Sie sind auf Seite 1von 13

Arid Zone Journal of Engineering, Technology and Environment, April, 2017; Vol.

13(2):184-196
Copyright Faculty of Engineering, University of Maiduguri, Maiduguri, Nigeria.
Print ISSN: 1596-2490, Electronic ISSN: 2545-5818, www.azojete.com.ng

INFLUENCE OF THERMAL GRADIENT ON GAS TURBINE COMBUSTOR WALL


USING IMPINGEMENT/EFFUSION COOLING TECHNIQUES: CHT CFD
PREDICTIONS
A. M. El-jummah1*, U. Ibrahim2, G. E. Andrews3 and J. E. J. Staggs3
1
Department of Mechanical Engineering, University of Maiduguri, Maiduguri, Nigeria
2
Department of Mechanical Engineering, Ramat Polytechnic Maiduguri, Maiduguri, Nigeria
3
Energy Research Institute, School of Chemical and Process Engineering, University of Leeds, UK
*
Corresponding authors e-mail address: al-jummah@hotmail.com,
GSM: +234(0)8037803678
Abstract
Internal wall heat transfer relevant to impingement/effusion cooling techniques was investigated using conjugate
heat transfer (CHT) computational fluid dynamics (CFD) with ANSYS Fluent and ICEM commercial software.
This work concentrates on the development of CHT CFD design procedures that are applicable to combustor
wall and turbine blade heat transfer optimisation in gas turbine (GT). It specifically modelled and compares two
configuration which are specifically relevant to the impingement and effusion holes density n (m-2) and is the
ratio of the hole pitch X2. The configurations investigated are equal and unequal impingement and effusion
holes density n (m-2), respectively, whereby in each case the variation in the number of cooling holes were
carried out. The ratio of impingement and effusion number of holes/m2 (or hole density) n, investigated were
impingement/effusion: 4306/4306 and 1076/4306, respectively. The geometries were for impingement wall,
hole pitch X to diameter D, X/D ratio of ~ 11 but different number of holes N for both n geometries, at a
constant offset effusion wall, hole X/D of 4.7 of the same N for both the two configurations. The model
geometries have a constant impingement gap of 8 mm with both impingement and effusion walls at 6.35 mm
thick Nimonic - 75 material and were computed for varied air mass flux G from 0.1 - 0.94 kg/sm2. Symmetrical
applications were employed in modelling each of the geometry, whereby for the impingement hole, only quarter
of one hole was modelled, while for the effusion side the holes were either quarter or half modelled. The two n
geometries were computed with k - turbulence model using standard wall functions, which also applies to all
G. The predicted locally surface X2 (or hole square area) average heat transfer coefficient (HTC) h values
compared with with previously published experimental data showed good agreement. The reduced internal gap
flow recirculation with reduced heat transfer to the impingement wall caused by increased number of effusion
holes for unequal n impingement/effusion cooling design, shows that wall thermal gradient are higher than that
found for equal n.
Keywords: Impingement/effusion walls, conjugate heat transfer, design optimisation, hole density.

1. Introduction
The requirement for higher combustor outlet and turbine inlet temperatures, essentially
requires that cooling design optimization controls the development of gas turbine (GT)
system. This should be based on the need to inprove on the GT cycle thermal efficiency, as it
critically depends on the increased GT combustion temperatures. The combination of
impingement and effusion cooling offers both the best adiabatic film cooling effectiveness
and the best overall wall cooling effectiveness in GT engine (El-Jummah et al., 2017), which
this work investigates using the impingement/effusion cooling techniques.
Impingement/effusion cooling as Figure 1a show (El-Jummah et al., 2016b, 2017), is one of
the most effective cooling system for GT combustor and turbine blades walls, it enables
lower coolant mass flow to be used (Andrews et al., 1988). The impingement/effusion array
of holes, are usually designed based on the equal and unequal number of holes N or hole
density n (m-2) of the GT cooling system: which is the number of air hole per hole area ( X2),
where X is the hole pitch as shown in Figure 1b. In impingement/effusion internal wall heat

184
Arid Zone Journal of Engineering, Technology and Environment, April, 2017; Vol. 13(2):184-196
ISSN 1596-2490; e-ISSN 2545-5818; www.azojete.com.ng

transfer, the variation of n is most significant for the effusion film cooling and a high value of
n may be required for best overall film cooling effectiveness (Al Dabagh et al., 1990,
Andrews et al., 1988). A large number of effusion holes is advantageous for effusion cooling
effectiveness, whereas a large number of impingement holes may not be necessary and a
lower number could reduce manufacturing costs (El-Jummah et al., 2016b, 2017). A very
large number of effusion holes approaches the ideal film cooling of transpiration cooling
using a porous Nimonic - 75 wall hence the optimum could be as many as can be
manufactured (Andrews et al., 1985).
Presently, two major design configurations have been investigated using conjugate heat
transfer (CHT) computational fluid dynamics (CFD) procedures (El-Jummah et al., 2016b,
2017). The first is the situation in which there is equal n: 4306/4306. Here, an impingement
jet hole was offset half a pitch relative to an effusion hole, with the jet hole located between
four effusion holes so that an impingement jet flow is at the centre of four effusion holes. The
second is the unequal n: 1076/4306 configuration, in which the number of impingement jets
holes was reduced relative to the effusion holes, implying that the impingement holes were
offset 1/4 a quare pitch relative to the effusion holes. This indicates that for the 1076/4306
geometry is a fixed number of jet holes with the number of effusion holes increased,
signifying that the location of the impingement jet hole was at the centre of six effusion 10
10 holes. This work also concentrate on understanding the aerodynamics and wall conductive
heat transfer in the choice of geometries with effusion number of holes that have no
implications of structural strength, but did not model the effusion only film cooling.
Therefore, the influence of internal wall total heat transfer will be critically looked at using
the CHT CFD procedures.
In impingement/effusion cooling techniques as Figure 1 (a & b) shows, most workers have
used one impingement jet hole for every effusion air hole (Rogers et al., 2016, Shi et al.,
2016). However, a high number of effusion holes are required to optimise the effusion film
cooling. This is not required (a large n) for impingement cooling and the number of holes
used for the same impingement wall porosity has a relatively low effect on impingement heat
transfer, as the significant part of the effect is mainly due to the impact of cross-flow (El-
Jummah et al., 2017). In this work, the influence of cross-flow is by the discharge of the
impingement jet air through the effusion holes. Currently, the effusion to impingement hole
number were varied with ratio the investigated: 4/1 for unequal n which potentially give an
additional benefit in the effusion jet flow acting as suction on the surface impingement jet
flow, which is expected to increase the heat transfer. The aerodynamics, heat transfer and
thermal gradients CFD predictions for equal and unequal n of varied mass flux G geometries
will also be compared.
Tables 1 and 2 shows the dimensional parameters that were used in creating the model
geometries and the boundary conditions used in the model computations. Table 1 are the
geometrical parameters that were experimentally used by El-Jummah et al., (2016b, 2017)
whereby, column 2 and 4 were for the equal n (4306/4306) and column 3 and 5 were for the
unequal n (1076/4306) geometries. Also shown is Table 2: the flow conditions for all air G
(0.1 - 0.94 kg/sm2) that were used to analytically calculate the impingement jet (or effusion
hole) velocity Vj (or Vh) and the Reynolds number Rej (or Reh), which were found using

185
El-jummah et al.: Influence of thermal gradient on gas turbine combustor wall using
impingement/effusion cooling techniques: CHT CFD predictions. AZOJETE, 13(2):184-196 ISSN
1596-2490; e-ISSN 2545-5818, www.azojete.com.ng

Equations 1 and 2, respectively. The values in Tables 1 and 2 were the conditions used in
modeling the geometries and numerically calculating the models, respectively. A constant
isothermal effusion wall measured with embedded thermocouples as Figure 2 (a & b) shows
and as used by Al Dabagh et al. (1990), Nazari (1991) and El-Jummah et al. (2016b, 2017)
was also employed in this work. Square arrays of 10 and 5 holes for the impingement wall
and 10 only for both equal and unequal n of the effusion wall were investigated. For optimum
performance of impingement/effusion cooling, the effusion wall should have a low blowing
ratio (BR), which means a low jet velocity and this in turn requires a low relative pressure
loss P/P. The design requirement for this is a low X/D with relatively large effusion holes.
In contrast, the impingement wall requires a high jet velocity for best effusion wall backside
cooling and this requires a higher X/D (El-Jummah et al., 2017). The combination of the
impingement X/D of 11 and effusion X/D of 4.7 walls, has been found experimentally to be a
practical combination in terms of the overall cooling effectiveness with a practical wall
pressure loss at low overall coolant mass flux (Al Dabagh et al., 1990). The aim was to
develop CHT CFD design procedures that could be used in combustor and turbine blade heat
transfer optimisation and is the reason for this work.

( ) ( ) ( )

( ) ( )

(a) The experimental test rig (b) The flat wall cooling setups and the flow scheme

Figure 1: Impingement/effusion heat transfer cooling systems

186
Arid Zone Journal of Engineering, Technology and Environment, April, 2017; Vol. 13(2):184-196
ISSN 1596-2490; e-ISSN 2545-5818; www.azojete.com.ng

Table 1: The Models Geometrical Parameters Table 2: The Computational Flow Conditions
Variables Impingement Effusion n ( m-2)
n ( m-2) 4306 1076 4306 4306 Impingement Effusion
Array 1010 55 1010 10 10 G
2
(kg/sm ) 4306 1076 4306 4306
D (mm) 1.41 2.88 3.27 3.27 Vj Rej Vj Rej Vh Reh Vh Reh
X (mm) 15.24 30.48 15.24 15.24 (m/s) ( 103) (m/s) ( 103) (m/s) ( 103) (m/s) (103)
L/D 4.50 2.21 1.94 1.94 0.94 114.5 10.98 109.9 21.53 21.7 4.83 21.7 4.83
X/D 10.80 10.58 4.7 4.7 0.77 93.4 8.96 89.7 17.56 17.7 3.94 17.7 3.94
Z/D 5.67 2.70 2.40 2.40 0.63 76.3 7.32 73.3 14.35 14.5 3.22 14.5 3.22
A (%) 0.67 0.70 3.61 3.61 0.50 60.6 5.81 58.2 11.39 11.5 2.55 11.5 2.55
0.30 36.4 3.49 34.9 6.84 6.9 1.53 6.9 1.53
0.10 12.2 1.17 11.7 2.29 2.3 0.51 2.3 0.52

Figure 2: Impingement/effusion test plates with their thermocouple location


1.1 Review of Studies on Impingement/Effusion Experimental Techniques
Andrews et al. (1988) and Andrews and Nazari (1999) have determined the overall cooling
effectiveness for impingement/effusion cooling for equal number of impingement and
effusion holes. They used a hot gas cross-flow test rig with 770 K gas temperature and room
temperature effusion air, which gives a density ratio of 2.7. This same effusion cooling air
feed which is typical of gas turbine combustor cooling was used in the present work. Also
shown in their work, was that at low coolant mass flow rates, the internal wall cooling
dominated the overall cooling. This conclusion was also found in a recent CHT CFD model
of an impingement/effusion cooling hot gas cross-flow effusion wall design by El-Jummah et
al. (2016b), for equal and unequal geometries. Figure 1a show the experimental equipment
that was used previously (Al Dabagh et al., 1990, El-Jummah et al., 2017) and is the same
facility that has been modelled for the current investigations using CFD CHT models. It
consists of an air supply to a thermally insulated plenum chamber feed to the impingement
holes.
The experimental test rig (El-Jummah et al., 2017) also consists of 152.4 mm square Nimonic
- 75 impingement hole test wall that was bolted to the plenum chamber. There was an 8 mm
impingement gap formed using a 8 mm thick Teflon spacer flange between the impingement
and the effusion (or target) walls. The Teflon spacer had a low thermal conductivity and this
minimized conductive heat transfer between the two metal walls. The target effusion wall

187
El-jummah et al.: Influence of thermal gradient on gas turbine combustor wall using
impingement/effusion cooling techniques: CHT CFD predictions. AZOJETE, 13(2):184-196 ISSN
1596-2490; e-ISSN 2545-5818, www.azojete.com.ng

was also Nimonic - 75, which is a common combustor metal wall material. Both walls were
instrumented with six grounded junction mineral insulated thermocouples, spaced at 25.4 mm
intervals (El-Jummah, 2015), brazed to the target Nimonic - 75 effusion wall, with the
thermocouple junction flush with the impingement jet wall. The transient response of these
thermocouples was used to determine the locally surface averaged heat transfer coefficient
(HTC) h using the lumped capacity method, which is also the method used in this current
investigation. The thermocouples were located on the centreline between the impingement
and effusion holes and thus were at the most remote location relative to the high local
convective cooling of the effusion holes (Andrews and Nazari, 1999). The number of holes
was a key variable, as for the same wall porosity A and relative pressure loss P/P, which
this work investigate using CFD CHT.

1.2 The Experimental Test Walls and Imposed Hot-side Conditions


Conduction heat transfer within the wall smoothens out the strong gradients in the surface
convective heat transfer, as the Biot numbers Bi for all conditions are less than 0.2. The
thermocouples thus measures the surface averaged temperature for effusion cooling only and
are usually located at the lowest local convective heat transfer position and thus the results
are expected to be slightly low as the thermocouples are placed in the metal on the centreline
between the effusion holes (Andrews et al., 1985). The target wall thermocouples as in
Figure 2 (El-Jummah, 2015) have been used to determine the locally surface averaged h
using the lumped capacitance method (Abdul Husain and Andrews, 1990). The thermally
insulated wall has been electrically heated in the absence of any impingement coolant flow,
to about 80 oC and then the air plenum and test wall hoisted off the heater and the
impingement/effusion air flow established. The transient cooling of the target wall by the
impingement flow enabled the surface averaged h to be determined. In this CHT CFD
prediction, the wall temperature on the heated side of the wall was held at a constant 80 oC, as
it would be on a steady state heat transfer test rig.

2. Computational Methods
2.1. Computational Grid Model
Figure 3 (a & b) are the control volume for the 4306/4306 and 1076/4306 geometries, which
were used to model the grid geometries shown in Figure 4 (a & b), respectively and were
based on the values shown in Table 1 above. ANSYS ICEM meshing tool (El-Jummah et al.,
2015b, 2016b) was used in the grid generation processes and for each ratio of n, six
geometries were modelled which represent the G values shown in Table 2. In each n case and
for each G, the number of cells generated critically depends on the value of G, hence, the
higher the G value implies the higher the total number of cells generated in the model and
was also dependant on the increased number of holes n in each case. Also, for the cells to
capture the effects of flow that includes flow separation, recirculation and reattachment, cells
refinement has to be applied in the hole as the flow velocities were increased which increases
the quantity of cells. This indicates that the n ratio of 1076/4306 with G of 0.94 kg/sm2 has
the highest number of cells generated and n of 4306/4306 with G of 0.1 kg/sm2 has the lowest

188
Arid Zone Journal of Engineering, Technology and Environment, April, 2017; Vol. 13(2):184-196
ISSN 1596-2490; e-ISSN 2545-5818; www.azojete.com.ng

number of cells in the geometries modelled. In addition to those, the X/D and the size of the
plenum chamber also contributes to the increased size of the cells, but as the flow in the
plenum is expected to be uniform, increased here is insignificant.
2.1. Numerical Analysis
ANSYS Fluent CFD tool was used in carrying out the numerical calculations for the
geometries modelled shown in Figure 4 using the flow conditions that were listed in Table 2.
The boundary conditions used include: an imposed hot side wall temperature Tw of 360 K,
which was similar to that used experimentally, density of 1.225 kg/m3, kinematic viscosity
of 1.476 10-5 m2/s and varied plenum inlet velocities (G/). The CFD CHT computations
were carried out using the wall function standard k - turbulence model that have been
validated (El-Jummah et al., 2015b), for which the first cell size near the target wall was kept
at y+ ~ 35 for all the G values. Table 2 show that at the lowest G of 0.1 kg/sm2, the Reynolds
number Re in the impingement holes were laminar and for all the effusion holes the Re were
also laminar which were for most G. However, the sharp edged inlet to the holes would create
flow separation and turbulence hence the flow was treated as turbulent (y+ ~ 35) in the
current CFD CHT investigations.

Figure 3: Symmetrical elements of the computational domain

Figure 4: The impingement/effusion model grid geometries

189
El-jummah et al.: Influence of thermal gradient on gas turbine combustor wall using
impingement/effusion cooling techniques: CHT CFD predictions. AZOJETE, 13(2):184-196 ISSN
1596-2490; e-ISSN 2545-5818, www.azojete.com.ng

( )
( )

( )

( )

( )
( )
( )

The application of Equations 3 - 6 in the computational tool shows the significance of


predicting at strategic zones the conjugate heat transfer data. Equation 3 helps in predicting
both local or surface average cooling heat transfer coefficient on the impingement and
effusion walls, while Equation 4 predict the hole wall data. Equation 5 normalises the heat
transfer data on the surface walls, which also show the influence of cooling effectiveness.
Higher values indicates better cooling and lower indicates worse cooling and too high could
affects the Biot number, hence the wall lumped capacity. Equation 6 helps to determine the
gap flow and wall temperatures; it helps one to know the quantity of heat being extracted
from the hot side of the effusion (or target) wall.

3. Results and Discussion


3.1. The Predicted CFD Validation Case
Equation 3 was used in predicting the surface average heat transfer coefficient (HTC) h on
the effusion wall and was combined with Equation 4 which predicted the effusion hole
surface hX (El-Jummah et al., 2014, 2015a). The use of Equation 4 was in order to calculate
the correct surface h (El-Jummah et al., 2015b) in the hole and the addition of the predicted h
and hx gave the overall surface h as Figure 5a show. The use of Equation 4, gives a precise
average h in the hole cross sectional area over X2 region, as the h in Equation 3 is for the
entire test walls area inputs in the CFD. The h used in Equation 3 which normalises and
lumped up the cooling surface average NuPr-0.33 predictions was further used and was found
from the Nusselt number Nu of Equation 5 (El-Jummah et al., 2014) and the results is shown
in Figure 5b. Both Figures 5a and b were used as validation case in order to justify the
objective of the present CHT CFD design optimization procedures. El-Jummah et al., (2016b,
2017) showed that the reason for these HTC additions was based on the diagonal flow
velocity through the effusion holes that was shown in Figure 6i (a & b). The surface average
Nu was determined over the area X2 cooled by the impingement jets while the surface
average hx was the area X2 of the effusion hole, as in both cases the cooling was based on the
holes in test walls. Experimentally, the work by Andrews et al., (1985, 1988) and El-Jummah
et al. (2017) applied the methods used Equations 3 - 5 in measuring h + hx and Nu data, these
were compared as a function of G and Re in Figure 5 (a & b) of the present predictions.
The comparison of the predicted CHT CFD and the measured data showed good agreement

190
Arid Zone Journal of Engineering, Technology and Environment, April, 2017; Vol. 13(2):184-196
ISSN 1596-2490; e-ISSN 2545-5818; www.azojete.com.ng

(a) Influence of HTC h on G (b) Influence of Nusselt number Nu on Re


Figure 5: Comparison of CFD predictions with measured data for both n model geometries

and within the error bars for all the geometries modelled. The best agreement was for the
4306/4306 geometries as all the data were predicted closely with the measured one for both h
and Nu. But for 1076/4306 geometries, there were slightly higher predictions at high G or Re
and slightly low at low G or Re but both were within the acceptable error of 10 %. This
indicates that the present CFD tool could be good optimization software for the design
analysis of impingement/effusion system.

(ii) Predicted contour plots of TKE, k (m2/s2)

(iii) Predicted contour plots of Nusselt number


Noting that test walls are the effusion or target wall
(i) Predicted velocity (m/s) pathlines in the gap
Figure 6: Comparison of predicted cooling aerodynamics for varied n at G of 0.94 kg/sm2

191
El-jummah et al.: Influence of thermal gradient on gas turbine combustor wall using
impingement/effusion cooling techniques: CHT CFD predictions. AZOJETE, 13(2):184-196 ISSN
1596-2490; e-ISSN 2545-5818, www.azojete.com.ng

3.2. Effects of Velocity and Turbulence on Heat Transfer


Figure 6i (a) show the velocity pathline of 4306/4306 and 6i (b) of 1076/4306 geometries for
G of 0.5 kg/sm2 in a 3D format. Both show that the air jet interaction of the impingement
flow (El-Jummah et al., 2016b) disappears with the effusion hole, which is more for
1076/4306 than for 4306/4306 and is because of fewer jets hole influence in 1076/4306. This
is the reason why in Figure 6ii (a) the pick turbulence is at the jets point and is higher than the
pick for 4306/4306 of Figure 6ii (b), hence the cooling effects at the effusion wall surface
also follow similar trend as that of the predicted Nusselt number Nu of Figure 6iii (a & b)
show. Also, this should be partially because of differences in Reynolds number, as 4306/4306
has lower jets Reynolds number than 1076/4306 and it significantly controls the Nu.
Unfortunately, most investigators (Al Dabagh et al., 1990, Cho and Rhee, 2001, Cho et al.,
2008, El-Jummah, 2015, Hollworth et al., 1983, Hong et al., 2007, Miller et al., 2014, Rhee et
al., 2003, Rhee et al., 2004), only show the aerodynamics influence of an
impingement/effusion cooling system in a 2D plane, which could be inadequate for better
revelation of data. However as the geometry is usually complex, the aerodynamics could
better be shown in a 3D (Figure 6ia).

(i) Effects of thermal gradients in the gap and test walls (ii) Effects of HTC on impingement & effusion walls
Figure 7: Comparison of predicted influence of cooling for varied n at G of 0.94 kg/sm2

(i) Surface thermal gradients on the effusion walls (ii) Thermal gradients in the eff. walls for G of 0.94 kg/sm2
Figure 8: Comparison of CFD predicted surface and locally T* for varied n model geometries

192
Arid Zone Journal of Engineering, Technology and Environment, April, 2017; Vol. 13(2):184-196
ISSN 1596-2490; e-ISSN 2545-5818; www.azojete.com.ng

The sucking of impingement air jets flow by the effusion holes helps to increase the heat
transfer as heat in the hole are also being extracted plus the wall surface extraction, this
enhances the heat transfer on the effusion wall backside. Figure 5a shows that jets sucking
could only be enhanced if the number and size of the effusion holes are more and smaller
than the impingement holes, which could be the weakness in 1076/4306 as the HTC was
lower than that of 4306/4306 even though not significant. Figure 6i (a & b) shows that the
effects of reversed jets on the impingement wall could affect the effusion wall cooling as the
heat that was sucked in the jets is now being used to heat the jets wall. This assertion is
clearly seen in Figure 7i (a & b) which was predicted using Equation 6 as normalized
temperature T*, it shows that jets reversal heats up the impingement wall as Figure 7i (a)
better shows and as Figure 7ii also shows. Figure 7ii reveals that for both the n geometries
significant heating (or HTC, h) of the impingement jet wall occurs to about ~ 20 % of the
effusion wall, which affects the cooling effects of the test wall. Even though, the trend of the
heating of the jet wall of both n is the same but the unequal n show lower heating than does
the equal one and this was attributed to the sucking effects of the effusion holes. Figure 7i
shows that the heating effects on the impingement wall for n of 4306/4306 is uniformly
distributed, which may be the reason why the surface average HTC for this geometry is
higher than for the 1076/4306 geometry.

3.3. Thermal Gradients in the Gap and on the Effusion Test Walls
Equation 6 shows the normalized temperature T* and have been applied in predicting the
thermal gradients in the impingement gap airflow and in the test walls (El-Jummah et al.,
2016a, b, 2017). Figure 7i shows the contour plots thermal gradient in the plane between the
jets points and along the effusion hole centre for both equal and unequal n. Figure 7i (a)
shows that thermal gradient in the effusion wall was the influence of the jets, as the wall in
the jets plane show better cooling, while that in the plane of the effusion hole, the cooling is
mostly sucked and concentrated at the hole boundaries. But in the impingement gap, the
extracted heat for both plane show uniform cooling and this is expected to heat the
impingement wall uniformly. Also, for the unequal n of Figure 7i (b), severe heat was
reversed back to the impingement wall for both the planes: even though there was significant
cooling in the effusion holes which mostly occupied the hole and should be the reason why
the surface average normalized T* for this n is higher than for equal n in Figure 8i.
Figure 8 shows the thermal gradient as a function of G on the effusion wall surface and in the
effusion wall. Figure 8i gave the surface average T* that excludes the hole surface and 8ii is
the local T* within the wall that was between the jets and also in line with the jets. Figure 8i
show that 4306/4306 geometry has better cooling overall at the surface, which justifies that
the sucking of the coolant air in the hole is not much, this was the opposite for the 1076/4306
geometry. To further show the influence of coolant sucking in the effusion: Figure 8ii show
that the locally T* for 1076/4306 n between jets was the highest but in line with the jets
whereby the sucking effect was not pronounced show lower thermal gradients. But for the
equal n geometry there was similar cooling effects shown for the thermal gradients between
and in line with the jets, this certifies that with impingement jet serving one effusion hole, the

193
El-jummah et al.: Influence of thermal gradient on gas turbine combustor wall using
impingement/effusion cooling techniques: CHT CFD predictions. AZOJETE, 13(2):184-196 ISSN
1596-2490; e-ISSN 2545-5818, www.azojete.com.ng

sucking effect was not severe. Therefore, it concludes that the higher the HTC the lower the
thermal gradients, which signifies that cooling is effective for gas turbine internal wall.

4. Conclusion
The predicted surface average heat transfer coefficient and normalized Nusselt number
agreed with the impingement/effusion internal wall heat transfer measured data, for both
equal and unequal n model geometries of all G and are within the measured uncertainty. This
indicates that the CHT CFD numerical procedures were adequate for the design of gas
turbines cooling typical of the impingement/effusion techniques.
The 1076/4306 combination shows that the heat transfer was reduced by about 10 % as
compared with 4306/4306 number of holes. This show that reducing the number of
impingement jet has little benefit as compared with using n of the 4306 m-2 impingement and
effusion holes.
For the 4306/4306 geometries, the overall internal wall cooling were worse affected by
heating of the impingement jet wall, which could be the reason why the unequal n
impingement/effusion cooling showed closely related cooling with equal n. This was the
reason why the effusion wall thermal gradients for the 4306/4306 impingement/effusion
cooling gave lowest surface average value, which is the same within the solid wall between
the jets, even though locally. But in-line with the jets, was the reversed as the impingement
jets pressure for the 1076 m-2 was higher than that for the 4306 m-2 hole density and so the
locally thermal gradient was higher for 4306/4306.
Nomenclature
A Hole porosity = {(/4) D2}/X2 Vj Impingement jet mean velocity, m/s
D Impingement air hole diameter, m Vh Effusion hole mean velocity, m/s
G Air coolant Mass flux, kg/sm2 Kinematic viscosity, m2/s
h Heat transfer coeff. (HTC), W/m2K X Hole to hole pitch, m
kf Thermal conductivity of fluid, W/mK Z Plate to plate gap, m
L Test wall metal thickness, m Subscripts
q'' Heat flux, W/m2
n Number of jet or hole/unit s. area, m-2 h Hole
N Number of upstream rows of holes j Jet
Nu Nusselt Number w Wall
Density of air, kg/m3 Coolant
P Impingement wall pressure loss, Pa f Fluid
P Coolant supply static pressure ( 1bar) m Mean
Pr Prandtl number L Local
Re Reynolds number X Pitch
T Coolant temperature, 288K
T* Normalized mean temperature

References
Abdul Husain, RAA and Andrews, GE 1990. Full Coverage Impingement Heat Transfer at
High Temperature. ASME Paper 90-GT-285: 1-12.

194
Arid Zone Journal of Engineering, Technology and Environment, April, 2017; Vol. 13(2):184-196
ISSN 1596-2490; e-ISSN 2545-5818; www.azojete.com.ng

Al Dabagh, AM., Andrews, GE., Abdul Husain, RAA., Husain, CI., Nazari, A. and Wu, J.
1990. Impingement/ Effusion Cooling: The Influence of Number of Impingement Holes and
Pressure Loss on the Heat Transfer Coefficient. Transactions of the ASME -
Turbomachinery, 112: 467 - 476.
Andrews, GE., Asere, AA., Hussain, CI. and Mkpadi, MC. 1985. Transpiration and
Impingement/Effusion Cooling of Gas Turbine Combustion Chambers. ISABE and AIAA
7th Propulsion Joint Specialist Conference, ISABE 85-7095: 794 - 803.
Andrews, GE., Asere, AA., Hussain, CI., Mkpadi, MC. and Nazari, A. 1988.
Impingement/Effusion Cooling: Overall Wall Heat Transfer. Proceedings of the ASME
International Gas Turbine and Aeroengine Congress, 88-GT-290: 1 - 9.
Andrews, GE. and Nazari, A. 1999. Impingement/Effusion Cooling: Influence of Number of
Holes on the Cooling Effectiveness for an Impingement X/D of 10.5 and Effusion X/D of 7.0.
Proceedings of the GTSJ International Gas Turbine Congress, Vol. II, IGTC TS-51: 639 -
646.
Cho, HH. and Rhee, DH. 2001. Local Heat/Mass Transfer Measurement on the Effusion Plate
in Impingement/Effusion Cooling Systems. Transactions of the ASME - Turbomachinery,
123: 601 - 608.
Cho, HH., Rhee, DH. and Goldstein, RJ 2008. Effects of Hole Arrangements on Local
Heat/Mass Transfer for Impingement/Effusion Cooling With Small Hole Spacing.
Transactions of the ASME - Turbomachinery, 130: 1 - 11.
El-Jummah, AM. 2015. Impingement and Impingement/Effusion Cooling of Gas Turbine
Components: Conjugate Heat Transfer Predictions. The White Rose UK: A Ph. D Thesis,
University of Leeds.
El-Jummah, AM., Abdul Hussain, RAA., Andrews, GE. and Staggs, JEJ. 2014. Conjugate
Heat Transfer Computational Fluid Dynamic Predictions of Impingement Heat Transfer: The
Influence of Hole Pitch to Diameter Ratio X/D at Constant Impingement Gap Z. Transactins
of the ASME - Turbomachinery, 136 (12): 1 - 16.
El-Jummah, AM., Andrews, GE. and Staggs, JEJ. 2015a. CHT/CFD Predictions of
Impingement Cooling With Four Sided Flow Exit. Proceedings of the ASME Turbo Expo,
GT-42256: 1 - 12.
El-Jummah, AM., Andrews, GE. and Staggs, JEJ. 2015b. Conjugate Heat Transfer CFD
Predictions of Metal Walls with Arrays of Short Holes as used in Impingement and Effusion
Cooling. Proceedings of the GTSJ International Gas Turbine Congress, IGTC TS - 266: 1 - 9.
El-Jummah A. M., Andrews G. E. and Staggs J. E. J. 2016a. Impingement Jet Cooling with
Ribs and Pin Fin Obstacles in Co-flow Configurations: Conjugate Heat Transfer
Computational Fluid Dynamic Predictions. Proceedings of the ASME Turbo Expo:
Turbomachinery Technical Conference and Exposition, GT- 57021: 1 - 15.
El-Jummah, AM., Andrews, GE. and Staggs, JEJ. 2016b. Impingement/Effusion Cooling
Wall Heat Transfer: Conjugate Heat Transfer Computational Fluid Dynamic Predictions.
Proceedings of the ASME Turbo Expo: Turbomachinery Technical Conference &
Exposition, GT-56961: 1 - 14.

195
El-jummah et al.: Influence of thermal gradient on gas turbine combustor wall using
impingement/effusion cooling techniques: CHT CFD predictions. AZOJETE, 13(2):184-196 ISSN
1596-2490; e-ISSN 2545-5818, www.azojete.com.ng

El-Jummah, AM., Nazari, A., Andrews, GE. and Staggs, JEJ. 2017. Impingement/Effusion
Cooling Wall Heat Transfer: Reduced Number of Impingement Jet Holes Relative to the
Effusion Holes. Proceedings of the ASME Turbo Expo: Turbomachinery Technical
Conference and Exposition, GT - 63494: 1 - 14.
Hollworth, BR, Lehmann, G and Rosiczkowski, J 1983. Arrays of Impinging Jets with Spent
Fluid Removal Through Vent Holes on the Target Surface, Part 2: Local Heat transfer.
Transactions of the ASME - Engine Power, 105: 393 - 402.
Hong, SK., Rhee, DH. and Cho, HH. 2007. Effects of Fin Shapes and Arrangements on Heat
Transfer forImpingement/Effusion Cooling With Cross-flow. Transactions of the ASME -
Heat Transfer, 129: 1697 - 1707.
Miller, M., Natsui, G., Ricklick, M. and Kapat, J. 2014. Heat Transfer in a Coupled Impinge-
Effusion Cooling System. Proceedings of the ASME Turbo Expo, GT-26416: 1 - 10.
Nazari, A. 1991. The Impingement/Effusion Cooling of Gas Turbine Combustor Walls. PhD
Thesis, University of Leeds.
Rhee, DH., Choi, JH. and Cho, HH. 2003. Flow and Heat (Mass) Transfer Characteristics in
an Impingement/Effusion Cooling System With Crossflow. Transactions of the ASME
Turbomachinery, 125: 74 - 82.
Rhee, DH., Nam, YW. and Cho, HH. 2004. Local Heat/Mass Transfer With Various Rib
Arrangements in Impingement/Effusion Cooling Systems With Crossflow. Transactions of
the ASME - Turbomachinery, 126: 615 - 626.
Rogers, N., Ren, Z., Buzzard, W., Sweeney, B., Tinker, N., Ligrani, P., Hollingsworth, K.,
Libertore, F., Patel, R., and Moon, H-K., 2016. Effects of Double Wall Cooling
Configuration and Conditions on Performance of Full Coverage Effusion Cooling.
Proceedings of the ASME Turbo Expo GT- 56515: 1 - 10.
Shi, B., Li, J., Li, M., Ren, J. and Jiang, H. 2016. Cooling Effectiveness on a Flat Plate both
Film Cooling and Impingement Cooling in Hot Gas Conditions. Proceedings of the ASME
Turbo Expo GT - 57224: 1 - 10.

196

Das könnte Ihnen auch gefallen