DcfSci J. Vol38, No. 4, October 1988.

pp363-380

Advanced Combustor Liner Cooling Technology for Gas Turbines

Aspi R. Wadia
General Electric Company, Cincinnati, Ohio45215

ABSTRACT

This paper briefly reviews some of the work on advanced liner cooling techniques - specifici~lly laminated porous wall cooling, angled-multihole (effusion) cooling ant1 composite metal matrix liner cooling. The concept definition, heat transfer design procedure and design problems including key materials and fabrication considerations associated with each basic concept will be reviewed. Published rig and engine experience of aircraft engine manufacturers and reseirch organizations will be cited. Some logical extensions of the current liner cooling schemes are suggested for future applications.

. INTRODUCTION
The projected designs for future aircraft engines are very aggressive in terms of thrust-to-weight, rotor speed and temperature goals'. Conventional wall cooling methods (e.g. film convection cooling) are incapable of providing satisfactory durability without using excessive amounts of cooling air, which in turn severely restricts air available for temperature pattern control. Engine envelope demands, especially for small aircraft engines, further burden the designer by mandating foldback (reverse flow) combustor designs, which reduce engine length and weight but substantially increase the combustor surface area to be cooled. Although attention to cooling problems has traditionally been focused on the HP turbine for obvious reasons, there is a need to view combustor liner cooling as a technical problem of equal importance if projected performance goals and durability improvements are t o be achieved.
Received 19 April 1988

363

364

Aspi R U'adia
2. LINER COOI.INC SCHEMES

The principal schemes representing both the current and future candidate approaches to combustor liner wall cooling are shown conceptually in Fig. 1. Simple slot film cooling has received the most attention in terms of application and fundamental studies. While the significant body of research work has greatly enhanced the understanding of film flow behaviour, significant reduction of cooling flows in pure slot film cooled combustors has not been forthcoming. This is attributed in part to the difficulty in maintaining film integrity in a turbulent environment characterized by recirculating flows. Also, the insulating film prin~arilyaffects the convective component of heat transfer and has little effect on the radiative heat transfer from the luminous gas. colladay2 was the first to recognize that significant reductions in coolant flow could be achieved if the heat sink capability inherent to the coolant is more fully utilized in the active mode prior to injection. This led to the development of the film-convection cooling system. Further improvements in cooli~ig usage rcquires a departure to more effective air mass transfer cooling schemes, such as transpiration o r effusion cooling. The heat transfer development work leading to incorporation of such systems into conlbustors is reviewed in the following sections.

3. LAMINATED POROUS WALL COOLING

3.1 General Description
Improvements in metal joining techniques has led to the development of fabricated multiple-laminate porous structures. One such structure, ~ a m i l l o y ~ * , consists of several diffusion-bonded, photoetched metal sheets shown in Fig. 2. This cooling scheme has considerable design flexibility relative to flow resistance control and optinlizatior~ of heat transfer performance. Design variables include hole size and spacing, laminate thickness, number of laminates, pedestal height, pedestal diameter and spacing. Large internal heat transfer surface area to volume ratio and hence improvcd thermal effectiveness levels (defined as actual coolant temperature rise divided by ideal rise) can be achieved by reducing all the dimer~sions that the internal structure approaches so that of a fine capillary mesh. The reduction in the di~nensions however iimited by is passage clogging considerations and by fabricatiorl limitarioos.

3.2 Heat Transfer Analysis

One of the key inputs in the heat transfer analysis of Ittnlinated porous wall structures is the knowledge of the internal (matrix) heat transfer coefficient. 'The heat transfer data is generally obtained by testing simple disk specinle~ls,nlou!ited in a fixture and heated by a high intensity radiant energy source3. Some typical hcat transfer trends for selected I,amilloy configurations releva;,; t:, con~bustorliner
'Lamilloy is a registered trademark of General Motors Corporation.

Cornbustor Liner Coolingfor Gas Turbines

a d s = = 7
SLOT FlLM COOLING ( a1

FILM-CONVECTION COOLING (CONVECTIVE CHANNEL

FILM-CONVECTION COOLING (IMPINGEMENT (c1

&4iFdb&&&
mULTlPLE DISCRETE HOLE FlLM COOLING (EFFUSION COOLING (d

TRANSPIRATION COOLING

(el
HOT GAS CERAMIC

COMPLIANT LAYER -POROUS METAL METAL SHEET

(f

COMPOSITE MATRIX

2in x 2 in CERAMIC BRICKS

COMPOSITE MATRIX

Figure 1

Combustor liner cooling schemcs.

Aspi

H Wadia

standard array

Figure 2.

Typicill g c o r ~ ~ c t r i c r r i ~ n g c ~ n c t f t I i ~ ~ n i ~ ~porous w ; ~ l l ~ ol i~tcd ktrucrurc

a ) THERMAL EFFECTIVENESS DATA

b ) MATRIX TEMPERATURE DATA

11

- 0.4;
0.2

. , . " , . , , , , ,
1.0 2.0 3.0 4.0 5.0 91 OWlNG RATE, G, ILglc .-nt') 6.0

.Cotnbwtor Liner Coolingfor Gas Turbines

application3are shown in Fig. 3. The internal 'heat exchanger' effectiveness parameter (q,.) is especially important since liner cooling requirements will be inversely proportional to it. Fig. 3 indicates the possibility of developing a unique geometric col~figurationthat approaches the ideal transpiration cooling effectiveness of unity. The difference in the performance trends of the configurations is due to differences in the holelpedestal geometry of the specimen.
3.3 Mechanical Design Considerations

While transpiration cooling potentially represents the most thermodynamically efficient approach to combustor cooling, practical implementation of the method has been hampered by limitations of porous materials. The most glaring drawback of this cooling technique has been its lack of structural strength and resistance to oxidation. Considerable difficulty in predicting or controlling local permeability has also been encountered, and susceptibility to foreign particle clogging coritinues to be a problem. Laminated porous wall structures3 have been fabricated from nickel-andcobalt-based superalloys in addition to a full spectrum of other materials, including liaselloy X and Haynes 188. The structure requires very thin sheet stock ranging from 0.01 to 0.025 in.
A typical laminated porous wall structure (Lamilloy) manufa~turing sequence3 is illustrated in Fig. 4. The process is complex and expensive.

3.4 Engine Experience

From the data published in the open literature, two aircraft engine companies appear to.be the leaders in the application of porous wall structures. Allison Gas ~urbines;,~ used Lamilloy combustors for several engine models; has the most recent application being for the T8001ATE109 (LHX) 1200 hp turboprop engine.

368

Aspi R Wadia

Rolls Royce has developed their own technology: using 'Transply' and rig and engine tests have dispelled doubts concefning the blockage of transpiration cooled systems by airborne particles, and Spey combustors incorporating sections of Transply have achieved flight certification for commercial airline operations. 4. ANGLED-MULTIHOLE EFFUSION COOLING
4.1 General Description

A more practical arrangemefit, but one that still approximates a porous wall, is a wall perforated by a large number of small holes as shown in Fig. l(d). Ideally, the individual holes should be large enough to remain free of blockage by dirt, but small enough to prevent excessive penetration of the airjets. The cooling performance is controlled by a relatively few geometric parameters, namely : sheet thickness, hole size, spacing and plunge angle. In theory, the effusion cooling concept can approach transpiration cooling performance in the limit as the hole diameters and spacing-to-diameter ratios are reduced to zero. Parametric studies3 have shown that when 'realistic minimum' hole sizes, spacings and plunge angles are considered effusion cooling performance lies between film-convection cooling and laminated porous wall cooling. The principal advantage of this approach lies in its inherent simplicity, ease of n~anufacturing cost-effectiveness, a major factor under and present economic conditions. 4.2 Heat Transfer Analysis

The modeling of heat transfer in effusion cooling systems is largely empirical. The basic working relationships between cooling flow rate, ,hole geomctry and cooling performance can be established from the work of Wadia and earl^^. Although the work described in ref. 6 is focused on the airfoil leading edge problem, the modeling of internal wall heat transfer processes is equally applicable to the combustor liner problem. Further experiments with respect to the internal wall heat transfer characteristics are underway at the University of Leeds under the direction of ~ n d r e w s and their results confirm the importance of the hole approach heat transfer. The cooling performance trends for an anglcd, multihole wall can best be illustrated by a parametric study presented in 'I'able 1 using a model very similar to that described in ref. 8 with the appropriate boundary conditions. Table 1 shows the cooling flow requirements for several alternate effusion cooling configurations compared with that for a laminated porous wall design. While this comparative study was made3 for the TF41 combustor liner, the results can be considered as at least representative and can be used to iilustrate the influence of the principal geometric variables on effusion cooling performance. The following key observations are summarized below. The best practical effusion geometry (a = 20". d = 0.015 in.) requires about 60
I(,I. ~ . ~ - I I ~ . I I,IIIII~III! flow IIoI~~
,.

t1bi111 IIIIV:, I ~ I I I I I I I ~ I I ~ ~II~ I \v;IIIl 1 i 1 I~III% lilt* ~II I II~; ~ t IIIII'I~!III I I ~ ~ ~ I Ii1lvi111; 111~-w;III t l i i t . h ~ ~ r sIII~.IC;I:,~.~ II IIIII~, s 1I1t- I X IIII~V I(.I~IIIII,IIII.II~:. 11y ;I~I~IIII 11t.1 1 0 t.111 ( ~ : ~ I I ~ ~ } ! I I I - ~ A~ ~ ~ I S I I S SI! ;III(I I! V~*I:,II:, I;, t.11.. ) , '1'111~ I V ~~ I t.1It.t I 01 I I ~ I I ;III}!I(* ~. is

Combustor Liner Cooling for Gas Turbines

Table 1
b

369

Effusion cooling - performance trends (refs. 3 and 7)

2

Configuration
L
A

Description

Cooling flux Hole spacing Number to holes requirement to diameter (per liner) Ibls-in2 ratio
0.010 0.0160 4 .O 7.12 6.812 11.052

Wall temperature gradient-"F

302
145

Laminated porous wall (baseline)

a = 20deg d=0.015 r = 0.060

a = 45 deg d = 0.015 T = 0.060

a = 26 deg d 0.030 r = 0.060 a = 45 deg d = 0.030 r = 0.060 a = 20 drg d = 0.015 r = 0.030

0.0247

5.90

16,117

146

0.0250

5.78

4.199

150

0.0389

4.73

6,255

153

E F
G

0.0214

6.24

14,390

75

a = 45 deg d=0.015 r = 0.030

a = 20deg d = 0.030 r = 0.030 a = 45 deg

0.0332

5.13

21.306

75

0.0332

5.05

5,488

R
78

0.0513

4.14

8.172

d = 0.030 r = 0.030
a - effusion hole plunge angle (measured from horizontal)

d - effusion hole diameter (in.)

r - liner wall thickness (in.)

significant indicating the importance af active cooling in the effusion cooling system. Decreasing the hole angle from 45" to 20" (hole length-to-diameter ratio increases from 5.7 to 11.7) reduces the cooling flow requirements by about 35 per cent (configuration A versus B, C versus D, etc.). The significant cooling performance advantage of small holes is partially offset by the rccluirement to drill significantly srorc l)ol~b (ubout 2-1/2 titiles Inore IIOICS for ~ I I C cast CI~CCI). 1101 tlral tctllpcraturc Wall gradients in effusion-cooled liners are considerably lower than in laminated porous walls with the same overall thickness (configurations A, B, C, D versus L).
1 3 Shaped Holes .

Sll'rl)cJ Itoles lrave been ellcclively used in iurbioc uirloil cooling applicalio~ls lo climinirte the problem of low film cooling effectiveness associ;~tcdwith lift-off. Two l 1 ~ l ; ~ l l ~ I~I ~ * I I Iof i ~ ~ c v e ~ ~ 111111 'lilt-ull' ' il I C i l l u s t ~ ~ ~ l c ~ l 3. d IS t~~~$~l t 111 la~g.

HOLE DETAIL

FOS =FLAT DOWNSTREAM FUS = FLAT UPSTREAM

Figure 5.

Shaped hole design configuratio~~s Goldbtrin (Rcf. 9). (h) Pappel (Ref. 1 0 ) . (a)

The improved performance of shaped holes relative to cylindrical holes is illustrated in Fig. 6. The shaped hole contributes to improved lateral spreading of the greater. coolant flow so that the film cooling effectiveness between holes is sig~~ificantly The larger discharge area of shaped holes reduces the injection velocity of the coolant flow resulting in improved effectiveness. The lower effective blowing ratio causes the flow to stay nearer to the wall with less pcnctration into the mainstream.
XID

CYLINDRICAL

HOLES

SHAPED HOLES

XI9

Figure 6.

Conlparison df cylindrical and shaped hole cooling effectivenc c f f c c f i v c ~ ~ ~b )cc111reJ i ~ i tfill11IWI~I~II~ ( ~s, ~ ~-(~I-I.I~VC*II~-~~,-

:%

(a)

loc

Combustor Liner Cooling for Gas Turbines

37 1

While most published data on effusion liners has been for straight cylindrical holes, it may be possible t improve the thermal performance by using shaped holes. o Trade-off studies between the improvement in performance and manufacturing cost are needed to fully evaluate the shaped hole concept for liner cooling.
4.4 Mechanical Design Considerations

Of the three liner cooling concepts described in this paper, effusion cooling is the simplest and most cost-effective to implement using either laser o r electron beam technology in conjunction with computer controlled manufacturing3. Messer Griesheim's (Germany) Electron Beam Drilling Machine (EBDPULS) is capable of drilling 20 holes (0.02 in. dia, 0.175 in. long) per second through both nickel and cobalt steels at angles down to 20 from the surface. Similar hole drilling speeds and quality are achieved by pulsating low divergence Nd : YAG lasers at very high peak powers with little or no thermal distortion. The manufacturing process is simple, since only a single thickness material need be rolled or formed to the final combustor shape. All holes, including both cooling and combustion air hoies, can then be drilled at thesame time.

4.5 HigIEngine Experience

One of the users of this concept has been Allison Gas Turbines. An excellent back-to-back comparison of the laminated porous wall cooling concept with the effusion cooling concept is presented in the paper by Mongia and eider^. A 135" segment of effusion liner was used for the primary zone wall region of an ATDEIGMASOO reverse-flow combustor manufactured using laminated'porous walls. The effusion patch tested had three-hole densities in equal areas of the segment giving cooling air fluxes of 0.0055,0.007 and 0.01 lb/in2 sec., respectively, at sea level design point. The baseline laminated porous wall configuration has a coolant flux of 0.0056 lblinlsec. Wall thermocouple readings, for the burner tested to a 2598F combustor exit temperature are shown in Fig. 7. This test indicates that for the same coolant flux, the effusion liner runs about 200F hotter than the laminated porous wall configuration, and it takes approximately 45 per cent more coolant flow for the effusion liner to attain the same temperature ;IS the porous wall liner. However, the estimated through-the-wall temperature gradient and the axial temperature gradient in the effusion cooling system is about 75 and 70 per cent less, respectively, relative to the porous wall concept, indicating significant life improvement benefit of the effusion liner.
5. COMPOSITE MATRIX LINER COOLING
5.1 General Description
'Two typical composite nlatrix cooling scl~cnics" r ~ c l i o w ~ ~ 1~1gs. ; s 111 I(l) a~ltlI(g). The canlpliant layer is inclt~dctlin t l ~ cI I \ ; I ~ I ~ Xto iti.1 irs i\ 511iii11 ~c*lic.vil~~: 11it.tl111111 Oclwcc~l ceritnlic and the solid metal support structure. lllc

'-FUEL
n n n n n n n

NOZZLE I TYPICAL) LAM~LLOY0.0056 O TR UE BARREL (UNROLLED)

R O T S U B MO C - , EFFUSION -LAMLLOY 0.0055 .010.6056

COOLANT FLUX (lb/s/~n?) T P O C L SOE WALL OD THERMOCOUPLEo 1 1 22 READINGS FF) 0l 1391

n n n n n n n n n n A n n " n n , l 1420

:=

FUEL NOZZLE (TYPICAL)

l 1148 l 1275 l 1400
- -

l 1600

'-

'

01590

l5 1 11

01250 17 13

C MUT R O TR O B SO UE BARREL (UNROLLED)

B R E OUTLET TEMPERATURE = 2598'F UNR

LAMILLOY

EFFUSION

1001

0.004 0.008 0.012 LOCAL COOLANT FLUX (lb/s/in2) LAMILLOY EFFUSION

Figure 7,

ATDElGMA SOU rcscrvc flow conil>ustortcsl rcsults with I;~mi~iatcd porous wall and effusion cooling".

The

solid surface or

t h e c o n d u c t i o n a n d backside c o n v e c t i o n c o o l i n g

only

design

i s t h e simplest c o n f i g u r a t i o n T h e entire

cooling

process

and t h e

compliant layer t o

design p o i n t as i l l u s t r a t e d in ref. 11.1 i s c o n d u c t i o n from t h e hot side w a l l t h r o u g h t h e c e r a m i c ! the m e t a l shell support.

from a
mechanical
b I

T h e i n t e r r u p t e d surface may be required below

or

t r a n s p i r a t i o n c o o l e d concept as i l l u s t r a t e d

if

the ceramic-compliant layer interface is t o b e which, oxidation

in ref. 4' m a i n t a i n e d a t or

1750F, Above

of

t h e c o m p l i a n t layel., espcci;~lly t h e
. ~ I ~ O ~ ) C t~l Ic i

portion
\ i g ~ i ~i c l ; ~

i ~ t l i ; i t * c ~ ~ t i t c CC.I;IIII~C, 10 l a very small a m o u n t temperature can

w i l l I)~.COIIIL- i l p1oIi1c111. ;I ~ i ~ ; ~ ~ i t . ; i ~ ~ ~WIIII

of

t h r o u g h f l o w air,

the ceramic-conipliilnt layer interface

he more easily

r n ; i i ~ ~ l ; i i n e t;it t h c d c s i r c d I c v c l . l

Combustor Liner Cooling for Gas Tlrrbines

5.2 Heat Transfer Analysis

The convection-cooling concept can be optimized using one-dimensional heat balance equations at selected discrete locations (primary zone, dilution zone. etc.) around the combustor liner. The objective here is to design the cooling configuration (c.g.,ceramic coating thickness, compliant layer thickness, etc.) that satisfy the thermal durability requirements (e.g. interface temperatures between pad and ceramic coating. thcrmal wall gradients, etc.) with minimum cooling airflow. The boundary conditions are empirical in nature. The selection of the ceramic and compliant pad thickness iscritical to the success of the convection cooled concept. Using a typical geometry and airflow distribution far a NASA reverse flow combustor (Fig. 3 in ref. 12), the effect of wall thickness on the temperature distribution through the lower walls of the combustor liner for the primary zone was calculated using the method proposed in ref. 13. The limiting hot streak temperature was assumed to be stoichiometric. For constant metal structure and ceramic coating thicknesses, Fig. 8 shows that the ceramiclpad interface temperature increases by about 200F, when the pad thickness is increased from 0.025 to 0.1 in. However, for constant metal structure and pad thicknesses, Fig. 9 shows that the ceramiclpad interface temperature decreases by about 500F by increasing the ceramic thickness from 0.025 to 0.1 in. From an idealistic heat transfer viewpoint,

I-

1850
1650

0.025 0.050 0.075 PAD THICKNESS (in)

0.100

A = W C ceramic wall temperature = TWC-B ceramiclpad interface temper.ature 0 = 7WB-M padlmetal interface temperature X = 7WM metal wall temperature t,,,,, = 0.04 in. t , ,=0.1 in. ,
Figure 8. Effect of compliant layer thickness on the wall temperatures fc

convection cooled design.

k !
I -

2.200 1950

1700 1
1
I

1450 0

0.025 0.050 0.075 COATING THICKNESS (in)

01 0 .0

A = TWC ceranlic wall temperature = TWC-B ceramiclpadinterface temperature 0 = TWB-M padlmetal interface temperature x = TWM metal wall ternperature t ,,, = 0.04in. ,,,t = 0.05 in.
Figore 9. Effect of ccrilniic coi~tingtliicknc\\ on t h ~ willl tcl1ipcr;lturc ror the convcctiot~-cool4 design.

COATING THICKNESS Zone 1, lowerwall, W = 0.78% , A = W C coating hot side wall temperature = W C - B ceramiclpad interface temperature 0 = W M coolant side wall temperature
Figure 10.

Effect f coi~ri~ig Iliickncss on the Inycr tetiipcr;rturcs li)r co~npositc niittri> porous wall cooling (tcsig~i'.

Combustor Liner Cooling for Go3 Tnrhines

the convection-cooledconcept would require the rninin~um possible pad thickness and the nlaximum possible coating thickness. However, thicker coatings will be subjccted to greater wall thermal gradients, including cracks much sooner due to cyclic variations, reducing the overall life of the component. The composite matrix transpiration-cooled concept can atso be evaluated by using the procedure developed in ref. 8 and assuming a thexmal effectiveness level of 0.5. For the GMASOO combustor liner, Mongia and ReiderQising this analytical technique have shown a 38.7 per cent reduction in coolant flow relative to a similar laminated porous wall design. Figure 10 shows the effect of coaling thickness on the padlcerarnic interface tcnlpcrature in the primary zone for the transpiration-cooled concept. Increesing the coating thickness from 0 025 to 0.1 in. reduces the padJceramic interface tenlperature from 21 10 to 1749OFwhilethe gradient across the coating increases from 483 to 1803F. Figure 11 illustrates the asymptotic effect of variations in the flow rate on the wall temperatures in the primary zone of the transpiration-c-oled concept.
5.3 Mechanical Design Considel-ations

Mechanical design studies of the composite matrix concept are described c l s e w t ~ e r e ~ ~ ~ ' ~ studies outline the application of the ceramics to the combustor These . liners by plasma spraying Yttria Stabilized Zirconin (YSZ) on a conlpliant nickel alloy substrate. The compliant metal substrate is designed to yield at relatively low levels of stress, thereby absorbing the differential expansion which develops between the mclal and the ceramic as the material is heated. 'The compliant mctal substrate is made from randomly oriented fibers which are sintered for strength. A large number of physical and mechanical properties are obtainable by controlling product alloy, fiher diameter, porosity and sintering conditions.

FLOW RATE ( X )

Zone 1, lower wall t,= 0.1 in., t, = 0.05in.. t , W = 1 1.46Ibls ,

0.04 i?.

n=
0
=

A = NVC'coating hot side wail temperature

TWC-I3 padlceramic interface temperstyre M M coolant side wall temperature

Flgure 11. Effect of variation in the flow rate on the layer temperatures of a NASA type reserve flow annular combustor linear wailx.

5.4 RigIEngine Experience

NASAtLewis has provided the leadership for the research"." and devell of the composite matrix liner. Using Brunswich Technetic's (Florida) p BRUNSBOND Hoskins-875 compliant pad, an experimental combustor was fat and tested" at NASA. The composite consisted of a 0.025 in. thick Hastelloy : substrate, 0.06 in. thick compliant pad and 0.06 in. thick YSZ ceramic coatin; convective coding was used during the test. The test results show that the composite matrix liner required 35 per ct coolant flow than a similar laminated porous wall liner. The outer liner appe, indicate good short term durability with no indication of spalling while the inn, did encounter spalling of the ceramic as a result of stresses within the coating The combustor was capable of providing higher exit temperatures than cu possible with conventional cooling schemes and further increases in temperatti be achieved by incorporating transpiration cooling to the metal ceramic interf
6. O'T11ER CONCEPTS

Liner cooling concept! section.

her than those described above are

hrieflv dtcd

To meet the challenging durability goals of the NASAJE' combustor prc General Electric Company and Pratt, and Whitney have devised new configurations which incorporate advanced film plts impingement cooling. GE's patented design concept'6 utilizes a double-walled 'shingle liner' that co of a load carrying 3600 turning which supports individual heat shield or shingles shingles are segmented axially and circumferentially to reduce stress and provide life. The support liner, in addition to supporting shingles, provides impinge cooling to the shingle. Details of the 'support foot' configuration and the method to control leakage between adjacent shingles is illustrated in great detail in ref

HOT GAS

COOLING FLOW

-\:I

H
Figure 12. Segmented, counter-parallel FINWAL.1. cooling scheme'

Combustor Liner Coolingfor Gas Turbines

Pratt and Whitney's patented design concept1' called 'FINWALL' also has a segmented construction and counter-parallel wall cooling. Fig. 12 shows the convectivelfilm cooling technique used by the liner. Airflow enters through slots in the cold wall and flows upstream and downstream in discrete cooling passages. The coolant is discharged on the hot side to form an additional protective film. Coniplete details of the application of this technique including test results are presented in ref. 18. Garrett's advanced liner cooling concepts'y are summarized in Fig. 13. Their experiments indicate that cold side convection, augmented by rectangular fins and combined with film cooling (Fig. 13(a)) offers the largest potential reduction in coolant-flux requirements relative to the other configurations in Fig. 13. ROUGH SURFACE/FILM

HOT GAS

(TWO TEST PANELS)

(a)

IMPINGEMENT/EFFUSION

(b) ANGLED EFFUSION COOLANT

SMOOTH SURFACE/FILM

BASELINE
COOLANT

HOT GAS
Figure 13.

HOT GAS

(el
Liner cooling configurations testcd at GarrettiY.

Aspi R Wadia
7. SUMMARY

Figure 14 summarizes4 the relative cooling air, requiremknt and some limi quantity manufacturing ;costs of the combustor liners reviewed in this pap Technically, the composite matrix liner demands the least amount of cooling air 2 is most suitable for high performance engines where cost may not Iw iin issue. Effus~ cooling appears to be the most iittractivc from a cost-cffcctivc viewpoint and s~tggcsts atlcquacy rklat its application i l l many of the currcat production c~tgiilcs to present design goals. For the near term growth, laminated porous wail and conlposite matrix lin will.suffice. In the long term, cqrbon-carbon appears to be the only lightweight mate1 capable of withstanding close to stoichiometric conditions. iiowevcr, the oxidati and inspection problems associated with carbon-carbon materials rnust be addrcs: before it is viable for usc in combustor liners.

FlLM
FlLM CONVECTION

-----

-----I

EFFUSION LAMILLOY COMPOSITE MATRIX

0 0.5 1.o RELATIVE COOLING AIR REQUIREMENT

FILM

EFFUSION LAMILLOY COMPOSITE MATRIX

m
0
1.O

2.0

RELATIVE PRODUCTION C O S T S ( LIMiTED QUANTITY)

Figure 14. Cooling air requirements and production cost comparison of co~nbustor liner cooling schemes4.

Combustor Liner Cooling for Gas Turbines

379

The technology status of the three liner concepts can bc summarized as follows : Transpiration designs are well developed and in use; effusion designs iirr being developed and are close to being applied; and composite matrix liners are largely in the experimental stage.

ACKNOWLEDGEMENTS would like to thank General Electric Company for permission to publish this paper. REFERENCES Cifone, A.J. & Krueger, E.L., Combustion. technology : A navy perspective, AIAA Paper No. 85-1400, July 1988.
2. Colladay, R.S., Importance of combining convection with film cooling, AIAA

Paper No. 72-8, January 1972. 3. Nealy, D.A., Reider, S.B. & Mongia, H.C., Alternate cooling configuration for gas turbine combustion systems, AGARD, Propulsion and Energetics Panel, Paper No. 25, May 1985.
4. Mongia, H.C. & Reider, S.B., Allison combustiorl research and development activities, AIAA Paper No. 85-1402, July 1985.

5. Lefebvre, A.H., Gas Turbine Combustion, (McGraw-Hill, New York), 1983.

6. Wadia, A.R. & Nealy, D.A., to be published in ASME Trans, Journal of Turbomachinery, (1988). 7. Andrews, G.E., Alikhanizadeh, M., Asere, A.A., Hussian, C.I., Azari, M.S.K. & Mkpadi, M.C., ASME Trans. Journal of Turbomachinery, 108 (1987), 283-289.
8. Wadia, A.R. & Nealy, D.A., Development of an engineering design model for airfoil leading edge film cooling, ASME Paper No. 85-GT-120, March 1985.

9. Goldstein, R.J., Eckert, E.R.G. & Burgraff, F., Intl. J. of Heat and Mass Transfer, 17 (1974), 595-607. 10. Papell, S.S., Vortex Generating Flow Passage Design for Increased Film Cooling Effectiveness, US Patent No. 4, 529, 358, 1985. Acosta, W.A. & Norgren, C.T., Small gas turbine combustor experimental study

- compliant metallceramic liner and performance evaluation, AIAA Paper No.

86-1452, June 1986.

12. Norgren, C.T. & Riddlebaugh, S.M., Advanced liner cooling techniques for gas turbine combustors, AIAA Paper No. 85-1290, July 1985.
13. McAdams, W.H., Heat Transmission, (McGraw-Hill, New York), 1954.
14. Venkat, R.S. & Roffe, G., Testing of Felt-Ceramic Materials for Combustor Applications, NASA-CR 168103, April 1983.

15. Ercegoric, D.B., Walker, C.L. & Norgen, C.T., Ceramic composite liner material for giis turbine combustors, AIAA Paper No. 84-0363, January 1984.

380

Aspi R Wadia

16 Burrus, D.L., Chahrour, C.A., Foltz, H.L., Sabla, P.E., Seto, S.P. & Taylor J.R., E' Combustion System Component Technology Development Report ! NASA-R82AEB401, (NAS3-20643). November 1982.

17. Greene, W., Tanrikut, S. & Sokolowski, D., Devclopmcnt and opcratin
characteristics of an advanced two stage combustor, AlAA Paper No. 82-0191 January 1982. 18. Tanrikut, S., Marshall, R.L. & Sokolowski, D. E.. Irnprovcd conihusto durability -segmented approach with advanced cooling techniques, AlAA Papc NO. 81-1356, July 1981.

19. Meyers, G., Van Der Geest, J., Sanborn, J. & Davis, F.. comparison of avnnccd cooling concepts using color therrnography, AIAA Paper No. 85-1289, July 1985.1