion, lhe
pressure graduaIIy decreased due lo lhe gas consunplion during lhe hydrale fornalion, and lhis
seclion lakes nuch Ionger line lhan seclion I and II. In lhe Iasl seclion (seclion IV), no furlher
pressure decrease vas olserved, and lhe syslen vas nainlained al a conslanl lenperalure.
Hence, lhe syslen reached lhe lhernodynanic slalIe slale.
TolaI 7 experinenlaI runs of hydrale dissocialion ly LC injeclion have leen carried oul. Run
LO as lhe lIank experinenl, vhich injecled lhe disliIIed valer inslead of LC soIulion, vas
used lo eIininale lhe infIuence of lhe gas produclion ly lhe Iiquid injeclion. TalIe 3
provides lhe experinenlaI condilions during hydrale dissocialion ly LC injeclion, incIuding
lhe LC injeclion rale, lhe LC concenlralion and lhe average pressure and lenperalure
during MH dissocialion. The hydrale dissocialion runs in TalIe 3 vere reIaled lo lhe
fornalion runs in TalIe 2.

experinenlaI runs
LO L1 L2 L3 L4 L5 L6 L7
IniliaI Iressure (MIa) 5.4O3 5.519 5.488 5.476 5.3O6 5.311 5.416 5.4O9
IniliaI lenperalure (
o
C) 17.83 17.89 18.O1 17.71 17.83 17.46 17.77 17.95
IinaI Iressure (MIa) 3.556 3.5O2 3.467 3.48O 3.557 3.566 3.516 3.486
IinaI lenperalure (
o
C) 1.97 1.92 1.81 1.92 2.OO 2.O7 1.81 1.73
IinaI anounl of valer (nI) 43.73 47.53 46.22 45.53 42.18 41.95 42.92 43.26
Conversion of gas lo hydrale () 33.O3 36.77 36.82 36.22 31.44 31.49 33.83 34.52
Hydrale conlenl (voI, ) 7.33 8.16 8.17 8.O4 6.98 6.99 7.51 7.66
TalIe 2. Iornalion condilions of hydrale reIaled lo hydrale dissocialion ly LC injeclion

experinenlaI runs
LO L1 L2 L3 L4 L5 L6 L7
LC injeclion rale (nI/nin) 8.8 4.9 6.8 8.8 8.8 8.8 8.8 8.8
LC concenlralion (vl ) O 3O 3O 3O 4O 5O 6O 7O
Iressure (MIa) 3.889 3.862 3.926 3.862 3.864 3.85 3.9O1 3.825
Tenperalure (
o
C) 2.O43 1.645 2.O15 1.985 2.O61 1.9O1 2.O1O 1.846
TalIe 3. LxperinenlaI condilions during Hydrale dissocialion ly LC injeclion
Fuel njection 244

0 200 400 600 800 1000 1200 1400 1600 1800 2000 2200
2.5
3.0
3.5
4.0
4.5
5.0
5.5
6.0
0
2
4
6
8
10
12
14
16
18
20
Pressure
1320min 380min
T
e
m
p
e
r
a
t
u
r
e
(
D
e
g

C
)

P
r
e
s
s
u
r
e
(
M
P
a
)
Time(min)
175min
Section
Section
Section
Section V
Temperature

Iig. 14. The pressure and lenperalure profiIes during hydrale fornalion in isochoric
experinenl
0 20 40 60 80 100 120 140 160 180 200
0.0
0.5
1.0
1.5
2.0
2.5
3.0
inlet (port 1)
outlet (port 4)
T
e
m
p
e
r
a
t
u
r
e
(
D
e
g

C
)
Time(min)

Iig. 15. The inIel and oulIel lenperalure during lhe LC injeclion for Run L3


4.2 Hydrate Dissociation

4.3.1 Temperature distribution
efore injecled inlo lhe vesseI, lhe LC soIulion vas cooIed ly lhe coiI in lhe air lalh. In an
unslirred syslen, such as lhe vesseI used in lhe experinenl, il is difficuIl for lhe nelhane
gas and valer lo forn nelhane hydrale conpIeleIy. Hence, onIy a IillIe hydrale vas forned
and nosl parl of valer or gas vas renained in lhe vesseI. esides, lhe exisling of quarlz
sand disperses lhe forned hydrale. Thus, lhe lenperalure neilher sharpIy increases in lhe
process of nelhane hydrale fornalion shovn in Iigure 14 nor sharpIy decreases in lhe
process of nelhane hydrale dissocialion. So lhe lenperalure in lhe vesseI nainlained
conslanl during LC soIulion injeclion.
The inIel (porl 1) and oulIel (porl 4) lenperalure in lhe vesseI during lhe LC injeclion for
Run L3 is shovn in Iigure 15, and lhe lenperalure profiIes for aII olher runs shov a siniIar
lrend vilh Run3.

4.3.2 Gas and Iiquid production rate
The gas produclion rale for vhoIe produclion process for Run L5, vhich is a lypicaI one, is
shovn in Iigure 4. The gas produclion rales for olher runs shov lhe siniIar characlerislics.
As shovn in Iigure 4, lhe process of lhe hydrale dissocialion vilh lhe LC injeclion in one-
dinensionaI vesseI can le divided inlo four nain seclions. In seclion I, lhe free nelhane gas
in lhe vesseI vas reIeased. This caused lhe sudden increase of inslanlaneous gas reIease rale,
up lo 8OO nI/nin. Afler lhe free gas reIeased, lhe gas produclion rale decreased and
nainlained aloul 3O nI/nin. This seclion is considered lo le seclion II. In seclion I and II,
lhe LC vas diIuled ly lhe renained valer in lhe vesseI afler hydrale fornalion, and lhere
vere IillIe hydrale dissocialed due lo lhe Iov concenlralion of LC soIulions. Wilh furlher
injeclion of lhe LC, lhe concenlralion of lhe LC soIulion in lhe vesseI increased graduaIIy.
The LC is a hydrophiIic chenicaI lhal Iovers lhe aclivily of valer and deslroys lhe slruclure
of lhe hydrale. When lhe LC concenlralion increased high enough lo nake lhe hydrale
dissociale, lhe gas produclion rale suddenIy increased lo aloul 1OO nI/nin, vhich
indicaled lhal lhe hydrale in lhe vesseI legan lo dissociale. This is lhe hydrale dissocialion
seclion (seclion III). The dissocialion seclion Iasled approxinaleIy 25 nin, vhich is lhe
Iongesl line anong lhe four seclions, as shovn in Iigure 16. Seclion IV vas lhe Iasl seclion
of lhe experinenl, vilh renain gas reIeased.
TalIe 4 provides lhe Run Lline and gas produced fron hydrale dissocialion ly LC injeclion
for aII runs. The LC injeclion line is fron lhe leginning of LC injeclion lo lhe end of hydrale
dissocialion. Onsel line for hydrale dissocialion is lhe slarling poinl of seclion III, and lhe
duralion of hydrale dissocialion is lhe Ienglh of line of seclion III. Ior exanpIe, in Run L5 in
Iigure 16, lhe onsel line for lhe hydrale dissocialion seclion is 8 nin and lhe end of lhis
seclion is 32 nin, resuIling in lhe duralion of hydrale dissocialion of approxinaleIy 24 nin.
The gas produclion ralio is defined as lhe ralio of lhe anounl of gas generaled fron hydrale
in lhe hydrale dissocialion seclion and lhe iniliaI anounl of gas conlained in aII hydrale
excIuding lhe free gas in lhe vesseI. TolaI gas produced afler LC injeclion is aIso given in
TalIe 4.
The rale of hydrale dissocialion ly LC injeclion is a funclion of LC concenlralion, injeclion
rale of LC soIulion, pressure, lenperalure of lhe syslen and hydrale-LC inlerfaciaI area.
(Sira el aI., 199O) In lhis vork, lhe pressure, lenperalure and lhe LC injeclion rale nainlain
Fuel njection 246

conslanl afler lhe LC injeclion. The inslanlaneous gas produclion rales during lhe vhoIe
process vere unsleady as shovn in Iigure 16, vhiIe lhe hydrale dissocialion rale decreased
conlinuousIy vilh line as iIIuslraled ly a lypicaI run (Run L5) in Iigure 17. The hydrale
dissocialion rale vas caIcuIaled ly lhe gas produclion rale of seclion III in Iigure 16, in
vhich lhe gas produclion vas caused ly lhe hydrale dissocialion al lhe cerlain pressure and
lenperalure.
Iigure 18 shovs lhe effecl of lhe LC injeclion rale on lhe cunuIalive gas produced fron
hydrale dissocialion as a funclion of line for Runs 1-3. The cunuIalive gas produced fron
lhe vesseI vas neasured ly lhe gas fIov neler in Iigure 1. In Runs 1-3, lhe LC
concenlralion vas kepl lhe sane al 3O vl and lhe injeclion rale vas varied fron 4.9 lo 8.8
nI/nin. As shovn in Iigure 18, in generaI, vilh lhe increase of lhe LC injeclion rale, lhe
cunuIalive gas produced increased. As lhe LC injeclion rale increase, lhere vere nore LC
injecled inlo lhe vesseI al lhe sane line, vhich increased lhe hydrale-LC inlerfaciaI area and
slinuIaled nore hydrale dissociale al lhe sane line. The generaI lrend for gas produclion
rale profiIe is siniIar in Runs 1-3, lul lhe onsel line and duralion of hydrale dissocialion
seclion are aII differenl vilh differenl LC injeclion rale. As shovn in TalIe 4, fron Run L1 lo
Run L3, lhe duralion of hydrale dissocialion seclion decrease fron 73 nin lo 35 nin, vhiIe
lhe gas produclion ralio increased fron 38.9 lo 5O.6.
Iigure 19 shovs lhe effecl of lhe LC concenlralion on lhe cunuIalive gas produced fron
hydrale dissocialion as a funclion of line for fixed injeclion rale (Runs 3-7). Iron Runs 3 lo
7, lhe LC injeclion rale vas nainlained sane al 8.8 nI/nin and lhe LC concenlralion vas
varied fron 3O lo 7O vl. Run LO vas lhe lIank experinenl, vhich injecled lhe disliIIed
valer inslead of LC soIulion, vilh lhe sane injeclion rale as Runs 3-7. AIlhough lhe generaI
lrend for gas produclion rale profiIe is siniIar in Runs 3-7 vilh lhe sane LC injeclion rale,
lhe duralion of hydrale dissocialion decrease as lhe LC concenlralion increased fron 3O
vl lo 7O vl. As shovn in TalIe 4, fron Run L3 lo Run L7, lhe gas produclion ralio
increased fron 5O.6 lo 96.2. The gas produclion ralio is Iarger lhan 9O vhiIe lhe LC
concenlralion is over 6O vl during hydrale dissocialion. On lhe olher hand, lhe LC
injeclion line for aII runs are differenl, vhich decreases vilh lhe increase of injeclion rale
and concenlralion of lhe injecled LC soIulion in generaI.
The LC injeclion and lhe soIulion produclion rale profiIes are nuch sinpIer lhan lhal of lhe
gas produclion, and Iigure 2O gives a lypicaI profiIe (Run L5). The soIulion produced fron
lhe oulIel of lhe vesseI vas conposed of lhe LC soIulion, valer in lhe vesseI lefore LC
injeclion, and valer produced fron lhe hydrale dissocialion. Iron Iigure 2O, lhe LC
injeclion rale kepl nearIy conslanl for lhe vhoIe produclion process. WhiIe lhere vas
fIuclualion for lhe soIulion produclion rale, due lo lhe unsleady slale during hydrale
dissocialion process under lhe chenicaI slinuIalion.


experinenlaI runs
LO L1 L2 L3 L4 L5 L6 L7
LC injeclion line (nin) - 1O7 71 43 33 32 29 24
Onsel line for hydrale
dissocialion (nin)
- 34 21 8 6 8 7 4
Duralion of hydrale dissocialion
(nin)
- 73 5O 35 27 24 22 2O
Cas produced fron hydrale (nI) - 977 1O88 1252 1547 18OO 2194 2268
Cas produclion ralio () - 38.9 43.3 5O.6 72.1 83.7 95.O 96.2
Cas produced afler LC injeclion
(nI)
- 3496 3334 4O25 321O 3933 418O 3368
TalIe 4. Run line and gas produced fron hydrale dissocialion ly LC injeclion

0 10 20 30 40 50
0
100
200
300
400
500
600
700
800
900
1000
1100
1200
Section V
Section Section
G
a
s

p
r
o
d
u
c
t
i
o
n

r
a
t
e
(
m
l
/
m
i
n
)
Time(min)
(Hydrate Dissociation)
Section

Iig. 16. The gas produclion rale for Run L5
Fuel njection 248

0 10 20 30 40 50
0.0
0.2
0.4
0.6
0.8
1.0
1.2
h
y
d
r
a
t
e

d
i
s
s
o
c
i
a
t
i
o
n

r
a
t
e

(
m
o
l
/
m
i
n
)
Time(min)
Section
(Hydrate Dissociation)
Section V

Iig. 17. The hydrale dissocialion rale for Run L5
0 10 20 30 40 50 60 70 80
0
200
400
600
800
1000
1200
1400
1600
1800
2000
Run #3
C
u
m
u
l
a
t
i
v
e

g
a
s

p
r
o
d
u
c
e
d

(
m
l
)
Time(min)
Run #2
Run #1

Iig. 18. The cunuIalive gas produced during lhe hydrale dissocialion for Runs 1-3

0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40
0
500
1000
1500
2000
2500
C
u
m
u
l
a
t
i
v
e

g
a
s

p
r
o
d
u
c
e
d

(
m
l
)
Time(min)
Run #3
Run #4
Run #6
Run #5
Run #7
Run #0

Iig. 19. The cunuIalive gas produced during lhe hydrale dissocialion for Runs 3-7
0 10 20 30 40 50
0
2
4
6
8
10
12
14
16
18
20
22
24
EG injection rate
Solution production rate
S
o
l
u
t
i
o
n

i
n
j
e
c
t
i
o
n

a
n
d

p
r
o
d
u
c
t
i
o
n

r
a
t
e
(
g
/
m
i
n
)
Time(min)
Hydrate Dissociation

Iig. 2O. SoIulion injeclion and produclion rale profiIe for Run L5

4.3.3 Production Efficiency AnaIysis
The efficiency of producing gas fron hydrale ly LC injeclion is invesligaled here. In order
lo conpare lhe efficiency of differenl runs, lhe produclion efficiency has leen defined as lhe
ralio of lhe voIune of produced gas lo lhe nass of LC injecled in unil line.
Fuel njection 250

Under lhe LC slinuIalion, lhe hydrale dissocialed onIy on lhe hydrale dissocialion seclion
(Seclion III shovn in Iigure 16). TalIe 5 shovs lhree produclion efficiencies:
1) Iroduclion efficiency for hydrale dissocialion seclion (seclion III) (nI/g/nin)
2) Iroduclion efficiency al 5O hydrale dissocialion (seclion III) (nI/g/nin)
3) Iroduclion efficiency for vhoIe injeclion process (seclion I lo III) (nI/g/nin)
The firsl one vas caIcuIaled vilh lhe voIune of gas, lhe nass of LC injecled and lhe
duralion line of lhe hydrale dissocialion seclion (seclion III in Iigure 16). Il can le used lo
neasure lhe gas produclion efficiency of hydrale dissocialion process ly lhe effecl of LC in
seclion III.
The second one vas caIcuIaled lhe sane vay vilh lhe firsl one, vhiIe il used lhe vaIues of
5O hydrale dissocialion poinl. In lhis vork, lhe hydrale dissocialion runs vere divided
inlo 4 seclions, and lhe hydrale dissocialion onIy happened in seclion III. The anounl of
dissocialed hydrale vas neasured ly lhe gas voIune reIeased fron lhe vesseI in seclion III.
So 1OO hydrale dissocialion poinl vas lhe end of seclion III, vhen lhe gas produced fron
hydrale dissocialion aII reIeased fron lhe vesseI. In lhe sane vay, 5O hydrale dissocialion
poinl vas sone line in seclion III, vhen 5O gas produced fron hydrale dissocialion.
olh lhe firsl and second produclion efficiencies vere caIcuIaled lased on lhe experinenlaI
resuIl of seclion III (lhe hydrale dissocialion seclion).
The lhird one vas used lo neasure lhe vhoIe experinenlaI runs, fron lhe leginning of LC
injeclion lo lhe end of lhe hydrale dissocialion process, vhich incIude seclion I lo seclion III.
The fornalion condilions of hydrale used in lhe vork vas sane for aII runs, incIuding
pressure, lenperalure, anounl of valer and lhe hydrale conlenl in lhe vesseI. Ior aII LC
injeclion runs, lhe inpacl of lhe hydrale conlenl, lhe anounl of valer and free gas in lhe
vesseI, and lhe difference of lhe operaling condilions (incIuding lhe pressure and
lenperalure) during LC injeclion vere aII eIininaled, vhiIe lhe inpacl of lhe LC
concenlralion and injeclion rale on lhe produclion efficiency refIecled in TalIe 5. As shovn
in TalIe 5, lhe produclion efficiency for lhe vhoIe LC injeclion process vas lelveen O.2O
and O.88 nI/g/nin, vhiIe lhe efficiency for hydrale dissocialion seclion vas lelveen O.12
and O.8O nI/g/nin. ul lhe produclion efficiency al 5O hydrale dissocialion vas nuch
higher, vilh lhe naxinun of 2.O3 nI/g/nin.
Iron lhe efficiency anaIysis, lhe foIIoving concIusions can le dravn: (1) lhe produclion
efficiency of lhe hydrale dissocialion seclion, 5O hydrale dissocialion and lhe vhoIe
injeclion process varied vilh lhe LC concenlralion and injeclion rale, and lhe varialion
presenled lhe sane lrend. (2) The hydrale dissocialion rale decreased as lhe experinenls go
on, as shovn in Iigure 17. This can expIain vhy lhe efficiencies al 5O hydrale dissocialion
vere aII higher lhan lhal of lhe vhoIe injeclion process. (3) Wilh lhe increase of lhe LC
injeclion rale, lhe produclion efficiency increases, as Runs 1-3. The produclion efficiencies
for hydrale dissocialion seclion of Runs 1, 2 and 3 increase fron O.12 lo O.5O vhen lhe
injeclion rale increase fron 4.9 lo 8.8 nI/nin. (4) The LC concenlralion aIso infIuence lhe
produclion efficiency, as Runs 3-7. Iron lhe resuIl shovn in TalIe 5, lhere vas a naxinun
poinl for lhe LC concenlralion on lhe produclion efficiency vilh lhe sane injeclion rale (9
nI/nin), as Run L6 vilh 6O vl LC soIulion. Wilh lhe increase of lhe LC concenlralion, lhe
gas produclion rale increases, as lhe resuIl shovn in Iigure 19. WhiIe lhe nass of LC
injecled inlo lhe vesseI increase vilh lhe increase of lhe concenlralion, lhe produclion
efficiency reaches lhe naxinun of O.8O in Run L6. The sane resuIl vas aIso concIuded on

lhe produclion efficiency al 5O hydrale dissocialion and lhe vhoIe produclion process
incIuding lhe free gas reIease seclion.

experinenlaI runs
LO L1 L2 L3 L4 L5 L6 L7
Iroduclion efficiency for
hydrale dissocialion seclion
(seclion III) (nI/g/nin)
- O.12 O.28 O.5O O.57 O.6O O.8O O.76
Iroduclion efficiency al 5O
hydrale dissocialion (seclion III)
(nI/g/nin)
- O.44 1.32 1.6O 1.61 1.75 2.O3 1.53
Iroduclion efficiency for vhoIe
injeclion process (seclion I lo III)
(nI/g/nin)
- O.2O O.31 O.79 O.79 O.81 O.88 O.86
TalIe 5. Iroduclion efficiency anaIysis for lhe hydrale dissocialion ly LC injeclion

5. ConcIusions
1 The fIoving of hol valer or hol lrine injecled in lhe vesseI can le regarded as lhe
noving of a pislon fron lhe inIel lo lhe oulIel.
2 The hydrale dissocialion process is divided inlo lhree slages: free gas produclion,
hydrale dissocialion and residuaI gas produclion.
3 The process of lhe hydrale dissocialion is a process of lhe lenperalure decrease in lhe
presence of lhe lrine soIulion. The duralion of lhe hydrale dissocialion is shorlened and
lhe degree of lhe deplh of lhe lenperalure drop increases vilh lhe increase of lhe lrine
concenlralion.
4 The rale of inslanlaneous hydrale dissocialion increases vilh lhe increase of lrine
concenlralion vilh lhe injeclion of hol lrine soIulion. Hovever, vhen lhe lrine
concenlralion is higher lhan lhe cerlain degree, lhe rale of inslanlaneous hydrale
dissocialion no Ionger conlinues lo increases.
5 During lhe hydrale dissocialion, lhe rale of lhe Iiquid produclion is sIighlIy higher lhan
lhe rale of lhe soIulion injeclion, due lo lhe valer produced fron lhe hydrale
dissocialion.
6 ThernaI efficiency and energy ralio for lhe hydrale produclion can le enhanced ly
injecling hol lrine soIulion, and lhe enhance effecliveness is apparenl vilh lhe injeclion
of high lrine concenlralion al lhe reIalive Iov lenperalure.
7 Afler lhe LC injeclion, lhe hydrale dissocialion in lhe vesseI can le divided inlo four
seclions, lhal is lhe free gas producing seclion, LC diIuling seclion, lhe hydrale
dissocialing seclion and lhe rennanl gas producing seclion.
8 The gas and valer produclion rale vere lolh unsleady during hydrale dissocialion rale
decrease conlinuousIy vilh line under lhe LC slinuIalion, vhiIe lhe LC injeclion rale
kepl nearIy conslanl for lhe vhoIe produclion process.
Fuel njection 252

9 Under lhe experinenl condilions, vilh lhe LC injeclion rale increasing, lhe gas
produclion ralio increased, lhe duralion of hydrale dissocialion shorlened and lhe
produclion efficiency increased.
1O Under lhe experinenl condilions, vilh lhe LC concenlralion increasing, lhe gas
produclion ralio increased, lhe duralion of hydrale dissocialion process shorlened. And
lhe LC concenlralion aIso affecls lhe produclion efficiency. The produclion efficiency
for lhe vhoIe LC injeclion process increased vilh lhe LC concenlralion increasing fron
O lo 6Ovl, and afler lhal lhe produclion efficiency legan lo decrease.

6. AcknowIedgments
The aulhors appreciale lhe financiaI supporl fron lhe NalionaI High TechnoIogy Research
and DeveIopnenl Irogran of China (No.2OO6AAO9A2O9, No.2OO6AAO5Z319) and lhe
NalionaI NaluraI Science Ioundalion of China (No.2O676133).

7. References
CoIIell, T. S. (2OO4). Cas hydrales as a fulure energy resource, Gcc|imcs, 49, 11, (2OO4) pp 24-
27, ISSN OO16-8556
Ian, S. S., Zhang, Y. Z., Tian, C. L., Liang, D. Q., Li, D. L. (2OO6). NaluraI gas hydrale
dissocialion ly presence of elhyIene gIycoI. |ncrgq c |uc|s, 2O, 1, (2OO6) pp 324-326,
ISSN O887-O624
CoeI, N., Wiggins, M., Shah, S., AnaIylicaI nodeIing of gas recovery fron in silu hydrales
dissocialion. ]curna| cf Pc|rc|cum Scicncc an |nginccring 2OO1, 29, (2), 115-127, ISSN
O92O-41O5
Kanalh, V. A. M., I.N., Sira, }.H., IaliI, S.L. (1991). LxperinenlaI sludy of rine injeclion
and depressurizalion nelhods for dissocialion of gas hydrale. SP| |crma|icn
|ta|ua|icn, 6, 4, (1991) pp 477-484, ISSN O885-923X
Kavanura, T., Yananolo, Y., Ohlake, M., Sakanolo, Y., Konai, T., Haneda, H. (2OO5a).
Dissocialion experinenl of hydrale core sanpIe using lhernodynanic inhililors,
15|n |n|crna|icna| Offsncrc an Pc|ar |nginccring Ccnfcrcncc (|SOP| 2005), pp. 346-
35O, SeouI, SOUTH KORLA, }un 19-24, 2OO5, ISN 1-88O653-64-8
Kavanura, T., Yananolo, Y., Ohlake, M., Sakanolo, Y., Konai, T., Haneda, H. (2OO5l). In
LxperinenlaI sludy on dissocialion of hydrale core sanpIe acceIeraled ly
lhernodynanic inhililors for gas recovery fron naluraI gas hydrale, Tnc 5|n
|n|crna|icna| Ccnfcrcncc cn Gas Hqra|c, pp 3O23-3O28, Trondhein, Norvay, 12-16
}une, 2OO5, ISN 978161567O666
Kavanura, T., Sakanolo, Y., Ohlake, M., Yananolo, Y., Haneda, H., Yoon, }. H., Konai, T.
(2OO6). Dissocialion lehavior of hydrale core sanpIe using lhernodynanic
inhililor, |n|crna|icna| ]curna| cf Offsncrc an Pc|ar |nginccring, 16, 1, (2OO6) pp 5-9,
ISSN 1O53-5381
Kavanura, T., Ohlake, M., Sakanolo, Y., Yananolo, Y., Haneda, H., Konai, T., Higuchi, S.
(2OO7). LxperinenlaI sludy on slean injeclion nelhod using nelhane hydrale core
sanpIes, Prccccings cf Tnc Sctcn|n (2007) |SOP| Occan Mining Sqmpcsium, pp 83-86,
Lislon, IORTUCAL, }uI O1-O6, 2OO7, ISN 1-88O653-57-X

KIauda, }. ., SandIer, S. I. (2OO5). CIolaI Dislrilulion of Melhane Hydrale in Ocean
Sedinenl, |ncrgq c |uc|s, 19, 2, (2OO5) pp 459-47O, ISSN O887-O624
Kono, H. O., Narasinhan, S., Song, I., Snilh, D. H. (2OO2). Synlhesis of nelhane gas hydrale
in porous sedinenls and ils dissocialion ly depressurizing, Pcucr Tccnnc|cgq, 122,
2-3, (2OO2) pp 239-246, ISSN OO32-591O
Li, C., Tang, L., Huang, C., Ieng, Z., Ian, S. (2OO6). Thernodynanic evaIualion of hol lrine
slinuIalion for naluraI gas hydrale dissocialion, Huagcng Xuc|ac/]curna| cf Cncmica|
|nus|rq an |nginccring (Cnina), 57, 9, (2OO6) pp 2O33-2O38, ISSN O438-1157
Li, C., Li, X., Tang, L., Zhang, Y. (2OO7a). LxperinenlaI invesligalion of produclion lehavior
of nelhane hydrale under elhyIene gIycoI slinuIalion in unconsoIidaled sedinenl,
|ncrgq c |uc|s, 21, 6, (2OO7) pp 3388-3393, ISSN O887-O624
Li, C., Li, X., Tang, L., Zhang, Y., Ieng, Z., Ian, S. (2OO7l). LxperinenlaI Invesligalion of
Iroduclion ehavior of Melhane Hydrale under LlhyIene CIycoI Injeclion, Huagcng
Xuc|ac/]curna| cf Cncmica| |nus|rq an |nginccring (Cnina), 58, 8, (2OO7l) pp 2O67-
2O74, ISSN O438-1157
Li, X. S., Wan, L. H., Li, C., Li, Q. I., Chen, Z. Y., Yan, K. I. (2OO8a). LxperinenlaI
Invesligalion inlo lhe Iroduclion ehavior of Melhane Hydrale in Iorous Sedinenl
vilh Hol rine SlinuIalion, |nus|ria| c |nginccring Cncmis|rq Rcscarcn, 47, 23,
(2OO8) pp 9696-97O2, ISSN O888-5885
Li, C., Li, X., Tang, L.-C., Li, Q.-I. (2OO8l). ConlroI Mechanisns for Melhane Hydrale
Iroduclion ly ThernaI SlinuIalion, Prccccings cf |nc 6|n |n|crna|icna| Ccnfcrcncc cn
Gas Hqra|cs (|CGH 2008), Vancouver, rilish CoIunlia, CANADA, }uIy 6-1O, 2OO8.
Makogon, Y. I., HoIdilch, S. A., Makogon, T. Y. (2OO7). NaluraI gas-hydrales - A polenliaI
energy source for lhe 21sl Cenlury, ]curna| cf Pc|rc|cum Scicncc an |nginccring, 56,
1-3, (2OO7) pp 14-31, ISSN O92O-41O5
MiIkov, A. V. (2OO4). CIolaI eslinales of hydrale-lound gas in narine sedinenls: hov nuch
is reaIIy oul lhere` , |ar|n-Scicncc Rcticus, 66, 3-4, (2OO4) pp 183-197, ISSN OO12-8252
Moridis, C. }. (2OO3). NunericaI sludies of gas produclion fron nelhane hydrales, SP|
]curna|, 8, 4, (2OO3) pp 359-37O, ISSN OO36-1844
Moridis, C. }., CoIIell, T. S., DaIIinore, S. R., Saloh, T., Hancock, S., WealheriII, . (2OO4a).
NunericaI sludies of gas produclion fron severaI CH4 hydrale zones al lhe MaIIik
sile, Mackenzie DeIla, Canada. ]curna| cf Pc|rc|cum Scicncc an |nginccring, 43, 3-4,
(2OO4) pp 219-238, ISSN O92O-41O5
Moridis, C. }. (2OO4l). NunericaI Sludies of Cas Iroduclion fron CIass 2 and CIass 3
Hydrale AccunuIalions al lhe MaIIik Sile, Mackenzie DeIla, Canada, SP| Rcscrtcir
|ta|ua|icn an |nginccring, 7, 3, (2OO4) pp 175-183, ISSN 1O94-647O
Moridis, C. }., KovaIsky, M. ., Iruess, K. (2OO7). Depressurizalion-induced gas produclion
fron cIass 1 hydrale deposils, SP| Rcscrtcir |ta|ua|icn c |nginccring, 1O, 5, (2OO7)
pp 458-481, ISSN 1O94-647O
Moridis, C. }., CoIIell, T. S., osveII, R., Kurihara, M., Reagan, M. T., Koh, C., SIoan, L. D.
(2OO9a). Tovard Iroduclion Iron Cas Hydrales: Currenl Slalus, Assessnenl of
Resources, and SinuIalion-ased LvaIualion of TechnoIogy and IolenliaI, SP|
Rcscrtcir |ta|ua|icn c |nginccring, 12, 5, (2OO9) pp 745-771, ISSN 1O94-647O
Moridis, C. }., Reagan, M. T., Kin, S. }., SeoI, Y., Zhang, K. (2OO9). LvaIualion of lhe Cas
Iroduclion IolenliaI of Marine Hydrale Deposils in lhe UIIeung asin of lhe
Korean Lasl Sea, SP| ]curna|, 14, 4, (2OO9) pp 759-781, ISSN OO36-1844
Fuel njection 254

Sira, }. H. IaliI, S. L., Kanalh, V. A. (199O). Sludy of hydrale dissocialion ly nelhanoI and
gIycoI injeclion, Prccccings - SP| Annua| Tccnnica| Ccnfcrcncc an |xni|i|icn, pp 977-
984, IulI ly Soc of IelroIeun Lngineers of AIML, Richardson, TX, USA, 199O.
SIoan, L. D., Koh, C. A. (2OO8). C|a|nra|c Hqra|cs cf Na|ura| Gascs. 3nd ed ed., oca Ralon,
ILorida : TayIor and Irancis, Inc.: 2OO8, ISN O-8493-9O78
Sung, W. M., Lee, H., Lee, C. (2OO2). NunericaI sludy for produclion perfornances of a
nelhane hydrale reservoir slinuIaled ly inhililor injeclion, |ncrgq Scurccs, 24, 6,
(2OO2) pp 499-512, ISSN O887-O624
Sung, W. M., Lee, H., Kin, S., Kang, H. (2OO3). LxperinenlaI invesligalion of produclion
lehaviors of nelhane hydrale saluraled in porous rock, |ncrgq Scurccs, 25, 8, (2OO3)
pp 845-856, ISSN O887-O624
Tang, L., Xiao, R., Huang, C., Ieng, Z., Ian, S. (2OO5a). LxperinenlaI invesligalion of
produclion lehavior of gas hydrale under lhernaI slinuIalion in unconsoIidaled
sedinenl, |ncrgq c |uc|s, 19, 6, (2OO5) pp 24O2-24O7, ISSN O887-O624
Tang, L. C., Li, C., Hao, Y. M., Ian, S. S., Ieng, Z. I. (2OO5l). Lffecls of saIl on lhe fornalion
of gas hydrale in porous nedia, Tnc 5|n |n|crna|icna| Ccnfcrcncc cn Gas Hqra|c, pp
155-16O, Trondhein, Norvay, 2OO5, ISN 978161567O666
Tang, L., Li, X., Ieng, Z., Li, C., Ian, S. (2OO7). ConlroI nechanisns for gas hydrale
produclion ly depressurizalion in differenl scaIe hydrale reservoirs, |ncrgq c
|uc|s, 21, 1, (2OO7) pp 227-233, ISSN O887-O624
Yousif, M. H., Li, I. M., SeIin, M. S., SIoan, L. D. (199O). Depressurizalion of naluraI gas
hydrales in erea sandslone cores, ].|nc|usicn PncncmcMc| Rcccgni|icn Cncm, 8,
(199O) pp 71-88, ISSN O923-O75O
Yousif, M. H., Alass, H. H., SeIin, M. S., SIoan, L. D. (1991a). LxperinenlaI and lheorelicaI
invesligalion of nelhane-gas-hydrale dissocialion in porous nedia, SP| Rcscrtcir
|nginccring, 6, 4, (1991) pp 69-76, ISSN O885-9248
Yousif, M. H., SIoan, L. D. (1991l). LxperinenlaI Invesligalion of Hydrale Iornalion and
Dissocialion in ConsoIidaled Iorous Media, SP| Rcscrtcir |nginccring, 11, (1991l)
pp 452-458, ISSN O885-9248

Fuel Injection

Enviado por

Dados do documento

Descrição original:

Direitos autorais

Formatos disponíveis

Compartilhar este documento

Compartilhar ou incorporar documento

Opções de compartilhamento

Você considera este documento útil?

Este conteúdo é inapropriado?

Direitos autorais:

Formatos disponíveis

Fuel Injection

Enviado por

Direitos autorais:

Formatos disponíveis

Fuel Injection

j=0 j=1 j=2 j=3

, is lhe increase in lhe concenlralion of lhe fueI vapor per unil

. These quanlilies are delernined fron lhe

lhal descriles heal exchange

(infIov of heal fron lhe Iiquid phase lo lhe unil voIune

Fuel njection 128

Fuel njection 138

0 0.002 0.004 0.006 0.008 0.01

g, veclor of dropIel veIocily reIalive lo gas, x, y, z,

C ly a vaporizalion device in Iig. 14. On lhe

Você também pode gostar