Lefebvre A.H., Ballal D.R. Gas Turbine Combustion: Alternative Fuels and Emissions

Подождите немного. Документ загружается.

340 Gas Turbine Combustion: Alternative Fuels and Emissions, Third Edition

We begin the heat-transfer calculation by computing C

. Since m > 1.3, we use

Equation 8.44:

036

010=













−

.(Re )()

w,adw

0010

0 157

0 0261

3261018 1176

508036

(. )(

×−

−

W/m

1926 1176

)

().=−T

From the previous example,

794 460 0 0032=−,. ,

W/m

229

100

13 715=













−.,,

W/m

921 810 400=−

W/m,,

KTT

12 12

21667

−

=−,( ).

Substitution of these expressions for C

, R

, C

, R

, and K

1–2

into Equation 8.25 yields

KK.

1283 1265==

A comparison of the results for T

with and without ﬁlm cooling shows that,

for the conditions chosen in this example, the use of ﬁlm-cooling air produces a

decrease in

of 357 K. It may also be of interest to note that, for the same condi-

tions, an increase in luminosity factor from 1.7 to 4.0 would produce an increase in

of approximately 80 K, whereas a decrease in ﬁlm-cooling effectiveness from

0.789 to 0.700 would increase T

by approximately 66 K.

The results of numerous calculations of lm-cooled wall temperature, car-

ried out on various representative types of gas turbine combustors [11], show

that changes in operating conditions have the same general effect on wall

temperature as for uncooled liners [2]. In particular, it is found that liner-wall

temperatures increase with

1. An increase in pressure

2. An increase in inlet temperature

3. A decrease in air mass ow

4. An increase in liner size

5. A decrease in fuel hydrogen content (see Figure 8.5)

Figure 8.5 also illustrates the benecial effect of lm-cooling air in reduc-

ing liner-wall temperatures for a General Electric F101 combustor. In this g-

ure, the dashed lines represent the results of calculations of uncooled wall

Heat Transfer 341

temperatures, whereas the full lines are drawn through points representing

measured values of lm-cooled wall temperature for different fuel types

ranging from JP4 to DF2. The inuence of fuel hydrogen content on ame

emissivity was accounted for by combining Equations 8.11 and 8.15 to obtain

g2bg

(%H=− −









−−

197440

20515

exp)() .

PqlT

(8.45)

In Figure 8.5, the calculated values of T

are generally higher than the corre-

sponding measured values, as would be expected due to the neglect of inter-

nal lm cooling. Only at low power conditions, where the errors incurred

through neglect of internal wall cooling are partially balanced by the assump-

tion of 100% combustion efciency in the combustion zone, do the measured

and calculated wall temperatures roughly coincide.

8.9.4 Film Cooling with Augmented Convection

Studies have shown that a substantial reduction in the lm-cooling airow

requirement can be achieved by augmenting the convective heat transfer

1400

Calculated

Experimental

F 101

Dash

Take-off

Cruise

Idle

1300

1200

1400

1300

1200

Liner wall temperature, K

1100

1000

900

700

600

500

11 12 13 14

Fuel H

content, percent

Figure 8.5

Comparison of measured and predicted values of wall temperature for a GE F101 combustor.

(From Lefebvre, A.H., AFWAL-TR-84-2104, 1985.)

342 Gas Turbine Combustion: Alternative Fuels and Emissions, Third Edition

on the coolant side of the liner [14,15]. Figure 8.6a illustrates the use of a

double-walled construction to provide lm cooling and augmented external

convection.

Further augmentation of C

by roughening the inside surfaces of the double-

walled slot (e.g., by chemical etching) has resulted in additional improve-

ment of the total cooling effectiveness [16].

8.9.5 impingement Cooling

Another method of augmenting the conventional lm-cooling process

is by means of an impingement lm system, as illustrated in Figure 8.6b.

This is similar to the convection system described above, except that the

double-walled passage is blocked at its upstream end, and the outer wall in

the double-walled region is perforated. The advantage of the method derives

from its use of cooling air to serve a dual purpose. First, the air is shaped into

multiple small jets that provide impingement cooling to one section of the

liner wall, and then the jets merge to form an annular sheet that operates in

a conventional lm-cooling mode to cool a further section of the liner wall.

Another advantage of impingement cooling is that the impingement jets can

be positioned to provide extra cooling on liner hot spots [17].

Hot gas

Cooling air

Spacer

(a) (b)

Hot gas

Cooling air

Hot gas

Cooling air

(c)

Hot gas

Cooling air

(d)

(e)

Hot gas

Cooling air

Figure 8.6

Alternative types of wall cooling: (a) combined lm cooling and augmented external convec-

tion, (b) multijet impingement combined with lm cooling, (c) transpiration cooling, (d) effu-

sion cooling, (e) effusion cooling combined with lm cooling.

Heat Transfer 343

Impingement cooling requires a double-wall construction with attendant

penalties in terms of cost and weight. Another drawback stems from the

fundamental difference in temperature between the two walls. This causes

problems of differential expansion that can lead to buckling of the inner wall

if the local hot spots become too severe. Moreover, the high heat-transfer

coefcients that are normally associated with impingement cooling cannot

be realized in full, because the lm of air formed on the inner wall from

the upstream air jets tends to reduce the efcacy of the impingement cooling

from the downstream air jets.

8.9.6 Transpiration Cooling

An ideal wall-cooling system would be one in which the entire liner was

maintained at the maximum temperature of the material because cooler

regions would represent a wasteful use of cooling air. The method that

comes closest to this ideal is known as transpiration cooling, whereby the

liner wall is constructed from a porous material that provides a large internal

area for heat transfer to the air passing through it, as illustrated in Figure 8.6c.

Since the pores are uniformly dispersed over the surface of the wall, the tiny

air jets emerging from each pore rapidly coalesce to form a protective layer

of cool air over the entire inner surface of the liner. In this way, the convective

heat transfer from the hot gas to the liner walls can be drastically reduced,

resulting in substantial economies in lm-cooling air.

Even if the protective lm of air is completely successful in preventing the

hot gases from making physical contact with the inner liner wall, the latter

will still be exposed to intense radiation from the ame. The only means by

which this heat can be removed is by transfer to the coolant air during its

passage through the porous wall. This means that, in addition to acting as a

porous medium, the wall must also have good heat-transfer properties and

be of adequate thickness. A problem this poses is that, in order to form a

stable boundary layer on the inner surface of the wall, the coolant ow should

emerge with as low a velocity as possible, whereas for maximum heat transfer

within the wall, a high velocity is required.

8.10 Practical Applications of Transpiration Cooling

Although transpiration cooling is potentially the most efcient method of

liner cooling, its practical implementation has been hampered by the limita-

tions of available porous materials. The porous materials developed to date

have failed to demonstrate the required tolerance to oxidation, which has led

to the small passages becoming blocked. These passages are also prone to

blockage by airborne debris.

344 Gas Turbine Combustion: Alternative Fuels and Emissions, Third Edition

The search for more practical methods of increasing the internal heat

transfer within the liner wall has led to the development of multilaminate

sheets from which a “quasi-transpiration” cooled wall can be manufactured.

The two best-known examples of this approach are Transply and Lamilloy,

as described below.

8.10.1 Transply

Transply is the name given to a composite liner structure developed by the

Rolls Royce Company [18]. It is produced by brazing together two or more

laminates of a high-temperature alloy containing a multiplicity of intercon-

necting ow passages, as illustrated in Figure 8.7. These passages are designed

for maximum heat transfer between the wall material and the air owing

through them. After cooling the wall structure, the air emerges in the form

of numerous tiny jets owing through evenly spaced holes on the hot gas side

of the liner wall. The air jets are deected downstream, gradually forming

an insulating blanket between the hot gas and the wall. Thus, the liner is

cooled partly by the air passing through the wall, and partly through the

leaving air acting as a protective lm.

Cold-side

laminate

Middle

laminate

Hot-side

laminate

Figure 8.7

Constructional features of Rolls Royce Transply. (From Wassell, A.B., and Banghu, J.Κ., ASME

Paper 80-GT-66, 1980. With permission.)

Heat Transfer 345

An important requirement is that the alloy selected for the sheets should

have good brazing characteristics, to ensure the mechanical integrity of the

liner structure, and high oxidation resistance at normal operating tempera-

tures. Intergranular oxidation of the thin-walled sections would reduce the

strength of the structure, whereas excessive oxide formation on the internal

surfaces would lead to partial blockage of the cooling passages and hence to

local overheating.

A Transply combustor was developed for the RR Spey Mk 512 engine and

entered service in the ΒA 1-11 aircraft. It achieved a 70% reduction in cooling

airow requirement relative to the conventional lm-cooled Spey combus-

tor. This large saving in cooling air allowed the liner internal airow distribu-

tion to be optimized for minimum pollutant emissions.

8.10.2 Lamilloy

In the United States, the Allison Engine Company has pursued a similar

development of quasi-transpirate materials and has produced a multilaminate

porous structure, known as Lamilloy, which is fabricated from two or more

diffusion-bonded, photoetched metal sheets, as illustrated in Figure 8.8.

In general, Lamilloy has the same high cooling potential and the same

practical problems and limitations as Transply. Both materials require fur-

ther development to reduce manufacturing costs and achieve better process

control. A more basic deciency is their lack of mechanical strength in com-

parison with more conventional cooling methods such as, e.g., machined and

rolled rings. This drawback is likely to militate against their application to

large annular combustors.

At the time of their emergence and initial development in the 1970s, both

Transply [18] and Lamilloy [19] were viewed as replacements for the various

lm-cooling congurations then in widespread use. Today, with the advent of

new cooling concepts, along with major advances in manufacturing techniques,

the competition is far more keen. It will, therefore, be of great interest to see

what the future holds for these and other quasi-transpiration cooling schemes.

Figure 8.8

Lamilloy construction and airow path. (From Nealy, D.Α., Gas Turbine Combustor Design

Problems, Hemisphere, Washington, DC, 151–85, 1980.)

346 Gas Turbine Combustion: Alternative Fuels and Emissions, Third Edition

8.10.3 effusion Cooling

Another, and perhaps the simplest approach to a practical form of transpira-

tion cooling, is a wall perforated by a large number of small holes, as shown

in Figure 8.6d. Ideally, the holes should be large enough to remain free from

blockage by impurities, but small enough to prevent excessive penetration of

the air jets. Provided that the jet penetration is small, it is possible to produce,

along the inner surface of the liner, a fairly uniform lm of cooling air. If, how-

ever, the penetration is too high, the air jets rapidly mix with the hot gases

and provide little cooling of the wall downstream.

Effusion cooling can be applied to all or any portion of the liner wall, but

because it is somewhat lavish in its use of cooling air, it is best used for treat-

ing local hot spots in the liner wall. Another useful role is in improving the

effectiveness of a conventional lm-cooling slot. As the lm of air from this

slot moves downstream, its temperature gradually rises due to entrainment

of the surrounding combustion gases. Eventually, it becomes so hot that it

starts to heat the liner wall instead of cooling it. If effusion cooling is applied

before this point is reached, the injection of cold air into the lm enables it

to maintain its cooling effectiveness for a further distance downstream. A

typical arrangement is shown in Figure 8.6e.

Few published data are available on the performance of conventional effu-

sion cooling systems. This is not a serious omission because recent develop-

ments in AEC have rendered them virtually obsolete.

8.11 Advanced Wall-Cooling Methods

The above discussion on wall-cooling methods has concentrated largely on

concepts that fall into the general category of lm cooling. This is appropriate

because almost all of the combustors now in service employ lm cooling in

one form or another. However, few of the combustors being designed today

for application to high-performance gas turbines use lm cooling, except per-

haps in local areas, such as the liner dome. At the present time, interest is

focused mainly on two new and widely different approaches—AEC and tiled

walls. These new concepts merit the designation “advanced,” partly because

of their potential for signicant reductions in cooling-air requirements, but

also because much more service experience with these devices must be gained

before they can be regarded as orthodox designs.

8.11.1 Angled effusion Cooling

With conventional effusion cooling, the holes are drilled normal to the liner

wall. The advantages to be gained from drilling the holes at a more shallow

angle are twofold, as described by Dodds and Ekstedt [20].

Heat Transfer 347

1. An increase in the internal surface area available for heat removal.

This area is inversely proportional to the square of the hole diameter

and the sine of the hole angle. Thus, for example, a hole drilled at

20° to the liner wall has almost three times the surface area of a hole

drilled normal to the wall.

2. Jets emerging from the wall at a shallow angle have low penetration

and are better able to form a lm along the surface of the wall. The

cooling effectiveness of this lm also improves as the hole size and

angle are decreased.

From this brief description of AEC, it is clear that its practical implemen-

tation is highly dependent on an ability to accurately, consistently, and

economically manufacture large numbers of oblique holes of very small

diameter. Advances in laser drilling have made this possible, and AEC is now

regarded as a viable and economically acceptable cooling technique. At the

present time, the lower limit on hole diameter is about 0.4 mm, whereas the

lowest attainable hole angle is just below 20°.

AEC is perhaps the most promising contender among the various advanced

combustor cooling techniques that are being actively developed for the new

generation of industrial and aeronautical gas turbines. It is a technique that

can be used locally, to supplement other forms of wall cooling, or it can be

applied to the entire liner. It is used extensively on the GE 90 combustor,

where it has reduced the normal cooling-air requirement by 30%.

The known drawbacks to AEC include an increase in liner weight of around

20%, which stems from the need for a thicker wall to achieve the required

hole length and to provide buckling strength. Cost is also an important con-

sideration, and this is allied to the need to drill the holes consistently and

to specication in a production environment. Other concerns include the

“repairability” of AEC liners and their durability. These issues can only be

fully resolved by extensive service experience.

Future developments in AEC will tend to focus on the optimization of hole

geometry, a topic that cannot be separated from hole manufacture. A diffus-

er-shaped expansion at the exit portion of the hole has been shown to improve

cooling effectiveness [21], presumably because the lower exit velocity reduces

the penetration of the air jet into the hot gas stream. As an extension of this

approach, lateral expansion of the hole exit (fanning) improves the lateral

spreading of the jet to give better coverage of the wall surface. However, a

cost effective method of producing shaped holes has yet to be developed.

8.11.2 Tiles

The principle of using tiles is well established for large industrial engines where

size and weight are of minor importance and it is thus practicable to line

the combustor with refractory bricks to reduce the heat ux to the liner wall.

348 Gas Turbine Combustion: Alternative Fuels and Emissions, Third Edition

Refractory bricks are clearly too heavy and cumbersome for application to aero

and most industrial engines, but metallic tiles offer an attractive solution. The

V2500 engine is now in service with a tiled combustor, as shown in Figure 8.9,

and P&W is also using tiles on its radially staged combustor for the PW4000.

The method of construction is to mount a large number of tiles on a support

shell. The shell is protected by the tiles, but the tiles themselves are exposed

to the hot combustion gases. This effectively decouples the mechanical

stresses, which are taken by the shell, from the thermal stresses, which are

taken by the tiles. Methods for alleviating the thermal load on tiles involve vari-

ous heat-transfer features, such as multiple pedestals (see below) on their rear

surface to enhance the convective heat transfer. Metered air ows over the

pedestals and is then ejected at the ends of the tiles to form a protective lm

over their front surface. Thermal barrier coatings (TBCs) can also be used to

give extra protection to this surface.

Tiled walls offer the following advantages:

1. The tiles can be cast from blade alloy materials having a much higher

temperature capability (>100°C) than typical combustor alloys

2. Since the combustor shell remains at a uniform low temperature,

relatively cheap alloys can be used

3. The low shell temperature minimizes thermal growth relative to the

combustor casing

4. Maintenance time and cost is reduced because changing a tile is

simpler than repairing a liner

5. Signicant reduction in cooling-air requirement

Figure 8.9

P&W V2500 tiled combustor.

Heat Transfer 349

The main drawbacks of tiled combustors are:

1. A substantial increase in weight

2. The various ports of entry for the combustion and dilution air-

streams are difcult to modify during combustor development

3. Difcult to scale down the tile attachment features for application to

small engines

8.12 Augmented Cold-Side Convection

The rate of convective heat transfer on the cool side of the liner can be increased

by the use of ns, pedestals, and ribs, or any other form of secondary surface

that increases the effective area for heat exchange. Secondary surfaces can-

not be 100% efcient because a temperature gradient must exist along each

protuberance, whatever its physical form, to allow heat to be conducted away

from its base.

Pedestals have been used on heat shields in the dome area of RR RB211 com-

bustors for many years, and the RR Tay combustor also features a pedestal

head. Pedestals are also used to augment convective heat transfer on the

cold sides of wall tiles. Finned outer surfaces have featured successfully in

a number of industrial gas turbines, but only recently has their use become

more widespread. Usually, the ns or ribs are arranged to run longitudinally,

but circumferential ns are also used at the expense of a small increase in pres-

sure loss. Dutta et al. [22] have described a catalytic combustor that employs

continuous round wire welded to the outside of the liner. The design is based

on data from Norris [23] and Evans and Noble [24], which show an average

threefold augmentation in convective heat transfer when compared with

similar geometries without the wires.

More detailed information on the use and performance of various extend-

ed-surface congurations, including ribs, ns, and pedestals, may be found

in Gardner [25] and Lohmann and Jeroszko [26].

8.13 Thermal Barrier Coatings

One attractive approach to the problem of achieving satisfactory liner life is

to coat the inside of the liner with a thin layer of refractory material. A suit-

able material of low emissivity and low thermal conductivity could reduce