Applied Thermal Engineering 94 (2016) 364–374

Contents lists available at ScienceDirect

Applied Thermal Engineering

j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / a p t h e r m e n g

Thermally induced stresses on radiant heating tubes including the

effect of fluid–structure interaction
Martin Hellenkamp *, Herbert Pfeifer
Department for Industrial Furnaces and Heat Engineering, RWTH Aachen University, Aachen, Germany

H I G H L I G H T S

• Temperature and stress distribution on a radiant heating tube are examined numerically.
• The numerical model is validated with experimental results.
• The stress calculations conducted are based on elastic material parameters.
• Critical areas of the component can be identified also with elastic material parameters.
• Mainly responsible for the stresses occurring are temperature differences on the tube.

A R T I C L E I N F O A B S T R A C T

Article history: High temperatures as well as local temperature peaks and their gradients within the individual compo-
Received 13 March 2015 nents of industrial furnaces are a challenge for the durability of the furnace. In order to optimize furnace
Accepted 17 October 2015 component design and thereby increase the life expectancy of each component being considered, the
Available online 10 November 2015
exact determination of local temperature distributions is indispensable. This necessitates a coupled ex-
amination of fluid flow and thermal processes, including the calculation of the resulting stresses for the
Keywords:
considered component. In order to introduce the so-called fluid–structure interaction (FSI) in the field
Radiant tubes
of plant engineering and industrial furnace engineering, a radiant heating tube was selected for calcu-
Thermal stresses
Temperature gradients lations concerning temperature and stress distribution, since radiant heating tubes are especially subject
Thermal expansion to considerable thermal and mechanical loading.
Fluid–structure interaction © 2015 Elsevier Ltd. All rights reserved.

1. Introduction
Industrial furnaces are thermoprocessing facilities, which are –
especially for high-performance facilities – often equipped with gas
burners.
For direct firing of an industrial furnace the material is in direct
contact with the combustion gases. This leads to scaling on the
product surface, and can also lead to a chemical change on the
surface. To avoid these effects, the combustion gas and process gas
atmosphere can be separated by means of radiant heating tubes.
Here, the heat transfer between the combustion gas and the process
gas takes place indirectly. Radiant tubes are for example used in fur-
naces for hot-dip galvanizing, strip flotation furnaces for the
recrystallization annealing of copper strips, or in furnaces for the
heat treatment of aluminum.
The heat transfer in fuel fired radiant heating tubes results from
internal and external heat transfer and thermal conduction in the ma-
terial of the radiant heating tube. The heat transfer on the inner side
* Corresponding author. Tel.: +49 241 80 25948; fax: +49 241 80 22289.
E-mail address: hellenkamp@iob.rwth-aachen.de (M. Hellenkamp).
of the tube results from the flame propagation and convection in the
tube. In a diathermic furnace atmosphere, for example nitrogen (N2)
or hydrogen (H2), the heat transfer on the outer surface of the radiant
heating tube depends on the convective heat transfer of the radiant
tube’s surface to the furnace atmosphere and the radiation ex-
change with the furnace walls. These act as auxiliary heating surfaces,
which in turn redistribute part of the heat received from the radiant
tubes to the furnace atmosphere by convection (Fig. 1).
Depending on their field of application, industrial furnaces are
exposed to different thermal loading. For example, facilities for alu-
minum heat treatment are operated in a temperature range of 150
to 600 °C, steel heat treatment facilities require temperatures above
1000 °C, and in the ceramic industry temperatures of 1300 °C are
not uncommon.
These different furnace temperatures result in considerable
demands on the deployed materials, since corrosion resistance and
mechanical strength properties decrease drastically with increas-
ing temperatures [2].
Furnace radiant heating tubes are a good example for the thermal
loading conditions to which furnace components are subjected. This
is confirmed by numerous failure analyses in literature concern-
ing heating tubes [3–8].

http://dx.doi.org/10.1016/j.applthermaleng.2015.10.080
1359-4311/© 2015 Elsevier Ltd. All rights reserved.
 M. Hellenkamp, H. Pfeifer/Applied Thermal Engineering 94 (2016) 364–374 365

Fig. 1. Heat transfer mechanisms on a p-shaped radiant heating tube (acc. to [1]).

In addition to the high process temperatures, the high local tem-
perature gradients in the components of the furnace result in
significant stresses in the materials used.
A better understanding of the fluid flow and heat transfer
in an industrial furnace can be achieved with the assistance
of computational fluid dynamics (CFD). For example, Hornig
et al. calculate the temperature distribution on the surface of
radiant heating tubes in a strip flotation furnace for recrystalliza-
tion annealing of copper at a temperature of approx. 750 °C
(Fig. 2) [9].
In this case for the simulation a constant heat flux is assumed
as a boundary condition at the surface of the tubes. Thus the areas
of different temperatures characterize areas of different heat trans-
fer conditions on the outer surface.

Fig. 2. Surface temperatures on p-shaped radiant heating tubes [9].

366 M. Hellenkamp, H. Pfeifer/Applied Thermal Engineering 94 (2016) 364–374

Fig. 3. Fluid–structure interaction as one-way coupling (acc. to [14]).

1.1. Fluid–structure interaction
The numerical results in [9] show that high temperature gradi-
ents are to be expected on the surface of radiant tubes, which can
lead to defects and failure in the long run. The defects, i.e. modi-
fied material structure due to plastic deformation, can subsequently
be detected by material testing [5].
An early identification of critical zones can be achieved using
the finite elements method (FEM), which calculates local stresses
and strains of the radiant heating tube [10–13].
However, this generally requires experimental data for the tem-
perature distribution on the radiant heating tube. Unfortunately, the
experimental determination of surface temperatures with respect
to ambient conditions and process parameters is seldom practica-
ble in an actual industrial facility.
Therefore, the design of radiant tubes is usually carried out with
highly simplified models which for example assume as a constant
heat flow density or even a homogeneous temperature distribution.
With these models, however, the exact description of the indi-
vidual load contributions, which accumulate to an effective total load,
both with regard to their cause and in terms of their contribution
to the total load, remains unclear.
In this paper the calculation results of an innovative numerical
model are presented. For the development of the numerical model,
the results of the numerical flow simulation are applied as input
boundary conditions for the structural mechanics simulations. This
method is referred to as fluid–structure interaction (FSI) and is an
innovative approach for the description of coupled problems. The
so developed model provides a holistic view of the relevant pro-
cesses that occur during the operation of a radiant tube. The thermo-
fluid dynamics and mechanical factors are taken into account.
Since the deformation of the radiant heating tube can be assumed
to be small, no relevant effects on the fluid flow are to be ex-
pected within and outside of the radiant heating tube. The fluid–
structure interaction can therefore be realized as one-way coupling
(Fig. 3).
With this model it is possible to investigate the stresses occur-
ring within the material of the radiant tube with regard to their
magnitude, their causes and the contribution of each loading type.
Furthermore, the numerical model enables the identification of
highly loaded and therefore critical positions during the design phase
of a thermal processing plant with radiant heating tubes. Therefore
an optimization with regard to the stress formation can be carried
out, thus increasing the durability of the components.
2. Experimental setup and simulation
The numerical model is validated by comparing simulated to
measured surface temperatures. These temperatures are mea-
sured at specific points using thermocouples on the surface of a
radiant heating tube in a test station. For the visualization of the
surface temperature, the decoiled length of the radiant heating tube
is considered (Fig. 4). The thermocouples are located on the top and
the bottom of the tube at the positions indicated in the figure.
The test station consists of a closed furnace room with a me-
tallic p-shaped radiant heating tube that is heated with a recuperative
burner. The heat sink due to material transported through a furnace
is represented by air-cooled heat sinks in the test stand’s ceiling and
floor (Fig. 5). It conduces to the examination of a thermal gradient
generation and hence arising thermally induced stresses in the in-
vestigated radiant heating tube.
The CFD simulations are carried out with the commercial sim-
ulation program ANSYS Fluent v14.0. The turbulence model chosen
throughout the simulations is the realizable k-ε turbulence model.
Radiation both inside and outside the radiant heating tube is cal-
culated with the “Discrete Ordinates” approach considering the gas
inside the burner and the tube as a participating medium. The gas/
air mixture’s combustion is simulated using the “Non-Premixed-
Combustion” model.
Fig. 4. Position of thermocouples.
 M. Hellenkamp, H. Pfeifer/Applied Thermal Engineering 94 (2016) 364–374 367

Fig. 5. CAD model and inner view of the test stand (acc. to [15]).

The CAD model shown in Fig. 5 is the basis for the geometry of
the radiant tube simulation in a furnace. The radiant heating tube,
the furnace chamber, the cooling tubes and a part of the thermal
insulation coating are included in the mesh of the model in order
to ensure that the model geometry corresponds to that of the test
stand as closely as possible.
The simulated burner geometry with inlets for natural gas and
air is shown in Fig. 6, as well as the exhaust gas outlet of the heating
tube. In the experiment the exhaust gas is used for recuperative air
preheating leading to an air temperature of 424 °C.
Within the scope of this paper, the results for the bilateral cooling
of the radiant heating tube through the combined use of the upper and
lower heat sinks are presented, which corresponds to the design case.
The boundary conditions for the design case are given in Tables 1–4.
The calculation of resulting stresses and strains (CSD – compu-
tational structural dynamics) is performed using ANSYS Mechanical
v14.5. In this case the walls are modeled using shell elements and
a perfectly elastic material behavior is assumed. The finite element
analysis of the heating tube is based on the same surface mesh as
the CFD simulation. This is necessary, as the node temperatures cal-
culated by ANSYS Fluent are transferred 1:1 to the CSD mesh. The
calculations evaluate the von Mises equivalent stress, assuming a
steady state and considering gravity effects.
The link between CFD and CSD is implemented as a one-way cou-
pling. This is admissible, since the resulting deformation is equivalent
to only approx. 1%, and no significant interdependences have to be
taken into account.
3. Results
The energy introduced into the system by the fuel stream is
80 kW. After combustion of the natural gas–air mixture, the heat
is transferred by convection and radiation to the radiant heating tube.
The radiant tube in turn emits an output of 59 kW to the furnace
chamber, while 33 kW remains in the exhaust gas. From the furnace
chamber the majority of the output, namely 57 kW, is transferred
to the cooling pipes. The wall losses are 2 kW.
Recuperative air preheating (12 kW) reduces the exhaust gas
losses to 21 kW. This corresponds to a combustion efficiency of 74%.
The exhaust gas exits the computational domain via the exhaust
stream, Fig. 7.
The experimentally and numerically obtained temperature pro-
files corresponding to the boundary conditions stated above are
shown in Fig. 8 as a function of the decoiled tube length. It can be
seen that the simulation tends to predict higher temperatures on
the lower side of the heating tube than on the upper side. In the

Fig. 6. Inlets and outlets of the numerical model (acc. to [15]).

368 M. Hellenkamp, H. Pfeifer/Applied Thermal Engineering 94 (2016) 364–374

Table 1
Boundary conditions for inlets and outlets.

Zone Type Mass-flow Temperature

Inlet natural gas Mass-flow inlet 0.00172 kg/s 30 °C

Inlet air Mass-flow inlet 0.030327 kg/s 424 °C
Inlet upper cooling Mass-flow inlet 0.319947 kg/s 32 °C
Inlet lower cooling Mass-flow inlet 0.318848 kg/s 32 °C
Outlet exhaust gas Pressure outlet
Outlet upper cooling Pressure outlet
Outlet lower cooling Pressure outlet

Table 2
Boundary conditions for the radiant heating tube.

Parameter Value

Internal emissivity 0.85

Wall thickness 4 mm
Fig. 8. Comparison between experimental data and numerically simulated tem-
Shell conduction
perature profiles on the upper and lower sides of the radiant heating tube with
bilateral cooling.

Table 3
Boundary conditions for the furnace chamber.
experiment, by contrast, temperatures measured on top are higher
Parameter Value than below. Nevertheless, the qualitative temperature profile is well
Heat transfer coefficient 25 W/(m2 K) reproduced by the simulation results.
Free stream temperature 30 °C The average surface temperature of the radiant heating tube is
External emissivity 1 681 °C. Thereby the average deviation between experimentally and
External radiation temperature 25 °C
Internal emissivity 1
numerically determined temperatures is 29 K at the measure-
Wall thickness 250 mm ment points on the upper side and 39 K on the lower side. The
standard deviation from the mean is 10 K on the upper side and 19 K
on the lower side (Fig. 9).
Table 4
Boundary conditions for the cooling pipes. 3.1. Temperatures and heat flux inside the radiant heating tube
Parameter Value
In Fig. 10a the velocity distribution of the flow in the symme-
Internal emissivity 0.85
try plane of the radiant heating tube is shown in the plan view. It
Wall thickness 2 mm
can be seen that it comes to a homogenization of the flow velocity

Fig. 7. Energy balance and computational domain (acc. to [15]).

 M. Hellenkamp, H. Pfeifer/Applied Thermal Engineering 94 (2016) 364–374
Fig. 9. Average deviation between simulation and experiment with bilateral cooling.
behind the reaction zone. In the area of the first elbow, the fluid is
decelerated and redirected around the bends by the outer wall. This
leads to a low flow velocity in the outer region and an accelera-
tion of the gas on the inner wall due to the corresponding static
pressure distribution. After passing through the elbow, the flow sepa-
rates from the inner side of the bend, thereby resulting in a
recirculation area. This is also apparent from the illustration of the
streamlines in Fig. 10b. After the fluid velocity homogenizes in the
second half of the right leg, the same can be observed for the re-
direction of the flow through the second elbow. A part of the exhaust
gas then leaves the computational domain through the recuperator,
while the rest recirculates and mixes with fresh gas and once more
passes through the radiant heating tube. Here also flow separa-
tion and a recirculation area on the inside of the elbow can be
observed.
The temperature distribution on the surface of a radiant heating
tube is influenced both by internal and by external parameters. The
internal factors determine the local energy flux density of the flame
and the exhaust gas to the inner tube wall.
Fig. 11 shows the temperature distribution inside the radiant
heating tube in the symmetry plane as well as along the stream-
lines. The area of the reaction zone with maximum temperatures
of around 1750 °C is clearly visible. It can be seen that it comes to
a homogenization of the flow velocity behind the reaction zone. In
areas of low flow rates (see Fig. 10), a reduced gas temperature is
determined, which is evident especially in the recirculation areas
behind the first and second elbow.
The heat flux density on the inner wall of the radiant tube is
shown in Fig. 12 for the top (a) and bottom (b) side. Noticeable is
Fig. 10. Flow velocity in the radiant heating tube [15]: (a) on the center cross section and (b) for 3D streamlines.
369
the outlet region of the heating tube, where almost no heat is trans-
ferred. This is due to the installation situation of the tube, which
is installed in the furnace through a hole in the heat-insulated
furnace wall.
Since the radiation intensity is proportional to the fourth power
of the temperature, it is dominant especially in the reaction zone.
Similarly, the proportion of radiation predominates in areas with
low flow rates, such as behind the first elbow and in the right part
of the outlet region (Fig. 12b). In areas of high flow velocities or
high turbulence intensities, the convective heat transfer domi-
nates, which can be seen by the low radiation contribution especially
at the end of the reaction zone and in the left part of the outlet
area. On average, the proportion of thermal radiation is approxi-
mately 50%.
3.2. Temperatures and heat flux outside the radiant heating tube
Fig. 13 shows the portion of the total heat flux on the outer wall
that is transmitted by radiation. Since the atmosphere is not recir-
culated in the furnace, the radiation portion is dominant and makes
up on average over 90%. On the upper surface of the heating tube,
the impact of the recirculation area is clearly visible: the propor-
tion of radiation here is due to the additionally reduced convective
heat transfer of almost 100%.
On the inner sides of the P the proportion of radiation is also
affected by the mutual shielding from the environment, which is
why the radiation contributes slightly lower with approximately 87%.
The occurring internal and external influences determine the re-
sulting temperature distribution on the surface of the radiant heating
tube, which is shown in Fig. 14 for the tube exterior. Due to the in-
sulation layer on the wall of the test furnace, the heat dissipation
in the outlet area of the radiant heating tube is hindered, leading
to high temperatures. The temperature drop at the outlet of the
radiant tube, however, results from an excessive cooling of the
exhaust gas leaving the computational domain in addition to back
flow effects. This cannot be explained physically and is attributed
to boundary effects of the modeling, which is why this effect will
not be evaluated any further.
High temperatures on the tube wall also occur behind the re-
action zone. These are mostly located on the lower side of the tube,
as on this side the resulting heat removed by the cooling pipes is
less than on the upper side. In the further course of the flow the
tube surface cools down continuously in accordance with the inside
gas temperature (see Fig. 11b). Due to the small distance between
the two legs of the heating tube, the heat dissipation to the
370 M. Hellenkamp, H. Pfeifer/Applied Thermal Engineering 94 (2016) 364–374
Fig. 11. Temperatures in the radiant heating tube [15]: (a) on the middle cross section and (b) for 3D streamlines.
environment between these surfaces is hindered, which leads to
higher temperatures on the inner thighs (Fig. 14c and d).
The mean temperature on the outer surface of the radiant heating
tube is 681 °C, whereas on the tube inside it is on average 4 K higher.
3.3. Thermomechanical loads
Based on the results of the thermo-fluid dynamic calculations,
the resulting stresses in the radiant heating tube for the design case
are examined below. The von Mises equivalent stress is evaluated
for this purpose. The contributions of the different loading types
to the occurring stresses are also addressed.
3.4. Radial temperature gradient through the wall of the tube
The heat transfer from the outer wall of a radiant heating tube
to a commodity to be heated is preceded by a conductive heat flux
from the inner side of the wall to the outer side, as a function of
the temperature gradient through the wall. With the approxima-
tion of a straight tube this can be calculated to be
(T − T )
Q = −2π ⋅ l ⋅ κ ⋅ a i .
r a loose bearing in the first bend, the stresses can be reduced by an
ln a
ri
where Q is the heat flux, l is the length, Ta is the temperature
outside, Ti is the temperature inside, κ is thermal conductivity, ra
is outside radius, and ri is inside radius.
For the temperature difference between the inner and outer wall
the following applies
Q r action, i.e. the coupling between the fluid and structural calculations.
Ta − Ti = − ⋅ ln a .
2π ⋅ l ⋅ κ ri
A transmitted heat flux of 59 kW through the wall (see Section
3) and a decoiled tube length of about 4 m result in a temperature
difference of about 4 K between the inner and outer wall of the
radiant heating tube. This temperature difference causes thermal
expansion of different magnitudes. As a result, compressive stresses
occur on the inside and tensile stresses on the outer wall.
When calculating the resulting stresses with a numerical sim-
ulation model of a hollow cylinder, whereby a heat flux density
corresponding to that of the experiments is defined as a boundary
condition, the results show that inside and outside maximum stresses
in the range of about 8 MPa arise (Fig. 15).
The magnitude of the stresses is proportional to the tempera-
ture gradient. Therefore, for larger heat fluxes due to higher burner
output, a higher stress through the radial temperature gradient is
to be expected.
3.5. Gravity
For a homogeneous surface temperature of 700 °C and a burner-
side mounting of the radiant heating tube, the influence of gravity
on the resulting stresses becomes evident especially in the con-
nector behind the second elbow. Due to bending and torsion forces,
stresses in the amount of approximately 20 MPa arise here (Fig. 16a).
With a further support of the radiant heating tube by means of
order of magnitude to about 2.5 MPa (Fig. 16b).
3.6. Strains and stresses of the heating tube as a result of fluid–
structure interaction
Fig. 17 shows the stresses in the radiant heating tube on the upper
side and the bottom side, as a result of the fluid–structure inter-
For a better overview, the already known temperature distribu-
tions from Fig. 14 are given.
The stress peaks marked in the figure are due to different phe-
nomena. The stresses at position (1) are due to the sharp drop
in temperature at the transition from furnace space to thermal
 M. Hellenkamp, H. Pfeifer/Applied Thermal Engineering 94 (2016) 364–374 371

Fig. 12. (a) Heat flux inside the heating tube [15] and (b) proportion of radiation compared to total heat flux at the inner wall [15].

insulation, thus causing different magnitudes of thermal expan-
sions that result in tensions in the transition area.
The stresses at positions (2) and (4) are due to the temperature
differences between the first and the second leg. The higher mean
temperature of the first leg causes higher thermal expansion rel-
ative to the second leg. This has to be compensated by the two bends,
thus leading to stresses.
Since in the inner region of the radiant heating tube the first and
second leg are in radiation exchange with each other while the heat
dissipation to the surroundings is limited, the temperature in this
area is higher than on the outside. This applies particularly to the
second leg, since here the reaction zone has no direct influence on
the temperature distribution. Therefore, the resulting increased stress
level occurs primarily at (3).

Fig. 13. Proportion of radiation compared to total heat flux at the outer wall [15].
372 M. Hellenkamp, H. Pfeifer/Applied Thermal Engineering 94 (2016) 364–374

Fig. 14. Temperatures of the outer wall [15].

The stress peak at (5) results from the locally high temperature
of the tube at the reaction zone in contrast to the surrounding areas.
As a result different expansion potentials and thus stresses are also
generated here.
Based on the calculated temperature drop of the exhaust gas
leaving the computational domain, very high stresses are also cal-
culated in this region. As mentioned before, this is due to the
proximity to the boundary conditions at the outlet and therefore
is not relevant (see Section 3.2).
3.7. Evaluation of causes of stresses
The stress distribution in the radiant heating tube is deter-
mined by four factors. The temperature gradient through the tube
wall has a subordinate influence, as it leads to stresses in the
range of 8 MPa. Gravity affects primarily the region of the second
manifold with stresses of up to 20 MPa, but can be almost com-
pletely avoided by a reasonable support of the radiant heating
tube. Temperature differences on the surface of the radiant heating
tube have the biggest influence on resulting stresses. Thereby the
average temperature difference of 20 K between the inside and
outside of each leg lead to maximum stresses of 90 MPa, while
the drop in temperature of 40 K from the first to the second leg
results in stresses of 100 MPa (Fig. 18).
Since these values are not necessarily present at identical po-
sitions, and the resulting stresses can be reduced in part by
superposition of tensile and compressive stresses, the calculated
maximum stress for the design case is not equal to the sum of these
values, but to a value of 150 MPa.
It should be noted that the stress calculations shown are
based on purely elastic material parameters. Since at high tem-
peratures the strength of the material is severely limited, plastic

Fig. 15. Stress in the wall of the tube due to a radial temperature gradient [15].
 M. Hellenkamp, H. Pfeifer/Applied Thermal Engineering 94 (2016) 364–374
Fig. 16. Stress by gravitation [15]: (a) for burner-side mounting and (b) with additional loose bearing.
deformations will occur for the considered application tempera-
tures as well.
4. Conclusion
An important result of this work is an understanding of the causes
(thermal/mechanical) of the stresses occurring within the radiant
heating tube. The radial temperature gradient through the wall, which
leads to a stress of 8 MPa, plays a minor role. If the radiant tube is
supported at both ends, the influence of gravity may also be ne-
glected, since it causes stresses of only 2.5 MPa. Mainly responsible
for the relevant stresses occurring on the radiant tube are local tem-
perature differences on the wall surface of the tube and the
temperature gradient from the first to the second leg of the tube. The
resulting stress levels are equal to 90 MPa and 100 MPa, respectively.
High stresses arise close to the inlet of the radiant tube, where
a temperature drop occurs at the transition between the thermal
insulation of the furnace and the furnace chamber. The tempera-
ture differences and thus the different values of thermal expansion
of the first and second leg mainly have an effect on the two elbows
and thus lead to high stresses in these areas. Local temperature dif-
ferences can be observed particularly between the inside and outside
of the second leg of the radiant tube as well as in the reaction zone,
where they affect the development of stresses.
The identified temperature differences and gradients can be di-
rectly linked to various internal and external factors that affect the
flow of energy from the flame and the exhaust gas to the inner wall
of the radiant heating tube and the outer heat transfer at the outside
373
of the radiant tube. For a low-stress operation of radiant heating
tubes, it is therefore necessary to first optimize the flow in the in-
terior of the tube in such a way that flow separation and local
recirculation areas are largely avoided or at least reduced. Further-
more a high recirculation ratio throughout the heating tube is
desirable.
For the interpretation of the results, it is important to know that
the stress calculations conducted are based on elastic material pa-
rameters. This leads to local stress peaks of up to 150 MPa, which
significantly exceeds creep rupture strength in the considered tem-
perature ranges and thus would result in a short-term failure of the
component during actual operation. Due to plastic deformations as
well as long-term creep effects occurring, however, a reduction of
stresses is achieved in reality, so that the calculated stress peaks
will not occur.
In order to take plastic deformation into account, knowledge of
the relevant material properties to describe the plastic behavior of
the examined material is required. In addition, detailed creep models
are needed, if this effect is to be included in the analyses as well.
The purely elastic stress analysis is nevertheless of impor-
tance, as critical areas of the component can already be identified
in the design phase and therefore appropriate measures can be taken
for optimization.
Acknowledgements
The IGF Project No. 15698 N of the Research Association for In-
dustrial Thermoprocessing Equipment e.V. (FOGI) and of the Research
374 M. Hellenkamp, H. Pfeifer/Applied Thermal Engineering 94 (2016) 364–374
Fig. 17. Stress on the radiant heating tube [15].
Fig. 18. Causes of stress [15].
Association for the Mechanical Engineering Industry (FKM) was pro-
moted via the AiF in the program for the promotion of industrial
research (IGF) by the Federal Ministry for Economic Affairs and
Energy due to a resolution of the German Bundestag.
References
[1] R. Bölling, M. Hellenkamp, W. Lenz, N. Giesselmann, H. Pfeifer, Verbesserte
Vorhersage thermisch induzierter Spannungen in Industrieöfen durch die gekoppelte
Simulation von Fluid-Struktur Wechselwirkungen. VDI-GVC Fachausschusssitzung
“Hochtemperaturtechnik”, Magdeburg, 2010.
[2] European Standard, EN 10095: heat resisting steels and nickel alloys, 1999.
[3] S. Chakrabarti, S. Sarkar, S. Kumar, G.R.P. Singh, M.D. Maheshwari, Failure of
radiant tubes in a continuous galvanising furnace. In: Tata Search (2003), pp.
399–403.
[4] K.H. Illgner, J. Sundermann, Investigation of high-temperature corrosion on
M-type radiant heating tube, Gaswärme Int. 56 (2007) 297–299 (in German).
[5] V. Palaniyandi, R.B. Allsup, R. Moser, Failure analysis of HP-40 (mod Nb) furnace
radiant tube. Materials Science & Technology 2008, Pittsburgh, 2008.
[6] H. Pohland vom Schloß, K. Lucka, H. Köhne, High temperature corrosion of
metallic radiant heating tubes in exhaust gas, Gaswärme Int. 56 (2007) 499–502
(in German).
[7] A. Ul-Hamid, H.M. Tawancy, A.-R.I. Mohammed, N.M. Abbas, Failure analysis
of furnace radiant tubes exposed to excessive temperature, Eng. Fail. Anal. 13
(2006) 1005–1021.
[8] K.B. Yoon, D.G. Jeong, Oxidation failure of radiant heater tubes, Eng. Fail. Anal.
6 (1999) 101–112.
[9] K. Hornig, R. Bölling, H.-J. Odenthal, H. Pfeifer, K. Berns, K. Schmitz, Numerical
and physical fluid flow simulation of a strip flotation furnace for non-ferrous
metal strips, Gaswärme Int. 52 (2003) 256–262 (in German).
[10] G. Dini, S.M. Monir Vaghefi, M. Lotfiani, M. Jafari, M. Safaei-Rad, M. Navabi, et al.,
Computational and experimental failure analysis of continuous-annealing
furnace radiant tubes exposed to excessive temperature, Eng. Fail. Anal. 15
(2008) 445–457.
[11] M.A. Irfan, W. Chapman, Thermal stresses in radiant tubes due to axial,
circumferential and radial temperature distributions, Appl. Therm. Eng. 29
(2009) 1913–1920.
[12] M.A. Irfan, W. Chapman, Thermal stresses in radiant tubes: a comparison
between recuperative and regenerative systems, Appl. Therm. Eng. 30 (2010)
196–200.
[13] H. Pohland vom Schloß, Investigations on high-temperature corrosion of metallic
radiant heat tubes in heat treatment plants (in German) (Doctoral thesis), RWTH
Aachen University, Aachen, Germany, 2008.
[14] W. Lenz, R. Bölling, H. Pfeifer, Einfluss inhomogener Temperaturverteilungen auf
mechanische Belastungen von Ofeninnengehäusen. Aachen-Freiberger-
Magdeburger-Kolloquium, Aachen, 2009.
[15] M. Hellenkamp, Analysis of stress in a metallic P-type radiant tube using the
fluid-structure interactions method (in German) (Doctoral thesis) RWTH Aachen
University, Aachen, Germany, 2013.