LIBRARY
Michigan State
University

 

 

 

This is to certify that the
thesis entitled

DESIGN VARIATIONS AND IMPROVEMENTS OF
PERSPIRABLE SKIN

presented by

JIN ZHANG TAM

has been accepted towards fulfillment
of the requirements for the

MS. degree in Mechanical Engineering

 

 

A

 

Major Professor’s Signature
a‘3/-2,/«='W0

Date

MSU is an Affirmative Action/Equal Opportunity Employer

-.-.-—-—c-u-

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

r

.03329 2
AbGa‘LZUd

 

 

 

 

 

 

 

 

 

 

 

 

 

5108 K'IProlecc8-Pres/CIRCIDateDue.indd

DESIGN VARIATIONS AND IMPROVEMENTS OF PERSPIRABLE SKIN

BY

JIN ZHANG TAM

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

MASTER OF SCIENCE

Mechanical Engineering

2010

ABSTRACT

DESIGN VARIATIONS AND IMPROVEMENTS OF PERSPIRABLE SKIN

By
JIN ZHANG TAM
An autonomous deforming and reusable self-cooling thermal protection system
(TPS) was designed and modified for surfaces operating in extreme
temperatures such as the surface of the space shuttle. This newly designed
structure, called ‘Perspirable Skin,’ contains cores and deformable rings that are

fitted into Reinforced Carbon-Carbon (RCC) composite skin. The cores and

deformable rings are made of a composition of pure ZrW203 (zirconium
tungstate) and Zr02 (zirconia), which forms a Functionally Graded Material

(FGM). The choice of ZrWzoa is due to its highly negative coefficient of thermal

expansion (CTE) in a wide range of temperatures. When the temperature
increases, the core would shrink while the deformable ring would have an out-of-
plane deformation to create a large gap. The compressed coolant gas onboard is
blown through this gap onto the surface. The coolant gas is expected to mix with
the surface air to eliminate the frictional heating. The geometry and stacking
sequence of the cores as well as the deformable rings were designed by Finite
Element Method (FEM). Possibilities of other negative coefficient of thermal

expansion materials were discussed and explored.

To my dear parents and Jomay.

ACKNOWLEDGEMENTS

I would like to express my gratitude to my advisor Dr. Patrick Kwon for giving me
the opportunity to work on this Perspirable Skin project and for his invaluable
supervision and support throughout the course as a Master’s student at Michigan
State University. I would also like to thank Dr. Dahsin Liu and Dr. Alfred Loos for

being on my thesis committee.

I am delighted to have such a valuable opportunity to work with a group of very
knowledgeable colleagues on this project. I would like to give thanks to Li Sun
and Basak Oguz who have been the pioneers to explore and have laid a good
foundation for the research work. I would also like to thank Mingang Wang,
Mayeul de La Tour du Pin and Matt Lempke for their assistance in conducting lab

experiments and contributions to the research work.

Finally, I would like to thank my family and friends for their countless support and
encouragement. They have helped me through difficult times and they are
always with me through the ups and downs. Without them, I would not have been

able to finish this task. Thank you very much!

iv

TABLE OF CONTENTS

LIST OF TABLES ................................................................................................. vi
LIST OF FIGURES .............................................................................................. vii
INTRODUCTION .................................................................................................. 1
CHAPTER 1: THE SPACE SHUTTLE - lN-DEPTH REVIEW .............................. 3
Thermal Protection System ............................................................................... 3
Reinforced Carbon-Carbon (RCC) Panel .......................................................... 7
Operating Conditions ...................................................................................... 11
Thermal Loading Conditions ........................................................................... 15
CHAPTER 2: RESULTS FROM PREVIOUS STAGE ......................................... 20
Design Concept .............................................................................................. 20
Functionally Graded material (FGM) ............................................................... 24
Powder Compaction ........................................................................................ 26
Core Design .................................................................................................... 30
CHAPTER 3: CURRENT STAGE - DEFORMABLE RING DESIGN .................. 34
New Concept .................................................................................................. 34
New Approaches ............................................................................................. 35
Layer Patterns ................................................................................................. 39
Cut-outs .......................................................................................................... 44
Mechanical Advantage Mechanism ................................................................ 51
Spring Design ................................................................................................. 55
CHAPTER 4: ALTERNATIVE DESIGN SOLUTIONS ......................................... 57
Concentric Rings ............................................................................................. 57
Slider Bar ........................................................................................................ 62
Segmented Plate ............................................................................................ 63
CHAPTER 5: DISCUSSION AND FUTURE WORK ........................................... 69
Discussion and Conclusion ............................................................................. 69
Future Work .................................................................................................... 71
BIBLIOGRAPHY ................................................................................................. 73

LIST OF TABLES

Table 1. Usage of different tiles and blankets [2]. ................................................. 5
Table 2. Operating conditions of the materials underneath the RCC panels. ....... 9
Table 3. Different compositions of FGMs. .......................................................... 27
Table 4. Negative CTE materials [5]. .................................................................. 30
Table 5. Design parameters for FGM core. ........................................................ 33
Table 6. CTE functions for COMSOL simulations ............................................... 38

Table 7. Simulation data of designs with lever mechanism and spring design. .. 52

Table 8. Material properties and dimensions for concentric ring design [25]. ..... 58

Table 9. Concentric ring simulation results [25]. ................................................. 59
Table 10. Diametric comparison of the type 1 slot concentric ring [25]. .............. 61
Table 11. Material properties of lnconel 718 [5] .................................................. 62

vi

LIST OF FIGURES

Figure 1. Surface temperature distribution of the space shuttle [11] ..................... 4
Figure 2. TPS tile distribution on the space shuttle [2] .......................................... 6
Figure 3. RCC panels on leading edge of wings, nose and chin (NASA). ............ 7
Figure 4. RCC wing panel [8] ................................................................................ 8
Figure 5. RCC on the nose [3]. ............................................................................. 8
Figure 6. RCC on the wing [10]. ........................................................................... 9
Figure 7. Cross section of RCC [7]. .................................................................... 10
Figure 8. Oxidation process of the RCC panel [10]. ........................................... 11
Figure 9. Gas temperatures from the CFD solution [11]. .................................... 12
Figure 10. Surface temperatures from CFD solution [11]. .................................. 13
Figure 11. Surface pressures from CFD solution [11] ......................................... 14
Figure 12. STS-5 mission data of space shuttle Columbia [12]. ......................... 15
Figure 13. Heating rate profile [10]. .................................................................... 17
Figure 14. Major heat flux consideration during re-entry [14]. ............................ 18

Figure 15. Schematic drawing of the fundamental design concept. (a) When the
skin is closed at room temperature; (b) opened at working temperature [5]. ...... 20

Figure 16. Conceptual idea of the working principle of the perspirable skin when

coolant gas is blown through the gap to the surface ........................................... 21
Figure 17. High temperature turbine blade [19]. ................................................. 22
Figure 18. Simulation model with ZrW203 (top surface: 30,000 W/m2; bottom
surface: 100 °C) .................................................................................................. 23
Figure 19. Illustration of functionally graded material. ........................................ 24

vii

Figure 20. Uniaxial powder compaction .............................................................. 28

Figure 21. Sintering temperature curve for ZrWZOB. ........................................... 29
Figure 22. CTE measurements of different samples produced in the laboratory
(refer to Table 3 for the compositions #1 — 6). .................................................... 31
Figure 23. Final design of the FGM core with channels for coolant gas to flow
through [5]. ......................................................................................................... 32
Figure 24. ABAQUS simulation of the perspirable skin [5]. ................................ 33
Figure 25. (a) New design concept with the deforming ring; (b) out-of-plane
deformation to increase the gap distance. .......................................................... 34
Figure 26. Meshing the deforming ring into pixels. ............................................. 36
Figure 27. COMSOL simulation of the core design using Table 6 data. ............. 38
Figure 28. COMSOL simulation of the fundamental design concept of deforming
nng. .................................................................................................................... 41
Figure 29. Slant layers design which opens a bottom gap for more coolant gas
flow. .................................................................................................................... 42
Figure 30. New slant layered design of the deforming ring. ................................ 43
Figure 31. Cut-out design with a wedge to overcome bottom gap issue. ........... 44
Figure 32. Cut-out designs analyzed. ................................................................. 45
Figure 33. Gap distance analysis for two cut-outs. ............................................. 47

Figure 34. Top and bottom gap distance comparisons for the four cut-outs
designs. .............................................................................................................. 48

Figure 35. Parametric study on the effect of the diameter has on gap distance. 50

Figure 36. Top gap result of the parametric study. ............................................. 50
Figure 37. Bottom gap result of the parametric study. ........................................ 51
Figure 38. Design with application of lever mechanism. ..................................... 53
Figure 39. Alternative design using the lever mechanism. ................................. 54

viii

Figure 40. Spring design with wedge t0p ............................................................ 55

Figure 41. Conceptual design of the concentric ring (left: design at room

temperature; right: design at working temperature) [25]. .................................... 58
Figure 42. Illustration of the slider bar (left: misaligned holes at room temperature;
right: the holes align at working temperature) [25]. ............................................. 62
Figure 43. Initial design concept of the segmented plate. ................................... 64

Figure 44. (a) CAD drawing of hexagonal plate by NX; (b) simulation of opening
at working temperature using ABAQUS ............................................................. 65

Figure 45. (a) Conceptual design of the square plate; (b) views of the plate
opening at working temperature. ........................................................................ 67

INTRODUCTION

When a space shuttle is returning to earth from its mission in space, the
temperatures on different regions, such as the nose and the leading edges of the
wings of the outer surfaces of the space shuttle would be extremely high. These
surface temperatures could reach as high as 1,650 °C [1], occurring at 20
minutes before landing [1]. These high temperatures are caused by the surface
friction of the shuttle and atmospheric gases when the space shuttle is trying to

slow down to the landing speed from a velocity exceeding 25,000 mph [1, 2].

The space shuttle orbiter is built mainly with 2024-T6 aluminum and graphite-
epoxy [3]; both materials can only withstand a temperature up to 175 °C. A
Thermal Protection System (TPS) is a crucial component to the shuttle because
it prevents heat from transferring to the shuttle structure [2, 3], which eliminates
the possibility of the structure degradation and thus protects the astronauts on

board.

The goal of this “Perspirable Skin” project, which is supported by Air Force Office
of Scientific Research (AFOSR), is carried out in an effort to design and fabricate
the prototype of an intelligent and automatic self-cooling system to the TPS that
are exposed to the highest temperatures during the crucial stage of the re-entry.
This project is divided into several stages. At the beginning stage, exploration of

suitable materials and a working design have been achieved. The motivation of

this thesis is focused on the second stage of the project, which is to modify and
experiment with other possible designs to create an out-of-plane gap deformation.
The objective is to achieve a larger gap opening through the incorporation of an

additional deformable ring to the current design.

CHAPTER 1

THE SPACE SHUTTLE - IN-DEPTH REVIEW

THERMAL PROTECTION SYSTEM

The orbiter operates at extreme temperatures as low as -157 °C while in outer
space; during re-entry the orbiter is oriented with the nose angled up about 40
degrees from the horizontal. This stage would last for about 25 minutes. In
between 45 and 80 km above the ground, the space shuttle must endure
tremendous heating, reaching a temperature as high as 1,650 °C [2, 3]. Other
than these two extreme conditions, the orbiter has to withstand local heating from
the engines, motors, and thrusters on board. Due to the wide variety of
temperatures that the space shuttle experiences in its mission, the TPS is

composed of different materials that have been carefully chosen and tested.

In an earlier NASA report, which dated back to 1989, it was revealed that the
surface temperatures distribution of a space shuttle as shown in Figure 1,
reaches a maximum temperature of 1,440 °C [3]. A more recent investigation
confirmed that the maximum temperature the space shuttle would experience
could reach 1,650°C [2]. Therefore, the materials that made up the TPS have
different material properties in order to handle these complex and extreme

thermal loading conditions.

«.3 2000F

 
 
 

 

 

 

 

 

 

 

.«.-. 1800F
1700F 1300]: 1.. 1700F
1600F 1800F
1700F 1800F “ . 1600F
18'0': 1600': - v" ' '1 1500F
1900 . ”W "' t; 4‘ L2000F
4”” rag—r... " 2300F
H ,w ' I _____
. C] ‘
L.....l .
Thermal maps ,
from flight data I
I 1200F
500F
600F 900F
00F 600E 4OOF 700
900 8 800F
1000': » 9001: 40.0F
90 _F
700F
900F
1400F I \ 1200F
1700F 1100F900F 8.00F 9 OF 800F 400F1000F 700.:

< 2200F (2100 -> 2200F)

Figure 1. Surface temperature distribution of the space shuttle [11]

The TPS has been improved over the years, current TPS uses 3 types of tiles, 2
types of insulation blankets, and the RCC as shown in Figure 2. There are about
24,300 tiles and 2,300 flexible insulation blankets covering the outer surface of
the orbiter [2]. These tiles and blankets cover different parts of the orbiter
depending on their properties and the respective heating conditions on different
locations of the orbiter. Each of these tiles or blankets has different thermal

properties and can withstand different temperatures as listed in Table 1 below.

Table 1. Usage of different tiles and blankets [2]

 

Category

Description

 

High-Temperature
Reusable Surface
Insulation (HRSI) Tiles

They have a black coating to evacuate the heat
during re-entry. They can protect the covered
areas up to 1,260 °C.

 

Low-Temperature
Reusable Surface
Insulation (LRSI) Tiles

They are the white tiles covering the upper
surfaces of the orbiter and reflect the solar heat
gain from the sun. They can withstand
temperatures up to 650 °C.

 

Felt Reusable Surface
Insulation (FRSI) Blankets

They cover the upper section of the payload doors
and the inboard sections of the upper wing
surfaces. They also cover about 25% of the whole
vehicle up to 370°C.

 

Reinforced Carbon-
Carbon (RCC)

 

 

They cover the areas on the nose, chin, and
leading edge of the wings where temperatures
can be as high as 1,260 °C. They can withstand a
high temperature of 1,770 °C while holding the
structure.

 

There are other types of tiles and blankets that also can be used on the space

shuttle. Fibrous Refractory Composite Insulation (FRCI) tiles can be used to

replace HRSI tiles in some areas. Another type of blanket, Advanced Flexible

Reusable Surface Insulation (AFRSI), can replace most of the LRSI tiles in order

to achieve the flight requirements more effectively [2].

 

 

Reinforced Carbon-Carbon (RCC)

- High-Temperature Reusable Surface Insulation (HRSI) Tiles

“ Low-Temperature Reusable Surface Insulation (LRSI) Tiles

.
. . -:
=-'3

'."-£-‘

’1‘: .(

Felt Reusable Surface Insulation (FRSI) Blankets

   

Metal or Glass

Flexible Insulation Blankets

' .
.«.-
' 5-;
W t .
.. : “é-
- _~ , "-
' -' .3.
_:"..u
J. - '..l
_u- ....

Figure 2. TPS tile distribution on the space shuttle [2]

 

REINFORCED CARBON-CARBON (RCC) PANEL

Some of the hottest areas on the surface of the space shuttle during the re-entry
are the nose cap; the chin, which is directly beneath the nose cap; and the
leading edge of the wings [4, 5]. Temperatures in these areas can get as high as
nearly 1,700 °C. These areas are covered by RCC, which is a gray composite,

along with lnconel foil insulators and quartz blankets [6] as shown in Figure 3.

 

Figure 3. RCC panels on leading edge of wings, nose, and chin (NASA)

The leading edge of the wings have 22 RCC panels installed and they are
approximately 6.3 mm thick; on the nose cap the thicknesses of these panels are

between 6.3 mm and 12.7 mm as shown in Figure 4 below.

' . F. . 'W' ;’m.;§.... -..,..,.

 

Figure 4. RCC wing panel [8]

These panels are manufactured by Lockheed-Martin’s Missile and Fire Control
Facilities. The public has very little information on the material properties and
performance on the RCC panels as they are classified materials. RCC panels
have relatively high thermal conductivity of approximately 70 W/m/K. As a result,
additional insulation blankets such as the Cerachrome are needed to stop the
heat transfer behind the RCC panels. There are also other alloys and metals

such Al 2024 and titanium used to support the structure [7, 8].

 
     
   
    
 
 

INTERNAL

ncc "035 c" msuurtou \

   
 
  
   
  
 

  
 
  

CLOSEOUT \4,, ncc T-SEAL
(Ll2200) 4 42,474 A SI N
Orl‘RCI-I‘Z - . -_'~.’_g‘ja SEAL

    
  

   

or FRCI-l?

Figure 5. RCC on the nose [3]

RCC
T-SEAL

RCC WING

INCONEL 718 PANEL

INCONEL 601
CERACHROME
INSULATION

  
 
 
  
 
  
 

-___oGAL4v TITANIUM
. SPAR FITTING

 

 

INCONEL 7I8
AB-312/SAFFII. THERMAL BARRIER

 
 

Figure 6. RCC on the wing [10]

In Figure 6, the lnconel 718 and lnconel 601 foils are used to stop the heat
transfer from the outermost surfaces into the structural support and interior of the
space shuttle. Table 2 gives the details of their respective operating

temperatures.

Table 2. Operating conditions of the materials underneath the RCC panels

 

 

 

 

Maximum service Approximate
AHOY Usage temperature (°C) meltigg poinfi’C)
IN 718 Clevis, spanner beam 980 1370
IN 601 Spar insulation foil 1,090 1370
Al 2024 Wing spar N/A 650

 

 

 

 

 

 

Carbon/carbon composites without any treatment and coating begin to oxidize at
around 500 °C [9]. Oxidation of the carbon/carbon composites greatly jeopardize

the safety and integrity of the space shuttle. In order to protect the carbon/carbon

composites, a RCC panel is coated with a layer of silicon carbide (SiC) and then
it is vacuum infiltrated with a phenolic resin to fill the voids in between. Due to the
different coefficients of thermal expansion of the two materials, an additional
sealant of sodium silicate glass is coated on top to prevent crack formation and
oxidation [7, 8]. Figure 7 shows the cross section of a RCC panel with the 3

layers of material.

sillcate glass

+- sodium
. -- '1

silicon carblde
conversion coating

 

carbon-carbon
composite

 

Figure 7. Cross section of RCC [7]

Despite the coating and sealant, oxidation of the RCC cannot be prevented. The
degradation of the RCC panel begins when the micro-cracks are formed on the
coating as shown in Figure 8. The oxidation process worsens as more and more
oxygen penetrates the coating and reacts with the carbon/carbon composite, and
finally with intense friction and ablation of the air outside, the coating would be

peeled off, exposing the carbon/carbon composites to the atmosphere.

IO

New RCC Increasing Interfacial Oxidation SIC Debond
With Flight Exposure

    

Craze cracks Small amounts of Increasing oxidation Silicon carbide

provide an oxygen penetrate areas decrease the "craze squares"

oxidation path to the coating and adherence of the may become

the silicon carbide react with the silicon carbide dislodged by

to carbon carbon matrix and "craze squares" to external forces

interface fibers, resulting in a the substrate (airloads, impact,
decrease in strength acoustics, flexure)

Figure 8. Oxidation process of the RCC panel [10]

OPERATING CONDITIONS

One of the tasks of this thesis is to perform a more in-depth study to understand
the actual operating conditions of the space shuttle during a re-entry. In the
previous work, the operating conditions were unclear and the thermal loading
conditions, temperatures, and many other parameters were estimated. Some of
the finite element method simulations were approximations of idealistic

assumptions.

After the tragedy of the Columbia space shuttle disintegration on February 1,
2003, NASA conducted a thorough investigation and studies on the re-entry. The
report of the Columbia tragedy has provided a lot of simulated results, which
have provided a better understanding of the actual conditions during a re-entry.

In the computational fluid dynamics (CFD) solution in Figure 9, which simulated

the re-entry of the space shuttle, the vehicle enters the atmosphere with an angle
of attack of 39.59 degrees, at an altitude of 69.319 km. The velocity of the space

shuttle reached a Mach number of 22.91, which is equivalent to 6.76 km/s.

 

  
  

W

0 1300 2600 3900 5200 I 6500 7800 9100 104001170013000
T (F) -460 1880 4220 6560 8900 11240 13580 15920 18260 20600 22940

Figure 9. Gas temperatures from the CFD solution [11]

The surrounding gas temperatures in the simulation revealed that gas near the
nose and chin of the space shuttle could reach a temperature over 10,000 °C
[11]. The high temperature would be caused by the friction between the air and

the surface of the space shuttle at great velocity.

12

The report also simulated the surface temperatures and the surface pressures on
the space shuttle in Figure 10 and Figure 11 respectively. Because the space
shuttle was entering the atmosphere at approximately 40 degrees to the
horizontal, most of the high temperature and pressure regions occur at the belly

and the lower surfaces of the wings.

T (K) T (°F)

2 1900 2960
"“ 1740 2672
1580 2384
1420 2096
.. 1260 1806
, 1100 1520
" 940 1232
780 944
Q”; _ 620 656
460 368
300 80

 

T (K) T (°F)

. . 1900 2960
‘ 1740 2672
1580 2384
1420 2096
~ , 1260 1806
.- 1100 1520
' 940 1232
780 944
620 656
460 368
300 80

   

Figure 10. Surface temperatures from CFD solution [11]

I3

P (Pa) P (psf)
3600 752
3240 67.7
2880 60.2
2520 52.6
2160 45.1
........_, 1800 37.6
' - 1440 30.1
1080 22.6
720 15.0
360 7.52
0 0

 

 

P (Pa) P (psf)
3600 75.2
3240 67.7
2880 6042
2520 52.6
2160 45.1
1800 37-6
1440 30.1
1080 2246
. 720 15.0

360 7.52
o 0

   
 

Figure 11. Surface pressures from CFD solution [11]

It can be observed from these two figures that the highest temperatures and
pressures locations fall on the nose, chin, and the leading edge of the wings
reaching approximately 1630 °C and 3000 Pa. Most of the lower surface of the

space shuttle had a temperature over 1000 °C and 1000 Pa.

THERMAL LOADING CONDITIONS

WIth little understanding of the actual operating conditions during the re-entry of
the space shuttle, previous finite element method simulations were based on the

assumption that the outside air temperature was a constant at 760 °C. This

assumption was based on the fact that ZrW203 dissociates at a temperature

around 800 °C. In order to produce more accurate simulation results that are as
close to the actual condition as possible, it is important to study the thermal

loading conditions during the re-entry.

5x19‘ 5x105

     

 

 

 

 

 

 

so
me FLIGHT DATA
X
1‘ 3°3 1‘ x104 0 ALTITUDE
D VELOCITY
4 o ANGLE OF ATTACK
129 12s- 0.0”. — NOMINAL 44 -4
1o __ 1° _ 3' ANGLE or ATTACK
1330 - 3 8 3
3 5 q {3' a:
W E. < . m‘
E 8 " In 8 P t o t O
- O f. a on can 3 . _ 3
I: I: I5 8 I-
8 I- o -I ..I
JG-gerflzor ~2§~2<
w o
> z
<
‘ t- ‘ I- 3%
10 t' VELOCITY a 4 1 J 1
o ..
2 - 2 *- un ° . o
D ‘4
|+PEAI< HEATING-b “a“ °o °
REGION aha o
0 _ 0 . I l L 1 I L 1 . ‘3‘: ° o.o _l o J o
o 2 4 e 8 10 12 14 16 18 20 x102
TIME, sec
REENTRY TIME, osec Toucuoowu TIME,
1821sec

Figure 12. STS-5 mission data of space shuttle Columbia [12]

I5

In the data collected from the mission STS-5 with space shuttle Columbia in 1982,
as shown in Figure 12, a re-entry process lasts for approximately 1,800 seconds
where there are about 800 seconds in which the space shuttle undergoes intense
heating. The exact phenomenon that is occurring during the re-entry is too
complicated to predict. It can be caused by a combination of the viscous
dissipation in the boundary layers, the thermal radiation, and transition to
turbulence and instabilities on the curve walls or the strong shock layers and

thermo-chemical effects [13].

The re-entry process started at an altitude of 120 km. Between 120 km and 80
km, the density of the air increases as the altitude decreases, but the velocity of
the space shuttle remains steady. The peak heating region starts when the space
shuttle is between 80 km and 45 km of altitude. Most of the kinetic energy is
dissipated into heat, resulting in the velocity of the orbiter being greatly reduced

by friction and drag.

To study the thermal loading conditions, there are three hypotheses assumed in
order to simplify the problem. Referring to Figure13, the maximum heat flux,
which occurs at the peak heating time during the re-entry, is relatively steady.
Therefore, one can assume that a steady state is reached between the absorbed

heat and the emitted heat during that period of time.

  
 

 
 
     

 

 

 

 

 

 

 

   
    

 

 

 

 

0 0,0 Q0 (5“
Q o c
‘9“: 4,8“ .oé‘bozqf. 45\
9 '4‘ s ‘ a Q~ .
\‘9 60¢) 91* o°\‘° o‘ \. \
o§1§°Xé°Q¢Q 00* . §®§§é§§°\§’.\o‘o
8° 00 Q0 \>° 9' ‘ %\ ‘00-'43“ *0 40b
*\&° &Q\°\%‘SFSCJ¢O$°\O\ $‘Z\§ $2050"
<~ <‘ 0‘ a V 6* o x
/ to 90 90x9 ‘b V» a" 5‘0 (‘5 6:14?
1.0 I"- Reference Heating
_|Il. Temperature (°C)
0.9 1'- 1650
'8 0'8 1315
% 0.7 III- I 1 980
E 815
h. 0 6 i- .
2° ' 650
- o.5 -- """ 480
g 33.1“: 315
°‘ 0" " it? 150
4? 0.3 -- f4 38
§ 0
I 0.2 “I" 423355
0.1
0.0 l L .l l l 2. l l l l _1_ l l L 1
0 360 660 960 1260 1560

Time (El + sec)

Figure 13. Heating rate profile [10]

According to [13], most of the heat is stopped near the outer surface of the space
shuttle. There are layers of insulations behind the RCC layer; therefore, another
assumption made in simplifying the problem is to neglect the conductive heat flux

that goes into the surface coating in Figure 14.

The last assumption neglects the radiation flux going into the TPS system
because as the space shuttle travels at a speed below 8 km/s, the radiation flux
from the atmosphere (radiation flux in (1) in Figure 14) is very small compared to

the convective heat flux.

17

Free stream ., , , n cive
..4- RadiatIon CO ve t

flux in (1) flux Boundary layer
or Shock Layer

Radiation
"WW9. .

     

Surface Coating

RCC

ooooo

.......
'''''''

Backup material

Figure 14. Major heat flux consideration during re-entry [14]

WIth the three assumptions mentioned above, the remaining heat transfers that
are left to be considered in Figure 14 are the convective flux and the radiation
flux out. According to [13, 15, 16], these two heat fluxes have to be at equilibrium,
which implies that the convective flux in must be balanced by the radiation flux
out. As a result, the convective heat flux that is exerted on the surface of the

space shuttle can be calculated by the radiation formula shown in Equation (1).
4
Q = 8 x 0' X T (1)
In Equation (1):
Q is the heat flux emitted from the surface, unit: W/m2

8 is the emissivity of the surface

18

0' is the Stefan-Boltzmann constant, 0' = 5.67 x 10'8 W-m'ZK'4

T is the surface temperature, which is in Kelvin scale
From the first assumption, the maximum local heating temperature on the
surface of the space shuttle is approximately 1,650 °C. Together with the

emissivity of the wall that is around 0.85, the total radiation heat flux is

Q = 666,000 Wim2 (2)
The highest peak heating reported by the NASA [17] was 60 Btu/ftzls. This is

equivalent to 681,391.6 W/m2. This value when compared to the approximated

radiation flux has about 2.26% error; therefore, the approximation found using

Equation (1) is fairly accurate.

According to the same report [17], the theoretical model actually predicted that
less than 5% of the heat is really absorbed and transferred by the wall into the
vehicle. Taking this prediction into account, the actual local heat flux that goes

through the wall is approximately
~ 2
Q ~ 33,300 W/m (3)

For the convenience of finite element simulation, the heat flux is taken to be

30,000 W/mz.

19

CHAPTER 2

RESULTS FROM PREVIOUS STAGE

DESIGN CONCEPT

After knowing the limitation and the structure of the current TPS and RCC, it is
important to develop a new self-cooling system that can enhance the durability
and reliability of the space shuttle. In the beginning stage of this project
“Perspirable Skin”, Sun and Oguz [5, 18] proposed the design of the perspirable
skin to be a multifunctional material system, which is arranged in a “peg and

hole” fashion.

RCC

 

Figure 15. Schematic drawing of the fundamental design concept. (a) When the
skin is closed at room temperature; (b) opened at working temperature [5]

20

The cores in the perspirable skin would be in contact with the RCC under normal
room temperature and conditions. During a re-entry, when the surface is heated
up to a high temperature, the cores with negative CTE would shrink in order to
create small gaps which are sufficient to allow coolant gas to pass through to the

surface of the space shuttle.

Air Flow
Coolant gas

 

',. .2.-

Figure 16. Conceptual idea of the working principle of the perspirable skin when
coolant gas is blown through the gap to the surface

The idea of the perspirable skin was developed from the combination of the
human perspiration system and the existing cooling system applied to high
temperature turbine blades. In the human body, perspiration is the most effective
method for regulating the body’s temperature. In human skin, there are
numerous sweat pores; when the temperature of the body gets higher than
normal 37.0 °C (98.6 °F), the sweat pores expand and the sweat glands
underneath the skin produce sweat. The sweat is mainly water and some
minerals. As it is evaporated from the surface of the skin, heat is removed from

the skin and, thus, the body is cooled down.

21

This natural phenomenon is also adapted in a high temperature turbine blade
cooling system. In advanced gas turbine blades, repeated rib turbulence
promoters are built into each blade through which compressed cooling air flows.
The compressed cooling air is ejected to the surface of the blade through tiny

holes and slots. The cooling air forms a thin film that cools and protects the blade

   

from the heat [19].
Tip cap holes
Squealer tip Film cooling holes
”7.14.,
., .
O .
- ; . I
. . , I
. . I Trailing edge
. I , : slots
. : . .
Blade : f 1 :
platform . ; ' I
I
I

 

\

 

 

: 2 Dovetail
I I I
Cooling air

Figure 17. High temperature turbine blade [19]

Such a cooling system has proven to be very effective in advanced high

temperature gas turbine blades; nonetheless, it cannot be applied directly to the

space shuttle. Surfaces of the space shuttle orbiter would have to endure
extreme temperatures, high pressures during re—entry, and a vacuum while
orbiting in deep space. Thus, holes for cooling gas flow must be able to open and
close, in response to the temperature and condition changes, like sweat pores in

human skin.

The highest working temperature for the perspirable skin is estimated to be

around 750 °C to 780 °C because the compound ZrWzoa would dissociate at

around 777 °C [5].

Surface: Temperature (°C) Max: 795-35

 

 

 

 

 

 

0.02 700
0.015
600
0.01
t ‘ 500
0.005
" ' 400
0 .
300
-0.005 5
-0.01 200

 

 

 

 

1
0 0.005 0.01 0.015 0.02 0.025 0.03 0.035 Min'gg77

Figure 18. Simulation model with ZrW203 (top surface: 30,000 W/m2; bottom
surface: 100 °C)

23

In a simulation performed with ZrW203, the upper boundary condition was set to

have a heat flux of 30,000 W/m2 applied to the surface; the bottom boundary

condition was set with a constant temperature of 100 °C. The result gave a

promising prediction, which agreed to the estimations made.

FUNCTIONALLY GRADED MATERIAL (FGM)

Functionally graded material (FGM) is a composite material that is characterized
by its gradual change in the composition over its volume. As the composition
changes gradually, functionally graded material is non-homogeneous and all the
material properties vary with the location in the material. Figure 20 illustrates the

difference between a homogeneous material and a FGM.

      

 

Homogeneous material Functionally graded
material

Figure 19. Illustration of functionally graded material

In the laboratory, most of the effort has been focused on manufacturing FGM
from zirconia (ZrOz) and aluminum oxide (AI203) [20], as well as ZrW203 from

ZrOz and tungstate (W03) [21]. ZrW203 has been chosen to be the primary

24

material for the core design because it exhibits a negative coefficient of thermal
expansion, i.e., it shrinks when it is heated up. This special feature is really

suitable for the application in the perspirable skin design.

There are several different ways to produce FGM proposed by researchers
around the globe. Lambros et al. have developed FGMs using ultraviolet (UV)
irradiation, in conjunction with Poly(ethylene co-carbon monoxide) (ECO), which
is a material that degrades rapidly under UV light [22]. Pei et al. explored the
fabrication technique utilizing laser beam to melt the alloy substrates, together
with additive metals to form FGMs [23]. The mentioned techniques have
drawbacks in which the materials have specific properties; such as materials

reactive to UV light. Such limitations have prohibited the application of these

methods to the fabrication of the perspirable skin, as the ZrOz and WO3 powders

are rather inert and do not exhibit such properties.

Mott et al. have studied and opened one promising method in fabricating FGMs,

using an ink jet printing technique. The technique is similar to 3D rapid
prototyping which the powders used in the experiment were ZrOz and AI203, The
printer ejected “inks” were made of Zr02 and Al203 solutions [24]. This technique

has revealed the possibility of fabrication FGMs with multi-dimensional gradients
in a controlled and simple setup, despite the fact a one-dimensional FMG was

fabricated for the research work.

25

The FGM that is fabricated in the laboratory is a one-dimensional FGM, where
the material properties would vary along the thickness of the material only. After
a study to examine the many different fabrication methods for FGMs, the powder
compaction technique is chosen to fabricate the core for the perspirable skin.

Details of the technique will be discussed in the next section.

POWDER COMPACTION

Powder compaction is chosen to be the fabrication method for the perspirable
skin for several reasons. This method is commonly used in industries that deal
with ceramic, metal, or pharmaceutical powders because of its simplicity and
accuracy. Powders are first compacted using die and punch, and then a sintering
process may be needed to form the final products. The final products normally
achieve the desired dimensions. Powder compaction would also be an economic

method as little or no machining is required after the final products are fabricated.

The powder compaction consists of 3 major steps: preparing and mixing the
powders, compressing the powder to the desired shape, and sintering the
compacted samples. The first step involves preparing the powders by breaking
agglomerates into smaller, fine particles in a mixing tumbler with mixing beads.
This process ensures that the powder particles loosely flow during the
compaction in the following steps. Once the powder is fully crushed into fine

particles, correct proportion and an amount of different powders are weighed and

26

mixed together (refer to Table 3). This process is important as the composition of

the powders would control the material and thermal properties of the final product.

In order to produce ZrW203, the Zr02 and W03 powders are mixed in the

stoichiometric ratio of 1:2 (mass ratio is mwoaz mzm =1:0.266) [5]. The other

ZrOzNVO3 mixtures with other mass ratios are shown in Table 3 below.

Table 3. Different compositions of FGMs

 

 

 

 

 

 

 

 

 

 

 

 

- - WO3IZr02 mass ZrW203IZr02 volume Final
Comggzltlon ratio In reactant ratios in the slntered relative
powder sample density

1 0.15921 20:80 77%

2 0.26421 30170 79%

3 0.3821 38.9261 .1 80%

4 0.59321 51 .4248.5 82%

5 1.09631 70230 83%

6 2.30721 90210 84%

 

The second step of the process is compressing the well-mixed powders into the
desired shapes using the conventional cold, uniaxial compaction. This step
involves applying a pressure of 70 MPa to the die and punches, which are made
of high grade steel, along the vertical axis. This compacting process is done at

room temperature and the green compact is close to the desired dimensions.

The compaction process is illustrated in Figure 18 shown below. The mixed

powder is transferred to the die with the bottom punch inserted in it. The top

surface is flattened before the insertion of the top punch. This is to ensure that

27

the top punch would have maximum surface contact with the powder and evenly

distribute the pressure during the compaction.

Top Punch—fl

Die —1
Powder
Bottom Punch

 

 

 

 

 

Filling Flattening Entry 0' Compaction Ejection
top punch

 

 

 

Figure 20. Uniaxial powder compaction

The compaction is achieved by applying 70 MPa on the top punch by a MTS
lnstron machine. Lastly, the green compact is ejected from the die from the
bottom and would normally have a density varying from 50% to 70%, depending

on the pressure applied.

Since the green compacts from the compaction process are not fully dense, the
third step of the process, which is sintering, is often essential to produce denser
final products that can reach a density of 95% or higher. Sintering is a thermal
process that bonds the powder particles together at high temperatures by solid-
state atomic transport phenomenon. The green compacts are heated to high
temperatures yet below their respective melting points, and the sintered products

would exhibit a much higher density and strength. The typical sintering curve for
ZrW203 is shown in Figure 21. The balanced chemical reaction formula for the
process is shown as follows:

ZrOz + 2WO3 9 ZrWan (4)

28

Since the ZrW203 will decompose at high temperature, the green compact is

kept in a platinum crucible during the sintering process. The quenching process

at the end of the sintering is to prevent the decomposition of Zr1N203 [18].

After the three processes, the final parts are often functional. However, while
the dimensions could just be close to the desired dimensions, sometimes,
additional finishing processes might be needed. These processes include
machining, grinding, or polishing; these additional processes are essential in
order to achieve dimensional accuracy and specific features where powder

compaction technique alone cannot achieve.

 

1200-

 

j

..I
O
O
O
l

800 -
600 -
400 -

200 -

Sintering temperature (°C)

 

O I l v I I I ' l '
0 2 4 6 8 10
Sintering time (hours)

 

Figure 21. Sintering temperature curve for ZrW203

29

CORE DESIGN

Among various fabrication methods and materials, the material that should be
chosen for the perspirable skin must meet the extreme conditions. There are
several materials that exhibit negative CTE property, Table 4 below shows some
of the materials:

Table 4. Negative CTE materials [5]

 

 

 

 

 

 

 

Material CTE range (10'6 K4) Temperature range (K)
ThP207 -9.3 to 4.3 573 - 480

up207 -7.1 to 3.1 773 - 1773
ZrV207 —11.1 to 15.4 380 - 875
ZrW208 -13.2 to -6.5 3 - 1050
HfW208 -129 to -6.4 3 - 1050

 

 

 

 

Among these materials, ThP207, UP207, and ZrV207 are not suitable for building the

cores for the perspirable skin due to their low CTE ranges; and at higher temperatures
they would lose their negative CTE property. In addition, their working temperature
ranges do not meet the requirement of a surface material for the space shuttle. Both
Hafnium (Hf) and Zirconium (Zr) meet the required working temperature range
and exhibit a good negative CTE range. However, Hf is a rare material; and,
therefore, Zr would be chosen as the best fabricating material for the perspirable

core.

The CTE measurements of ZrW203 and other varieties of ZrOzNV03

compositions are shown in Figure 22. One very interesting phenomenon

30

observed is the sudden drop in CTE values for all the compositions except for

ZrOz. This huge shrinkage in CTE values occurred at about 150 °C in all the
mixtures of ZrOzNVO3 is due to the phase change of W03. As the amount of

WO3 in the mixture increases, the shrinkage would be bigger. If the phase

change regions are ignored, the CTE values of these compositions behave as

constants for a wide range of temperatures.

CTE Measurement

 

 

 

 

1.50505
1-0054’5 ‘7 ' . -1 ;.;: :;:;*. ::::..::; -; ;;.. . . : . 2 ...." ,.-_, : .-. " "
5.00E-06 _ _ 2
§ 0.00E+00 .4 """" 3
2 . g . -x-—4
g -5.00E-06
I! i -I—"5
0 4.00505 , .. __ -6
_1.50E-05 '__ _ yr;-,2-- . ._ ‘wmwmnzrwzos
4444-44-44sz
0 100 200 300 400 500 600 700

Temperature [00]

Figure 22. CTE measurements of different samples produced in the laboratory
(refer to Table 3 for the compositions #1 — 6)

After a series of studies in simulating the core designs in a commercial FEM
software ABAQUS, the final geometry of the core design for the perspirable skin
took the form shown in Figure 23. The core consisted of 5 layers with their
respective properties shown in Table 5. In the simulation, the top layer of the
core was exposed to a temperature of 760 °C and the bottom layer was set to a

31

constant temperature of 100 °C. The top part of the core shrank isotropically in

conjunction with the shrinkage in the surrounding RCC, generating a total gap

width of approximately 100 um (100x1015 m).

As the lower part of the cylindrical core was made of composite materials with
smaller negative CTE values, the shrinkage that occurred in the bottom was
relatively minute when compared to the top. Due to this small shrinkage, the
bottom part of the care would remain in contact with the surrounding RCC skin.
This design feature enabled the core to attach to the system while opening a gap
near the surface. The four semi-annular channels were incorporated into the core

to allow coolant gas flow during operation.

 

Bottom

Figure 23. Final design of the FGM core with channels for coolant gas to flow
through [5]

32

Table 5. Design parameters for FGM core

 

 

 

 

 

 

 

 

 

 

 

 

Layer erzouzmz The'mf'. Young's Modulus Thickness
number olume ConductIVIty (GPa) (mm)
Ratio (WIm/K)
1 100100 0.9672 4.12 :t 0.13 3
2 88:12 1.0696 3.65 i 0.13 1
3 76:24 1.1346 31810.10 1
4 64:36 1.2099 2.71 :l: 0.08 1
5 52:48 1.4531 2.24 :I: 0.08 2

 

(layer 1: top layer; layer 5: bottom layer)

The ABAQUS simulation shown in Figure 24 represented one quarter of the
design because the setup was axis-symmetric. The result illustrated that a gap
was formed on the top part surface of the care while remaining in contact with the

RCC skin in the bottom part.

RCCsMn

 

Gap for by shrinkage

Figure 24. ABAQUS simulation of the perspirable skin [5]

33

CHAPTER 3

CURRENT STAGE - DEFORMABLE RING DESIGN

NEW CONCEPT

mm a proven success in the fabrication and the simulation of the care, this stage
of the project aims to enhance the design to increase the gap distance. The new
design concept involves the introduction of a deformable ring. The deformation of
the ring generates an out-of-plane motion, which is believed to create a bigger

opening.

 

(a) (b)

Figure 25. (a) New design concept with the deforming ring; (b) out-of-plane
deformation to increase the gap distance

The main focus of this thesis would be to investigate different approaches and
combinations to achieve the desired out-of-plane deformation of the deforming

ring and the fabrication methods that are suitable for the proposed methods.

34

WIth the knowledge of fabricating FMG by powder compaction, the study is
conducted to investigate the possibilities of arranging the powders differently
within the deforming ring to create a smart material, which produces the desired

deformation.

NEW APPROACHES

At the beginning stage of the project, core designs were restricted to flat layers of
different composite materials. This was driven by the limitation of the fabrication
method because the existing lab equipment and technology are only capable of
achieving horizontal layers of powder deposition and compaction. However, such
a limitation would circumscribe the deformation of the core to a linear and one-
dimensional contraction. In order to further increase the gap opening, out-of-

plane deformation and other solutions must be explored.

In order to generate an out-of-plane deformation in the deforming ring, the
powder distribution must be irregular and more versatile. The approach to start
this new design concept is to divide the cross-section of the deforming ring into
fine squares named “pixels". Depending on the amount of shrinkage desired,
each pixel would then be assigned a material property based on the CTE value
of the material (refer to Table 3). When all the pixels have been assigned a

material property, adjacent pixels that have the same material properties would

35

be coerced to form the same material. A simplified example of the resulting

design of the deforming ring is shown in Figure 26.

Zirconium
tungstate
wear-94444154 - ~81~aqa~n+4 r

4

I l b.- Material 6
be

 

 

 

 

Material 5

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

i '\
Material 2 _ _ /1 Material 4
“ ‘ l . ‘
Material 5 / Mater '3' 1 Material 3
Cross-section of the deforming ring
- Zirconium
Material 2 Materlal 5 tungstate Material 6

    
      

Material 5 Material 1 Material 3 Material 4

Axis of symmetry
Attached to RCC

Flgure 26. Meshing the deforming ring into pixels

36

be coerced to form the same material. A simplified example of the resulting

design of the deforming ring is shown in Figure 26.

Zirconium

tungstate
M_a_a..san#— aaaaeaw .
/

I l Material 6
a...
r" .

r—‘d—M
‘ 1

 

 

 

 

' Material 5

 

 

 

._ _ Material 4

/l/

' 1L
3 T ’5 _ i
: Material 5 / Material 1 \ Material 3

J. ~---.-.~ w-n.-.,.....,,. A

 

? Material 2

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

o lessees. or the ogrear‘..a"‘r‘r'r;'g‘"“

Material 5 Zirconium

tungstate Material 6

Material 2

  
 

   
  
     

Material 5 -
Material 1 Material 3 Material 4

Axis of symmetry
Attached to RCC

Figure 26. Meshing the deforming ring into pixels

36

Previously, finite element simulations were done using FEM packages ABAQUS
and ANSYS. In order to simulate continuously varying material properties that
are dependent on temperature within the geometry, the finite element models
were divided into layers with fixed material properties and then glued together as
a whole. Another limitation that these packages inherit is the limited capability of
computing the coupled physics of structural and thermal analyses simultaneously;
therefore, the actual computations had uncoupled the physics that might lead to

slight deviations in accuracy of the results.

New FEM commercial software, COMSOL Multiphysics, is adopted in order to
accommodate the new concept design. There are three main advantages of
COMSOL over the other commercial packages: the abilities for solving multiple
coupled physics simultaneously, processing properties that are functions of other
variables, and the compatibility with MATLAB. The perspirable skin project
involves the structural mechanical and the thermal heat transfer analyses that
are coupled together, where the mechanical properties such as coefficient of
thermal expansion depend on the temperature. The user can assign multiple
physics problems and COMSOL will solve the FEM solutions simultaneously.

This will guarantee that the solutions would be realistic and accurate.

The second advantage of COMSOL over other FEM packages ties to its ability to
compute multiple physics problems simultaneously. Rather than entering a

constant value for properties such as CTE or thermal conductivity the user can

37

input an equation that is a function of other parameters, such as temperature.
Referring to Figure 22, CTE values that are extrapolated from the linear region of
each material are entered to COMSOL as a function of temperature (T) shown in
Table 6 below.

Table 6. CTE functions for COMSOL simulations

 

 

 

 

 

 

 

 

 

Composite Type CTE linear curve fit equation (K4)

1 CTE = 2E-09 X T + 4E-06

2 CTE = 1E-09 x T + 2E-06

3 CTE = 1E-09 x T + 8E-07

4 CTE = -1E-23 x T - 8E-07

5 CTE = -1E-22 X T - 3E-06

6 CTE = -2E-22 x T - 6E-06
ZrW208 CTE = 1E-22 x T - 7E-06
ZrOz CTE = 2E-09 X T + 8E-06

 

 

 

 

As linear as they may seem in Table 6, the CTE values are affected by the high
working temperatures, which are as high as over 700 °C. This phenomenon is

also valid in other material properties such as Young’s Modulus.

0.005 """'o.01,—""_l 0.015

    
   

._. llll

...........

0.005 .
\0 01 . d

0.015
Figure 27. COMSOL simulation of the core design using Table 6 data

38

COMSOL Multiphysics is tested with the data from Table 6 and verified against
the results from previous simulations performed using ABAQUS or ANSYS.
Figure 27 shows a quarter of the core design from the previous stage of the
project. The gap distance generated by the deformation of the core is compared
to previous results and showed a consistency in the displacement. The deformed
shaped also matched to the previous work. This verification and validation of the
simulation results computed by COMSOL support the migration to this new

commercial package.

Last but not least, COMSOL is capable of compiling the CAD and FEM models
into MATLAB script files. This feature adds the ability to perform a parametric
study to the design process. When the FEM file is converted into the dot-m file
(.m extension), which is accessible in MATLAB, parameters can be edited and/or
rewritten as variables. This would allow MATLAB to run the simulations varying
the values of these parameters to produce the results of a given range of inputs;
which minimizes the effort in editing the CAD and FEM files many times in

COMSOL.

LAYER PATTERNS

The study of the deforming ring concept took several steps and it is still in the
process of exploration and experimentation. With the assistance of COMSOL

Multiphysics, the first concept of the deforming ring kicked off with the

39

composition shown in Figure 26. The initial design is based on the horizontal,
rectangular distribution of composite powders for simplicity. The simulation result
produced by COMSOL revealed a gap formation at the top surface of the
deforming ring; while the bottom was still in contact with the core. This formation
of the powder arrangement met the desired deformation, which retained contact
with the core at the bottom while opening a larger gap at the top. Figure 28
shows the COMSOL simulation of the cross-section of the deforming ring (the

core is omitted in the simulation to reduce computational time).

The drawback of such design is that at least four semi-annular channels must be
machined to the bottom of the core as shown in Figure 23. This would add
additional complexity to the design and thus increase the production cost of the
care due to the machining process. Another drawback with this design is the

limited cross-section area of the flow passages.

In order to achieve a better cooling effect, the air flow passage must be widened
to increase the volume of coolant gas flowing to the surface of the perspirable
skin. The four semi-annular channels could only provide'minimal clearance for
the gas to flow through. New designs are considered to create greater
displacement in the bottom, which would provide feasible solutions to the two

mentioned disadvantages of previous core design.

40

Surface: second principal stress [Pa] Max: 6.178e7 Max: 3.205e-5
Boundary: total displacement [m] x 107 x 105
x 10'

 

 

-

VT

.01
4t.
w .
. 'l
, _ ' Ili- .
. . . r 1.“ j
4 , ,4; ,
'_:‘,,.‘.‘. ‘i.p.;uh.4.,. 44‘
u .;.:,..1&m;nn. I1: .-

1'l"

 

 

 

 

 

 

 

 

 

 

 

0.005 0.006 0.007 0.008 0.009 0.01 0.011 0.012 0.013 Min: -7.052e6 Min:0

Figure 28. COMSOL simulation of the fundamental design concept of deforming
ring
By understanding the shrinkage behavior, the new powder formation is arranged

in slant layers as shown in Figure 29. The top and bottom most layers (zone 1

and 4) are assigned with er203 materials to produce the largest shrinkage. The

middle zone 2 and zone 3 are assigned with composite type 6 and type 5
respectively (Table 3). The black lines mark the non-deforrned shape in which the
core and the deforming ring are in contact. When the design reaches the working
temperature, the top surface is exposed to a high temperature of 760 °C while

the bottom surface is maintained at 100 °C. This temperature gradient would

41

cause the top portion of the deforming ring to bend upward (outward deformation

if considered in 3-dimensions). This great deformation is achieved by the

shrinkage of the large amount of 2rw203 material in zone 1. The change in

materials through zone 2 and 3 help to create a gradual reduction in CTE values,

in which the stresses formed between zones are minimized. Despite the lower

temperature, zone 4 has used ZrW208 to create gap formation at the bottom.

:Axis of symmetry 2 1

 

 

Non-deformed geometry

Figure 29. Slant layers design which opens a bottom gap for more coolant gas
flow

In order to explore the possibilities to further widen the gaps, the finite element
model has been converted into MATLAB script file for parametric study. The aim
is to investigate the correlation between the cross-section area of zone 1 and gap
distances at the top and the bottom. In Figure 30, the diagrams show the
boundary dividing zone 1 and 2 moves in different directions and gradually
generates different cross-section areas. The left shows the transformation of
zone 1 diminishing and eventually zone 2 takes over; whereas the right illustrates

that zone 1 has shrunk to the side wall.

42

 

 

 

 

 

Figure 30. New slant layered design of the deforming ring

The results of the parametric study reveal an inspiring discovery: opening a
larger top gap would result in a smaller gap at the bottom. From the left

transformation, deformation of the deforming ring is minimal when there is no

2rw203 material on the top section. As a result, the top gap distance is reduced,

but the bottom gap distance is increased with less deflection of the ring. The
transformation on the right illustrates that the deformation is increased and thus
bringing more displacement at the top gap. However, this has an adversarial
effect to the bottom gap, where the gap distance is reduced because of the

higher deflection of the deforming ring.

This inverse phenomenon has brought the understanding that solely changing
the powder arrangement cannot achieve the desire goal. The geometry of the
deforming ring must be re-examined and designed to increase the top gap

distance while maintaining a significant clearance at the bottom.

43

CUT-OUTS

Many other ideas and arrangements of the powder layers have been explored
before, but none of these solutions has proven a significant increase to the gap
distance both at the top and bottom. The major issue in the design remains:
increasing the gap distances while providing significant clearance between the

core and deforming ring at the bottom.

To effectively solve the issue of the clearance at the bottom of the opening is
achieved by altering the geometry of the deforming ring as shown in Figure 31
below. A wedge is cut off the bottom corner of the deforming ring to eliminate the
effect of narrowing the clearance at the bottom when the deforming ring bends

upward.

 

 

 

 

 

 

 

Figure 31. Cut-out design with a wedge to overcome bottom gap issue

44

This wedged arrangement of the deforming ring, however, does come with a
reduction on the overall deformation of the ring; the top gap distance is slightly
affected by this proposition. This is a result of the reduction in the amount of

negative CTE materials that make up the deforming ring.

To compensate the effect of the wedge cut-off, semi-circular cut-outs are added
to the deforming ring geometry as shown in Figure 31. The purpose of
incorporating cut-outs is to reduce the strength of the structure by creating higher
stress points; these locations will be the weak spots for larger deformation. To
determine how many semi-circular cut-outs should be put into the geometry or
how large each cut-out should be, four different designs are used to investigate
the number of cut-outs required for the geometry; and MATLAB in conjunction

with COMSOL are used to perform the analyses.

 

 

 

 

 

One cut-out Two cut-outs

 

 

 

 

Three cut-outs Five cut-outs

Figure 32. Cut-out designs analyzed

45

For each of the above designs, a parametric analysis is performed to investigate
the relationship between the cut-out radius and the generated gap distances.
Coincidently, all the four analyses produce a similar trend in both top and bottom
gap distances as shown in Figure 33. It must be emphasized that the bottom gap
results shown in Figure 33 and 34 do not include the wedge cut-off design;

therefore, the resulting bottom gap distance is minute.

46

1 .0567

1 .0566

1.0565

Gap distance [m]

1.0563

1 .0562

1.0561
0

8.466

8.465

8.464

8.463

8.462

8.461

8.46

Gap distance [m]

8.459

8.458

8.457

8.456
0

X10

 

T

I

T

 

I

Top gap distance VS fillet radius

\

J

 

 

X10

2
Fillet radius [m]

Bottom gap distance VS fillet radius

4X10'4

 

T

Y

I’

I

I

T

I

 

l

j-

T

L

T

1

 

 

Figure 33. Gap distance analysis for two cut-outs

2
Fillet radius [m]

47

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1 a
1.06 X 0 Top gap distance
1.059 ~
1.058 - -----------------------------------
2 1.057 4
m _______________________________________
0
g 1.056 ~
,‘g’ 1 circle fillet
g 1.055 ~ - - -2 circle fillet
t9 - -- «3 circle fillet
1.054 - 5 circle fillet
1.053 ~
1.052 .. .........
1'0510 i 2 3 4 x 104
Fillet radius [m]
.4 .
10 X 10 Bottom gap dlstance
945 . 1 circle fillet
—- - -2 circle fillet
-— -- -3 circle fillet
9 r 5 circle fillet
E.
0
‘c’
g 8.5 r _______________________________________
.9
‘O
O.
(U
(9 3 .
7.5 4 _________________________
70 i 2 3 4 x10“4

Fillet radius [m]

Figure 34. Tap and bottom gap distance comparisons for the four cut-outs
designs

48

In Figure 34, the results from the parametric analyses from MATLAB are
combined and compared. The top and bottom show the comparisons of the top
and bottom gap distances respectively. In the top gap distance comparison, the
design with 3 cut-outs achieved the largest distance followed by 2 cut-outs;
whereas 1 cut-out and 5 cut-outs designs showed relatively smaller gap
distances. The 1 cut-out design on the other hand has the largest gap distance at
the bottom gap analysis, followed by 2 cut-outs, 3 cut-outs, and 5 cut-outs in

descending order of gap distance.

Despite the fact that 3 cut-outs and 1 cut-out designs have distinctive gap
distance in the top and bottom gap analyses respectively, they do not achieve a
well overall performance in both cases. On the contrary, the design with 2 cut-
outs shows a relatively large gap distances in both top and bottom analyses. This

desired overall performance has made 2 cut-outs the most favorable design.

From the analyses, it is noted that more cut-outs such as the design with 5 cut-
outs do not necessary prove to be effective. This result can account for the
reduction in the materials. As more cut—outs are made to the geometry, more
negative CTE materials would be removed; the overall deformation would be
reduced. The result of this reduction in deformation is small displacements in
both the top and bottom gaps. The previous two analyses have revealed and

confirmed that the design with 2 cut-outs has the best performance in generating

49

great gap distance. Additional analysis is carried out to investigate relationship of

the diameter of the cut-outs and the gap distance.

 

 

 

Figure 35. Parametric study on the effect of the diameter has on gap distance

The results of the analysis are shown in Figure 36 and 37; they have revealed

that to achieve large openings in top and bottom locations, the optimum diameter

of the cut-outs should be 2.5 mm.

 

 

 

 

X 10'4 Top gap distance VS circle radius
1.1 . , , .

1.09 r ..

1.08 - /' _
E
8
g 1.07 - .
.9
'O
3 /,.4
0

1.06 ,4/ -

1.05 r a

1.04 1 I l l 1 -3

0 0.5 1 1.5 2 2.5 3 x10

Circle radius [m]

Figure 36. Top gap result of the parametric study

50

X 10 Bottom gap distance VS circle radius
14 4 r . . 4

 

12— -

Gap distance [m]
an

 

 

,-
I-

2o 0.5 1 1.5 2 2.5 3 x10”3

Circle radius [m]

 

Figure 37. Bottom gap result of the parametric study

MECHANICAL ADVANTAGE MECHANISM

In the previous tvllo sections, two approaches of changing the arrangement of the
powder layers and introducing weak spots to facilitate more deformation in the
deforming ring have been explored. Regardless of these efforts to create a
greater opening to allow coolant gas to flow through, the achievements of the
gap distances are not appealing. Up to this point, the design concept remains at
utilizing the negative properties of the materials to generate out-of-plane

deformation upon temperature rise. However, with the limited amount of

51

deformation that can be created by the temperature gradient, designs should

incorporate other mechanisms to enhance and assist the deformation.

Mechanism is commonly used in a variety of applications; examples of
mechanisms include lever, pulleys, gears, and screws. In the following designs,
the lever mechanism is deployed to the deforming ring structure. The purpose of
applying a mechanism to the structure is because of the benefit that the
mechanical advantage brings to the design. Mechanical advantage generally
refers to the factor by which the mechanism is capable of multiplying the applied

force.

For this application, instead of the applied force being multiplied, the
displacement of the deforming ring is multiplied. When the location of the force is
applied due to the deformation of the structure (analogous to the “effort point”),
displacement would take place; through the principle of mechanical advantage,
the displacement at the target site (analogous to the “load”) would be multiplied.
Figure 38 below illustrates the displacement at the “effort point” got multiplied,
creating a greater displacement at the “load”.

Table 7. Simulation data of designs with lever mechanism and spring design

 

 

 

 

 

 

 

 

 

Thermal loading condition Temperature (°C)
Top surface 760
Bottom surface 100
Material type CTE linear curve fit equation
1 = ZTWZOB y = 1E-22X - 7E-06
2 = ZrOz y = 2E-09x + 8E-06

 

 

 

 

52

Surface: von Mises stress [Pa] Max: 3.55393

 

 

 

 

 

0.02
Large
Small displacement
0.015 , .
, dlsplacement
I
0.01 I
l
e 42
0.005
0 . - 41.5
l
I u in II
-0005 I effort po t
\
_0_01 AXIS of symmetry

 

 

 

 

0 0.005 0.01 0.015 0.02 0.025 0.03 0.035 Min:1.647e4

Figure 38. Design with application of lever mechanism

In both designs shown in Figure 38 and 39, it is noticed that the structures
marked with type 2 are assigned with pure ZrOz material. This material has

positive CTE and is needed to generate the necessary expansion for the

activation of the lever mechanism. In the alternative design illustrated in Figure
39, the Zr02 material is placed at the bottom region where the temperature is
maintained at 100 °C. The expansion of the ZrOz material is minute since the

elevation in temperature is mild; however, this expansion is multiplied and thus

creating a huge deformation in the top region. Combining the shrinkage from the

53

ZrW203 material making up the top zone, the overall gap distance acquired is

larger than the results achieved previously.

 

 

 

 

 

  

 

    

Surface: von Mises stress [Pa] Max: 4.10597
x 107
4
0.02
Large
displacement 3-5
0.015
0.01
,2.5
0.005 I
L .12
0 T i
Small h 1'5
dis Iacement
-0005 Core p i 1
-0.01 0.5

 

 

 

 

0 0.005 0.01 0.015 0.02 0.025 0.03 0.035 Min:458.303

Figure 39. Alternative design using the lever mechanism

The application of mechanical advantage of lever mechanism to assist the
deformation brings a larger gap distance to facilitate better coolant gas flow.
Nonetheless, the serious shortcoming of this mechanism is the weakened
structural configuration, which jeopardizes the integrity of the whole thermal
protection system. In the designs shown, they require a large fraction of the
structural to be voids; this would lead to a very thin structure of about 2 - 3 mm
thick. In addition, high stress points would make the structure of the deforming

54

ring more fragile and prone to fracture under the extreme working conditions of

the space shuttle.

SPRING DESIGN

Parallel to the designs with a mechanical advantage mechanism is the spring
design shoWn in Figure 40. The thermal loading conditions of the simulation and
the material properties of the design follow the data used in Table 7, which is

shown in the previous section of this chapter.

Surface: temperature [°C] M3)“ 765-932

 

  
  

 

 

 

 

 

 

 

 

x 10'3
8 Axis of 70°
symmetry
7 ............ ..
600
6
5 '- ‘500
4
1 r ‘400
2 .4 .
3
2 300
1
200
0

 

 

 

 

100
X 10'3 Min: 99.85

Figure 40. Spring design with wedge top

55

Different from other design investigated, the spring design does not operate in
conjunction with the core. The spring design is developed with the intention to
replace the FGM core designed in the preceding stage of the perspirable skin
project. The working principle relies on the arrangement of positive and negative
CTE materials in alternating sequence. Under the working temperature gradient,
positive material would push laterally, while negative material would shrink. This
coupling effect would generate a contraction in the vertical direction, thus

lowering the top core section downward to open the gap.

Nonetheless, given the complexity of the structure, the spring design has
encountered great difficulties in the fabrication. The design comprises voids that
are difficult to achieve by powder compaction technology alone. The other
unfavorable concern regarding fragility of the structure has also reduced its

sustainability.

56

CHAPTER 4

ALTERNATIVE DESIGN SOLUTIONS

CONCENTRIC RINGS

In the pursuit of the most promising design for the perspirable skin project, many
other great ideas have come along and been studied. They began just as
conceptual ideas, with the continuous effort from the research group; they were
explored and evaluated through FEM and CAD simulations. Although they show
strong potentials to becoming plausible solutions to the problem, they pose
physical constraints in aspects such as fabrication and actual operation. Perhaps
these designs remain theoretical because of current limitations, the
advancements technologies and manufacturing might turn the theoretical into
practical. The design of concentric rings is one of such proposed conceptual

designs.

Inspired by Mayeul de La Tour du Pin who was the exchange student for the
research group, he introduced the pioneer idea of concentric rings and cores.

The outer ring of the design has a cut-through slot with negative CTE material;

the core, on the other hand, has positive CTE materials such as ZrOz shrink

fitted into the outer ring. The illustration and the properties of the concept are

shown in Figure 41 and Table 8 respectively.

57

As one can notice, the assignments of the materials are in a reversed
configuration, the working principle of the concentric ring is different from that of
the previous designs. When the design reaches the working temperature (about

760 °C), the inner core would expand according to the nature of the positive CTE

materials, such as ZrOz. Opposite to this deformation, the outer ring, which is

made of negative CTE, would shrink and tightly wrap around the core. The
resulting gap would be formed at the slot as the expanding core pushes outward

against the outer concentric ring.

Negative
Ring CTE materlal

Slot
Outer Inner

gap ' gap

Positive CTE
material

Figure 41. Conceptual design of the concentric ring (left: design at room
temperature; right: design at working temperature) [25]

Table 8. Material properties and dimensions for concentric ring design [25]

 

 

 

Inner core 7.5 mm 5 GPa 0.3 7 x 10'6 K'1
. ID: 7. -
Outer rlng OD: 1552:!“ 5 GPa 0.3 -7 x 10'6 K 1

 

 

 

 

 

 

 

58

Table 9. Concentric ring simulation results [25]

 

 

 

    

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Desig_11_ Type Gap Type AT = 600 °C AT = 800 °C
Outer Gap 482 pm 636 pm
Slot 1
Inner Gap 280 pm 370 um
Outer Gap 503 pm 626 um
Slot 2
Inner Gap 196 pm 260 um
Outer Gap 466 pm 617 um
Slot 3
Inner Gap 232 pm 307 pm
Outer Gap 330 pm 437 um
Slot 4
Inner Gap 150 pm 174 um
Outer Gap 142 pm 282 um
Slot 5
Inner Gap 296 pm 386 um

 

 

 

 

 

 

There are altogether five proposed designs of the concentric ring, each with a
slot cut in different orientation and angle. The designs and their respective

ANSYS simulation results are shown in Table 9. Two computations are

performed at temperature gradients of 600 °C and 800 °C in an effort to

59

 

investigate how much the temperature change would have on the simulation
results. The average percentage increase in the gap distance is about 36% when
the temperature gradient is increased from 600 °C to 800 °C. This study has
proved that temperature fluctuation has a predominant influence in the gap

distance.

Reviewing all the results, the design with the first type of slot in the concentric
ring yields the best overall results in the outer and inner gap distance. Individually,
designs with type 2 slot and type 5 slot achieve maximum gap distance in outer
and inner gap analyses respectively; however, they do not retain consistency in
bath temperature gradients. Furthermore, the type 1 slot design has the simplest
yet consistent performance in both gap distances over a wide range of

temperatures.

Further parametric study is performed to verify the effect of the size (diameter) of
the concentric ring and core on the gap distances, and the results are presented
in Table 10. The gap distances vary proportionally with the diameter of the
concentric ring and core; the consistency of the increase is observed at bath

temperature gradients.

Despite the promising results of the simulations, the concentric ring design
suffers several major defects that inhibit the advocacy of this design. Regardless

of the minute change in the overall dimensions of the design upon heating, the

60

overall size of the design is expected to increase and thus imposes tremendous
stress on the RCC skin around the concentric ring. Knowing that the RCC has
nearly zero CTE, the concentric ring and the core would not have sufficient

clearance for the heated deformation and is subject to cracking or fracture.

Closely related to the dimensional effect of the design upon heating, concerns
have arisen regarding the contact surface between the core and the concentric
ring. At working temperature, the core will expand and is tightly embraced by the
concentric ring, which shrinks at the same temperature. The huge pressure
induced by this shrink fitted configuration creates a lot of friction at the interface
where these two components meet. Operation of this gap opening mechanism
demands that the concentric ring be able to slide around the core to open the
gap at the working temperature. If this motion is obstructed by friction, a fracture
might occur at the concentric ring or core, jeopardizing the integrity of the safety
of the space shuttle.

Table 10. Diametric comparison of the type 1 slot concentric ring [25]

 

  

 

 

 

 

 

 

 

 

 

 

- AT = 600 °C AT = 800 °C
.11.
rIRratio (mm) 7.5/15 10I20 15130 7.5/15 10I20 15130
Outer Gap 482 pm 656 pm 984 pm 636 pm 872 pm 1,308 um
Inner Gap 280 pm 376 pm 564 pm 370 pm 502 pm 752 um

 

61

 

SLIDER BAR

The slider bar design is proposed as an alternative that requires further
experimentation and simulation to verify its possibilities. From the literature
review of the space shuttle, a special metal called lnconel 718 [7] is used in
building the ribs of the wings. The properties of this material shown in the Table

11 below and the conceptual idea of the alternative mechanism is illustrated in

 

 

 

 

 

Figure 42.
Table 11. Material properties of lnconel 718 [5]
Maximum service temperature 980 °C
Approximate melting temperature 1370 °C
Coefficient of thermal expansion (average) 13 x 1? m/K
Thermal conductivity 11.4 W/m/K
Density 8,200 kg/m3

 

 

 

 

Exterior skin
of the wing \,

/

Moveable
end

Fixed end 51.!

\

Figure 42. Illustration of the slider bar (left: misaligned holes at room
temperature; right: the holes align at working temperature) [25]

62

The working principle of the slider bar simplifies the structural design for

composite materials like ZrW203 and ZrOz. Figure 42 shows that the lnconel 718

strip is attached to the interior of the skin of the wing. At room temperature, the
hole on the lnconel 718 strip and the hole on the exterior skin do not overlap one
another, creating a tight seal. When the temperature increases, the expansion of
the lnconel 718 strip would extend along the curvature of the skin. The two holes

would align and create an opening for the coolant gas to flow through.

The uniquely high service temperature and melting temperature of lnconel 718
open the exploration to different special metals and alloys. lnconel 718’s
capability of withstanding extreme temperature is the key ingredient to activate
this innovative idea. For a simple calculation using the facts of lnconel 718 from
Table 11, and Equation (5), a 30 cm long lnconel 718 operating at a temperature

gradient of 600 °C would create a gap of size of 2.34 mm.

M = ExATxCTE (5)

SEGMENTED PLATE

Currently, there is a new direction in which the research group is heading for a
new opening mechanism for the perspirable skin, the concept of segmented plate
design shown in Figure 43. The segmented plate design evolves from the

principle of buckling in solid mechanics. Although current manufacturing

63

techniques and machining equipment may not be able to fabricate this complex
geometry, simulations in ABAQUS have shown sustainable results to continue

the pursuit in this novel design.

/
RCC/f \_\

5. \Zero CTE material
.' \

Positive CTE
material in radial /

direction / _

 

    
     
  

   

.....

.; .75 , Negative CTE material
in circumferential
direction

Figure 43. initial design concept of the segmented plate

At mentioned previously, the working principle of the segmented plate design
relies on the buckling effect, which is triggered by the rise in the temperature.
One of the difficulties imposed by the design requirement is the multi-dimensional
FGM properties of the segmented discs. Indicated by the two different arrows in
Figure 43, each segmented disc is made of a multi-dimensional FGM, which has
a positive CTE property along the radial direction of the disc. Whereas in the
circumferential direction, the material exhibits a negative CTE property. The
center core is made of material with a nearly zero CTE material such as the

composite type 4 material in Table 6.

64

As the temperature increases, heat would trigger the segmented discs to expand
along the radial direction, but shrinkage occurs in the circumferential direction.
While the core remains at its initial size, the discs extend toward the center,
compressing the core while creating gaps between adjacent discs. At the working
temperature, the expansion of each disc has just achieved sufficient pressure
and buckling would occur when the core is pushed downward. Gaps would attain
their maximum distances at this point while the whole segmented plate would be
maintained at this configuration until the temperature goes down. The whole

design is expected to reverse the process as the temperature is reduced.

 

Figure 44. (a) CAD drawing of hexagonal plate by NX; (b) simulation of opening
at working temperature using ABAQUS

Using the concept established in the pioneer design, Figure 44 (a) shows a
modified model of the segmented plate in a hexagonal configuration. In addition

65

to the hexagonal shape of the core and 6 trapezoidal discs; hinges are
introduced to the edges of the discs to provide greater mobility, as well as easier
installation to the RCC skin around the trapezoidal plate. The simulation of the
hexagonal plate performed in ABAQUS is shown in Figure 44 (b) at the working
temperature. It must be clarified that the simulation does not comprise all the

features of a multi-dimensional FGM due to the limitation of the FEM package.

At the preliminary stage of the design, the hexagonal plate has demonstrated a
potential solution to open gaps with much larger distances than preceding
designs. The addition of hinges to the edges of the trapezoidal discs is expected
to reduce the friction and stress induced at the interfaces between the discs and
the RCC skin. This promotes a smoother and easier buckling effect as the disw
press against the hexagonal core. However, this improvement comes with
increased complexity of the design and demands dedicated machining for details

such as the chamfers and fillets.

Continuous improvements have been made to perfect the segmented plate
design as the research group pursues solutions in this direction. The
development of a square segmented plate design is anticipated to further
advance the gap distance while aiming to reduce the pressure exerted on the
components of the segmented plate. The CAD simulations in Figure 45

demonstrate the working principle of the proposed idea.

66

 

(b)
Figure 45. (a) Conceptual design of the square plate; (b) views of the plate
opening at working temperature

The new feature to this version of the plate design is the proposed sliding slot
and pin mechanism at the connection of the discs and core. Slots are cut into the
core where the pins at the protrusions of the discs ride. As the radial expansion
of the discs creates a buckling effect, the core would be pushed downward and
the pins would slide along the slot to open wide gaps around the corners of the

core, simulated in Figure 45 (b).

The slot and pin mechanism is proposed to produce larger openings by the
smoothly guided mechanism and to facilitate the recovery of the discs and core

back to their respective original positions during cooling down. The advantage of

67

the slot and pin mechanism grants more mobility as the components can adjust
freely within the guided tolerance. Along with the increase in complexity of the
features, dedicated machining and tighter dimensional tolerance have become a

bigger challenge.

68

CHAPTER 5

DISCUSSION AND FUTURE WORK

DISCUSSION AND CONCLUSION

Throughout the course of various design considerations and improvements to the
perspirable skin, the major properties and behaviors of the FGM deforming ring
are observed and documented. The parametric study of the layered patterns and

cut-outs of the deforming ring, together with the mechanical advantage

mechanism and spring designs, revealed two key features of the ZrW208 - ZrOz

composite material: the shrinkage has a dependency on the thickness and the

deformed geometry can be determined by the pattern orientation of the layers.

The layer patterns and the cut-out designs proved that the amount of deformation
is directly proportional to the thickness (amount) of negative CTE material. If
there are more negative materials in the constituents of the design, the
deformation achieved by applying the same working temperature is greater than
that of a design with fewer amounts of negative materials. From the layer
patterns analyses, a greater deformation is a result of putting more negative

materials on the top surface where the temperature is higher.

Secondly, intelligently manipulating the patterns and geometry of the deforming

ring would control the deformed geometry, as well as the magnitude of

69

deformation. Stacking different CTE materials in different orientations and layers
can achieve a desired deformation; and distinctively introduce voids and/or weak
spots to reduce the stiffness can also help to attain better deformation. The four
new design concepts have proven that creating higher stress concentration

points would assist in the deformation by reducing the stiffness of the structure.

However, it must be noted that excessive removal of material or inappropriate
orientation of the powder may jeopardize the overall result of the design. Too
many voids and/or higher stress concentration points increase the likelihood of
fracture of the deforming ring, which might lead to a catastrophic consequence.
Therefore the balance between the magnitude of deformation and the
manipulation of the removal of material should be carefully optimized to achieve
the best results. In addition, the locations of where these voids and stress
concentration points should be introduced would also affect the overall

performance of the deforming ring.

At the current stage, incorporating mechanisms such as a lever, spring, and
other opening features may face various difficulties in structural design,
fabrication, and machining capability. However, the increase in gap distances
has been well approved by the simulations, supporting the possibilities to
continue exploration in this direction. The ever-changing advancements in
technologies and equipment and the designs in virtual space may one day be

transformed into reality.

70

FUTURE WORK

This intermediate stage of the perspirable skin project focuses on exploring
different design variations and improvements to perfect the opening mechanism
to allow more and better flow of coolant gas of the TPS for the space shuttle.
Therefore, much of the effort is put into inspiring, innovative designs and ideas,
followed by computer simulations and data analyses of the results. CAD and
FEM packages allow a fast and accurate visualization of the concepts and
prediction of the designs at operating conditions. Nonetheless, these virtual tools

do not reflect real-time and actual performance of these proposed ideas.

Another aspect that has been excluded from the simulations is the effect of the
fluid mechanics of the flow of coolant gas has in the overall performance of the
design. The exclusion is to confine the project scope to establish the functional

geometry of the mechanism of the perspirable skin.

Based on the existing knowledge and findings, together with the placement of
new equipment, as the perspirable skin project moves onto the next stage, more
experiments can be performed to investigate the hypotheses and simulations.
Some of these designs can be fabricated using the powder processing
techniques with the current machining capability. The recent installment of the
triaxial machine may help to produce geometries like the hexagonal or square

segmented plate designs.

71

During the development of the project, there is a strong need for more
understanding and a capability to manipulate multi-dimensional FGMs. Research
findings showed that only a small fraction of the science of FGMs had been
studied and this remains as a scientific field that is yet to be cultivated. Existing
technologies and methods for manufacturing FGMs are limited to some specific
types of materials, and often these methods are costly and restricted to
laboratory scale. Knowing and gaining the capability to cheaply and accurately
produce multi—dimensional FGMs can certainly assist the success of the

perspirable skin project.

72

[1]

[2]

[3]

[4]

I5]

[6]

[7]

[3]

I9]

[10]

[11]

I12]

BIBLIOGRAPHY

Angelo J. A., 1999, The dictionary of space technology, New York, NY:
Facts on File, pp. 211-25.

NASA Facts, 2006, “Orbiter Thermal Protection System”, FS-2004-09-
014-KSC (Rev. 2006), pp. 1-6.

Cleland G. G., lannetti F., 1989, "Thermal protection system of the space
shuttle", NASA contractor report NASA CR-4227, pp.1-17

Richard S., 1980, "Orbiter Protective Tiles Assume Structural Role", Aviat
Week & Spa Tech., Feb.25, pp. 22-24.

Oguz, B., 2009 “Design and processing of perspirable skin through
numerical analysis”, MS Thesis at Michigan State University.

Savage G., 1993, Carbon—carbon composites, New York, NY, Chapman &
Hall, pp. 37—80 & pp. 322-350.

Opila E., Jacobson N., and Jerrnan G., 2004 (Rev. 2006), “Columbia
Tragedy. High Temperature Materials Chemistry and Thermodynamic
Considerations of the Breached Wing Leading Edge", Journal of Failure
Analysis and Prevention, pp. 102-110.

NASA facts, 2006, “Reinforced Carbon-Carbon (RCC) panels“, FS-2004-
01—001-KSC (Rev. 2006), pp. 1-2.

Gee S. M., Little J. A., 1991, “Oxidation behavior and protection of
carbon/carbon composites”, Journal of Materials Science, Vol. 26, pp.
1093-1100.

NASA, 2003, “Working Scenario”, Columbia Accident Investigation Board
Report, Vol. 2.

NASA, 2003, “Follow the TPS”, Columbia Accident Investigation Board
Report, Vol. 4.

K0 William L., Fields Roger A. (NASA), 1987, “Thermal Stress Analysis of

Space Shuttle Orbiter Subjected to Reentry Aerodynamic Heating”, NASA
Technical Memorandum.

73

I13]

[14]

[15]

[16]

[17]

[18]
[19]

[20]

[21]
[22]

[23]
[24]

Boulahia B., Belkhiri M., Afrid M., 2008, “Simulation du phenomene de
relaxation dans un ecoulement reactif parietal derriere un choc fort",
Revue des Energies Renouvelables, pp. 890-102.

Laub B., Venkatapathy E. (NASA), 2003, “Thermal Protection System
Technology and Facility Needs for Demanding Future Planetary Missions",
International Workshop on Planetary Probe Atmospheric Entry and
Descent Trajectory Analysis and Science.

Fritsche B., Roberts T., Romay M., lvanov M., Grinberg E., Klinkrad H.,
1997, “Spacecraft disintegration during uncontrolled atmospheric re-entry“,
Second European Conference on Space Debris.

Thornton Earl A., 1992, “Thermal Structures: Four Decades of Progress”,
Journal of Aircrafts, Vol. 29, No. 3.

NASA, 2008, “Columbia Crew Survival Investigation Report", p. 2-24.

Sun L., 2009, “Multi-functional materials by powder processing for a
thermal protection system with self-cooling capability: Perspirable skin“,
PhD. Dissertation at Michigan State University.

Han, J. C., Dutta, 8., 2001, “Recent Developments in Turbine Blade
Internal Cooling,” Heat Transfer in Gas Turbine Systems, ed., R. J.
Goldstein, Ann. NY. Acad. Sci, 934, pp. 162-178.

Sun L., Sneller A., Kwon P., 2008. “Fabrication of Alumina/Zirconia
functionally graded material: from optimization of processing parameters
to phenomenological constitutive models”, Mater. Sci. Eng. A, 488(1-2), pp.
31-38.

Sun L., Baldauf S., Kwon P., 2008, “Procession of ZI'W203 -Zr02
continuous Functionally graded materials by co-sintering Zr02 and
Zr02+W03 multi-layer compacts", ASME lntemational.

Lambros, J., Santare, M.H., Li, H., Sapna, G.l., 1999. “A novel technique

for the fabrication of laboratory scale functionally graded materials”, Exp.
Mech., to appear.

Pei Y. T., De Hosson J. Th. M., 2000, Acta mater 48(2000), pp. 2617-25.
Mott M., Evans J. R. G., 1999, “Zirconiazalumina functionally graded

material made by ceramic ink jet printing", Materials Science and
Engineering A271, pp. 344—352.

74

[25] de La Tour du Pin M., 2010, “Internship report: Design of a perspirable
skin”, Michigan State University.

75

I O I t.‘ 'u.‘

i If... '11.".017'..- .

 

ll
Ill
l