University of Massachusets - Amherst

ScholarWorks@UMass Amherst
Wind Energy Center Reports UMass Wind Energy Center
1976
Discussion Of Momentum Teory For Windmills
Forrest S. Stoddard
Follow this and additional works at: htp://scholarworks.umass.edu/windenergy_report
Tis Article is brought to you for free and open access by the UMass Wind Energy Center at ScholarWorks@UMass Amherst. It has been accepted for
inclusion in Wind Energy Center Reports by an authorized administrator of ScholarWorks@UMass Amherst. For more information, please contact
scholarworks@library.umass.edu.
Stoddard, Forrest S., "Discussion Of Momentum Teory For Windmills" (1976). Wind Energy Center Reports. Paper 20.
htp://scholarworks.umass.edu/windenergy_report/20
UNNUZSrrV OF MASSACHUSETE/AMHUZST
ENERGY ALTERNATIVES PROGRAM
UNIVERSITY OF MASSACHUSETTSIAMHERST
ENERGY ALTERNATIVES PROGRAM
DISCUSSION OF MOMENTUM THEORY FOR WINDMILLS
by
For r est S. Stoddard
U.Mass. Wind Furnace
Energy A1 t e r n a t i ves Program
Uni v e r s i t y o f Massachusetts
Amherst, Massachusetts 01002
TR/76/ 2
Appendix I V
DISCUSSION OF MOMEr4TUM THEORY FOR WINDMILLS
Molllenturn t heor y has been ex t ens i v el y used t o p r e d i c t t he r e l a t i v e per -
f ormance o f 1 i f t i n g p r o p e l l e r s and. r ot or s. ' 2 y 3 y 4 ) The per f or mance equat i ons
a r e a n a l y t i c a l l y and c oncept ual l y si mpl e; t h i s l eads t o qui c k i d e a l p r e d i c t i o n s
and compari sons. Of t en a phy s i c al " f e e l i n g " o r under st andi ng o f t he syst em can
be gai ned by ex er c i s i ng t h i s si ni pl e approach f i r s t , bef or e more complex bl ade
el ement and s t r i p t heor i es ar e used.
Thi s di s cuss i on concer ns t he r a mi f i c a t i o n s and i n t e r p r e t a t i o n o f t he
momentum t heor y expr essi ons f o r a t h r u s t i n g wi ndmi l l i n t he l i g h t o f obser ved
wi ndmi l l behavi or . It shoul d be kept i n mi nd t h a t t h e ( i d e a l i z e d ) a n a l y t i c a l
f or mul at i on depends on t he f o l l o wi n g s t r i c t assumpt i ons:
a )
d e f i n i t e st r eaml i nes e x i s t i n t he f l o w f i e l d ;
b) no f r i c t i o n a l l osses a r e pr esent ;
c )
t he i nduced v e l o c i t y i mpar t ed t o t he f r e e st r eam i s const ant over t he
ar ea o f t h e i d e a l i z e d r o t o r , o r act uat or .
The f o l l o wi n g wel l - known expr essi ons ar e t hus devel oped f o r t he ac t uat or shown: ( 4 s )
T = t h r u s t = 2p A [ Vo-v] v
( 1
Z
P = power = T( Vo- V) = 211 A [ v 0- v l v
I n non- di mensi onal f or m:
t h r u s t
T
C = c o e f f i c i e n t =
T
1 /2pAV0
2
power
Cp = c o e f f i c i e n t = -
P
3
1 / 2 d V0
and
a x i a l
a = i n t e r f e r e n c e = -- v
f a c t o r vo
The f l ow f i e l d f or t h e normal wi ndmi l l s t a t e i s :
I
v p o s i t i v e i n
t h e d i r e c t i o n
shown
I
where :
Act uat or !_
V o = f r ee st ream v e l o c i t y
Di sc V = f a r wake v e l o c i t y
ru
v = i nduced v e l o c i t y
T = t h r u s t (down wi nd)
P = power ( out of system)
FIGURE 1
Thi s i s a l s o c a l l e d t he wi ndmi l l brake s t a t e f o r r o t o r s si nce t he t h r u s t i s i n
t he same d i r e c t i o n as t he f r e e stream; t hus, t h i s s t a t e r epr esent s a r o t o r i n
aut or ot at i on o r v e r t i c a l descent. I n wi ndmi l l t er mi nol ogy, as shown above,
t h r u s t c oef f i c i ent , CT, and power c o e f f i c i e n t , Cp, ar e pos i t i v e, i n d i c a t i n g
t h r u s t ( o r dr ag) i n t he downwind d i r e c t i o n and power o u t o f t he system.
As shown by Wi l son & Lissaman, Reference 1, t her e ar e two ot her i mpor t ant
f l ow st at es det ermi ned by t he val ue o f i nduced v el oc i t y , v, wr i t t e n non-
di nl ensi onal l y as a x i a l i nt er f er ence f ac t or , a = v/Vo:
v = O
Normal Working St at e
(Windmi 11 Brake St at e)
- - - - - - -
.: <
0 - a - 0.5
. . . . . . - - . . - - - - . - . - - . - - - - - .
Ci,, C o o s i t i vc
T
. . - - - - . - . . - - - - . . . - . - - - - - - -
Propel 1 e r St at e
-
a ; O
. - - - - - - -- - - - - - - - . . . - - - - - - - - -.
Cp, CT negat i ve
- - - - . . - - - -. - - - . . . . - - . . - . . . . -
Zero Sl i p Case
a = O
- -. -. - - - - -. -
C P =C - I - = O
- - -- - - - - - - - - - -. - - - - - - -- - -
For a = 1/ 2, t he devel oped wake v e l o c i t y i s :
f a r
V = wake = Vo( l - 2a) = 0
" v e l o c i t y
Thi s c ondi t i on r epr esent s a p o i n t a t whi ch st r eaml i nes no l onger e x i s t .
Hence, t he momentum t heor y assumpt i on has been vi ol at ed, and t h i s s t a t e cannot
e x i s t as a momentum t heor y s ol ut i on.
Wr i t i n g CT and Cp i n t erms o f a:
Wi l son & Li ssamanget t he f o l l o wi n g p l o t (Ref erence 1 ) :
C~
I I
-
1. 5
a = v/V
0
Momentum t heor y
does n o t work
I
I
I
I
FIGURE 3
For a -1.0, t he f a r wake v el oc i t y , V .,, , i s s t i l l not def i ned by 11ionientu111 t heor y;
but t h i s s t at e can al s o be seen t o be t he pr opel l er ac t i ng a s a brake. That i s ,
power i s bei ng added t o t he f l ow t o cr eat e hi gh CT, o r hi gh t hr us t downwind.
Thi s would be t he case o f r ever si ng pr opel l er t hr us t on l andi ng.
These f l ow st at es r epr esent a ser i es o f gradual changes whi ch occur on
t he r o t o r as i nduced v el oc i t y i s changed uni f or ml y. Unf or t unat el y, because a
>
d e f i n i t e sl i pst r eam does not e x i s t f or a -112, momentum t heor y cannot be used t o
pr edi c t t he r o t o r performance i n t h i s r egi on. However, f rom hel i copt er aut o-
r o t a t i v e data, empi r i cal curves have been drawn, and t he f l ow st at es have been
documented. ( 29697) Expanding t he cur ve f rom t he precedi ng page, t he r o t o r
behavi or can be descri bed i n more det ai 1 :
Windmil 1 Brake St at e
Zero S l i p
I
Propel l e r St at e
Tur bul ent Wake St at e
FIGURE 4
The e l l ~p i r i c a l s t a t e boundari es ar e found f r om t y p i c a l h e l i c o p t e r descent
( aut or ot at i v e) perf orl l l ance d a t a p l o t t e d as f o l l ows. The absci ssa i s non-dimen-
s i onal r a t e o f cl i mb, To, and t he or di nat e i s non-di mensi onal i nduced v el oc i t y .
-
v, as def i ned bel ow:
wi t h - v - o
v = n 7 7 J r ; = JT/ L~A
[Note: Here t he si gn o f t he f r e e st ream v el oc i t y , Vo, i s opposi t e t o t h a t used
f o r wi ndmi 11 t heor y]
-
v
Dat a f r om I
[Momentum \l;i \ , , \ + - 5
t heor y does
fnot -- - - - - - -
Vor t ex Ri ng St at e (
Rot or o r Pr opel l er St at e
, . - .
FIGURE 5
, I
Wi ndmi l l Brake St at e I
1 I
1
( h e l i c o p t e r ascendi ng)
I I 1
- 3
I
I I
J - ~ L
- 1 0 1 2
3
-
Tur bul ent Wake St at e vo
The niomentuni equat i ons (shown here i n sol i d 1 i nes) f o r t he he1 i c opt er
perf or~i i ance case are:
(2' 8)
Pr opel l er o r Rot or St at e
CJindmi 11 Brake St at e:
It i s al so seen t h a t any poi nt on t he curve r epr esent s a val ue o f a:
Thus, t he Tur bul ent Wake St at e l i mi t i n g case occurs at :
-
v = 1.0, ( - ) vo = 2 , a = 1/ 2
And t he Vor t ex Ri ng St at e:
-.
( - ) \ 1.73, v = 1.3, a = - 75 (Ref. 7 )
-
V = 1.75, a - 1. 0 (Ref. 6)
The act ual val ues of t hr us t cannot be foundfroni Fi gur e 5, si nce t he
or i gi nal dat a were l o s t i n t he non- di mensi onal i zi ng process. However, t he
e x i s t i n g t e s t dat a coul d be used t o generate CT and Cp vs v/Vo empi r i cal p l o t s
t o appear i n Fi gur e 4.
The t e s t dat a r ef er r ed t o appear i n two d i s t i n c t groupi ngs; one i s a t
-
Vo = -1.75 and t he ot her i s j u s t below hover ( j u s t descendi ng). The c l us t er
around hover i ndi cat es t he part-power descent o f ( r ot ar y wi ng) f l i g h t vehi cl es ;
t he l e f t c l us t er i s f rom f ree-wheel i ng r ot or s ( i . . , Cp = 0 negl ect i ng p r o f i l e
- -
dr ag l osses) , and occurs a t v/Vo 2 1.0. Thi s v e r i f i e s t he momentum t heor y
pr edi c t i on t h a t Cp = 0 a t v/Vo = 1. 0 (see Fi gur e 3) .
I n t he Tur bul ent Wake St at e, t he sl i pst r eam expansi on i s ver y l ar ge, and
consi der abl e t ur bul ence and r e c i r c u l a t i o n e x i s t (Ref erence 2) . I n f a c t , t he
r o t o r act s as a l a r g e di s c per pendi cul ar t o t he f l ow. Thus, t he t h r u s t
c o e f f i c i e n t i s synonymous wi t h t he drag c o e f f i c i e n t o f a s o l i d d i s c :
C,,
1.5 - 2. 0. As a r e s u l t we woul d expect t he CT t o go up shar pl y as
a = v/V (synonymous wi t h p o r o s i t y of t h e d i s c ) i ncr eases. Hence, we woul d
0
expect hi gher val ues o f CT t han t hose pr edi ct ed by momentum t heor y.
The Vor t ex Ri ng St at e boundary i s n o t wel l est abl i shed, b u t has been
observed f o r a number of di f f er ent NACA r o t o r s (see Ref erence 6) and e a r l i e r
t e s t s (see Reference 7 ) . Thi s boundary a l s o marks t he change f r om p o s i t i v e
t o negat i ve power; t h a t i s , t he vor t ex r i n g s t a t e r epr esent s h e l i c o p t e r r o t o r s
i n powered a u t o r o t a t i o n ( f o r whi ch s i g n i f i c a n t t e s t dat a e x i s t ) as we l l as
r ever se t h r u s t i n g pr opel 1 er s ( t h e same f l o w f i e l d ) . Momentum t heor y cannot be
bel i eved i n t he range, 1/ 2 < a < 1.0, but t he power equat i on ( Equat i on 5 ) does
p r e d i c t t he change i n Cp (see Fi gur e 3) .
References
1. Wilson &Lissaman "Applied Aerodynamics of Wind Power Machines," Oregon
State University, July 1974.
2. Gessow & Myers, Aerodynamics of the Helicopter, Frederick Ungar Publishing,
1952.
3. Prandtl & Tietjens, Applied Hydro & Aeromechanics, Dover Publications, 1934.
4. Glauert, Elements of Aerofoil & Airscrew Theory, Cambridge University Press,
1959.
5. Glauert, "Windmills and Fans," Aerodynamic Theory, Vol IV, Durand, ed.,
Dover Pub1 ications, 1963, pp. 324-332.
6. Gessow, "Flight Investigation of Effects of Rotor Blade Twist on Helicopter
Performance in the High Speed and Vertical-Autorotative Decsent Conditions,"
NACA TN 1666, 1948.
7. Lock, Bateman, Townend, "An Extension of the Vortex Theory of Airscrews
with Applications to Airscrews of Small Pitch and Includivg Experimental
Results," British Aeronautical Research Comm. Reports & Memoranda, No. 1014,
1924.
8. Ham, "V/STOL Vehicle Aerodynamics and Dynamics," Course 16.50 class notes,
MIT, 1963.
UNIVERSITY OF MASSACHUSETTS/AMHERST
ENERGY ALTERNATIVES PROGRAM
UNIVERSITY OF MASSACHUSETTS/AMHERST
ENERGY ALTERNATIVES PROGRAM
AN APPROACH TO PRELIMINARY SYSTEMS OPTIMIZATION
OF THE NEW ENGLAND WIND FURNACE
F o r r e s t S. St o d d a r d
U.Mass. Wi nd Fur nac e
Ener gy A l t e r n a t i v e s Pr ogr am
U n i v e r s i t y o f Mas s ac hus et t s
Amher st , Mas s ac hus et t s 01002
TR/ 76/ 3
Appendi x V
AN APPROACH TO-,PR_F-~_T_M!~lA~Yv _S_Y;T_E_M_S- _O_PTJfl,ZA_TT_ION- _OJ_ T! 1.r- r\rLw. JNGLAND. !dI gn. 11 l?N?c r
--
1. Genera 1 Approach
The Wind Furnace system i s r epr esent ed i n Fi gur e 1. The di agram
i l l u s t r a t e s t h e t hr ee i ndependent c ont r ol var i abl es: p i t c h ( ) f i e l d
e x c i t a t i o n (PeXc), and l oad r esi st ance (RL). That is, f o r each s e t o f
i n p u t condi t i ons (Vo,B,,Pex, , R ) t her e w i l l be an out put energy, 1 ; ~ ~ .
The opt i mi z at i on t ask c ons i s t s o f f i n d i n g t h e r i g h t combi nat i ons of t hese
var i abl es whi ch y i e l d t he hi ghest out put f o r gi ven wi nd speed, Vo. Lat er ,
t he anal ysi s can be r e f i n e d t o g i v e dynamic and s t a b i l i t y improvements
gi ven Vo( t ) ( subj ect t o t he quas i - s t at i c assumpt i ons used i n d e r i v i n g
t he a n a l y t i c a l model s f or t he p r o p e l l e r and gener at or ) . Tr ansi ent behavi or
o f t he system may be at t empt ed f o r c e r t a i n equi l i br i um cases, as i s done
wi t h s t a b i l i t y d e r i v a t i v e s f o r f l i g h t systems.
Pr el i mi nar y, o r " st at i c, " systems opt i mi z at i on uses a n a l y t i c and/ or
semi - empi r i cal performance c h a r a c t e r i s t i c s of t he two subsystems, and
si mpl y matches oper at i ng poi nt s t o det er mi ne e q u i l i b r i u m condi t i ons.
Par amet r i c t orque-speed p l o t s w i l l be used i n t h i s anal ysi s; hence t he
i n t e r p r e t a t i o n of "mat chi ng t he char act er i st i cs, " i s t o equat e t h e out put
t or que o f t h e mechanical ( wi ndmi l l ) shaf t , t o t h e i n p u t t or que o f t h e gener at or .
The par amet r i c performance of t he gener at or has been measured i n
l abor at or y t e ~ t s ( ~ ) ( ~ i g u r e 2) .
For each p o i n t on t h i s p l o t an out put l oad
2
cur r ent , Icy w i l l det er mi ne t he out put power, I o RL -
Work i s cont i nui ng, t o
devel op an a n a l y t i c a l model f o r t h e gener at or and f o r ot her t ypes of
gener at or s.
I t i s underst ood t h a t l abor at or y t e s t s ar e i mpor t ant f o r t he
comprehension and v e r i f i c a t i o n of t he gener al a n a l y t i c a l t heor y.
Heat
Energy
T
genera t or
FIGURE 1
' exci t at i on
N - RPM
FIGURE 2
Uo changi ng
0~
Input Torque
(Windmill)
Torque
N - RPM o f wi ndmi 1 1
FIGURE 3
R PM
FIGURE 4
input (windmill)
R PM
FIGURE 5
The t ask r emai ni ng i s t o der i ve a r epr esent at i on o f t he pr opel l er
t orque/ speed c h a r a c t e r i s t i c f o r var i ous wind speeds (v,) and p i t c h al l ql es (I:,)
( Fi gur e 3) . The c h a r a c t e r i s t i c s can t hen be superposed, e.Lj. l s h a f t
T and i n t e r s e c t i o n poi nt s w i l l det er mi ne t h e par amet r i c oper at i ng
gener at or '
poi nt s o f t he system. [ Aot e t h a t t he s h a f t t orque, T ~ ~ ~ ~ ~ , must i ncl ude a
penal t y due t o mechani cal l osses i n t he st ep up gear i ng and bear i ngs,
t y p i c a l l y 25%.]
The system performance comput at i on w i l l e s t a b l i s h out put power ( or
energy) f or var i ous poi nt s al ong t he i n t e r s e c t i o n l ocus o f t hese t or que
c har ac t er i s t i c s . Thi s w i l l y i e l d t he combi nat i on of var i abl es whi ch gi ves
hi ghest out put ; e.g. whi ch combi nat i on o f RL, Pexc,
2
and fro gi ves (I,RL)
maximum. These ( quasi s t a t i c ) s ol ut i ons ar e f o r equi l i br i um condi t i ons and
f o r const ant wi nd speed. I t i s underst ood t h a t dynamic f act or s such as
gust ampl i t ude and frequency may d i c t a t e ot her oper at i ng condi t i ons t han
t he " hi ghest out put " det ermi ned by t h i s s i l ~ ~ p l e model .
I I. St a b i l i t y Consi der at i ons
The superimposed curves w i l l g i v e a q u a l i t a t i v e i n d i c a t i o n of t he
f i r s t or der " t i ght ness" o r s e n s i t i v i t y of t he system t o smal l per t ur bat i ons
f rom equi l i br i um. As an example consi der t he p l o t i n Fi gur e 4.
For t h i s example assume t he gener at or t or que c h a r a c t e r i s t i c i s f l a t
( i n s e n s i t i v e t o RW) . If t he oper at i ng equi 1 i bri um p o i n t ( i n t e r s e c t i o n wi t h
gener at or c h a r a c t e r i s t i c ) occur s at @, i t can be seen t h a t out put w i 11 be
i n s e n s i t i v e t o a wi de v a r i a t i o n i n RPM. That i s , t he f i r s t or der system
i s n e u t r a l l y st abl e, and w i l l e s t a b l i s h equi l i br i um f o r any RPM s e t t i n g
i n t he f l a t range. A ~ Q as wi ndmi l l RPM decreases (due t o over l oadi ng o r
gust s) t he gener at or c h a r a c t e r i s t i c must be changed t o keep f rom s t a l l i n g
t he system; l i kewi se, an i ncr ease i n i nput RPM w i l l cause t h e system t o
di ver ge. Thi s c o n d i t i o n i s s t a t i c a l l y unst abl e. I n r egi on@ t he system
i s s t a t i c a l l y st abl e, si nce a p e r t u r b a t i o n i n wi ndmi l l RPM w i l l cause an
opposi t e change i n excess t or que. The sl ope of t he c h a r a c t e r i s t i c i s
anal ogous t o spr i ng r a t e , and i f ot her dynamic ( hi gher or der ) i nf l uences
a r e i mpor t ant , t h e sl ope w i l l be anal ogous t o t he nat ur al f requency o f
v i b r a t i o n i n t h a t mode.
Now reexami ne t he assumpt i on of f l a t gener at or t or que; consi der Fi gur e
5. We have est abl i shed t h a t c o n d i t i o n a i s st abl e; i f i n p u t RPM decreases
t he t or que i ncr ement AT i s p o s i t i v e , and r eaccel er at es t he system t o
equi l i br i um. I t i s evi dent t h a t t h i s t or que i ncr ement a l s o depends on
t he sl ope of t he gener at or c h a r a c t e r i s t i c : v i z . , f o r l i n e a t h e AT i s
l a r g e r and t h e system t h a t much " t i g h t e r " ; and f o r l i n e @t h e AT has
decreased t o a smal l r e s t o r a t i v e quant i t y. Thus, f o r l i n e 4, t he system
i s s t a t i c a l l y unst abl e; i . e . t h e sl ope o f t h e out put c h a r a c t e r i s t i c i s
l a r g e r t han t he sl ope of t he i n p u t c h a r a c t e r i s t i c .
111. Approach t o Pr opel l er Sol ut i on
The approach f o r s ol ut i ons of t he p r o p e l l e r probl em i s t o devel op power-
speed and t orque-speed c h a r a c t e r i s t i c s usi ng nondi mensi onal q u a n t i t i e s t or que
and power c oef f i c i ent s and t i p speed r a t i o ; t h i s el i mi nat es t he dependence
on Vo o r on RPM. The nondi mensi onal p l o t can t hen be used t o devel op any
desi r ed c r os s - pl ot a t any desi r ed Vo t o be used i n mat chi ng t he system.
One o f t he e x i s t i n g comput at i onal r e s u l t s i s a p l o t of t h e power
c o e f f i c i e n t maxima, f o r var i ous t i p speed r a t i o s , f o r t he machine, see Fi gur e
6. Thi s cur ve r epr esent s t h e l ocus of a l l maxima of i n d i v i d u a l C cur ves
P
f o r var i ous val ues o f ro, as shown.
For a f i x e d p i t c h aachi ne, t he
performance c h a r a c t e r i s t i c would be as shown on t he dot t ed l i n e .
For
synchronous ( const ant speed) w i ndmi l l systems, t he power c oaf f i c i e n t v a r i a t i o n
wi t h i ncr easi ng wind speed i s det ermi ned f r om t h i s p l o t ( 2 ) .
The t or que can be der i ved f r om t h i s cur ve usi ng t he f ol l owi ng approach.
The d e f i n i t i o n s ar e:
power - P
- --
Cp = c o e f f i c i e n t 1 3
?PA",
p = densi t y o f a i r
@ sea l e v e l
A = pr opel l er di s c area
Vo = wi nd speed
t or que .r
' T = c o e f f i c i e n t = 1
2
p AVo R
wi t h:
t i p speed - szR
- -
= r a t i o
0
n = r o t a t i o n a l speed
R = r adi us o f p r o p e l l e r
A1 so:
Thus t he t or que c o e f f i c i e n t can be found d i r e c t l y f r om t he C si mul at i on
P
r es ul t s . The CT dependence on p i t c h w i l l l i k ewi s e be shown ( Fi gur e 7 ) .
When t h i s r e l a t i o n s h i p i s found, a map of poi nt s ( par amet r i c p l o t )
w i l l be cal cul at ed f o r each wi nd speed o f i nt er es t , and act ual t or que vs. RPM
curves can be generat ed (see Fi gur e 8 ) . There w i l l be r egi ons on t hese
p l o t s whi ch w i l l i ndi c at e aerodynamic non- - l i near phenomena, and w i l l pr obabl y
n o t be i mpor t ant equi l i br i um condi t i ons:
s t a l l - occurs where angl e of at t ack i s hi gher t han t he maximum
a) -
val ue f or mai nt ai ni ng st r eaml i nes f o r t he p a r t i c u l a r a i r f o i l o f
T = t or que
1 ocu s of 111ax inlulr~ va 1 ues
/
/
F I GU R E 7
N RPM
F I GURE 8
i nt er es t ; performance i s shar pl y degraded.
b)
r ever se f l o w - where l o c a l v e l o c i t y over bl ade i s from t he
t r a i l i n g edge f or war d; t hus w i l l pr obabl y n o t occl;r except i n
smal l por t i ons o f t he bl ade and f o r unusual condi t i ons; t h i s
c ondi t i on w i l l not show up except as a r api d decrease i n C
wi t h 11.
P
c )
negat i ve l i f t -- - t he wi ndmi l l act s as a pr opel l er as angl es o f
at t ack, 1 i f t , and t orque, become negat i ve.
I V . Resul t s o f Pr opel l er Si mul at i on
The comput at i onal r e s u l t s o f t he perf ormance ( s t r i p ) t heor y ar e shown
i n Fi gur es 9 and 10. I n Fi gur e 9 can be seen t he const ant p i t c h curves
used t o generat e t he l ocus o f C maxima shown i n Fi gur e 6 . As p i t c h angl e
P
(B,) i s decreased (e. g. angl e of at t ac k o f bl ade el ement s i ncr eased as shown
i n t he i ns et ) , t he r o t o r i s l oaded more and more. As t he angl e changes,
t he t i p speed r a t i o f or hi ghest power i ncr eases t o about 9 ( B eo = -2")
and t hen decreases t o t he desi gn t i p speed r a t i o o f 7. The decrease i n
power wi t h i ncreased p i t c h angl e r epr esent s t he c ont r ol phi l osophy advanced
for ' " r at ed
(26. 1 MPH); t h a t i s , shaf t speed i s kept const ant and C
P
decreases wi t h i ncr easi ng f r ee stream v e l o c i t y . Thi s phi l osophy i s al so
di scussed by utter,") ~o l d i n g , ( ' ) ~ o h r b a c k , ' ~ ) and ~ e u t s c h , ' ~ ) among ot her s.
A s i mi l a r cur ve i n terms of power ( wat t s) vs. RPM, i s shown i 1 1 Fi gur e 11.
These r e s u l t s were obt ai ned from wi nd t unnel t e s t s of a f i xed pi t c h, const ant
chord bl ade (Reference 7 ) . The cor r espondi ng power c o e f f i c i e n t c h a r a c t e r i s t i c
i s i ncl uded i n Fi gur e 9 .
I n Fi gur e 9, as p i t c h angl e i s decreased, t he power c h a r a c t e r i s t i c
becomes more and more sensi t i ve.
The l e f t mar gi n of each cl l r ve i s a s t a l l
boundary; a phenomenon whi ch occurs f i r s t a t t he t i p o f t he bl ade, si nce t he
NEWF desi gn i s cl ose t o optirnum t wi s t and t aper .
The i nduced v e l o c i t y
0 30i) 40U 500 GOO 7 00 800 900 1000 1100
R P I,.I
FTCIIRF ( 1 1 \
(downwash) d i s t r i b u t i o n al ong t he bl ade i s c h a r a c t e r i s t i c o f optimum
desi gns (see Fi gur e 12). Unt wi st ed, and const ant chord, bl ades show a much
l es s s e n s i t i v e s t a l l t h a t s t a r t s a t t he r o o t a t much srlialler p i t c h angl es
(5,6).
t hus, t he power c h a r a c t e r i s t i c o f an of f - desi qn bl ade w i l l , be f l a t t e r and
have smal l er C t han t h e NEWF bl ade (see Fi gur es 9 and 11) . Thi s p o i n t s
P
out t h e need f o r t horough s t a b i l i t y anal ysi s o f t he optimum momentum
exchanger system p h i 1 osophy adopted f o r t he NEWF desi gn.
Al so, from Fi gur e 9, t he power c h a r a c t e r i s t i c i s seen t o become
hi gher or der as s t a l l i s approached; t h a t i s , a t p i t c h angl es l es s t han
0, t he drop i n power wi t h i ncr easi ng t i p speed r a t i o ( o r RPCl f o r const ant
V ) i s much more severe. Thi s i s al s o bel i eved t o be a f unc t i on o f
0
bl ade desi gn, char act er i zed by i ncr eased dr ag and hi ghl y t wi s t ed sect i ons
becoming negat i ve l i f t .
I n Fi gur e l o i s p l o t t e d t he cor r espondi ng t or que c h a r a c t e r i s t i c ,
di scussed e a r l i e r . The non- l i near behavi or a t p i t c h angl es l es s t han 0"
i s al s o seen. But t he s t r i k i n g c h a r a c t e r i s t i c i s t h e l i n e a r i t y o f t he
t or que curves over t h e uns t al l ed r egi on. Thi s r epr esent s system s t a t i c
s t a b i l i t y f o r each const ant p i t c h s e t t i n q ( i n t he 1 i near r ange) .
Near maximum t orque, t he curves ar e ver y peaky, and may c o n s t i t u t e a
cat ast r ophi c s t a l l f o r smal l RPM per t ur bat i ons. Ty pi c al l y , t hese RPM
v a r i a t i o n s w i l l be caused by gust s; an i ncr ease i n wi nd speed moment ar i l y
decr easi ng t i p speed r a t i o , and vi ce- ver sa. However, when 1~ i s decreased f a r
enough t o push t he pr opel l er i n t o s t a l l , t her e i s al s o a bal anci ng e f f e c t
caused by t h e i ncr ease i n t or que a v a i l a b l ~ i n t he f r e e st ream (e. g. due t o
i ncreased wind v e l o c i t y ) . Theref ore, t he dynamic behavi or i s i mpor t ant , and
must be consi dered f o r a r e a l i s t i c s t a b i l i t y and ~ o n t r o l s ol ut i on.
More a n a l y t i c a l and exper i ment al dat a a r e needed t o es t abl i s h thc!
sl opes o f t he t or que cur ves a t t he absci ssa i n t e r s e c t i o n (e. g. r o t o r unl oaded) .
Thi s i s t he c ondi t i on whi ch woul d be reached i n t he event o f a s haf t o r
gener at or f a i l u r e . The wi nd gener at or w i l l speed up t o t he p o i n t a t whi ch
shaf t t or que i n p u t j u s t bal ances f r i c t i o n t or que, and c l a s s i c a l " wi ndmi l l i ng"
i s achi eved. ( 7)
[ Thi s i s not t he same as f eat her i ng, whi ch i s r epr esent ed
by t he o r i g i n , o r zer o RPM.]
V. Rot or Fl ow St at es
The var i ous r o t o r f l o w s t at es can be i d e n t i f i e d on t h i s c h a r a c t e r i s t i c .
( 595)
The Zero S l i p Case i s r epr esent ed by CT = 0 j u s t di scussed; and t h e Pr opel l er
St at e i s bel ow t he absci ssa, where CT i s negat i ve (power goes i n t o t he syst em).
As t he const ant p i t c h pr opel l er i s l oaded more and more, t he out put power (and
t h r u s t ) i ncr eases t o t he p o i n t where ( t or que x RPM) = power i s a maximum
( t or que al one i s not a maximum). Thi s p o i n t occurs, accor di ng t o momentum
t heor y, where t he average i nf l ow i s 1/ 3; and t he i deal C = . 5926. (2) I f
P
l oadi ng i s i ncr eased beyond t h i s , t he i nf l ow i ncr eases on t h e bl ade u n t i l t he
a i r f o i l s t a l l s ; power f a l l s of f and t h r u s t ( not shown) i ncr eases, as v e r i f i e d
by he1 i c o p t e r a u t o r o t a t i v e t es t s . ( 6) Thi s i s c a l l e d t he Tur bul ent Wake St at e,
and i s char act er i zed by t he absence of st r eaml i nes, severe buf f et i ng, and t he
qui ck f a l l - o f f i n devel oped power. The t ask o f t he c ont r ol system w i 11 be t o
pr event ent r ance i n t o t h i s oper at i ng st at e, even f o r t r a n s i e n t condi t i ons. I t
shoul d be underst ood t h a t f or hi gh val ues of fro (e. g. 12O) t he angl es o f
at t ac k a r e smal l t o begi n wi t h, and i t i s u n l i k e l y t h a t a s i g n i f i c a n t p o r t i o n
o f t he ( h i g h l y t wi s t ed) bl ade coul d ever be s t a l l e d . And t h e bl ade may never
r e a l l y ent er Tur bul ent Wake St at e.
Hence a p o s i t i v e out put t or que coul d be
expect ed f or ext r emel y l ow RPM ( o r u ) .
However, our di scussi on w i l l f ocus
on p i t c h angl es c l o s e r t o t he t or que and power peaks, where most r unni nq
i s done.
It i s i n s t r u c t i v e t o examine t he bl ade el ement di aqr ams f o r t hese
f l o w s t at es. From Wi l son and Lissman, Ref erence 5, Fi gur e 13 i s
r epr oduced.
From exami nat i on o f Fi gur e '1 3, t he "aver age" bl ade el ement s i t u a t i o n
can be descr i bed f o r r egi ons o f t h e power c h a r a c t e r i s t i c , Fi gur e 9 , as
di scussed. I n t h e Tur bul ent Wake St at e, power decr eases r a p i d l y as most
o f t he bl ade i s s t a l l e d , and t h r u s t i ncr eases r a p i d l y . I f t h e bl ade i s al l owed
t o r each s t a l l e d equi l i br i um, t he r e s t i n g p o i n t i s agai n t he zer o t or que
absci ssa. That i s , a f i x e d p i t c h wi ndmi l l can be d r i v e n i n t o t h e s t a l l e d ,
o r Tur bul ent Wake St at e, and t h i s w i l l cause t h e pr opel l er - gener at or
syst em t o become s t a t i c a l l y unst abl e; a s l i ght - i nc r eas e i n gener at or t or que
( o r l oad) wi 11 cause a r a p i d decr ease i n p r o p e l l e r 9PM and t h e t or que w i l l
dr op t o zer o. Most gener at or s have " r es i dual " t or que when unl oaded a t l ow
RPM. The const ant f i el di naper manent magnet gener at or , and t he r e s i d u a l
magneti sm i n separ at el y ex c i t ed f i e l d machi nes, w i l l be a l a r g e r es i dual
t or que. The s t a l l e d equi l i br i um, t h e r , wi l l be t he p o i n t a t whi ch t h e
Tur bul ent Wake St a t e shaf t t or que ( s l i g h t l y p o s i t i v e ) wi 11 j u s t be enough
t o bal ance t h e gener at or " r es i dual " t or que. 1 f ' t h e r es i dual magneti sm o r
permanent f i e l d i s t o o hi gh even f o r v er y l ow RPM's, t he wi ndmi l l can
a c t u a l l y come t o r e s t . I n pr ac t i c e, a machi ne such as t h i s i s sel dom seen
because i t woul d n o t be s e l f - s t a r t i n g ( wi t hout p i t c h c o n t r o l ) . Thi s c ondi t i on,
however, does e x i s t f o r c e r t a i n hi gh e f f i c i e n c y v e r t i c a l a x i s wi nd gener at or s. ( 10)
I t i s n o t expect ed t h a t t he Tur bul ent Wake St a t e oper at i on w i l l be us ef ul f o r wi nd
gener at or speed c o n t r o l s i nce t he bl ade l oadi ngs arld t h r u s t i ncr ease wi t h
r a p i d l y decr easi ng power.
Al so, t h e f l ow c ondi t i ons a r e unst abl e; hi gh
Bl ade E l e ~ ~ ~ e n t s f o r Var i ous i t ot or St at es ( f r o n ~ Ref. 5 )
Vo = f r e e st ream v e l o c i t y
v = i nduced v e l o c i t y ( a x i a l )
n = l o c a l angl e o f at t ac k
13, = p i t c h angl e a t bl ade el ement
( i ncl udes l o c a l t wi s t angl e)
l i f t
Vo
Prope 1 1 e r St at e
(Cp m"dC ar e negat i ve)
Zero S l i p Case
--
(C = C = 0)
P T
a = O
Windnii .- 11 St at e
(Cp and CT ar e p o s i t i v e )
Tur bul ent Wake St at e
- . -
(Cp i s negat i ve; CT i s p o s i t i v e )
a 7/2
v i b r a t i o n and b u f f e t t i n g can occur , as wi t h h e l i c o p t e r r o t o r s .
I t i s i n t e r e s t i n g t o a l s o consi der t he l ow RPM behavi or o f hi gh
s o l i d i t y pr opel l er s , such as t he Ameri can f an n i i l l . A s RPM i s decreased,
t he cor r espondi ng change i n downwash and angl e of a t t a c k i s much smal l er
f o r t hese machi nes; hence, l a r g e p o s i t i v e t or que i s s t i 11 produced a t
v er y l ow RPM, and t he machi nes can 1 i t e r a l l y never be s t a l l e d by
convent i onal l oads. Al so, i t p o i n t s o u t t he non- necessi t y o f v a r i a b l e p i t c h on
t hese machi nes ( s t a t i c t or que i s al r eady hi gh) . However, t hese machi nes
a r e gener al l y n o t i mpor t ant f or e l e c t r i c i t y gener at i on si nce t h e o v e r a l l
power c oef f i c i ent s a r e smal l , and optimum t i p speed r a t i o s ar e c l os e t o
u n i t y . ( 2)
A r ough i dea of t h e Tur bul ent Wake St at e boundary ( f o r l ow s o l i d i t y ,
hi gh speed wi nd gener at or p r o p e l l e r s ) can be gai ned f r om si mpl e t heor y.
From bl ade el ement t heor y ( negl ect i ng s l i p st r eam r o t a t i o n ) t he t i p
s e c t i o n i s :
where
4, = bl ade el ement angl e
The Tur bul ent Wake St a t e boundary i s t he p o i n t a t whi ch st ream1 i nes ar e no
l onger wel l - def i ned; t h i s occur s a t val ues o f a = v/ Vo = 1 / 2 . ( ~ ' ~ ' ~ ) Hence:
V ( 1- a)
-1 0
- t a n
I -
-1 1-a
1 = t an [-I = t a n
-1 1
40 n R u
[GI
Tur bul ent
Wake St at e
Boundary
Thus, t he bl ade el ement angl e ( o r i n f l o w angl e) a t t he t i p f o r Tur bul ent
Wake St at e t o occur depends on t i p speed r a t i o :
Ti p Speed Rat i o
Turbul ent Wake Boundary
9 R/ Vo
$o a t bl ade t i p
TABLE 1
Most a i r f o i l s approach s t a l l a t angl es of a t t a c k about 12". Thus, the
approximate pi t ch angl e t o produce t ur bul ent wake s t a t e i s simply:
BO I = m o I - a s t a l l
Turbul ent TWS
Wake St a t e
where + I is gi ven i n Tabl e 1 above, and as t al l 1 12".
-rws
Ref er ences
1. Hut t er , " Oper at i ng Exper i ence Obt ai ned wi t h a 100-kW Wind Power
Pl ant , " NASA TTF-15, 063, August 1973.
2. Gol di ng, The Gener at i on - - - . - o f E l e c t r i c i t y -. by Wind Power, -.. Phi l os ophi c al
Li br ar y , 1955.
3. Rohrback and Worobel , "Per f or mance Char ac t er i s t i c s o f Aerodynami cal l y
Optimum Tur bi nes f o r Wind Energy Gener at or s, " 31s t Annual Nat i onal
Forum, Ameri can Hel i c opt er s oc i et y , May 1975.
4. Deut sch, L. , Grunlman Aerospace Cor por at i on, per sonal comnuni c at i on,
19 March 1976.
5. W i l s on and Li ssnl an, "Appl i e d Aerodynami cs of Wind Power Machi nes ,"
Oregon St at e Un i v e r s i t y , J u l y 1974.
6. Gessow and Myers, Aerodynami cs .~ -- o f t he Hel i c opt er , -. Fr eder i c k Ungar
Publ i sher s, 1952.
7. St oddar d, Edds, " Fi nal Repor t on Wind Tunnel Test Program o f Model
Bl ades on a 200 wat t , 12 v o l t Wind Generat or, " Un i v e r s i t y o f Massachuset t s,
C i v i l Engi neer i ng Depart ment In-House Repor t , May 1974.
8. St oddard, " Di scussi on of Monientum Theor y f o r Wi ndmi l l s, " ( t h i s r e p o r t ) .
9. Edds, M., " Opt i mi z at i on of Out put Power o f an AC Synchronous Machi ne
by Var yi ng Ex c i t a t i o n and Load," M. S. Ocean Engi neer i ng Thesi s, t o be
publ i shed, June 1976.
10. Bl ackwel l , "The Ve r t i c a l Ax i s Wind Tur bi ne; How i t Works," Sandi a
Labor at or i es, Al buquerque, New Mexi co, SLA-74-0160, A p r i l 1974.
UN-TTY OF MASSACi"3SElTWAhllHERST
ENERGY ALTERNATIVES PROGRAM
h \
A ,
UNIVERSITY OF MASSACHUSETTS/AMHERST
ENERGY ALTERNATIVES PROGRAM
Pr el i ~ n i nar y Report on
OPTIMIZING THE WINDMILL ROTOR
Paul Lef ebvr e
and
Duane E. Cromack
U.Mass Wind Furnace
Energy A1 t e r n a t i ves Program
Uni v er s i t y o f Massachuset t s
Amherst, Massachuset t s 01002
TR/76/4
Appendi x V I
OPTIMIZING THE WINDMILL ROTOR
a b s t r a c t
Twel ve h o r i z o n t a l a x i s wi nd r o t o r syst ems a r e anal yzed by means o f comput er
s i mul at i on. The pur pose o f t h i s a n a l y s i s i s t o devel op a method o f des i gni ng
opt i mi z ed bl ades o f d i f f e r e n t r o t o r c o n f i g u r a t i o n s . The r e s u l t s o f t h e si mul a-
t i o n a r e t hen compared w i t h wi nd t unnel t e s t r e s u l t s .
i n t r o d u c t i o n
One o f t he p r i n c i p a l components of any wi n d mi l l i s t he momentum exchange
devi ce, o r r o t o r . Th i s dev i c e c onv er t s t he k i n e t i c ener gy o f a movi ng a i r
st r eam t o a more usabl e f or m o f power. Ther e a r e two bas i c c l a s s i f i c a t i o n s
f o r wi ndmi l l s : v e r t i c a l a x i s machi nes and h o r i z o n t a l a x i s machi nes. Thi s
r e p o r t deal s o n l y w i t h h o r i z o n t a l a x i s r o t o r s . Wi t h i n t h i s c at egor y a l l r o t o r s
t ak e t h e f or m o f an ai r s c r ew w i t h t h e d i f f e r e n c e bet ween desi gns bei ng i n t h e
chor d and t w i s t d i s t r i b u t i o n o f each bl ade, and t h e number o f bl ades err~pl oyed
f o r any gi v en r o t o r . The f unc t i onal r e l a t i o n s h i p o f t hese v a r i a b l e s det er mi nes
how e f f i c i e n t l y a r o t o r per f or n~s . Four d i f f e r e n t r o t o r sys tems a r e s t udi ed
and t h e i r per f or mance c a p a b i l i t i e s opt i mi zed. The r o t o r t ypes consi der ed ar e:
( 1 ) c ons t ant chor d, zer o t w i s t (CCZT); ( 2 ) l i n e a r t aper ed chor d, zer o t w i s t
(LCZT); ( 3) l i n e a r t aper ed chor d, l i n e a r t w i s t (LCLT); and ( 4 ) aer odynami cal l y
opt i mum chor d and t aper (OPT). The opt i mum bl ade i s t aken t o be t h a t a i r
scr ew c o n f i g u r a t i o n whi ch w i l l e x t r a c t t h e hi ghes t per cent age o f t h e power
a v a i l a b l e . I t has been shown t h a t t h e maximum power c o e f f i c i e n t o b t a i n a b l e
i s . 5926, and i n p r a c t i c e C w i l l i n v a r i a b l y be l e s s t han t h i s val ue. To
P
det er mi ne t he most advantageous chor d and t wi s t d i s t r i b u t i o n , i t i s necessary
2 -
t o combine momentum and annul us t heor y. The f o l l o wi n g e q u a t i ms ar e obt ai ned
and t he optimum bl ade can be desi gned accor di ngl y.
X = s i n +(2cos + - 1) / ( ( 1+2cos+) ( l - cos + ) )
BcCL ~ / 2 n Vo=4 s i n + (2cos + - 1) / ( 1+2 cos
$)
A si mpl e computer program was wr i t t e n t o f a c i l i t a t e t h i s desi gn s p e c i f i c a t i o n
process and i s l i s t e d i n Appendi x A. The program i s wr i t t e n i n For t r an f o r use
on t he Kronos t i me shar i ng system i n use a t t he Un i v e r s i t y o f Massachuset t s,
and shoul d be r e a d i l y adapt abl e t o any ot her system. The i n p u t i s l i mi t e d t o
f i v e l i n e s cont ai ni ng t he r o t o r r adi us, t he number o f bl ades, t he t i p speed
r a t i o and t he CL and a t h a t correspond t o CL/ CD maximum f o r t he a i r f o i l dat a
bei ng used. The f or mat i n each case i s F7.3.
The perf ormance c h a r a c t e r i s t i c s o f each o f t hese f o u r ai r scr ew conf i gur a-
t i o n s a r e det ermi ned f o r two, t hr ee, and f our bl aded systems. The power
c o e f f i c i e n t s c i t e d were obt ai ned usi ng a NACA 4415 st andar d roughness a i r f o i l
p r o f i l e and a desi gn t i p speed r a t i o of 7. Each val ue r epr esent s t h e maximum
power obt ai nabl e f o r t h a t gi ven r o t o r t ype under t hese condi t i ons.
Computer Si mul at i on: The computer model used t o p r e d i c t performance c a p a b i l i t i e s
i s one devel oped by Wi l son and Li ssamanY3 and modi f i ed f o r our purposes. Thi s
model uses bl ade el ement t heor y t o c a l c u l a t e t he oper at i ng c h a r a c r e r i s t i c s a t
each r a d i a l s t a t i o n . These c h a r a c t e r i s t i c s ar e t hen numer i cal l y i nt egr at ed
al ong t he bl ade t o oobt ai n t he c al c ul at ed r o t o r perf ormance. Ref er i ng t o
Fi gur e 1, t he f o l l o wi n g equat i ons may be obt ai ned:
LY. = @ - I)
t a n $= ( 1 - a) ( Vo / ( ( l + a l ) ; ! r )
Cy=CL cos@ +CD s i n4
cX = cL s i n+ -cD C O S ~ ( 6
From aer odynami c s t r i p t heor y, t h e t h r u s t and t or que on a d i f f e r e n t i a l bl ade
el ement a r e gi v en as
dT, = .5Bc p v r 2 Cydr
From momentum t heor y, t he t h r u s t and t or que become
dTm = ( 2 n r d r ) , ~ ~ ( V o - u ~ )
( 9 )
Def i ni ng t he downst ream a x i a l i n t e r f e r e n c e f ac t or , aw, as t wi c e t he a x i a l i n t e r -
f er ence f a c t o r a t t he r o t o r , equat i ons ( 7) and ( 9 ) may t hen be equat ed t o y i e l d
a r e l a t i o n between a and o t h e r known r o t o r par amet er s such t h a t
a = ( ~ c ~ ~ / ~ n r ) / ( s i n ~ ; + (BcCy/8. r ) )
( 1 1)
I n a s i mi l a r manner, i f t he angul ar v e l o c i t y downst ream i s assumed t o be t wi c e
t h a t f ound a t t he r o t o r , equat i ons (8) and ( 10) may be equat ed t o o b t a i n t he
angul ar i n t e r f e r e n c e f a c t o r i n t er ms o f known v a r i a b l e s . Thus:
a ' = ( Bc CX/ 4nr ) / ( s i n2j - ( BcCX/ 4nr ) )
( 12)
Enough i n f o r ma t i o n i s now a v a i l a b l e so t h a t t he f l ow c h a r a c t e r i s t i c s f o r
any gi v en bl ade el ement can be found. I f i n i t i a l val ues f o r a and a ' a r e
assumed ( a=. 05 and al =O) equat i on ( 4) can be used t o c a l c u l a t e 4 . Once @ i s
known equat i on ( 3 ) i s used t o f i n d a. For a gi ven val ue o f a , CL and CD can
be obt ai ned f r om pol ynomi al c u r v e f i t s o f t he aerodynami c dat a. Knowing t he
l i f t and dr ag c o e f f i c i e n t s , and $, equat i ons ( 5) and ( 6) a r e t hen used t o
f i n d Cy and Cx. Thi s al l ows f o r t he c a l c u l a t i o n o f a new val ue o f a and a '
whi ch ar e t hen used t o r epeat t he e n t i r e process. The i t e r a t i o n cont i nues
u n t i l convergence on a and a ' occur s.
I n p r a c t i c e i t was f ound t h a t a t l ow l o c a l speed r a t i o s t h e i t e r a t i o n
l oop woul d sometimes f a i l t o converge. Thi s i s due t o t he non- l i near
r e l a t i o n s h i p between a and a' , and t he l o c a l speed r a t i o ( Fi gur e 2) . Because
o f t h i s convergence probl em, i t was necessary t o f u r t h e r modi f y t he o r i g i n a l
program so t h a t each successi ve i t e r a t i o n used t h e mean val ue o f a and a '
f ound i n t he t wo pr evi ous passes. A c o r r e c t i o n f a c t o r was a l s o i nt r oduced i n
whi ch a t l o c a l speed r a t i o s o f .75 o r bel ow, a ' was mu l t i p l i e d by one quar t er
o f t he l o c a l speed r a t i o , and t h i s new val ue of a' used f o r t he nex t pass. Thi s
f a c i l i t a t e s t he convergence procedure.
*The c ons t r ai nt s used i n t he corr~put er s i mul at i on a r e as f ol l ows :
( 1)
Hub r adi us o f 10 per cent t he bl ade r adi us,
( 2) A st andar d a x i a l i nt er f er enc e model,
( 3) Al t i t u d e above sea l e v e l o f 200 f eet ,
( 4 ) No coni ng angl e used,
( 5)
T i p l osses modeled by Pr andt l ' s method,
( 6) No nub l os s model used,
( 7 ) A NACA 4415 St andard roughness a i r f o i l p r o f i l e was used,
( 8)
A t i p speed r a t i o o f 7 was used.
Computer Si nl ul at i on - Resul t s: - . - - - The power c o e f f i c i e n t s obtctint.d wi t h opt ~IIIUIII
r o t o r s o f two, t hr ee and f our bl ades ar e l i s t e d i n Tabl e 1. These r o t o r s
were desi gned f o r a t i p speed r a t i o o f 7 , a CL o f -914, a t an angl e o f at t ac k
o f 5. 57 degrees. The l a s t two f i gur es r epr esent t he p o i n t a t whi ch CL/ CD i s
maximum f o r an NACA 4415 st andar d roughness p r o f i l e .
The nex t two bl ade t ypes t o be consi der ed ar e l i n e a r chor d zer o t w i s t
and l i n e a r chord l i n e a r t wi s t . I n bot h cases t he f i n a l desi gn c o n f i g u r a t i o n was
f ound by maki ng a s er i es o f corr~put er r uns i n whi ch chor d and t wi s t were
syst emat i cal l y changed i n a s er i es o f appr oxi mat i ons, as were f i and X, u n t i l
0
a maximum power out put a t a t i p speed r a t i o o f 7 was reached. The chor d and
t wi s t appr oxi mat i ons t h a t were t r i e d ar e shown i n Fi gur e 3. I n or der t o o b t a i n
a consi st ent method o f desi gni ng bl ades o f t h i s t ype, t h e i r chor d and t wi s t d i s -
t r i b u t i o n s ar e l i s t e d i n Tabl e 2 i n t erms o f t he optinlum bl ade o f s i mi l a r desi gn
c ons t r ai nt s .
For example, i f an LCZT 3-bl aded r o t o r wi t h a r adi us o f 10 f e e t and a t i p
speed r a t i o o f 7 i s desi r ed, t he bl ade c o n f i g u r a t i o n i s f ound as f ol l ows . The
optimum bl ade f or t hese condi t i ons i s l a i d out . The chor d o f t he LCZT bl ade a t
100 per cent o f t he r adi us i s t aken as 90 per cent o f t he chor d o f t he optimum
bl ade a t t h a t poi nt . The chor d o f t he LCZT bl ade a t 10 per cent o f t he r adi us
i s t aken as 67 per cent o f t he chord o f t he optimum bl ade a t t h a t poi nt . As
t he chor d d i s t r i b u t i o n i s l i n e a r , t he di mensi ons of t he bl ade ar e now f i xed.
A l l t h a t remai ns i s t o s e t t he , - o t o t h a t s p e c i f i e d i n Tabl e 2.
The power c o e f f i c i e n t s obt ai ned f o r LCZT and LCLT r o t o r s o f two, t hr ee,
and f o u r bl ades a r e shown i n Tabl e 1.
The f i n a l bl ade t ype t o be consi der ed i s const ant chord zer o t wi s t . For
r o t o r s o f t h i s t ype t her e ar e t hr ee v ar i abl es t h a t w i l l a f f e c t perf ormance:
t h e r o t o r s o l i d i t y , t he fro s et t i ng, and t he number o f bl ades. The r e l a t i o n
between r o t o r s o l i d i t y and K~ f o r maximum power, and t he r es ul ~ i n g t i p speed
r a t i o , i s shown i n Fi gur e 4. These cur ves were generat ed by maki ng a s er i es o f
comput er r uns f o r a gi ven s o l i d i t y f o r whi ch Bo and X were syst emat i cal l y
changed u n t i l a maxirnun~ power was obt ai ned. The process was t hen r epeat ed f o r
a nurnber o f r o t o r s o f d i f f e r e n t s o l i d i t i e s .
These r e s u l t s were checked t o i ns ur e t h a t t hey appl y f o r a gi ven r o t o r
s o l i d i t y r egar dl ess o f whet her i t cont ai ns 2, 3, o r 4 bl ades. The anal y s i s
was t hen r epeat ed usi ng NACA 4418 a i r f o i l dat a i n or der t o check f o r consi st ancy
between s i mi l a r p r o f i l e s .
The power c o e f f i c i e n t s t h a t can be expect ed f r om const ant chor d zer o t wi s t
r o t o r s wi t h two, t hr ee, and f o u r bl ades ar e al s o shown i n Tabl e 1.
W i nd Tunnel Resul t s : I n an e f f o r t t o exper i ment al l y v e r i f y t he r e s u l t s o f
t h e computer si mul at i on, a wi nd t unnel t e s t program was est abl i shed. Si x model
bl ades were const r uct ed: two optimum bl ades and f o u r const ant chor d zer o
t w i s t bl ades. Thi s woul d a l l o w f o u r r o t o r systems t o be t est ed and a compari -
son t o be made between sever al bl ade t ypes, s o l i d i t i e s , and numbers o f bl ades.
The t e s t arrangement i s shown i n Fi gur e 5. The e l e c t r i c gener at or was l oaded
by means o f f o u r v a r i a b l e r es i s t anc e l oad banks. For any gi ven t e s t , t he l oad
appl i ed t o t he gener at or was syst emat i cal l y v ar i ed u n t i l t he combi nat i on o f f i e l d
e x c i t a t i o n and r esi st ance were such as t o produce maximum power. Thi s power
was det er mi ned by measuri ng t he vol t age and amperage t o t he 1 oad (P=VI ) and
t hen addi ng t o t h i s t he exper i ment al l y det er mi ned e f f i c i e n c y l osses o f t he
gener at or . ( see Fi gur e 6)
For each r o t o r t ype, a s er i es o f t e s t s were f i r s t conduct ed t o as c er t ai n
t h e po s e t t i n g f o r maximum power. That s e t t i n g was t hen used f o r a s er i es o f
r uns a t i ncr easi ng wi nd speeds. The i ncr ease i n maximum power wi t h wi nd speed
p a r a l l e l e d c l o s e l y t he expect ed cubi c r e l a t i o n s h i p and i s shown i n Fi gur e 6.
Al so shown i s t he r e l a t i o n s h i p between power and RPM as a f u n c t i o n o f appl i ed
l oad. Tabl e 3 cont ai ns t he power c o e f f i c i e n t s o f t hr ee r o t o r systems f o r
compari son wi t h t h e power c oef f i c i ent s pr edi ct ed by t he computer si mul at i on.
concl usi on
Exper i ment al v e r i f i c a t i o n o f t he computer s i mul at i on has n o t been compl et ed.
The di f f er ences i n power c oef f i c i ent s l i s t e d i n Tabl e 3 a r e q u i t e l a r g e wi t h
no consi st ent p a t t e r n di scer nabl e. Addi t i onal wi nd t unnel t e s t s ar e pl anned
i n or der t o o b t a i n a b e t t e r c o r r e l a t i o n between t h e o r e t i c a l and exper i ment al
r e s u l t s . Si nce t he computer model i s based on wel l - est abl i shed aerodynami c
t heor y, t h e r e s u l t s f o r comparat i ve purposes a r e c e r t a i n l y v a l i d .
The s e l e c t i o n o f t he r o t o r t ype, i .e. , optimum, LCLT, CCZT, et c. , depends
on t he p a r t i c u l a r a p p l i c a t i o n b u t more i mpor t ant , depends on t he method o f
bl ade manuf act ure and t he mat er i al t o be used. Thi s i s a probl em of c o s t
ef f ect i veness and not o f j u s t r o t o r perf ormance.
The probl em o f opt i mi z i ng may be vi ewed d i f f e r e n t l y by st andar di zi ng t he
power o f each r o t o r syst em i n t erms of t he r equi r ed r o t o r di amet er . Thi s i s
done i n t he f o l l o wi n g manner. For any proposed wi nd gener at i ng system t her e
w i l l be a r equi r ed power f o r a p a r t i c u l a r desi gn wi nd speed. For t h e desi gn
power and r a t e d wi nd speed, t he r equi r ed r o t o r di amet er can be c al c ul at ed
f r om t h e st andar d power equat i on as D = [8P/nC The r e s u l t i s t h a t
P
a t wo- bl aded CCZT r o t o r wi t h a di amet er 5. 27 per cent gr eat er t han a t en- f oot
,
di amet er t wo bl aded optimum r o t o r w i l l y i e l d equal power. Ther ef or e, t h e
choi ce t o be made i s between a t en- f oot optimum bl ade and a 10.53 f o o t CCZT
bl ade. The dec i s i on as t o whi ch bl ade t o use i s based on whi ch bl ade t ype
woul d have t h e l owest u n i t pr oduct i on c o s t f o r t h e number o f bl ades r equi r ed.
Tabl e 4 i ndi c at es t h e per cent i ncr ease i n bl ade di amet er needed f o r equal
power out put , st andar di zed t o t h e optimum bl ade shape.
It shoul d be st r essed t h a t t h e r e s u l t s c i t e d appl y r e a l l y o n l y t o NACA 4415
a i r f o i l p r o f i l e s , al t hough t he r e s u l t s obt ai ned usi ng a NACA 4418 p r o f i l e were
q u i t e s i mi l a r . Ot her bl ade sec t i ons need t o be anal yzed bef or e t h i s pr ocedur e
can be gener al i zed.
I t shoul d a l s o be kept i n mi nd t h a t t h e power c o e f f i c i e n t s f ound ar e f o r
p e r f e c t l y c onst r uc t ed bl ades. If a l i mi t e d number o f opti mum bl ades were t o
be produced, i t woul d be d i f f i c u l t t o o b t a i n t h e exact chor d and t w i s t d i s t r i -
b u t i o n needed. On t h e ot her hand, b e t t e r q u a l i t y c o n t r o l mi ght be expect ed f o r
CCZT bl ades because o f t he si mpl er c ons t r uc t i on.
Thi s st udy at t empt ed t o opt i mi z e t h e chor d and t w i s t d i s t r i b u t i o n o f f our
r o t o r t ypes. I t has been suggest ed t h a t f u r t h e r i ncr eases i n t he power c o e f f i -
c i e n t s mi ght be obt ai ned by such methods as v ar y i ng t h e a i r f o i l p r o f i l e al ong
t h e bl ade, o r by desi gni ng t h e bl ade i n such a way t h a t t h e CL/ CD r a t i o was a
f u n c t i o n o f t he r a d i a l s t a t i o n i ns t ead of hol di ng i t a t a const ant CL/ CD maximum.
TAGLE 1
POWER COEFFICIENTS OF TWELVE SIMULATED ROTORS
( L i s t e d i n or der o f magni t ude)
Type o f Rot or No. o f Bl ades Bo Power Co e f f i c i e n t
Optimum
Optimum
Optimum
LCLT
LCLT
LCLT
LCZT
L CZT
LCZT
CCZT
CCZT
CCZT
CCZT
TABLE 4
INCREASE OF BLADE DIAMETER NEEDED TO OFFSET LOSS OF
EFFICIENCY DUE TO USE OF A NON-OPTIMUM BLADE
% I ncr ease i n
Na Bl ades Rot or Type Power Co e f f i c i e n t Di amet er
@o
- - - -
2 Optimum 0.0 .439 0.00
2 LCLT 2.0 .428 1.28
2 LCZT
2 CCZT
3 Optimum 0.0 .463
3 LCLT . 1.5 .449
3 LCZT 4.5 . 423 4.62
3 CCZT 6.1 .415 5.62
4 Optimum 0. 0 .474 0.00
4 LCLT 1.5 .460 1.51
4 LCZT 4.5 - 430 4.99
4 CCZT 6.1 .425 5.61
FI GURE 2
0 I 2 3 4
ar5 6 7
L O C A L S P E E D R A T I O -
' V o
AX I AL AND ANGU L A R I NTERFERENCE
-
-
-
F ACT OR S vs LOCAL SPEED RATI O
a
a
I I I I
7 2
FIGURE 3
APPROXIMATIONS USED FOR THE COMPUTER
SIMULATION OF THE LCZT AND LCLT BLADES
LINEAR TWIST
APPROXIMATIONS
CHORD TWI ST
FI GURE 4
SOLI DI TY vs PI TCH ANGLE AND T I P
SPEED RATI O FOR CONSTANT CHORD
ZERO TWI ST BLADES
- - /' NACA 4415
P o
STD. ROUGHNESS
-
NACA 4418
V I I I I
.O 3 -0 6 .O 9 . I 2
SOLI DI TY , 0-
FI GURE 6
POWER vs R.F!M. FOR
2- BLADED OPTIMUM
ROT0 R
300 4 00 500 600 700 800 900
ROTOR R.P.M.
Nomencl a t ur e
Axi al i nt er f er ence f ac t or
Angul ar i nt er f er ence f ac t or
Wake ax i al i nt er f er ence f ac t or
Number o f bl ades i n a gi ven r o t o r
Local chord f o r a gi ven bl ade
Drag c o e f f i c i e n t
L i f t c o e f f i c i e n t
Power c o e f f i c i e n t
Coef f i c i ent o f f or ce i n t he di r ec t i on o f r o t a t i o n
Coef f i c i ent o f f or ce normal t o t he pl ane o f r o t a t i o n
Radius o f t he r o t o r
Radius o f a gi ven l oc al s t at i on
Axi al f l o r v el oc i t y a t t he r o t o r
Axi al f l ow v el oc i t y i n t he wake
Torque as gi ven by aerodynamic t heor y
Torque as gi ven by momentum t heor y
Thr ust as gi ven by aerodynamic t heor y
Thr ust as gi ven by momentum t heor y
Free stream vel oci t y
The r e l a t i v e v el oc i t y as seen by a moving bl ade element
The t i p speed r a t i o o f t he r o t o r devi ned as nR/Vo
The angl e between t he chord of t he bl ade element and t he r e l a t i v e
v el oc i t y
An equal angul ar def l ect i on, i n addi t i on t o bl ade t wi s t , gi ven t o
each bl ade el ement
n Angul ar v el oc i t y o f t he r o t o r
p
Densi t y o f t he f l u i d medium
Angl e between t he pl ane o f r o t a t i o n and t he r e l a t i v e v el oc i t y
e Angl e between t he pl ane o f r o t a t i o n and t he chord o f t he bl ade
element.
ref erences
1. Gl aver t , H., Aerodynan~ics Theory (I J. F. Durand, Edi t or - i n- Chi ef ) , Vol . I V ,
Di v i s i on L, J ul i us Spri nger, Ber l i n, 1935.
2. i b i d
3. Wi l son and Lissaman, "Appl i ed Aerodynamics o f Wind Power Machines," Oregon
St at e Uni ver si t y, J ul y 1974.
4. Putnam, P. C., Power from t he Wind, Van Nostrand Company, I nc. , New York,
1 948.
APPENDIX P
LNH
0 0 1 0 0 P??OG??AI.1 O?T < IPJP'JT J O' !T?' .!T
0 0 1 0 5 DI ME?I . SI C?I C E ( 1 5 )
0 0 1 1 8 ?EAD l r ?.
0 0 1 2 0 ?EAD l r B
0 0 1 3 8 ?EA9 l r X
0 0 1 4 3 FZAD l r CL
0 0 1 5 8 ?EAD l r A
0 0 1 6 0 1 F O! ? ! AA( F 7 * )
@0 1 7 E ??IN? 2
0 0 1 68 2 FC?MAT<//SXJ*?AD I US*r SYr * ? C ~ R * J ~ Y , * B C C L ~ / ~ ? I ~ * , ~ ~ , * ? H I * J ~ ~ J *EO*rS
0 0 1 9 8 +Xr *CH03D*/ )
0 0 2 0 0 ?1=3 1 4 1 5 9 2 0 5 4
0 0 2 1 0 FiL=O*
e 0 2 2 0 XL=E?*
0 0 2 3 0 ?E?!=a*
0 0 2 4 0 ?=60*
0 0 2 4 1 DA=?I *( ?**2*>
0 0 2 4 5 v=?. / i e .
0 0 2 5 0 DO 3 2 L-1 J 1 0
0 0 2 6 0 YL=YL+( Y/ l Bm>
0 0 2 7 0 P =? * P I / l EB*
E 0 2 8 8 ! ? L=! ? L+( ? / l a * )
0 8 2 9 3 ?E?=?E"+*l
0 0 3 0 0 DO 5 I = 1 ~ 6 0 0 C
0 0 3 1 0 P = ? - < * C l * ? I / 1 8 3 * >
0 0 3 2 0 XE=(SI?I(?>*(2**CCC(?>-l*>)/<(l * + 2 e * C O ~ ( ? > > * ( l * - CGS ( P ) ) )
00330 I F( ASZ( : <L- XE>*LT* * 2 5 > G O TO 6
0 0 3 3 2 5 COM?IP!!!E
0 0 3 3 4 6 DO 7 J - l r 4 D 2 2
0 0 3 3 6 ?=?- <. C! Ol *?I / l E! ?~>
00338 YE=<SI~l<??*(2~*C@S<?>-le > > / ( ( I * + ~ * * C O S ( Q ) ) * ( I * - COS ( ? ) ) )
0 3 3 4 8 IF<ASC. (YL-XE) eLT* e Z C 5 ) G O TO IO
0 6 3 4 2 7 COP!TIN'JE
00358 l e 9CCL=<4**sIN<?>*(2**CC5(?)-1 ) > / ( I * +2 * * CGS ( ? ) )
0 0 3 6 E ?=F*l EO * / ? I
0 8 3 7 8 EO=?-A
0 0 3 8 9 C,!(L)=BCCL/(E*CL*:' /(2 . *?I *?> 1
0 0 3 9 0 PDI?J? 2 8 1 ?. Lr ?E?~3CCLr ?r 93r C: i ( L>
0OGL34? 2 2 FC?.??Ar<SY, i 6. 2J 5?: , F LI . ~ ~ S : - , F E * ~ J S Y , F C * ~ J ~ Y ~ T S * ~ ~ ~ Y ~ F ~ * ~ >
0 0 4 1 P 3 2 C CPIT I P!','S
0 Q 4 2 0 S'.TY=CH(l ) +3 * * CH( 2 ) +3 * * CH( 3 >+2 * * C: E( 4 ) +3 * * Ct i ( 5 >+3 * * CH( 6 >+2 * * CF. ( 7 )
0 0 4 2 1 + +3 * r * C1 ! ( 8 > + 3 * * CH< 3 ) + C1 ? ( 1 0 >
a 3 4 3 0 8A?EA=( ( 3 **E*S!!:4) / S * >*3
0 0 4 4 0 S I GCL=BA?.EA/DA
m 4 5 3 ??I!?T 4 2 1 aA?z A
0 0 4 6 3 4 Q FOE?- ! AT( / / 3Yr *TCTAt '3LSC.E A?EA MINUS HU 3 OF * l ? = * r F 7 * 2 r * S n * F T * * )
0 8 4 7 8 !'?IN- 5 2 r YIGCL
0 G4 8 3 50 FO??IAT<//3?' ,*$ IG:.!A Tc 7 AL=* r F 6 * 4 )
0 0 4 9 8 END
2EP.DY
UNIVERSITY OF MASSACr:;JSETTS/'AMHW
ENERGY ALTERNATNES PROGRAM
UNIVERSITY OF F1ASSACHUSETTS/AMHERST
ENERGY ALTERNATIVES PROGRAM
FIELD COtiTROLLER FOR THE UMASS WI ND FURNACE
by
Dani el Handmann
U.Mass Ui nd Furnace
Energy Al t er nat i v es Program
Un i v e r s i t y o f Massachuset t s
Amherst, Massachuset t s 01002
TR/ 7 6 / 5
Appendi x V I I
FIELD CONTROLLER FOR THE UMASS WIND FURNACE
It has been deterrni ned experi l nental l y t hat III~X~II~UIII power out put o f t he
Lima El ec t r i c Generator can be achi eved i f t he f i e l d cur r ent suppl i ed t o t he
generat or f or a gi ven generat or speed f ol l ows t he curve shown i n Fi gur e 1.
The f unct i on of t he f i e l d c ont r ol l er i s t o suppl y f i e l d cur r ent al ong t h i s If
vs. RPM curve. To accompl i sh t h i s t ask t he f i e l d c ont r ol l er has f i v e basi c
components; ( r ef er t o drawi ng #SH-I, 03.01.01 " Fi el d Cont r ol l er , Bl ock Diagram")
a tachometer, a 6 b i t anal og t o d i g i t a l conver t er , a 64 word by 8 b i t semi-
conductor memory, an 8 b i t d i g i t a l t o anal og conver t er , and a pul se wi dt h
modulated t r ans i s t or swi t chi ng amp1 i f i e r .
The oper at i on o f t he system i s as f ol l ows: The tachometer i s geared t o
t he generat or and has a d.c. vol t age out put t h a t i s l i n e a r wi t h rpm. The
out put of t he tachometer i s convert ed f r om a vol t age l evel t o a 6 b i t bi nar y
number by t he anal og t o d i g i t a l conver t er . The 6 b i t bi nar y number can t ake
on i nt eger val ues between 0 and 63 (base 10) whi ch has t he e f f e c t o f break-
i ng down t he oper at i ng speed range o f t he generat or i n t o i ncrements o f 28.5 rpm.
The out put o f t he anal og t o d i g i t a l conver t er i s used t o address one o f t he 64
words i n memory. The cont ent s o f t he addressed 8 bi nar y b i t word appear a t t he
out put o f t he memory. The d i g i t a l t o anal og conver t er conver t s t he word from
memory i n t o a cur r ent l ev el between 0 and 2.6 m i 11 i-amps. Thi s cur r ent 1 eve1 i s
t he cont r ol si gnal f or t he pul se wi dt h modul ated t r ans i s t or swi t chi ng ampl i f i er .
The swi t chi ng ampl i f i er swi tches a 50 v o l t d.c. power suppl y on and o f f a t a
f i xed frequency (10kHz).
The wi dt h of t he pul se var i es l i n e a r l y wi t h t he cont r ol
si gnal f rom t he d i g i t a l t o anal og convert er. The wi der t he wi dt h o f t he pul ses,
t he hi gher t he average cur r ent suppl i ed t o t he generat or f i e l d .
The swi t chi ng
ampl i f i er al so moni tors t he amount o f cur r ent bei ng del i ver ed t o t he generator
f i e l d by means o f a feedback loop. The feedback si gnal i s sumed i nt o t he
cont r ol si gnal t o provi de cur r ent r egul at i on necessi tated by t he f l uct uat i on
of t he generator f i e l d ' s r esi st ance wi t h changes o f f i e l d temperature.
By bui l di ng t he system around a memory, a cer t ai n amount o f f l e x i b i l i t y
i s i ntroduced i n t o t he system. If, f or example, i t i s determined duri ng t est s
o f t he wind generator t hat a di f f er ent If vs. RPM curve should be used, t he
ol d memory can be unplugged and repl aced by a d i f f e r e n t l y programed one. For
t he f i r s t model o f t he f i e l d cont r ol l er , two 32 word memories were used
i nst ead o f one 64 word memory. Thi s provi des t he opt i on o f changing par t o f
t he If vs. RPM curve a t one-hal f t he cost . The If vs. RPM curve can al so be
shi f t ed hor i zont al l y and v e r t i c a l l y by adj ust i ng t r i mner potenti ometers i n -
cl uded i n t he c i r c u i t . The act ual out put o f t he f i e l d cont r ol l er i s shown
superimposed over t he i deal If vs. RPM curve i n Fi gur e 2.
.-.- ---- .. .
. --- - - - -. - . - - . - . .-.- - . -
I , . .
. -
.
.- - - --- ----
i
I
I
I
i
-- - -. . . .- -
I
!
---.. . ..-- ,. Tg . -- ... . -- . . . . - -
j
-7 -
-L ,.--..,
1
I
I ."
?< or
I
< , r- , y-I
..,
--. .
I-:.; c;, -.-
- . k i j
cr, 0- I..
, y --. 7-.
-
!
-
\
'- - L-C ' . : j --7 I
\ . LJ
0
;_. ,
I
-...
;
9 g- -- !
LLJ clj i - . 4 1
--A ,, -2 :
I
I
.\ / W ) c ; Z ,
g a t I c; .< + -A t E I I i
c i I
T oi lu 4 f f ( c * i 1
I
i
-.- i.
I
j
I
i
'
I
I ...
1
! j
a j
APPENDIX I : Determination of the Memory Bi t St ruct ure
In order t o determine the proper bi t pattern t o be programmed i nt o the
memory t o match the If vs. RPM curve, i t was necessary t o f i r s t determine
the output current of the f i el d cont r ol l er ci r cui t f or every possible value of
a memory word. Since each word contains 8 bi t s of information, each word can
have 256 di f f er ent values. To find the output current t hat would r es ul t f or
each of the 256 possible combinations of bi t s in a word, the following t e s t s
were run. ( r ef er t o drawing kSH-1-03.01.03, "Field Control 1 e r , Memory Simula-
t i on Test ") The memories were simulated with 8 s p d t switches t hat connected
each input of the di gi t al t o analog converter t o ei t her +5 vol t s ( l ogi cal 1)
or ground (l obi cal 0) . Then the switches were s e t f or a l l of the 256 possible
combinations of t he 8 bi t s , and with an ammeter, t he r esul t i ng current delivered
t o the generator f i el d f or each combination was recorded. Cornb-ining the
r es ul t s of the memory simulation t e s t with a l i s t i ng of the rpm range t hat each
memory word would cover, and the ideal If vs. RPM curve, the proper memory
contents t o generate the desired shape f or the I f vs. rpm curve were determined.
A copy of the b i t st r uct ur e i s shown in Table 1. The memories were programmed
using t he Si gnet i cs 8223 Programming Procedure A, a copy of which i s provided.
RPV RPN f:E MLMrJhY PIrhESS 89 MFMGhY CCvPJ'IFN 7 S
------- --- ---- - - - - - - - - - - - - - - - - - . - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
m 5 3 L 5 rl
0- - 29 0 . --.
0 0 0 0 - 3 0 2 . . 1
--- ---- .......................................
- 29-- 5 7 1
fi :-: ,,-, ;> ,
----.--------c---------.---------------.---------------------
57- - 86 2
0 0
---------------.-.---- .......................................
86-- 114 3 13 a G i-.I> i3 (3 ."J ,~I
...
-....-- .-------------.---.----.------------------------------
-
SlGNETlCS DI GI TAL 8000 SERIES TTL/MSI-8223
~- - ~-
8223 PROGRAMMING PROCEDURE
Tl~cs 111;1:3 I ~I , I ~ 1 ~ ' (r1o(ll;1n1rr11~0 I l y 11511111 ( : t ~r t ~< C11'1:trc)
Dr.v~cc:s PC1 73 S~IIC:, 01 S~II!CIIIIIII D~II,I~~II~.' , 300. 400 ~ 1 1
500 So~les P~ograni~~,crs. E,~cl, i ~ c r l o ~ ~ n s [he ~~r o~- t * dur e
outl~ned.
The 8223 Standard part 1s sll~l,p:d w~t l i al l outputs at
log~cal "0". To wrtte a log~cal "I" proceed as follow$.
Programming Procedure A ,
Simple Programming Procedure uslng "bench" Equ~pment
1. Start with pi n 8 grounded and VCC removed from
pi n 16.
2. Remove any load from the outputs.
3. Ground the Chip Enable.
4. Address the desired l ocat~on by apply~ng ground (1.e..
0.4V maximum) for a "0". and +5.OV (1.e. +2.8V
minimum) for a "1" at the address input l~nes
5. Apply +12.5V to the output to be programmed through
a 390 0hm210?~ resistor. Program one output at a tlme.
6. Apply +12.5V to VCC (pin 16) for up to 1 0 second. If
1.0second i s exceeded, the duty cycle should be l ~r n~t ed
t o a maximum of 25%. The VCC overshoot should be
l ~mi t ed t o 1.OV maxlmum. If necessnrv, a clamping
c~rcui t should t ~e used. The VCC currerlt rec;u~rement I S
400 mA maxi r n~~m at +12.5V. Several fuses can be pro-
grammed In sequence unt ~l 1.0 s~ of h~y h VCC time IS
accumulated before Irnposllig tlio duty cycle restrrction
NOTE. Normal pract~ce i n test f ~xt ur e l ~ y o u t should bt
followed. Lead lengths, particularly t o the
power supply, should t ~ e as short as possible. A
capaciror of 10 m~crofarads mlnlmum, connec-
ted from the +12.5V to ground, should be
located close to the unlt I h n g arogrammed.
7. Remove the proyrammingvoitaye from pi n 16.
8. Open the output.
9. P~oceed t o the next outpot and repeat, or change
address and repeat procedure.
10. Cont~nue unt ~l the entire b ~ t pattern 1s programme,.j
i nto your custom 8223.
Fast Programming Procedure - Programm~ny Procedure B
1. Remove VCC (open or ground pln 16).
2. Remove any load from the output.
3. Ground CE (pi n 151.
4. Address the word t o be programmed by applyirrg 5 volts
of a " 1" and ground for a "0" to the address l ~nr s (Solid
TTL l og~c levels are ok, but wtb sugyest buffer dl lvers or
Utilogic ORINOR yares for the atldress~ng).
MANUAL PROGRAMMER DI AGRAM
5. Al,()ly 12 5V to tht. c)lrtr..t~l to 111. rrroqr,~rr,ni~~tl rhrouc~ti
3 1 1 0 " l t \ l t l l l 1 ' 1 0 1 ) 1 . 1 1 1 1 0 1 1 l ' 0 1 1 1 1 l 1 1 1 ,If .I 111111.
6. Al)i)ly 4 17 5V to V(:C ( 1 1 1 r l I ( ; \ I t ~ r 7 5 50111:> LI ~I I I I 1\ 11>
vcc 0vt'l5/10111 1 0 1 0 vt1lts rl1;lY
7. Reduce, VCC to qr otlritl (or ol)c,rii ~ r ~ t l Il*nli)vc3 111t. lodtl
from the output.
8. lmmedialcly repedl stt!l~s 5 ' ,I~C! 6 for other otrtl~urs of
the samt wortl, or rel l t.at 4 t h~ouqt i 6 for a clhfterent
word. Con~~nue programming for a man of 1 second.
Then remove power for 4 seconds and cont~nue unl i l
the entire b ~ t pattern I S programmed
After programming the 8223, the unit should be checked
t o Insure the code is correct. If add~t ~onal fuses must be
opened, they may be programmed dur ~ng verification.
Fast Proyramm~ng Procedure - Programming Procedure C
Steps 1 through 5 are the same as ~n Procedure B
6 Apply a 5mS pulse to VCC (pin 16) Ll m~r the VCC
overshoot
7. Reduce VCC to 5 volts for 10 15uS and verlfy the fuse
opened (output I S now a "1" I f the hrt programmed go
on to the next b ~ i to be programmed I f the b ~ t d ~ d not
program then reuvce VCC to ground (or open1 for 1-5uS
and repea: rtea 6 .,nd 7 unt ~l the fuse proglams ( 1 second
total tlme maxi
8 Cont~nue progiammlng a: thls rate for 1 second Remove
all power from the device for 4 seconds then contlnue
programmlng procedure
VCC Waveform
BOARD LEVEL PROGRAMMING PROCEDURE
FOR THE 8223
The c h ~p select controls which 8223 IS be~ng programrced
when several PROPAS ar e collector OR'd To program In
t h~s manner the 071y changes requlred are
1 The 390 ohm res,stor IS reduced to 200$hm where N
IS the number of outl ~uts Ired together ( 2 < N < 12)
2 Reduce rnax fuse pui s~ w~dt h from 1 second max to
0 92 sec maz
. I.. I, V
d-~ - -.-T--r.--- - 7
7 -- I
P . 3
4
L I , I
I
1 t
L i .-- -.-A ,
&
S, - Sangle pol e 9 posttl on sw11ch
S2 t hr ough S 6 - song!c ool e 2 por~tlon s i v l l c h
5,
Two pole 3 posttmon s wi t c h i v! t h ground connect ed
10 t he r n , ~ : b e cor ?i on c l t he secl ! on c o n n e c t e d t o
VCC pl n 16 10 ~c f r om 5 v o l l s l o 1 2 . 5 V l h e
si v. l i h W, I I T C - i - : a: e~y ground VCCi and pos!ttons
1 a n 2 2 01 t h e - :' -L- s e c t cn connect ed t o 5 OV t o
pr ov~ac t h e nerJec 5 bol t s t o t h e output Tor
~ ~ ~ I ~ I C ~ T I O A
NOTE 1 Tne I S, . ' ~ ~ ? a ; , t o : azrnss omn 16 t o ground 8s
r eaul r ed t c el!mlna:e nctse f r om VCC.
UNNERSlTY OF MASSACi-:;ISm/MHm
ENERGY MTERNAtlVES PROGRAM
h C
4
UNIVERSITY OF MASSACHUSETTS/AMHERST
ENERGY ALTERNATIVES PROGRAM
A GENERAL DESCRIPTION OF THE BLADE-PITCH CONTROLLER
by
Bruce A. Caccamo
U .Mass. W i nd Furnace
Energy Al t er nat i ves Program
Uni v er s i t y o f Massachusetts
Amherst, Massachusetts 01002
TR/76/ 6
Appendix V I I I
A GENERAL DESCRIPTION OF THE BLADE-PITCH CONTROLLER
Abst r act
The bl ade- pi t ch c ont r ol l er provi des f o r maximum energy t r ansf er between
t he wind and t he wi ndmi l l under var yi ng wind condi t i ons. The purpose o f t h i s
paper i s t o def i ne t he necessary r egi ons o f cont r ol and t he methods whi ch
per mi t st abl e oper at i on t hrough bl ade- pi t c h cont r ol .
The bl ade- pi t ch c ont r ol l er pr ovi des f o r maximum energy t r ansf er between
t he wi nd and t he wi ndmi l l under var yi ng wi nd condi t i ons. I n or der t o under-
stand how t he c ont r ol l er does t hi s , we must f i r s t l ook a t t he ov er al l st r uct ur e.
The wi ndmi l l i s f r e e t o r ot at e i n yaw on a pl at f or m so t h a t t he bl ades
ar e always f aci ng i n t o t he wind. Thre ar e t hr ee bl ades, whi ch ar e c ont r ol l abl e
i n pi t c h. The r o t o r i s connected t hrough a speed-up dr i v e t o a 25 kW generat or.
Rel at i ve t o windspeed t her e ar e f our d i s t i n c t r egi ons o f p i t c h c ont r ol . I n
r egi on 1 (0-5 mph) t he p i t c h c ont r ol l er w i l l be c al l ed upon t o pos i t i on t he
bl ades a t an angl e o f at t ack whi ch w i l l pr ovi de t he gr eat est s t a r t i n g t orque.
For these wi nd speeds, t he p i t c h angl e i s set a t minus ei ght degrees. I n
r egi on 2, (5-26 mph) t he c ont r ol l er w i l l mai nt ai n t he pi t c h angl e a t minus
ei ght degrees. Thi s w i l l br i ng t he generat or up t o speed i n t he shor t est
possi bl e t i me and w i l l del i v er t he maximum power a t any gi ven wi nd speed. I n
r egi on 3 (26-45 mph) t he cont r ol l e r mai nt ai ns t he generat or speed a t 1800 RPM.
Wi t hout t he c ont r ol l er , t he generat or would overspeed and t r i p o f f t he l i ne.
I n r egi on 4 (above 45 mph) t he c ont r ol l er has t o f eat her t he wi ndmi l l bl ades.
Thi s amounts t o shut t i ng down t he system t o pr event dest r uct i on under hi gh wi nds.
Through computer anal ysi s, t he power out put curve has been r el at ed t o
p i t c h angl e through t he t i p speed r a t i o (see bl ade- pi t ch schedule, Fi gur e 1 ) .
The t i p speed r a t i o (nR/V) i s t he act ual cont r ol si gnal used i n t he pi t c h-
cont r ol c i r c u i t r y t o r epr esent t he desi r ed p i t c h angl e.
As t he f our cont r ol r egi ons ar e t ransversed, t he t i p speed r a t i o w i l l
i ncr ease f rom 0 t o 7 and t hen decrease t o 0. As wi nd speed i ncreases f rom
0-5 mph, t he t i p speed r a t i o i ncreases u n t i l i t reaches 7. Thi s poi nt
r epr esent s t he begi nni ng o f r egi on 2 where generat or RPM w i l l i ncr ease wi t h
i ncr easi ng windspeed. Throughout t h i s r egi on t he bl ade p i t c h w i l l be mai nt ai ned
a t ~i l i nus ei ght degrees whi ch i s equi val ent t o QR/ V = 7. A t 26 nlph t he generat or
w i l l reach r at ed RPM, and i i R/ V w i l l s t a r t t o decrease as wi nd speed i ncreases.
Thi s occurs because t he tachometer out put i s clamped a t 20 v ol t s . Thus, as
wi nd speed i ncreases, nR/V w i l l decrease and bl ade- pi t ch w i l l i ncrease u n t i l
reachi ng t he f eat her ed pos i t i on a t a windspeed o f 45 mph. I t shoul d be noted
t h a t t he cont r ol 1 er can be commanded t o f eat her sooner i f desi r ed (see Dwg.03.
02.01 ).
The p i t c h c ont r ol l er i s ac t ual l y a pul se- wi dt h modul ated ( PW) t r ans i s t or
swi t chi ng ampl i f i er . (see Dwg. 03.02.01) I t uses a DC power suppl y whi ch
t he ampl i f i er swi tches on and o f f a t a f i x ed frequency. The wi der t he wi dt h
of t he dr i v i ng pul se; t he hi gher w i l l be t he average cur r ent del i ver ed t o t he
motor; and t he f ast er t he motor w i l l t ur n. PWM provi des a cont i nuous f i n e
cont r ol o f t he p i t c h angl e.
Pul se-wi dt h modul at i on i s obt ai ned by comparing t he desi r ed p i t c h angl e
si gnal (nR/V) wi t h a t r i a n g l e wave. The t r i a n g l e wave i s obt ai ned by i n t e -
gr at i ng a square wave. The l engt h of t i me i n whi ch t he vol t age l evel o f t he
t r i a n g l e wave i s gr eat er t han t h a t of R/ V det ermi nes t he wi dt h o f t he d r i v i r ~ g
pul se (see Fi gur e 2) . A s i mi l a r compari son i s made between t h i s same t r i a n g l e
wave and a feedback si gnal whi ch r epr esent s t he act ual p o s i t i o n o f t he bl ades.
These two pul se- wi dt h modul ated si gnal s ar e t hen compared ( t hey ar e 180 degrees
o u t o f phase wi t h each ot her ) ; t hus, pr ov i di ng t he d r i v i n g si gnal f o r t he
motor. The mot or w i l l t u r n i n t he d i r e c t i o n whi ch w i l l equat e t he act ual
angl e wi t h t he desi r ed angl e.
, , , >, : nl . l , I . . .
I ENERGY ALTERNATIVE PROGRAM
UNIVERSITY OF MASSACHUSETTS
UNNERSITY OF MASACHUSElTS/AMHERST
ENERGY ALTERNATIVES PROGRAM
UNIVERSITY OF MASSACHUSETTS/AMHERST
ENERGY ALTERNATIVES PROGRAM
PRELIMINARY REPORT
THERMAL SYSTEMS
WIND FURNACE PROJECT
Jon G. McGowan
and
Ghazi Dar kazal l i
U.Mass. Wind Furnace
Energy Al t er nat i v es Program
Un i v e r s i t y o f Massachuset t s
Amherst, Massachuset t s 01 002
A p r i l 1975
TR/76/ 7
Appendi x I X
Work in the area of thermal systems can be divided into two categories:
analytical model ing work and experimental work.
The analytical modeling has
formed the basis for the experimental design and a summary of i t s key
points follows. (With the completion of the economic p a r t of t hi s work,
a more detailed technical report on t hi s subject will be issued. )
1. ANALYTICAL MODEL
A. Description of the Overall System Configuration
The analytical model i s based on a mathematical simulation using
a di gi t al computer t o determine the f easi bi l i t y and performance of
using w-ind heating systems for home heating and domestic hot water
demands. Also, the possi bi l i t y of combining the wind systems with a
f l at - pl at e solar collector sub-system i s investigated. The basic wind
energy input component for a l l systems i s a horizontal axis wind machine.
The performance of the heating systems, for a given s i t e and
weather conditions, i s studied as a function of the following key
system parameters: 1 ) the wind generator blade diameter, 2 ) the wind
generator tower height, 3 ) the si ze of the residential heating del ivery
system, 4) the si ze of the solar col l ect or, and 5 ) the si ze of the thermal
storage water tank. A detailed economical analysis of the t ot al cost ,
for each of the systems studied will be based on the assumption of mass
produced unit manufacturing. The description of the di fferent system
model s fol 1 ows.
Model I i s the simplest windpower system (Fig. 4. 1) . I t has no energy
storage and el ect ri cal energy i s delivered t o the house di rect l y from the
wind generator using el ect r i c baseboard heaters. In case there i s not
enough el ect r i c energy generated t o keep the house a t room temperature
an auxiliary heating system i s switched on.
If a t any time the el ect r i c
energy generated exceeds the house heating demands, the excess energy i s
used f or domestic hot water heating (optional).
( 2) Model 11.
In addition t o the wind generator t hi s system (Fig. 4.2) has a
water thermal storage and a liquid to a i r baseboard heaters. This
system's operational procedure i s similar t o Model I except t hat the
el ect ri cal energy from the wind generator i s transferred t o the storage
t a n k by means of resistance heaters. Heat then i s delivered from the
storage t o the house by ci rcul at i ng hot water through the baseboard
heaters.
13) Model 3-A.
This system has two energy sources and one thermal storage (Fig. 4. 3).
A f l a t pl at e solar col l ect or sub-system i s added t o the wind system
described in Model 11. This col l ect or sub-system consists of a f l a t
pl at e col l ect or, a water-to-water heat exchanger and a ci rcul at i ng pump.
Both wind and sol ar energy collected are delivered t o the storage tank
and then, using the same heat delivery system as in Model 11, thermal
energy i s transferred t o the house.
(4) Model 3-B.
This system i s similar t o Model 3-A except t h a t t hi s model uses two
separate thermal storage tanks (Fig. 4. 4), one f or wind energy and the
other f or sol ar energy.
In order t o operate the col l ect or a t i t s highest
efficiency, t hi s model gives pri ori t y t o delivering energy from the sol ar
st or age t ank t o t he house.
When t he t emperat ure of t he s o l a r st or age
i s l es s t han t he desi r ed l ower l i m i t , t hen t he syst em w i l l d e l i v e r energy
f r om t he wi nd st or age t ank.
The two st or age t anks ar e connect ed t o
each ot her v i a a mi xi ng pump whi ch i s used when t he t emper at ur e o f one
o f t he st or age t anks i s hi gher t han t he upper l i m i t .
( 5) Domesti c Hot Water Nodel .
Thi s model i s used, i n conj unct i on wi t h t he pr evi ous models, t o
i n v e s t i g a t e t h e p o s s i b i l i t y o f usi ng t he wi nd and s o l a r systems t o
pr ovi de t he domest i c hot wat er r equi r ement of a r e s i d e n t i a l home.
B. System Components Descr l pt i on
Si nce t he o v e r a l l system performance i s a f u n c t i o n o f each i n d i v i d u a l
component, a d e t a i l e d physi cal model and mat hemat i cal anal ysi s has been
devel oped f o r t he f ol l owi ng basi c components: t he e l e c t r i c wi nd gener at or ,
t he f l a t - p l a t e s ol ar c o l l e c t o r , t he t her mal wat er st or age t ank, t he
house, and t he baseboard heat exchangers. Al so, i t i ncl udes a d e r i v a t i o n
of t he maximum windpower del i ver ed by a wi nd gener at or as a f unc t i on o f
t he wi nd v el oc i t y , gener at or bl ade di amet er, and gener at or e f f i c i e n c y .
I n addi t i on, based on a v a i l a b l e dat a, t he usef ul s ol ar energy i s det ermi ned.
Al so, an energy bal ance equat i on f o r t he c ont r ol volume around a st or age
t ank i s der i ved. The hour l y heat i ng demands a r e cal cul at ed usi ng basi c
p r i n c i p l e s o f thermodynamics and heat t r ansf er , and st andar d ASHRAE
pr act i ce. The amount of heat t r ansf er r ed t o t he house by t he heat d e l i v e r y
system i s cal cul at ed as a f unct i on of t he s i z e and c h a r a c t e r i s t i c s o f t he
i n t e r n a l heat exchangers.
C. Di g i t a l Computer Anal y t i c al Model s
Because o f t he v a r i e t y o f model s t est ed and t h e number o f d i f f e r e n t
component c onf i gur at i ons t he computer model s i ncl udes t he f o l l o wi n g
program and sub-programs: 1) a mai n program whi ch s pec i f i es t he model,
t h e system conf i gur at i on, and t he i n i t i a l condi t i ons o f t he var i ous
components. Al so, t he program combines t he ot her sub-programs t o det er mi ne
t he system perf ormance ( Fi g. 4. 5) , 2) a dat a sub-program whi ch i ncl udes
t he weat her and s o l a r dat a f o r a gi ven s i t e , 3) a wi nd sub-program t h a t
c al c ul at es t he windpower avai l abl e, 4) a s o l a r sub-program t h a t c al c ul at es
t he usef ul s o l a r energy a v a i l a b l e t o t he house, 5) a l oad sub-program whi ch
c al c ul at es t he house heat i ng l oad, 6) a heat exchanger sub-program whi ch
cal cul at es t he energy del i v er ed by t he baseboard heat ers, and 7) a hot
wat er sub- r out i ne t h a t det ermi nes t he energy del i v er ed by t he system i n
a form o f domest i c hot wat er. The use of a s p e c i f i c sub-program depends
on t he system conf i gur at i on of t he model t est ed.
D. System Performance and Anal t y i c a l Resul t s
A summary o f t h e most i mpor t ant a n a l y t i c a l r e s u l t s f or t he var i ous
model s pr evi ousl y di scussed f ol l ows. The r e s u l t s ar e based on mai nt ai ni ng
t he house i n s i d e t emper at ur e a t 68OF. A l l systems a r e t est ed f o r bot h t he
average and model homes by var yi ng key system paramet ers such as:
1 )
wi nd gener at or bl ade d i arnet er
2) wi nd gener at or t ower hei ght
3) wat er st or age t ank s i z e
4) t he heat de1i ver. y system s i z e
5 ) s o l a r c o l l e c t o r s i z e
Sub - pmgt ~m
Tabl e 4.1 represent s a mont hl y sumnary of t he heat i ng l oad o f bot h
t he average and model halies based on t he weather data fro111 art ford,
Conn. Fi gur e 4.6 represent s t y pi c al monthl y val ues o f t he "average"
home energy requi rement s and t he i nput o f t he wi nd and sol ar systems.
The r epr esent at i ve windpower out put shown i n t he f i gur e i s based on a
32.5 f o o t bl ade di amet er wi nd generat or pl aced on an 80 f o o t hi gh tower.
The s ol ar energy i nput i s gi ven f o r a 200 square f oot f l a t pl at e sol ar
col l ect or , mounted v e r t i c a l l y on t he sout h wal l o f t he house. Due t o
t he i nt er ac t i on between t he system components, and ot her l osses, t he
amount o f sol ar and wi nd energy t he system del i v er s t o t he house i s
l ess than t he energy i nput s shown i n t he f i gur e. To determi ne t he exact
amount o f usef ul energy del i ver ed by t he system an hour-by-hour anal ysi s
i s performed usi ng t he anal y t i c al model descri bed pr evi ousl y.
To determi ne and compare t he performance o f t he pr evi ousl y descri bed
models, a parameter (Qaux/Qtotal ) i s defi ned as t he r a t i o o f t he energy
del i ver ed t o t he house from an a u x i l i a r y source t o t he t o t a l house heat i ng
energy requi rements. Throughout t he past year, a l ar ge number o f
canputer runs f or t he Har t f or d wi nd dat a and Bl ue Hi l l s , Mass. sol ar
dat a were performed. Thi s parameter serves t o measure t he energy performance
of t he var i ous systems, and as a maj or i nput t o f ut ur e system economic
st udi es.
Fi gur e 4.7 shows performance f o r t he wind gener at or system
wi t hout st orage (Yodel 1) as a f unct i on o f wind generat or bl ade di ameter
and tower hei ght ( l ower 1 i nes of each band r epr esent 100 ft. tower hei ght s) .
The performance o f t he windpower heat i ng system wi t h wat er thermal st or age i s
summarized f o r var yi ng st orage si zes and bl ade di ameters i n Fi gures 4.8
and 4.9.
Month Model House*
(No Basement)
SEPTEMBER 308.7
OCTOBER 826.1
NOVEMBER 184.1 .1
DECEMBER 2295.5
JANUARY 301 8. 6
FEBRUARY 2265.9
MARCH 2172.8
APRIL 1481.8
MAY
Tot al l oad f o r 1 5028.4 kWhr
Heat i ng Season
Average House* (SH-1)
(Model wi t h Basement)
721.6
1850.6
3503.8
4829.3
6272.1
4746.1
4581.6
3186.2
1822.2
31 91 3. kWhr
Tabl e 4.1 - Resi dent i al Heat i ng Requi rements ( f o r 6600 Degree Day Cl i mat e)
* These numbers mi ght be changed s l i g h t l y when t he f i n a l house
des c r i pt i on i s added t o t he program.
Q total
0. 6
NO ENERGY STORAGE
SYSTEM
TOWER HEIGHTS : 60 to I 0 0 ft.
"AVERAGE" RESIDENCE
(MODEL HOME) f
BLADE DIAMETER , FT.
The model l i ng o f t he combined wind and sol ar heat i ng system obvi ousl y
opens up t he possi b i 1 i t y o f var yi ng more system parameters, t hus i t takes
more graphs t o show t h i s system' s performance.
Fi gures 4.10 t o 4.17
summarize a ser i es of runs f o r t h i s system. Fi gur e 4.10 gi ves r es ul t s
f o r t hr ee house heat i ng l oads and var yi ng c ol l ec t or si ze, wi t h t he
system al so suppl yi ng 50 gal l ons o f hot water per day. Fi gur e 4.11
shows t he ef f ec t o f storage si ze and bl ade di ameter on system performance.
Fi gures 4.12 t o 4.14 present r es ul t s f o r 3 d i f f e r e n t tower hei ght s var yi ng
bl ade di ameter and c ol l ec t or area. Fi gures 4.15 t o 4.17 r epr esent another
way o f showing these r esul t s.
The two t ank storage system (Model 3B0 performance i s shown i n
Fi gures 4.18 t o 4.20. Fi gur e 4.18 al so shows t he performance o f sol ar - onl y
and wi nd-onl y heat i ng systems. From these ser i es o f anal y t i c al runs
i t i s apparent t hat t he added cost o f t he two tank system i s not
j u s t i f i e d by a l ar ge correspondi ng i ncrease i n system performance.
Wi t h t he i ncreased a v a i l a b i l i t y o f more wi nd and sol ar dat a a t t he
end o f t he 1975-76 heat i ng season, i t i s ant i ci pat ed t hat a ser i es o f
anal y t i c al runs usi ng t h i s new data w i l l be car r i ed out.
D. The Systems Economics Model
A det ai l ed cost anal ysi s of t he var i ous systems under consi der at i on
i s pr esent l y i n progress. Thi s model i ncl udes a det ai l ed cost anal ysi s
(present and f ut ur e pr i ces) of a l l t he system component and i n s t a l l a t i o n
cost s. The t o t a l cost o f t he wind o r wi nd and sol ar heat i ng systems w i l l
i ncl ude t he ac qui s i t i on cost s of a l l systems components, and t he cost o f
oper at i on and maintenance.
I t i s pl anned t o opt i mi ze system cost , var yi ng
Qaux
Q total
0.6
TOWER HEIGHT t 8 0 FT.
BLADE DIAMETER = 32.5 FT.
. .
' AVERAGE" HOME
MODEL HOME
(NO BASEMENT)
STORAGE SI ZE, GALLONS '
1 .o
0.8
Q aux
total
0.6
0.4
0.2
0
TOWER HEIGHTS =60t o100 ft.
STORAGE SI ZE = 2000 GAL.
HW = 5 0 GAL/DAY
D.32.5 FT. HT = 60 FT
H.X.= 50 FT., TANK = 2000 GAL. 1.0
.8
AUX
QTOTAL
.6
-
-
-
LOAD = 46193 k w h
LOAD= 32995 k w h
17 166 kwh
0 -;
I I 1
"AVERAGE " HOUSE
TOWER HEI GHT. s 80 FT.
ACOL = 200 F T . ~
WIND S SOLAR
STORAGE SI ZE , GALLONS
Q TOTAL
.6
' AVERAGE " HOUSE
HT = 40 FT
STORAGE = 2 0 0 0 GAL.
20 25 3 0 . 35 40
BLADE DIAMETE R, FT.
' AVERAGE" HOUSE
HT = 60 FT.
STORAGE 2000
BLADE DIAMETER , FT
-1
.8
QAUX.
TOTAL
-6
4
.2
7
-
-
-
-
ACOL
200 FT2
4 4 0 0 F T ~
-
-
600 F T ~
I I I I I
20 25 30 35 40
.8
AUX.
TOTA L
.6
20 25 30 35 40
BLADE DI AMETER, FT
Q
TOTAL .6
H. X. = 50 FT.
TANK = 2000 GAL.
ACOL = 200 F T . ~
- -- -
40 FT.
60 FT.
80 FT.
BLADE DIAMETER - FT.
.
LOAD = 32995 kwh
H.X. = 5 0 FT
TAN K = 2000 GAL.
ACOL = 400 FT.
TOTAL '.,
4
BLADE DIAMETER - FT.
1 .o
.8
Q~~~ 06
TOTAL,
.4
.2
0
LOAD = 32995 kwh
M.X. = 5 0
TANK = 2000 GAL.
ACOL = 600 FT.
BLADE DI AMETER - FT.
LOAD = 46193 K WH
Hr= 6 0 f '
600 FT2 (2000 GAL. )
---
800 F T ~ (2000 GAL)
------
WI ND ONLY (2C00G.l
200 ( 2000 GAL 1
200,( 2 :2000 GAL )
600, ( 2000 GAL)
4
600, (2 :2000 GAL)
BLADE DI AMETER, FT
LOAD = 46193 KWH
D = 32.5 FT.
HT. 60 FT.
ACO L = 600 FT'
1.0
.8
. -
P~~~
).TOTAL
.6
SI ZE OF STORAGE 2 ( WI ND) I N GALLONS
-
-
TANK I
-
*.
--
;2--
--
-
-
I I I I 1 I 1
' - - - - . - .
loo0 2000 3000 4000
LOAD = 46193 KWH
HT = 60 FT.
1.0
-8
'AM.
-
TOTAL
.6
D = 32. 5FT.
ACOL = 600 FT
.
. n
.
TANK 2
SIZE OF STORAGE ( SOL AR) GALLONS
t he i mpor t ant paramet ers t h a t a f f e c t o v e r a l l perfornlat~cc!.
2. EXPERIMENTAL SYSTEM
Under t h i s t ask o f t he work, t he t hermal systems group has desi gned
and purchased a l l t he components necessary f o r t he si mul at i on o f systems
1-3 i n St{-1. Previ ous q u a r t e r l y r epor t s have d e a l t wi t h t he d e t a i l s o f
t he exper i ment al system and w i l l n o t be r epeat ed here. Dur i ng t he
summer o f 1976, a compl et e t echni cal r e p o r t descr i bi ng t he background
and f i n a l d e t a i l s o f t he exper i ment al system w i l l be compl eted.
A t t h e pr esent t i me, due t o t he s h i f t i n g o f a v a i l a b l e t echni ci ans
t o ot her phases of t he p r o j e c t , t h e t hermal subsystems (such as s o l a r
col l ec t or s , st or age t anks, baseboard heat er s, et c . ) have n o t been
compl et el y i n s t a l l e d (and t est ed) i n SH-1. However, a l l t h e necessary
equi pment and pr el i mi nar y i nst r ument at i on has been obt ai ned and i s
expect ed t o be i n oper at i on ( f o r exper i ment al t e s t i n g ) by June o f 1976.
UNNERSrrY OF MASSACHUSETTS/AMHERST
ENERGY ALTERNATIVES PROGRAM
UNIVERSITY OF MASSACHUSETTS/AMHERST
ENERGY ALTERNATIVES PROGRAM
Pr el i mi nar y Report
WIND DATA COLLECTION AND ANALYSIS
WIND FURNACE PROJECT
by
Frank C. Kaminsky
U.Mass. Wind Furnace
Energy A1 t er nat i v es Program
Uni v er s i t y of Massachuset t s
Amherst, frlassachusetts 01002
TR76/8
Appendix X
Wind Dat a Co l l e c t i o n and Anal ysi s
The o v e r a l l o b j e c t i v e o f t h i s p o r t i o n o f t he r esear ch e f f o r t i s
t he devel opment o f s t a t i s t i c a l model s and comput er sof t war e t h a t can
be used i n es t i mat i ng t he p r o d u c t i v i t y o f a l t e r n a t i v e t ypes of wi nd-
d r i v e n gener at or s proposed f o r a gi ven l o c a t i o n . These model s and
computer programs a r e bei ng devel oped and t es t ed usi ng wi nd dat a f r om
t he New Engl and ar ea and desi gn dat a f r om sever al sour ces: The UMass
wi nd- dr i ven gener at or s bei ng ' const r uct ed and proposed as p a r t o f t he
t o t a l r esear ch e f f o r t i n t he devel opment o f a wi nd f ur nace; a l t e r n a t i v e
wi nd- dr i ven gener at or s bei ng const r uct ed and proposed by NASA; and
commer ci al l y a v a i l a b l e gener at i ng systems f r om Dunl i t e, Au s t r a l i a ,
El eckt r o, Swi t zer l and and Aero Power, U.S.A. The work i n t h i s p o r t i o n
of t he r esear ch e f f o r t i s pr ogr essi ng on schedul e and i s summarized
bel ow i n two p a r t s . The f i r s t descr i bes t he dat a c o l l e c t i o n s i t e s
t h a t have been es t abl i shed and t h e second descr i bes t h e e f f o r t t h a t
has been d i r e c t e d t owar d t he devel opment o f s t a t i s t i c a l model s and
comput er sof t war e.
Par t I : Dat a Co l l e c t i o n Si t e s
Usi ng equi pment f unds f r om t he wi nd f ur nace g r a n t and a d d i t i o n a l
f unds f r om ot her sour ces, f o u r dat a c o l l e c t i o n s i t e s have been est abl i shed
i n t he v i c i n i t y o f t he Un i v e r s i t y o f ! ~~assachuset t s/ Amher st . These s i t e s
have been oper at i ng cont i nuousl y si nce tF,c Summer of 1975 and w i l l pr o-
v i d e hour l y wi nd speed dat a t h a t ser ves as i n p u t t o comput er s i mul at i on
programs f o r eval uat i ng t he performance of t h e wi r ~r l f ur nace.
These
s i t e s a r e l ocat ed i n t he towns of Leyden, Nor t hf i el d, Shel bur ne and
Amherst, Massachusetts. Except f o r t he Arnherst s i t e , cont i nuous char t -
paper r ecor di nq systems have been i n s t a l l e d wi t h anemometer amps l ocat ed
on t owers and/ or t el escopi c masts. A t t he Amherst s i t e , average wi nd
speed i s recorded a t hal f - hour i n t e r v a l s usi ng a t r ansmi ssi on l i n e and
a mechani cal paper punch. The r e s u l t i s a paper t ape t h a t can be i n p u t
t o t he t i me shar i ng system a t t he Uni v er s i t y . Si nce t he anemometer cups
f o r t h i s system ar e l ocat ed a t t he s i t e o f t he home whi ch i s bei ng
const r uct ed as p a r t o f t h i s r esear ch ef f or t , t h i s i n p u t w i l l pr ovi de a
means o f eval uat i ng and compari ng t he si mul at ed performance o f t he wi nd
f ur nace t o t he act ual performance.
Four ot her dat a s i t e s have been est abl i shed t hr ough a cooper at i ve
e f f o r t between t he r esear ch team and p r i v a t e i ndi v i dual s al s o concerned
wi t h t he devel opment o f windpower systems. These i ndi v i dual s , Mr. R. Bl azej ,
Grassy Brook Vi l l age, Newfane, Vermont; Mr. D. Lawrence, Bar nst abl e Harbor,
Barnst abl e, Massachusetts; Mr . F. Harvey, M t . Equinox, Sky Li ne Dr i ve,
Vermont; and M r . G. Leary, Hol yoke Gas and El e c t r i c , have pr ovi ded f unds
and, wi t h assi st ance from t he r esear ch team a t t he Uni v er s i t y o f Massachusetts,
have est abl i shed dat a c o l l e c t i o n s i t es . These s i t e s have been oper at i ng
cont i nuousl y si nce t he F a l l of 1975 and w i l l pr ovi de addi t i onal dat a f o r
eval uat i ng t he performance of t he wi nd f urnace.
I n a d d i t i o n t o t he e i g h t s i t e s menti oned above, addi t i onal dat a has
been secured t hr ough t he gener osi t y of Brookhaven Nat i onal Labs, Long
I sl and, New York.
Thi s dat a, hour l y wi nd speed t aken si mul t aneousl y a t
s i x d i f f e r e n t hei ght s (f rom 37 ft. t o 400 f t . ) f o r two year s, w i l l pr ovi de
t he basi s f o r sever al s t a t i s t i c a l st udi es t h a t ar e bei ng conducted.
Through t he e f f o r t of sever al undergraduat e st udent s a t UMass, t he
dat a f r om a l l s i t e s i s bei ng anal yzed on a mont hl y basi s. A s o l i d dat a
base w i l l be avai l abl e, event ual l y, t o eval uat e t he perf ormance o f
a l t e r n a t i v e desi gns f o r t he wi nd f urnace i n d i f f e r e n t cont ext s, namely,
h i l l t o p l ocat i ons, coast al areas and v a l l e y l oc at i on.
6
Par t 11: S t a t i s t i c a l Models and Computer Sof t war e
Several computer programs have been. devel oped t o a s s i s t i n t he
anal y s i s o f wi nd dat a. Al t hough t hese programs w i l l be used p r i ma r i l y
i n t he ev al uat i on o f wi nd- dr i ven gener at or s under c ons t r uc t i on and bei ng
proposed by r esear cher s a t LlMass/Amherst, ever y at t empt has been made
t o make a l l computer programs s u f f i c i e n t l y gener al f or wi despread use.
Si nce many of t he computer programs ar e l engt hy, t hey w i l l n o t be i n -
cl uded as p a r t of t h i s pr ogr ess r e p o r t . Onl y a d e s c r i p t i o n o f t he pr o-
gram w i l l be pr ovi ded. A1 1 programs have been wr i t t e n i n t he For t r an
l anguage and ar e c u r r e n t l y bei ng processed usi ng t he Cont r ol Data Cyber
70-4 a t IIMass/Amherst. The aut omat i c p l o t t e r associ at ed wi t h t h i s
system i s used ext ensi vel y t o save t i me and e f f o r t . Each computer
program i s descr i bed b r i e f l y be1 ow.
Program I
Thi s program produces a v el oc i t y - dur at i on cur ve f o r i n p u t dat a
i n t he form of hour l y wi nd speed o r i n t he form o f a hi st ogr am o f hour l y
wi nd speed.
Program I1
Thi s program produces bot h a v el oc i t y - dur at i on cur ve and a power-
dur at i on cur ve f o r e i t h e r hour l y wi nd speed dat a o r a hi st ogr am o f hour l y
wi nd speed dat a. The user of t h i s program speci f i cs t he power out put
cur ve o f t he wi nd- dr i ven gener at or under consi der at i on i n e i t h e r a f unct i onal
form o r as a ser i es of pi ecewi se 1 i tnear appr oxi mat i on.
Program 11 I
Thi s program provi des t he user wi t h est i mat es o f t he parameters
o f t hr ee pr obabi l i t y densi t y f unct i ons t hat can be used t o model
average hour l y wind speed over a speci f i ed per i od o f ti me. The gamma,
log-normal and t he Wei bul l densi t y f unct i ons ar e used as pr obabi l i t y
models. The program then produces a v el oc i t y dur at i on curve o f t he
i nput dat a and di spl ays t he "goodness o f f i t " o f each model. Thi s
program pr ovi des t he basi s f o r pr oduct i on o f power-durati on curves
usi ng t he t hr ee mathemati cal models.
Program I V
Thi s program provi des an i nt er ac t i v e t i me shar i ng si mul at or f o r
eval uat i ng di f f er ent wi nd-dri ven generat ors i n a speci f i ed wind f i e l d
I n a manner s i mi l ar t o t hat used i n Program 11, t he user descri bes t he
power out put curve o f t he wi nd-dri ven generat or i n ei t her f unct i onal
form o r by a ser i es of pi ecewi se l i near approxi mati ons. The program
then operates t he wind generator hour-by-hour usi ng a data f i l e creat ed
by t he user. The energy produced on a d a i l y basi s i s then provi ded al ong
wi t h a monthl y summary. Wi th t h i s program, t he performance of numerous
wi nd-dri ven generat ors can be eval uated r api dl y f or a speci f i c l ocat i on.
I n addi t i on. t o t he development of computer software f or t he anal ysi s
and processi ng of wind dat a s t a t i s t i c a l st udi es ar e bei ng conducted i n
t hr ee areas:
1. Development o f confi dence i nt er v al s f o r power dur at i on curves.
2. Anal ysi s o f appr opr i at e pr obabi l i t y densi t y f unct i ons f or
descr i bi ng t he l engt h of calm peri ods, pr oduct i ve per i ods,
and f u r l i n g per i ods.
3. Development of mathemati cal procedures f o r produci ng vel oci t y-
dur at i on curves a t speci f i ed hei ght s usi ng a hi stogram of wind
dat a t h a t i s mixed. Mixed dat a occurs where a por t i on o f t he
f r equenci es were recorded a t one hei ght and another a t a second
hei ght , et c.
A paper descr i bi ng t he r es ul t s obt ai ned i n t he t h i r d area descri bed
above i s c ur r ent l y i n process and w i l l be avai l abl e shor t l y. The f i r s t
two areas descri bed above ar e under exami nati on and r es ul t s w i l l be
descri bed as t hey become avai l abl e.
UNNWITY OF MASSACr:;JSllS/AMHER!X
FNFRCV Al TFRNATNFC DEWTaAM
UNIVERSITY OF MASSACHUSETTS/AMHERST
ENERGY ALTERNATIVES PROGRAM
DYNAMICS OF NEW ENGLAND WIND FURNACE
by
Me r i t P. Whi t e
U.Mass. Wind Fur nace
Ener gy Al t e r n a t i v e s Program
Un i v e r s i t y of Massachuset t s
Amherst, Massachuset t s 01 002
TR/76/9
Appendi x X I
DYIUAMICS OF NEW ENGLANO WIND FIJKNACT
The f r equenci es and shapes o f t he l ower modes o f vi t ) , - at i on o f
t he s t r u c t u r e and gener at or system have been cal cul at ed. The p r i n c i p a l
dynamic l oads have been anal yzed and t h e i r i n t e r a c t i o n s wi t h t he
v i b r a t i o n modes est i mat ed. The r e s u l t can be r epr esent ed di agr amat i c al l y
as a mat r i x o f columns (modes of response) and rows ( l oadi ngs) .
General remarks
The ex t ent t o whi ch a p a r t i c u l a r mode o f v i b r a t i o n w i l l be ex c i t ed
by a p a r t i c u l a r l oadi ng system depends p r i ma r i l y on t he f o l l o wi n g
f ac t or s :
1 Load magni t ude
2. Degree o f coupl i ng between t he e x c i t i n g mechanism and t he
modal response. Bas i c al l y , t he e x c i t i n g f or ce must be capabl e
o f doi ng p o s i t i v e work dur i ng a c y c l e o f mot i on. For example,
an ext er nal f or ce a c t i n g on a p o i n t o f t he s t r u c t u r e whi ch has
a component o f mot i on p a r a l l e l t o t he f or c e does work and can
add energy t o t he system whi l e t he system moves.
3. Resonance: i n t he case of a per i odi c f or ce t he r e l a t i o n
between i t s f requency and t h a t of t he modal response i s i mpor t ant ,
t he c l os er t he agreement t he gr eat er i s t he r a t e of doi ng work on
t he system.
4.
System damping or t he a b i l i t y t o d i s s i p a t e v i b r a t i o n a l energy
compared t o t he r a t e a t whi ch enel yy i s suppl i ed. A st eady s t a t e
const ant ampl i t ude c ondi t i on i s reached when t he two r at es ar e
equal - work done per u n i t t i me equal s mect:anical energy l o s t per
u n i t t i me.
The s t r u c t u r a l system o f t he Windpower Furnace as c u r r e n t l y
desi gned does n o t cont ai n any s i g n i f i c a n t damping o r energy d i s s i p a t i n g
c a p a b i l i t y . I f i t t ur ns o u t t o be necessary, such mechanisms (damping
devi ces) can be i n s t a l l e d i n appr opr i at e l ocat i ons. Thi s shoul d be
avoi ded i f possi bl e because o f cost s o f such u n i t s and a l s o t he
necessi t y o f mai nt ai ni ng them over t h e l i f e o f t he system. A b e t t e r
s o l u t i o n f o r undesi r abl e v i b r a t i o n s i s t o avoi d resonances between
l oadi ngs and system v i br at i ons , perhaps by changi ng s t i f f n e s s e s o r
wei ght s o f components.
Modes o f Vi b r a t i o n Consi --- dered
1. Ve r t i c a l di spl acement : The f oot o f t he 10" s t eel pi pe t ower
mai nt ai ns cont act wi t h i t s base suppor t due t o t h e wei ght o f t h e whol e
system combined wi t h t he v e r t i c a l corr~ponent of t he 3000- l b f or c e i n
each o f t he f o u r st eel - r ope guys. The s t eel pi pe and t he gener at or
u n i t have a combined wei ght o f 4900 I b s . Thus t he t o t a l base compressi on
i s about 16,000 I b s and t h i s i s t he maximum val ue o f a v e r t i c a l dynamic
o s c i l l a t i o n f o r c e t h a t w i l l not cause u p l i f t . A more severe v i b r a t i o n
w i l l produce u p l i f t and t he r e s u l t w i l l be a l a r g e energy d i s s i p a t i o n
due t o r epeat ed i mpact s a t t he base. A v e r t i c a l v i b r a t i o n mode can be
produced by a v e r t i c a l dynamic l oadi ng, f o r example due t o bl ade unbal ance.
The ext reme case occur s i f a bl ade i s l o s t dur i ng r o t a t i o n .
The
c e n t r i f u g a l f or c e i n t h a t s i t u a t i o n i s
2
F = W r (,) / g
where W = bl ade wei ght , r = di st ance of i t s cent er o f g r a v i t y f r om t he
spi n axi s, (,J t he angul ar v e l o c i t y i n r adi ansl second, and g = g r a v i t y
accel er at i on.
At t h e maximum expect ed RPM, W = 7 7 I b, r = 4 ' , u = 17.5
and g = 32. 2. Then,
F = 3000 l b
F has a per i od T = 0. 36 sec. At l ower speeds t he dynami c f or ce i s
r educed and t he per i od l engt hened.
2A-2B. Ax i a l (horizonta1)displacement o f t h e gener at or u n i t
combined wi t h p i t c h i n g about a h o r i z o n t a l t r ansver se a x i s : Th i s mode
causes f l e x u r e o f t he suppor t i ng p i p e and changes o f l engt h o f t he guys.
Two modes have been c al c ul at ed: t he l ower i s p r i ma r i l y t r a n s l a t i o n wi t h
smal l p i t c h i n g angl es and has a per i od o f 0.7 second. The second i s
p r i ma r i l y p i t c h i n g wi t h smal l t r a n s l a t i o n s and has a per i od o f 0.05
seconds .
3A-3B. Lat er al t r a n s l a t i o n combined wi t h r o c k i n g about t he
l o n g i t u d i n a l a x i s : Ther e a r e two modes cal c ul at ed, t he f i r s t i s p r i ma r i l y
t r a n s l a t i o n wi t h a per i od of 0.7 seconds, t he second i s p r i ma r i l y r oc k i ng
wi t h a per i od o f 0. 03 seconds.
4A-4B. Tor s i onal ( about a v e r t i c a l a x i s c o i n c i d i n g wi t h t he suppor t
a x i s ) : Two c ondi t i ons have been consi der ed: i n t he f i r s t , t he base o f
t he suppor t i ng pi pe i s n o t r e s t r a i n e d i n t o r s i o n and t he onl y t o r s i o n a l
r e s t r a i n t i n t he system i s f ur ni shed by t he guys. Thi s mode has a per i od
of 1.66 seconds. I f t he base of t he suppor t i s r e s t r a i n e d i n t o r s i o n t he
per i od becomes 0. 12-0. 15 seconds. I n each case, t he gener at or u n i t i s
assumed t o be r i g i d l y f i x e d by a br ake t o t he suppor t i ng mast . Of cour se,
dur i ng oper at i on t he gener at or u n i t i s n o t so r es t r ai ned and t h e r e i s no
t o r s i o n a l o r r o t a t i o n a l mode o f ,i br at i on.
5. A i r f o i l f l e x u r e : Time dependent a i r f o i l l oads w i l l cause t he
a i r f o i l t o def l ec t , mai nl y i n i t s weakest d i r e c t i o e whi ch i s r oughl y
p a r a l l e l t o t he s pi n a x i s of t he syst em.
Thi s i s an i mpor t ant desi gn
consi der at i on si nce i t det ermi nes how much cl ear ance between t he a i r f o i 1
di s c and t he suppor t systc111 - s t eel pi pe and guys - hac t o ho l e f t ,
f o r t h i s bendi ng def or n~at i on.
The s ~ i ~ a l l e r t he f l e x u r e al l owance of t he
a i r f o i l bl ades t he hi gher can be t he poi nt o f at t achment o f t he guys
t o t he suppor t i ng pi pe. The hi gher t h a t at t achment t he s t i f f e r and
st r onger i s t he system as a whole, whi ch i s advantageous. The f l e x u r a l
f requency o f each bl ade i s est i mat ed, f rom measurements on a si mpl y
support ed bl ade, t o be about 3 - 4 cycl es per second when cant i l ever ed a t
t he hub.
Dynamic Loads
1. Bl ade unbal ance: Di f f er ences i n wei ght o r geometry o f
i ndi v i dual bl ades can gi v e r i s e t o unbal anced c e n t r i f u g a l f or ces normal
t o t he r o t a t i n g s h a f t ac t i ng a t t he hub. Thi s f or ce vect or r o t a t e s
wi t h t he t ur ni ng s haf t . Besi des pr oduci ng bendi ng st r esses i n t he
s haf t t h i s f or ce can al s o i n i t i a t e o s c i l l a t i o n i n Modes 1, 2B, 3A, 3B,
4A ( o r 4B). The extreme case i s , o f course, t h a t i n whi ch a bl ade i s
l o s t .
2. Wind gr adi ent and t ower shadow: Dur i ng a r o t a t i o n each bl ade
passes t hr ough a c y c l e of d i f f e r e n t wi nd v e l o c i t i e s . The wi nd gr adi ent
r e s u l t s from t he ground drag whi ch nor mal l y r e s u l t s i n i ncr easi ng wi nd
speeds wi t h el evat i on. Tower shadow i s due t o i nt er f er ence o f t he
suppor t i ng pi pe wi t h t he ai r f l ow. The r e s u l t i n g l i f t f or ce on each bl ade
i s a per i odi c f or ce so t h a t t he combined l oadi ng due t o a l l t hr ee bl ades
act ed on by t he wi nd i s a per i odi c f or ce more o r l ess p a r a l l e l t o t he spi n
axi s and a per i odi c coupl e about a hor i zont al ax i s normal t o t he spi n
axi s.
Bot h have t he same f requency, t hr ee t i mes t h e r o t a t i o n a l f requency
of t he shaf t .
The e f f e c t of t h i s l oadi ng i s t o cause hor i zont al
o s c i l l a t i o n s p a r a l l e l t o t he spi n a x i s and angul ar o s c i l l a t i o n s
about a hor i z ont al ax i s t hr ough t h e mass cent er and normal t o t he
s pi n ax i s .
3. Gusts and t ur bul ence: These g i v e r i s e t o non- per i odi c f or ces
a c t i n g on t he bl ades more o r l es s p a r a l l e l t o t h e spi n a x i s . There
woul d a l s o be some l oadi ng o f t he gener at or u n i t . The bl ades and t ower
have t o be desi gned f o r peak gust l oads. There i s u n l i k e l y t o be any
per i odi c response o f t h e system.
4. Gyr oscopi c- Cor i ol i s Forces: These r e s u l t f r om si mul t aneous
r o t a t i o n s about two per pendi cul ar axes, i n t h i s case t he spi n LI o f
t he shaf t , r o t o r , and bl ade assembl y on t he one hand, c oi nc i di ng wi t h
a r o t a t i o n n o f t he e n t i r e system about i t s v e r t i c a l ax i s , due t o a
change of wi nd d i r e c t i o n . For an cor r espondi ng t o 167 rpm and a
r easonabl e val ue o f i ?, f o r example 1 rpm ( = 0.1 r ad/ sec) , t he Co r i o l i s
ac c el er at i on a t t he t i p o f a bl ade i s smal l , about 29. The gyr oscopi c
coupl e o f t he whol e r o t o r , shaf t , et c. i s
c = I o)n
where I i s t h e mass moment of i n e r t i a about t h e s h a f t ax i s .
For a l l p e r i o d i c l oadi ngs t he per i ods o r ranges o f per i ods have
been cal cul at ed.
There has been no at t empt t o p r e d i c t modal responses
caused by t h e expect ed l oads si nce t he r e s u l t o f any such c a l c u l a t i o n i s
expect ed t o be h i g h l y undependabl e. When t he system i s i n oper at i on,
obser vat i ons w i l l i n d i c a t e whi ch of t he modes a r e ex c i t ed and t o what
ext ent . I f c u r a t i v e measures ar e needed, t hey w i l l be f a c i l i t a t e d by
knowi ng whi ch of t he dynamic l oadi ngs may be causi ng t he t r oubl e.
The f o l l o wi n g t a b l e shows ( h o r i z o n t a l l y ) t h e b.ari ous modes o f response
and ( v e r t i c a l l y ) t he sever al dynami c l oadi ng mechanisms - bot h wi t h t h e i r
per i ods o r f r equenci es. The squar es wi t h + cor r espond t o a pos s i bl e
model exci t ement by a p a r t i c u l a r l oad. An ++ i n d i c a t e s a pr obabl e
s t a t e o f e x c i t a t i o n .