Design and Analysis of Metallic Thermal Protection Systems

Suneeth Sukumaran*

Post Graduate Student, Department of Mechanical Engineering, SCT College of

Engineering, Thiruvananthapuram, Kerala, India.

and

Dr. S.H. Anilkumar

Assistant Professor, Department of Mechanical Engineering, SCT College of

Engineering, Thiruvananthapuram, Kerala, India.
*Corresponding Author

ABSTRACT

The thermal protection system used in the study is an active type structure (withstand

both thermal as well as structural loads) in the form of metallic corrugated sandwich. In

GPP like ANSYS and NASTRAN ( using 2D heat transfer equation ) have ability to do

thermal analysis in component level but the customized special software ( using 3D heat

transfer equation ) have capability to do thermal analysis of assembled structures by

giving basic properties (, c, k) of each component. But the main intricacy is the

variation of Thermal Contact Conductance (TCC) at metal to metal contact portions of

the brazed components. In this work contains transient thermal analysis of a unit cell of

MTPS and a parametric study by changing TCC value and finally reach an appropriate

value and validated with available experimental value and Compared it with a casting

model ( No TCC ) and collected the thermal data and transfer to ANSYS model. Finally

carry out both thermo-structural analysis and buckling analysis in ANSYS and bring out

the capability of thermo-structural analysis of ANSYS with the support of existing

customized software for evaluation of thermal field.

1
Biographical notes:

Suneeth Sukumaran earned his M.Tech degree in Mechanical Engg. (Machine Design)

from SCT College of Engineering, Thiruvananthapuram. Currently he is working as

Assistant Professor in Jyothi College of Engineering, Cheruthuruthy, Vettikkattiri (po)

Thrissur - 679 531. His research interests are design and analysis of thermal protection

systems for aerodynamic vehicles.

Dr. S.H Anilkumar is presently working as an Assistant Professor at SCT College of

Engineering, Thiruvananthapuram. He earned his PhD from National Institute of

Technology, Kozhikode and, undergraduate and graduate of College of Engineering

(B.Tech & M.Tech) Thiruvananthapuram. He is a life member of Indian Society for

Technical Education, Indian society for heat and mass transfer and Combustion Institute.

He is a member of International Association of Engineers and American Society of

Mechanical Engineers. During his 17 years of teaching career, he has taught various

courses for undergraduate and graduate programmes and developed course work for

graduate studies.

2
1. Introduction

Thermal Protection System (TPS) is a feature designed to protect the space vehicles from

the extreme thermal environment arising during atmospheric re-entry. The aerodynamic

heating is generated at the surface of an entering object due to the combination of

compression and surface friction of the atmospheric gas. There are mainly two types

TPS, one passive like silica tile and other active, that can withstand both thermal and

mechanical loads. The former has less density while latter has three times higher density

but highly reliable and needs less time for refurbishment. The thermal protection systems

used on the Space Shuttles are reinforced carbon/carbon (RCC) at the nose and wing

leading edges, high and low temperature reusable surface insulation (HRSI, LRSI) used

to cover the major portion of the Shuttle windward side, advanced flexible reusable

surface insulation (AFRSI) and coated named felt reusable surface insulation ( FRSI ) for

the leeward side. While not much can be done with respect to replacing RCC other than

develop new materials, there is good scope for developing new structures to replace the

LRSI tiles, AFRSI, and FRSI which occupy the major portion on the Shuttle as shown in

Figure 1.

Figure 1 : TPS distribution

3
For commercial viability, the TPS must contribute to minimizing life cycle costs to

enable delivery of commercial payloads at reasonable cost. For military applications, the

TPS must enable high performance, rapid response, and rapid turn around under adverse

conditions. The most extensive experience with reusable TPS is with the ceramic tile and

blanket TPS on the Space Shuttle orbiter. Although the orbiter TPS does an excellent job

of protecting the vehicle from aerodynamic heating, more than 40,000 work hours are

typically expended to refurbish and inspect the TPS between flights. The launch cost for

the space shuttle increases by about $10,000[1] for every pound of launch weight. So to

keep space expedition economically viable in 21st century we need to decrease the cost of

launching a space craft. One of the most expensive systems of a space vehicle is the

Thermal Protection System (TPS), which protects the vehicle from the high thermal loads

during re-entry; therefore it deserves some special attention.

Many methods of protecting the re-entry vehicles have been conceived and tried.

Approaches are constrained by material limitation, weight, cost and reliability

considerations. TPS generally increases weight, cost and complexity as the intensity and

duration of aerodynamic heating increases. In view of the above, the designers are

concentrating on the development of Metallic Thermal Protection Systems (MTPS),

particularly for reusable launch vehicle (RLV). The TPS used in the present study was

corrugated core sandwich Metallic Thermal Protection Systems (MTPS). The corrugated

core sandwich structure has more damage tolerant properties and load bearing capacities

[2].

4
2. Geometry of Corrugated-Web MTPS

The Corrugated web MTPS used for the present study consists of: (i) Top face sheet, (ii)

Bottom face sandwich, (iii) Truss-core and (iv) Thermal insulation. Metallic sheets are

used for the first three items. Ceramic insulation in the form of fibers is filled inside the

truss core space and the sandwich core to block the heat flow from top to bottom face

sheet as shown in the Figure 2. It is a one storied construction. The main design constraint

of MTPS is back wall temperature of Sandwich structure should be less than 375K for all

re-entry space vehicles at a required duration of 1000s of the flight during the re-entry

phase and this can be achieved by filling “Saffil” like insulation material inside the truss

core. The heat flux used in the study was 18 W/cm² [3]. The structure is symmetric with

respect to vertical plane (Z-plane) and one unit cell of 37.5 x 37.5 x 20 mm is considered

to obtain the thermal field data. The one unit cell of corrugated sandwich MTPS is shown

in Figure 3.

Figure 2: Geometry of MTPS

5
 Figure 3: Unit cell of MTPS

The panel is constructed of three materials PM2000 for top face sheet (1.0 mm thick),

truss Core (0.1mm thick) and bottom face of top sandwich (1.0mm thick). This model is

made by fixing corrugated web to the top and bottom face sheet by the process of

brazing.

Figure 4 : Non-brazed Casted Model

3. Finite Element Modeling of MTPS

This work describes a three dimensional finite element method based thermal modeling of

a metallic thermal protection system (MTPS). Analytical solutions are not possible with

6
this type of heterogeneous construction since it is a non-linear transient problem. The

formulation uses three types of element: (i) Shell element for metallic parts, (ii) 3D solid

element for insulation and (iii) Interface element for heat flow between metallic skins and

insulation. Then conducted transient thermal analysis and collected thermal field data at

each time step. Polynomial approximation is used to represent the through thickness

temperature profile in shell elements.

Finally transfer these temperature values to ANSYS model to conduct thermo-structural

analysis because stresses are considered equally important as the temperatures. Any

structural failure due to high stresses may call for replacement of failed ones, before the

next launch. Three types of elements: (i) 8-node shell element, (ii) 20-node solid element

and (iii) 8-node interface element are employed. Shell elements are used in the metallic

parts (Top face sheet, Truss core, Top and Bottom face sheets of sandwich). Solid

elements are used for the insulation parts, both in truss core space and core of bottom

sandwich. Interface elements are used in:

1. Between face sheets and truss core sheets (metal-metal contact).

2. Between face sheets and Saffil insulation.

Figure 5: Finite element modeling

7
The modes of heat transfer are: (i) Conduction through places of direct contact, (ii)

Conduction through the medium filling the interstitial space, (iii) Re-radiative heat

transfer across the interface. The heat flow through a common surface between two

contacting surfaces is modeled using the concept of interface element. The aerodynamic

heating ceased after 640s is shown in Table 1, but the heat transfer was not ceased. After

640s radiative heat transfer would be take place so give additional time up to 5640s for

heat transfer ceasing. The emissivity value used in the analysis is 1.

Table 1: Transient heat flux loading history.

The element exists at (i) metal to metal contact (ii) metal to saffil contact. In the present

study, concept of Thermal Contact Conductance (TCC) has been included in the

modeling for making it close to reality. Thermal Contact Conductance exists between

bodies in contact whether they are of similar or dissimilar materials. The contacting

surfaces correspond to the sides of two adjacent elements. Three combinations of

interface elements are possible in MTPS. (i) shell-solid, (ii) solid-shell, (iii) shell-shell.

Since high temperatures are involved material properties becomes function of

temperature that is also considered in the analysis. Figure 6 depicts typical temperature

8
profiles on either side of interface with a temperature drop ( T ) across the interface. The

heat flux [4] passing through the interface is expressed as:

q  tcc(T1  T2 ) (1)

Where T1 - temperature on the bottom surface of medium – I and T2 - temperature on the

top surface of medium- II, tcc - thermal contact conductance, the value of which is

difficult to specify. An experimentally evaluated one is more appropriate. But

experiments have to be conducted for the specific combinations and statistical values

have to be obtained.

Figure 6: Typical temperature profile

Alternately reasonable values may be obtained from existing literature (TCC= 5 x 10 -4

W/ mm2K). In the study conduct transient thermal analysis by four TCC values (5 x 10 4 ,

3 x 10-4 ,1 x 10-4, 0.5 x 10-4 W/ mm2K) and find thermal variables upto 1000s. The thermal

load is applied on the top surface. It is uniformly applied on the surface, the intensity

varies with time. In addition to thermal load mentioned above, re-radiation heat loss

condition is applied on the top surface. Also convective heat loss is applied on the top

surface, after aerodynamic heating is ceased. The convective heat transfer [5] is hc= 6.5 x

9
10-6 W/mm2. The time varying heat flux load (MACH12 trajectory) and thermal properties

are shown in Table 2 and Table 3.

Table 2: Thermal properties of PM 2000

Temperatue Thermal Density, Specific

conductivity, kg/mm3 Heat,
(K) W/mm K J/kg/K

300 0.0109 480

473 0.016 480
1. 7.18e-06
773 0.021 610
1023 0.022 680
1273 0.0255 740
1473 0.028 800

Table 3: Thermal properties of Saffil insulation

Temperature, Thermal Density, Specific

K conductivity, kg/mm3 Heat,
W/mm K J/kg/K

300 1.1e-05 900

372 1.21e-05 950
9.61e-
572 1.97e-05 1092
08
772 3.36e-05 1172
972 5.48e-05 1222

4. Results and Discussions

The positions for Temperature measurement is depicted in Figure 7. The transient thermal

analysis is carried out for four Thermal contact conductance (TCC) value and find

temperature at top and bottom face sheet. The selected location of MTPS is depicted in

Table 4.

10
 Figure 7: Positions for Temperature Measurement.

Table 4: Selected location across MTPS

Position Details

1 Outer surface of top PM

2000 skin, z = 20 mm

2 Bottom of PM2000 sheet

(back wall), z =0.0 mm

4.1 Transient Thermal Analysis Results

The four TCC values are: 1) A= 5 x 10-4 W/ mm2K 2) B= 3 x 10-4 W/ mm2K 3) C=1x

10-4 W/ mm2K 4) D=0.5 x 10-4 W/ mm2K. The variation of TFS and BFS temperature for

different values of TCC are depicted in Figure 8 and Figure 9.

Figure 8 : Top face sheet temperature for Different TCC.

11
 Figure 9 : Bottom face sheet temperature for Different TCC

For a MACH12 heat flux history ( Table 1) the heat transfer through thickness from TFS

to BFS was affected by TCC value. So the TFS temperature shows not much difference

but BFS temperature shows much difference as shown in Figure 8 and Figure 9, and

comparison with experimental value was shown in Table 5.

Table 5: Comparison with Experimental values

Maximum Maximum
Temperature Temperature
(K) (K) Difference
(K)
Experimental Special software(TCC=
0.5E-04)

TFS 1295 1312 17

BFS 575 601 26

This transient thermal analysis for different TCC value shows the maximum temperature

value which is most close to experimental value is for TCC= 0.5 x 10-4 W/ mm2K. For this

TCC value the maximum temperature at top face sheet is 1312K (at 370s) and maximum

temperature at bottom face sheet is 601K (840s) and conducted transient thermal analysis

of a non-brazed cast model (No TCC) to find the importance of thermal contact

12
conductance. In this model there is no need for giving thermal contact conductance at

metal-metal contact part because corrugated web is made as a single volume with top and

bottom face sheet. The Variation of top face sheet and bottom face sheet temperatures of

brazing model and non-brazed casted model (TCC=1) is shown in Figure 10 and Figure

11 respectively.

Figure 10 : Variation of TFS temperature

Figure 11: Variation of BFS temperature

The maximum TFS temperature for non-brazed casted model is 1260K (t=370s) and for

brazing model is 1304K (t=370s), i.e. temperature difference of 44K. The maximum BFS

temperature for non-brazed casted model is 976K (t=575s) and brazing model is 667K

(t=620s).

13
4.2 Thermo-Structural Analysis

Stresses are considered equally important as the temperatures any structural failure due to

high stresses may call for replacement of failed ones, before the next launch. The shell

element used in the special purpose software give top and bottom temperature of each

node other than mid surface temperature because of interpolation function used in the

software. So it is easily to use solid element for modeling of TFS and BFS in ANSYS and

give proper thermal gradient by giving top bottom node temperature. Transfer the nodal

temperature in the special software to ANSYS model. The elements used in ANSYS are

SHELL93 (web) and SOLID45 (TFS and BFS). The saffil insulation is not considered in

the structural modeling because Saffil cannot take structural loads and the structural

properties are not taken into consideration. Transfer the nodal temperatures at t=376

seconds (TCC=1) from special software to ANSYS with reference to the material property

data (Table 2-3). The Figure 12 shows thermal distribution at t=376s in ANSYS.

Figure 12: Temperature distribution at 376s (TCC=1)

14
The maximum temperature in TFS is at 376s (1034K) and temperature in BFS is 425K.

Then constrained displacements for all nodes in the four sides of BFS and conducted the

static type analysis give stresses due to temperature. Figure 13 and Figure 14 shows the

Von-Mises Stress distribution. The Figure 13 shows the stresses at position 1 and 2 are

352.492 M Pa and at position 3 is 604 M Pa. The allowable stresses in the corresponding

locations are 255 M Pa and 425 M Pa respectively. The stress values exceed the safe

limit according to allowable stress value at the corresponding temperature (Tables 6-7).

Figure 14 shows the stresses at position 4, 5 is 772.131 M Pa which exceed the allowable

stress limit.

Table 6: Structural Properties of PM 2000

Temperature Elastic Poisson’s Coeff. of

modulus, ratio Thermal
(K) MPa Expansion, /
0C

300 21.5e04 2. 10.7e-06

373 21.0e04 12.0e-06
523 19.9e04 3. 0.3 12.2e-06
773 17.9e04 12.5e-06
1273 13.5e04 15.1e-06
1450 12.0e04 15.45e-06

Table 7: Allowable Stress of PM2000

Temperature(K) Allowable Stress(MPa)

300 850

600 425

1200 255

15
 Figure 13: Von-Mises Stress

Figure 14 : Von-Mises Stresses

4.3 Buckling Analysis

Finally buckling analysis was done by giving nodal temperatures at 376s (TCC=1) and a

uniform pressure load of 100kPa at top surface of TFS and constrained the displacements

at 4 bottom sides. Generally 100kPa and 40kPa was considered for re-entry space

vehicles according to the aerodynamic force condition. The maximum deflection is

1.365mm at the edges of TFS is shown in Figure 15. The buckling modes are depicted in

Figure 16, Figure 17 and Figure 18.

16
 Figure 15 : Displacements in Buckling

Figure 16 : First Buckling Mode

The first buckling mode shows the buckling occurs at web for a critical buckling load of

36.1 kPa .The second buckling mode shows the buckling at centre of corrugated web for

critical buckling load of 919.98 kPa. The buckling modes showed that maximum

deflection is at corrugated web portions edges.

17
 Figure 17 : Second Buckling Mode

Figure 18 : Third Buckling Mode

5. Conclusion

A transient heat transfer analysis of MTPS followed by structural analysis has been

carried out to bring out the intricacies associated with TCC. Thermal contact

conductance which is a parameter that determines the heat transfer across the two

different surfaces that are brazed together is varied from 0.5E-4 to 5E-4 W/mm²K and

thermal profile across MTPS. It is interesting to note that for a heat flux history with

maximum value of 0.18W/mm² and for TCC= 0.5E-4 W/mm²K the maximum back wall

18
temperature of the MTPS is predicted as 601K at 840s as against maximum respective

value of 575K at 704s. As TCC value increases the back wall temperature increases and

for TCC = 5 E-4 W/mm²K, it is obtained as 835K at 590s. When compared to the limit

case of an integral MTPS (casted), the back wall temperature becomes 976K at 575s as

expected.

Structural analysis results on MTPS indicated that for 100kPa load corresponding to a

thermal profile for TCC=1E-4 W/mm²K, the critical buckling load as 35.6kPa with web

of the truss core in bending mode. Based on the von -misses stress the truss core has a

maximum stress 772MPa (476K) as against the strength of 520MPa. It is concluded that

a locally the stresses are exceeding the limit. However, for a distance of 5mm away from

the maximum stress zone, the value becomes 510MPa (501K) with a positive margin.

Present study indicated that the junction where truss core meets the skin needs additional

reinforcement in the form of beading and recommended for using C-type channel section

at the junction for supporting the truss core.

6. Future work

1) The main future work on this area is to develop software to link between special

3D software to ANSYS software for thermo structural analysis.

2) Thermo structural analysis of two storied MTPS by increasing web thickness.

References

[1] Satish K, Oscar M.Matrinez, Christian Gogu, Bhavani V.Shankar, Raphael T. Haftka
(May 2006), Analysis and Design of Corrugated-Core Sandwich Panels for Thermal
Protection Systems of Space Vehicles, AIAA/ASME//AHS/ASC Structures,

19
 Dynamics and Materials Conference, Newport, Rhode Island. Vol. AIAA 2006-1942,
pp. 1-18.

[2] Satish K, Bhavani V.Shankar, Christian Gogu, Raphael T.Haftka (April 2007),
Comparison of Materials for Integrated Thermal Protection Systems for Spacecraft
Reentry, AIAA/ASME/ASCE/AHS/ASC Structures, Structural dynamics and
Materials Conference, Hawaii, Vol. AIAA 2007-1860, pp. 1-21.

[3] Doualas Ow Stanley, Theodore Am Talay, W. Douglas Morrisa, Kathryn W Urster

(January 1991), Conceptual Design of a Next-Generation, Fully Reusable Manned
Launch System, AIAA/ASME/ASCE/AHS/ASC, Vol. AIAA 1991-0537, pp.1-14.

[4] Fukuoka, Luo Xing et.al (2009), Multi-Fidelity Analysis of Corrugated-Core

Sandwich Panels for Integrated Thermal Protection systems, Vol. AIAA 2009-0463,
pp.1-15.

[5] Reddy J. N (1985), An introduction to the finite element method, McGraw Hill, New
York.

[6] Timoshenko S. P and J.N Goodier (1970), Theory of elasticity, McGraw Hill,
Singapore.

20