[american institute of aeronautics and astronautics 46th aiaa/asme/sae/asee joint propulsion...

American Institute of Aeronautics and Astronautics

1

Temperature Measurements of Porous Media for

Transpiration Studies

G. Natsui1, Mark A. Ricklick

2, Cuong Q. Nguyen

3, J. S. Kapat

4

Laboratory for Turbine Heat Transfer and Aerodynamics

Center for Advanced Turbine Energy Research

University of Central Florida, Orlando, Florida, 32816

A method of measuring the local temperature of a porous wall is discussed.

Measurements are taken with temperature sensitive paint applied in thin coats to the wall.

This technique was validated on a 40PPI, 7% relative density aluminum porous coupon.

Measurements of discharge coefficients as well as downstream effectiveness data are

included to verify the flow through the porous wall was unaltered by applying the paint. A

maximum deviation in film-cooling effectiveness of 9% between the two cases with the

majority of data falling within 4% was found, very similar to the experimental uncertainty

of the rig. This excellent agreement between the repeated tests showed that by applying

thermal paint to a wall of such porosity does not significantly affect the flow exiting the wall

and hence the measurement technique can readily be applied to transpiration cooling studies

at this scale. Methods of filtering the temperature sensitive paint on the porous wall are

presented.

Nomenclature

T = Temperature

P = Pressure

U = Velocity

x = Distance downstream (from the trailing edge of coolant outlet)

M = Blowing ratio = (ρU)c/( ρU)s

DR = Density ratio

St = Stanton number

h = Length of transpiring surface in flow direction

Cd = Discharge coefficient

Greek symbols

ε = Porosity

η = Adiabatic film cooling effectiveness = (Tw-Ts)/(Tc-Ts)

ρ = Density

θ = Temperature difference for TSP calibration

Subscripts and Superscripts

w = adiabatic wall

c = coolant

- = laterally averaged

s = main flow

r = reference

t = total

1 Graduate Assistant, Department of Mechanical, Materials and Aerospace Engineering, Bldg 40, Room 307, UCF

2 Post Doctorate, Department of Mechanical, Materials and Aerospace Engineering, Bldg 40, Room 307, UCF

3 Graduate Assistant, Department of Mechanical, Materials and Aerospace Engineering, Bldg 40, Room 307, UCF

4 Professor, Material Mechanical and Aerospace Engineering Department, Building 40, Room 407, UCF

46th AIAA/ASME/SAE/ASEE Joint Propulsion Conference & Exhibit25 - 28 July 2010, Nashville, TN

AIAA 2010-6950

Copyright © 2010 by the American Institute of Aeronautics and Astronautics, Inc. All rights reserved.


2

I. Introduction

orous materials are being used in multiple diverse applications with many more yet to be introduced. With more

consistent pore sizes and pore structures, these materials can be incorporated into products more confidently by

designers. One very promising use of porous materials is in transpiration cooling for turbine blades. Ideal

transpiration cooling is equivalent to the case of film cooling in which the coolant hole spacing to diameter ratio is

taken theoretically to one resulting in a uniform coolant velocity exiting the perforated wall [1]. In reality designs

can only approach this assumption. The closest form of cooling to ideal transpiration is made realizable by

incorporating porous matrix into the cooled wall. Transpiration is so effective because it involves two forms of heat

transfer. As the coolant flows through the porous matrix it efficiently draws internal energy away from the cooled

wall by convection. After the coolant passes the wall it then coats the downstream exterior with a film, greatly

reducing the heat transfer from the hot fluid to the cooled wall.

Uses for transpiration cooling have been found in gas turbines, rocket nozzles, re-entry thermal protection

systems, fire protection suits, fusion reactors, and combustors [2-6]. A constant among all these applications is the

want of removing the most heat for the least amount of spent coolant. In reference to gas turbines, transpiration is

being looked at as the future method for cooling the first stage turbine blades more effectively with hopes of raising

the turbine inlet temperature above current capabilities [7]. A gas turbine’s efficiency increases with increased inlet

temperatures when operating at high pressure ratios hence the focus on raising this temperature. With heat fluxes on

the order of 2 MW/m2 and temperatures exceeding 1700 K in the first stage of gas turbines, current cooling

technologies are reaching a wall in cooling capability [8]. To further increase the efficiency of gas turbines a full

understanding of transpiration cooling is needed in order to increase these temperatures even higher.

Porous transpiration cooling provides many benefits when compared to the technology currently in its place, film

cooling. Film cooling provided a drastic increase in the allowable turbine inlet temperature when it was introduced.

Full coverage film cooling is however only an attempt at ideal transpiration cooling. Film cooling causes

detrimental vortical structures near the coolant holes due to having discrete points of injection. The kidney vortices

formed act to entrap the hot main flow and bring it in contact with the cooled wall. Conversely transpiration

provides a uniform coolant flow leaving the wall resulting in a much more two-dimensional flow field. Perhaps the

most beneficial aspect of transpiration is the amount of coolant needed to provide large cooling efficiencies.

Transpiration can provide the same cooling efficiency as film cooling with much less spent air [9].

In a study comparing film cooling with transpiration, Eckert and Cho [1] by showed that the dimensionless

temperature ratio is a function of the wall Stanton number and the blowing ratio.

�� − �� − �� = ��

With knowledge of the local Stanton number on a surface, a desired wall temperature can easily be achieved by

designers. This introduces the need to study heat transfer on and around the surface of a transpiring wall to obtain

these Stanton numbers. There are several established procedures for studying the cooling of a smooth surface

however many will not translate well to a porous surface. Napthalene sublimation [10] has been used extensively to

study cooling however is in no way applicable to a porous surface as it would continually alter the geometry. Using

encapsulated thermochromic liquid crystal (TLC) is another common approach [11]. It is an easily implementable

technique for studying temperature but again is not ideal when looking at a porous surface. TLC is very sensitive to

viewing angle, this would incur lots of error when measuring a porous surface, also the measurements are over such

small areas that TLC may not have a high enough spatial resolution to capture data.

When studying flat plate transpiration Jiang et al [12] experimentally measured the ‘local’ temperature

distribution of a porous wall by means of five thin K-type thermocouples. Although they were successful in

studying many important features, this technique of measuring temperature leaves much to be desired when

attempting to study a more complex geometry with transpiration cooling. Another method used to gather local

temperature distributions on a surface is the infrared thermal imaging technique (IRTIT). This is the method used

by Wang, Messner and Stetter [13] who decided IRTIT is the most effective method for evaluating the performance

of transpiration cooling after successfully studying a porous cylinder. There are however complications involved

with the technique. The use of infrared transmissible windows is required; sapphire, zinc selenide and calcium

fluoride are suitable yet this can sometimes place a limit on the size of a test section due to cost. When evaluating

the temperature with an IRTIT there is a compounding effect of the infrared window, the scanner position, the

P


surface emissivity and the stability of the infrared system on the accuracy of the measureme

prospective complications involved in the measurement.

The focus of the current work is to develop a method for evaluating a temperature distribution on the transpiring

surface of an arbitrary geometry such as a turbine blade. This will be done by the use of tem

(TSP). TSP has been shown to be an effective method of evaluating temperature by Liu et al

capable of not significantly altering the porosity of the material.

and has very fine spatial resolution. It is possible to calibr

calibrations. The effects of TSP on the geometry of the porous wall will be evaluated by comparing the discharge

coefficient as well as film-cooling effectiveness before and after the application of T

similar to that studied by Goldstein [9]

transpiring porous material. These comparisons will show the significance of the paint on the aerodynamic

performance of the wall. With confidence in this tec

applied in future studies of more compl

II. Problem Formulation and Methodology

A. The Subsonic film Cooling Wind tunnel Test Rig

A subsonic wind tunnel built for

transpiration. A schematic of the experimental test

supplies air at a rate of 4.7m3/s and yields a velocity of

closed loop configuration, the fan work

over a period of three hours until the mainstream temperature remains consta

from the tunnel balances out the blower work.

consisting of a honeycomb and three screens

nozzle leading to the inlet of the test section. The

Figure 1: Schematic of the subsonic wind tunnel and components


3

surface emissivity and the stability of the infrared system on the accuracy of the measurement introducing many

in the measurement.


surface of an arbitrary geometry such as a turbine blade. This will be done by the use of temperature sensitive paint

). TSP has been shown to be an effective method of evaluating temperature by Liu et al

capable of not significantly altering the porosity of the material. Unlike TLC, TSP is insensitive to viewing angle

and has very fine spatial resolution. It is possible to calibrate TSP remotely, taking away complications of in

The effects of TSP on the geometry of the porous wall will be evaluated by comparing the discharge

cooling effectiveness before and after the application of TSP. The flow setup is very

[9] with secondary air being injected into a hot main flow through a strip of

These comparisons will show the significance of the paint on the aerodynamic

With confidence in this technique of experimentally measuring temperature it can then be

applied in future studies of more complex transpiration cooling scenarios.

Problem Formulation and Methodology

The Subsonic film Cooling Wind tunnel Test Rig

A subsonic wind tunnel built for film cooling investigations was modified for the study of flat plate

xperimental test rig used in this study is shown in Figure 1. The 15

yields a velocity of approximately 54 +/- 3m/s at the test

fan work heats up the re-circulated air to about 66°C (340 K) at the test section inlet

three hours until the mainstream temperature remains constant. It steadies out once the heat loss

from the tunnel balances out the blower work. The blower outlet was connected to a flow-conditioning module

consisting of a honeycomb and three screens. The flow conditioning module was followed by a two

section. The test section has dimensions of 1.2 x 0.53 x 0.154

: Schematic of the subsonic wind tunnel and components

nt introducing many


perature sensitive paint

). TSP has been shown to be an effective method of evaluating temperature by Liu et al [14] and may be

Unlike TLC, TSP is insensitive to viewing angle

ate TSP remotely, taking away complications of in-situ

The effects of TSP on the geometry of the porous wall will be evaluated by comparing the discharge

The flow setup is very

with secondary air being injected into a hot main flow through a strip of

These comparisons will show the significance of the paint on the aerodynamic

hnique of experimentally measuring temperature it can then be

s was modified for the study of flat plate

. The 15-kW blower

m/s at the test section inlet. In this

circulated air to about 66°C (340 K) at the test section inlet

It steadies out once the heat loss

conditioning module

followed by a two-dimensional

x 0.154 m.


4

The test coupon is aluminum foam of 40PPI and 7% density. The coupon dimensions are 2 cm in the stream

direction and the thickness is 1.1 cm. The span of the coupon was 8 cm. There was a T-type thermocouple

embedded in the porous coupon to verify recorded wall temperatures. The test section consists of a 0.5in thick balsa

wood plate. Nitrogen gas at a temperature of 5 ± 1.5°C was used as the coolant in these experiments. This variation

is due to small fluctuations in the nitrogen supply temperature. The temperature of the coolant was controlled to a

fixed density ratio of 1.26. The plenum confined a perforated plate to break the jet and to promote mixing inside the

plenum, as shown in Figure 1. The plenum system and the bottom surface of the test-section were covered with 5 cm

thick fiberglass insulation. This fiberglass insulation along with the adiabatic balsa wood test section provided an

ideal adiabatic condition for testing. Adiabatic wall temperature distribution measurements immediately downstream

of the test coupon are made with the use of TSP. The coolant temperature is measured by means of a thermocouple

at the exit of the porous wall. Measurements are taken when the tunnel attains steady state which is achieved when

the rise in the mainstream temperature is 0.1°C or less over 10 minutes. While radiation loss is reasonably neglected

in the analysis, heat lost by means of conduction through the balsa wood is small and also neglected. The estimated

uncertainty in temperature measurements using TSP is ± 1°C shown by Liu [14].

Table 1: Test Parameters

Test coupon specifications

Porosity [ppi]

Relative density

40

7%

Dimensions [mm] 80 x 20 x 11

Main Flow Condition

Blowing ratio, BR 0.03, 0.06, 0.09

Temperature, Tm [oC | K] 66 | 339

Inlet total pressure, Pt_inlet [kPa] 102.5

Inlet static pressure, Ps_inlet [kPa] 101.1

Outlet static pressure, Ps_outlet [kPa] 101.1

Coolant Flow Condition

Temperature, Tc [oC | K] 5 | 278

Density ratio, DR 1.26

Discharge coefficients were calculated by measuring the coolant mass flow, the pressure in the coolant plenum,

and a fixed pressure in the main stream flow. Mass flow to the coolant supply was measured with an Omega FMA

1700/1800 mass flow meter. Discharge coefficients were measured at a constant freestream Reynolds number and

over a range of blowing ratios.

� = � �� = � �� ∗ �� 2� � − 1"#�� %�� &�� − 1'

TSP is basically a luminescent molecule suspended in a binder which can be calibrated against temperature. A

more detailed description of TSP can be found in [14]. When taking measurements with TSP it is common practice

to take multiple pictures to filter out the noise arising in the CCD. This is important when filtering the data.

The uncertainty in the calculation of the blowing ratios is estimated to be less than 5% over the range of

conditions being tested. The adiabatic wall temperatures measured by TSP were used to calculate effectiveness

values in each case. The effectiveness values had an average uncertainty of less than 3% with the highest value

calculated to be less than 9%. The uncertainties reported were estimated by following the procedure described by

Kline and McClintock [15].

B. TSP Calibration of a Porous Coupon

The porous test piece was first properly coated in TSP with the method shown by Liu [14]. A pump was then

used to evacuate a small chamber with the test plate inside to reduce the Biot number of the coupon being calibrated.

A small thermoelectric heater was used to bring the coupon to a constant temperature before each calibration point.

A 405nm light was used to illuminate the test piece and a CCD camera was used to record the images. The

temperature was monitored and pictures were recorded at five degree increments. A traditional calibration curve for

TSP could then be made for the porous section. Figure 2 shows the calibration curve of the porous specimen

alongside a calibration curve of a solid section. There is a difference between the two however this small variation


5

between two different calibrations is normal and of no concern. Differences are usually attributed to different

batches of paint and paint thickness.

Figure 2: Comparison of TSP calibrations

C. Processing Porous TSP

The main challenge when processing TSP on a porous surface is deciding the areas from which data will be

taken. Areas not on the surface would provide false data and areas with very large angles off normal to the camera

would provide unreliable data. Even though TSP is good at measuring from large viewing angles, angles greater

than 70º are still problematic as the noise to signal increases. The reference image shown at the top of Figure 3 can

be manipulated to make a filter which ignores poor data. One condition is that the standard deviation of the intensity

between each of the pictures must not be large. This requirement would eliminate the pixels with excessive amounts

of noise. This is usually due to looking at surfaces being recorded from very large viewing angles. Also the

reference intensity must fall within a certain range so that the CCD camera used to capture the image does not

saturate. This requirement blocks out all the data which is in a recess of the material which is not illuminated

properly. If both these criteria are met in the reference picture, that area may be used to take data. The desired area

is only the top surface and these filters ensure that paint which was applied to lower layers does not affect the

measurement. Hence the only added complexity of processing the information from a porous TSP measurement

arises in designing a filter to distinguish reliable information from poor data. This filter can be made by forcing two

simple conditions on the picture. Once this area is decided on a temperature profile similar to Figure 3 can be

calculated through normal TSP processing procedures. The temperature distribution in Figure 3 was applied to the

coupon by heating it on one side and shows the capability of TSP applied to a porous medium.

Figure 3: Top: Raw picture of intensity. Middle: Contours of intensity. Bottom: Filtered temperature profile

0

0.2

0.4

0.6

0.8

1

1.2

0 0.1 0.2 0.3 0.4 0.5

I/Ir

θ

Solid Calibration

Porous Calibration


6

III. Results and Discussions

A. Experimental Results

For TSP to be a valid technique of temperature measurements on a porous wall, it must not affect the behavior of

the body it is applied to. For sintered materials or very small pore sizes, TSP would cover a large percentage of the

open area on the porous wall making it impractical as a measurement technique. However for materials such as the

one being tested, 40PPI 7% relative density, the paint would cover an insignificant proportion of the total open area

implying the performance of the porous material would be unchanged by applying TSP. To evaluate the performance

of the porous material, discharge coefficients and downstream effectiveness were measured over a range of bowing

ratios. Discharge coefficients for both cases are shown in Figure 4.

Figure 4: Comparison of Cd for the 2 coupons

Downstream of the coupon η was measured to compare the effect of paint on end-wall temperature distributions.

Three blowing ratios were tested for each case, 0.03, 0.06, and 0.09. Figure 5 shows η downstream of the injection

location with “P” representing paint and “NP” representing no paint. This value η was obtained by averaging

laterally within the 1-D area of downstream effectiveness so it represents both η and η). Comparing similar blowing

ratios from Figure 5 one can see the differences between a painted and un-painted porous wall. There are differences

between the two runs however none of the differences appear to be systematic. That is, the differences appear to be

random and due more to experimental uncertainty rather than a result of altering the performance of the wall due to

painting. Figure 6 shows the recorded data plotted against other film literature. The flow scenario best matches that

reported by Goldstein [9] and the results also match closest even though the actual geometry of the transpiring

section is very different. Goldstein used a sintered stainless steel wall with much larger relative density. Results

shown from Seban [16] and Scesa [17] also match the data well and bound the currently recorded data. Typical for

transpiration literature or slot literature, the independent variable is x/Mh with h being the dimension of the plate in

the stream direction. The data gathered makes good sense when compared to other investigators, even though the

true goal is for it to be very repeatable this still adds to the confidence in these results. The data from the current tests

shown in Figure 6 which does not follow the rest of the current data is the higher blowing ratio of 0.09. Figure 8

shows the difference between measured values of η downstream of the injection location. This again shows the error

appears to be random and follow no systematic trend, indicating these differences seen are due to coolant

temperatures and pressures rather than altering the geometry of the wall.

0

0.02

0.04

0.06

0.08

0.1

0.12

0 0.01 0.02 0.03 0.04 0.05

CD

M

Paint

No Paint


7

Figure 5: End-wall η downstream of coupon

0

0.2

0.4

0.6

0.8

1

0 20 40 60 80 100 120 140 160

η

X (mm)

M=0.03 NPM=0.03 P

M=0.06 NPM=0.06 PM=0.09 NP

M=0.09 PGoldstein M=0.03

POROUS WALL


8

Figure 6: Comparison to literature

Figure 7: Example of η over porous strip and downstream

0.02

0.2

2

1 10 100 1000

η

X /Ms

Current Study

Goldstein

Scesa

Seban M=0.03

Seban M=0.09

0

0.2

0.4

0.6

0.8

1

0 20 40 60 80 100 120 140 160

η

X (mm)

M=0.03 P


9

Figure 8: % Difference between runs against stream location

The discharge coefficients from Figure 4 and the end-wall effectiveness values from Figure 5 were shown to be

very repeatable before and after application of TSP. If by applying paint to the wall, there is such a small impact on

its performance then TSP would seem to be a viable method for evaluating local temperature distributions on a

porous wall. In cases where the temperature distribution on the transpiring section is 2-D such as when coolant is

supplied by impingement on the backside, this would be a useful tool. Figure 7 shows the adiabatic film cooling

effectiveness over the test coupon and downstream. This is an example of the added information the capability of

measuring wall temperatures on a porous surface can provide.

IV. Conclusion

To verify that TSP can be used for measuring the local temperature distributions on a porous surface, the

aerodynamic performance both before and after the application of TSP was evaluated. A 40PPI, 7% relative density

aluminum foam coupon was used. Discharge coefficients and end-wall adiabatic film-cooling effectiveness were

measured over a range of blowing ratios on the coupons before and after TSP. The discharge coefficient curves of

both coupons collapsed to the same curve with an average difference between runs of 2%. The end-wall

effectiveness downstream of the porous strip also matched between coupons over several blowing ratios. The largest

difference between the two runs occurred far downstream, in the area of largest uncertainty and was 9%. Most points

for all runs fall within 4% of one another. By applying TSP to 40PPI 7% density metal foam, there will be no

significant effect on the aerodynamic performance that can be captured by typical lab tolerances. With this

technique, effectiveness profiles can be obtained for more complicated transpiration configurations.

0.0%

1.0%

2.0%

3.0%

4.0%

5.0%

6.0%

7.0%

8.0%

9.0%

10.0%

0 20 40 60 80 100 120 140 160

% D

iffe

ren

ce

X (mm)

M=0.03

M=0.06

M=0.09

POROUS WALL


10

References

[1] E. R. G. Eckert and H. H. Cho, Transition from transpiration to film cooling, Int. J. Heat Mass

Transfer, Vol. 37, Suppl. 1, pp. 3-8, 1994.

[2] F. Chen, W. J. Bowman, and R. Bowersox, Effect of Transpiration Cooling on Nozzle Heat Transfer,

J. Spacecraft and Rockets, Vol. 33, No. 3, pp. 453-455, 1996.

[3] D. Keener, J. Lenertz, R. Bowersox, and J. Bowman, Transpiration Cooling Effects on Nozzle Heat

Transfer and Performance, J. of Spacecraft and Rockets, Vol. 32, No. 6, pp. 981-985, 1995.

[4] Jury Polezhaev, The Transpiration Cooling for Blades of High Temperatures Gas Turbine, Energy

Convers. Mgmt, Vol. 38, No. 10-13, pp. 1123-1133, 1997.

[5] Hirotaka Otsu, Kazuhisa Fujita, and Takeshi Ito, Application of the Transpiration Cooling Method

for Reentry Vehicles, AIAA 2007-1209.

[6] E. Eugene Callens, Jr., and Robert F. Vinet, Design of Transpiration Cooled Thermal Protection

Systems, Technical Report Contract No. 32-4136-58019 and NASA(1997)-Stennis-17 and NAS13-

580, March 1999.

[7] James P. Downs, Kenneth K. Landis, Turbine Cooling Systems Design – Past, Present and Future,

ASME Turbo Expo 2009, GT2009-59991

[8] I. Leontiev, and A. F. Polyakov, The Thermal State of a Porous Wall under Conditions of

Transpiration Cooling, High Temperature, Vol 44. No. 1, 2006, pp. 99-107.

[9] R. J. Goldstein, G. Shavit and T. S. Chen, Film-Cooling Effectiveness With Injection Through a

Porous Section, 29 Nov.-4 Dec. 1964 ASME 64-WA/HT-30, pp. 1-11

[10] R. J. Goldstein, H. H. Cho, A Review of Mass Transfer Measurements Using Napthalene

Sublimation, Experimental Thermal and Fluid Science, 1995; 10:416-434.

[11] T. L. Chan, S. Ashforth-Frost, K. Jambunathan, Calibrating for Viewing Angle Effect During Heat

Transfer Measurements on a Curved Surface, Int. J. of Heat and Mass Transfer, 44 (2001) 2209-

2223.

[12] Pei-Xue Jiang, Lei Yu, Ji-Guo Sun, Jue Wang, “Experimental and numerical investigation of

convection heat transfer in transpiration cooling”, Applied Thermal Engineering, 24 (2004) 1271-

1289.

[13] J. H. Wang, J. Messner and H. Stetter, An Experimental Investigation of Transpiration Cooling:

Application of in Infrared Measurement Technique, International Journal of Rotating Machinery,

Vol. 9, No. 3, 2003, pp. 153-161.

[14] Quan Liu, A. K. Sleiti, and J. S. Kapat, Application of pressure and temperature sensitive paints for

study of heat transfer to a circular impinging air jet, Int. J. Thermal Sciences, Vol. 47 pp. 749-757,

2008.

[15] Kline, S., and McClintok, F., 1953, “Describing Uncertainties in Single-Sample Experiments”,

Mechanical Engineering, 75, pp. 3-8.

[16] R. A. Seban. “Heat Transfer and Effectiveness for a Turbulent Boundary Layer With Tangential

Fluid Injections. Journal of Heat Transfer, Trans. ASME, Series C, vol. 82, 1960, pp. 303-312

[17] S. Scesa, “Effect of Local Normal Injection on Flat Plate Heat Transfer,” PhD thesis, University of

California, Berkeley, Calif. 1954

[american institute of aeronautics and astronautics 46th aiaa/asme/sae/asee joint propulsion...

Documents