OPTIMIZATION OF INTEGRATED THERMAL PROTECTION …...Sizing of the structures can be conducted via...
Transcript of OPTIMIZATION OF INTEGRATED THERMAL PROTECTION …...Sizing of the structures can be conducted via...
AFRL-RX-WP-JA-2016-0325
OPTIMIZATION OF INTEGRATED THERMAL PROTECTION SYSTEM WITH VARIOUS INSULATING CORE OPTIONS (PREPRINT) Fang Jiang Utah State University Wenbin Yu Purdue University Zheng Ye Baker Hughes Ronald Kerans AFRL/RX Ming Y. Chen UES
19 October 2015 Interim Report
Distribution Statement A.
Approved for public release: distribution unlimited.
(STINFO COPY)
AIR FORCE RESEARCH LABORATORY MATERIALS AND MANUFACTURING DIRECTORATE
WRIGHT-PATTERSON AIR FORCE BASE, OH 45433-7750 AIR FORCE MATERIEL COMMAND
UNITED STATES AIR FORCE
REPORT DOCUMENTATION PAGE Form Approved OMB No. 0704-0188
The public reporting burden for this collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, gathering and maintaining the data needed, and completing and reviewing the collection of information. Send comments regarding this burden estimate or any other aspect of this collection of information, including suggestions for reducing this burden, to Department of Defense, Washington Headquarters Services, Directorate for Information Operations and Reports (0704-0188), 1215 Jefferson Davis Highway, Suite 1204, Arlington, VA 22202-4302. Respondents should be aware that notwithstanding any other provision of law, no person shall be subject to any penalty for failing to comply with a collection of information if it does not display a currently valid OMB control number. PLEASE DO NOT RETURN YOUR FORM TO THE ABOVE ADDRESS.
1. REPORT DATE (DD-MM-YY) 2. REPORT TYPE 3. DATES COVERED (From - To)19 October 2015 Interim 22 January 2009 – 19 September 2015
4. TITLE AND SUBTITLE
OPTIMIZATION OF INTEGRATED THERMAL PROTECTION SYSTEM WITH VARIOUS INSULATING CORE OPTIONS (PREPRINT)
5a. CONTRACT NUMBER FA8650-10-D-5011-0001
5b. GRANT NUMBER
5c. PROGRAM ELEMENT NUMBER 62102F
6. AUTHOR(S)
1) Fang Jiang –Utah State University
2)Wenbin Yu –Purdue University
(continued on page 2)
5d. PROJECT NUMBER 4347
5e. TASK NUMBER 0001 5f. WORK UNIT NUMBER
X058 7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) 8. PERFORMING ORGANIZATION REPORT NUMBER
1) Utah State UniversityOld Main HL, LoganUT 84322-4130
2 Purdue University 610 Purdue Mall West Lafayette, IN 47907-2045
(continued on page 2) 9. SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES) 10. SPONSORING/MONITORING AGENCY
ACRONYM(S)Air Force Research Laboratory Materials and Manufacturing Directorate Wright-Patterson Air Force Base, OH 45433-7750 Air Force Materiel Command United States Air Force
AFRL/RXCCM11. SPONSORING/MONITORING AGENCY
REPORT NUMBER(S)AFRL-RX-WP-JA-2016-0325
12. DISTRIBUTION/AVAILABILITY STATEMENTDistribution Statement A. Approved for public release: distribution unlimited.
13. SUPPLEMENTARY NOTESPA Case Number: 88ABW-2015-4969; Clearance Date: 19 Oct 2015. This document contains color. The U.S. Government isjoint author of the work and has the right to use, modify, reproduce, release, perform, display, or disclose the work.
14. ABSTRACT (Maximum 200 words)A finite element analysis (FEA) tool is developed for evaluating the effects of basic design features on a basic structure for anIntegrated Thermal Protection System (ITPS) panel. A set of practical optimization problems regarding different insulating coreoptions are solved by utilizing the commercial FEA software ANSYS with scripts written using ANSYS Parametric DesignLanguage (APDL). The core options represent five different layouts for the insulation layer: bonded/unbonded foam with bladestiffeners, bonded/unbonded foam with hat-section stiffeners, and simple bonded foam. A conventional design with parasiticinsulation tiles is also analyzed as a reference. Figures of Merit (FoMs) identifying the combination of load-bearing capabilityand the mass of the ITPS are defined. Using these FoMs, the optimization objective functions were created so as to considerboth the insulation performance and structural strength. Some examples of FoMs of each design candidate are optimized andcompared with each other to identify the best structural layout, in which the FoMs featuring on the effective bending stiffnessare calculated by SwiftCompTM, an efficient yet accurate tool for constitutive modeling tool of composite materials andstructures.
15. SUBJECT TERMSfinite element analysis (FEA); Integrated Thermal Protection System (ITPS); ANSYS; Figures of Merit (FoMs)
16. SECURITY CLASSIFICATION OF: 17. LIMITATIONOF ABSTRACT:
SAR
18. NUMBEROF PAGES
29
19a. NAME OF RESPONSIBLE PERSON (Monitor) a. REPORTUnclassified
b. ABSTRACTUnclassified
c. THIS PAGEUnclassified
Patrick Carlin 19b. TELEPHONE NUMBER (Include Area Code)
(937) 255-9800Standard Form 298 (Rev. 8-98) Prescribed by ANSI Std. Z39-18
REPORT DOCUMENTATION PAGE Cont’d 6. AUTHOR(S)
3) Zheng Ye - Baker Hughes 4) Ronald Kerans - AFRL/RX 5) Ming Y. Chen - UES
7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES)
3) Baker Hughes, 200 W Stuart Roosa Dr, Claremore, OK 74017 4) AFRL/RX, Wright-Patterson AFB, OH 45433-7750 5) UES, Inc., 4401 Dayton Xenia Rd, Beavercreek, OH 45433
Standard Form 298 (Rev. 8-98) Prescribed by ANSI Std. Z39-18
Optimization of Integrated Thermal Protection
System with Various Insulating Core Options
Fang Jianga and Albert Changb
Purdue University, West Lafayette, IN 47907-2045, USA
Zheng Yec
Baker Hughes, Claremore, OK 74017, USA
Wenbin Yud
Purdue University, West Lafayette, IN 47907-2045, USA
Ronald Kerans and Charles Tseng
UES, Inc., Dayton, OH 45432-1894, USA and
Materials and Manufacturing Directorate, Air Force Research Laboratory,
Wright-Patterson Air Force Base, Dayton, OH 45433, USA
Ming Y. Chen
Materials and Manufacturing Directorate, Air Force Research Laboratory,
Wright-Patterson Air Force Base, Dayton, OH 45433, USA
A �nite element analysis (FEA) tool is developed for evaluating the e�ects of ba-
sic design features on a basic structure for an Integrated Thermal Protection System
(ITPS) panel. A set of practical optimization problems regarding di�erent insulat-
ing core options are solved by utilizing the commercial FEA software ANSYS with
scripts written using ANSYS Parametric Design Language (APDL). The core options
represent �ve di�erent layouts for the insulation layer: bonded/unbonded foam with
blade sti�eners, bonded/unbonded foam with hat-section sti�eners, and simple bonded
foam. A conventional design with parasitic insulation tiles is also analyzed as a refer-
a Graduate Research Assistant, School of Aeronautics and Astronautics, Purdue University, AIAA Student Member.b Graduate Student, School of Aeronautics and Astronautics, Purdue University.c Research Engineer, Baker Hughes.d Associate Professor, School of Aeronautics and Astronautics, Purdue University, AIAA Associate Fellow.
1Distribution A. Approved for public release (PA): distribution unlimited.
ence. Figures of Merit (FoMs) identifying the combination of load-bearing capability
and the mass of the ITPS are de�ned. Using these FoMs, the optimization objective
functions were created so as to consider both the insulation performance and struc-
tural strength. Some examples of FoMs of each design candidate are optimized and
compared with each other to identify the best structural layout, in which the FoMs
featuring on the e�ective bending sti�ness are calculated by SwiftCompTM, an e�cient
yet accurate tool for constitutive modeling tool of composite materials and structures.
I. Introduction
Thermal Protection Systems (TPS) are the heat shields attached to the surfaces of high speed
air vehicles to limit the temperatures of underlying structure. In general, TPS approaches include
both ablative and reusable systems, depending on requirements, such as Apollo Avcoat ablator and
NASA LI-2200 Shuttle tiles, respectively [1, 2]. TPS concepts are also divided into categories of
passive, semi-passive, and active [3], in which the passive ones are regarded as the most weight
e�cient and generally the safest [4] and further classi�ed into load-carrying and nonload-carrying
TPS [5]. Because incident heating rates vary across a vehicle surface, di�erent types of TPS are
generally used on the same vehicle [4].
From the perspective of reducing the expense of orbital transportation, there is interest in
hypersonic vehicles with reusable TPS including Reusable Launch Vehicles (RLVs) [6, 7], miliary
spaceplanes [8], spaceplanes for tourism [9], space trucks [10], suborbital package delivery vehi-
cles [11], and hypersonic air breathing vehicles [12].
Many material candidates are proposed for TPS. Metallics are robust and waterproof, but heavy
and of either limited temperature capability or with poor environmental resistance [13]. The nose
cap and wing leading-edge of a spacecraft often reach the highest temperature. For these parts, some
C-C and SiC/SiC composites are used as they can sustain high temperature [14]. To keep the inner
temperature of the Space Shuttle Orbiter less than 450 Kelvin, reusable surface insulation tiles were
used primarily on the windward surface [15]. Development of improved ceramic TPS was an active
2
2 Distribution A. Approved for public release (PA): distribution unlimited.
topic at the NASA Ames Research Center for many years [16]. The Alumina Enhanced Thermal
Barrier (AETB) is a typical type of ceramic tile [17], which is usually used with Toughened Uni-piece
Fibrous Insulation (TUFI) coating and Reaction Cured Glass (RCG) coating. The latter versions
are believed to be signi�cantly stronger and more resistant to rain erosion than the original Shuttle
tiles [16], and to have improved dimensional stability at high temperatures � 2600 Fahrenheit and
above [13]. However, these are relatively poor structural materials [18].
A possible improvement might be an Integrated Thermal Protection System (ITPS) � a reusable
load-carrying passive insulation system. The entire ITPS would have some structural capability and
would carry load so the structure underneath it can be lighter. The materials considered for the
external surface of the ITPS are often Ceramic Matrix Composites (CMC) due to their excellent
thermal and structural performance [19]. Ceramic foams are used as insulating materials and can be
combined with CMC sti�eners as candidates for the insulation core. For the internal layer, which
is used as a part of the structure, polymer matrix composites (PMC) such as the T650-35 �ber
reinforced PMR-15 polyimide resin are regarded as candidate aeroshell structural material for their
good mechanical properties and light weight [20].
Sizing of the structures can be conducted via optimization [18, 21] that includes both thermal
analysis and structural analysis. For load-carrying passive ITPS, the change of insulation layer
thickness will in�uence structural weight, strength, and temperature simultaneously, which likely
results in groups of local optimum [21]. So the parameter selection for design variables for sizing
optimization is an issue [22]. FEA software is often utilized for analysis of the complex structures
such as honeycombs and corrugated sti�eners [23, 24]. In some cases, the thermal analysis model
can be simpli�ed from 3D to 2D and/or 1D after homogenization of the insulation layer [18, 25�27],
which is often followed by a static structural analysis only under the mechanical loading conditions
separately. Most recently, Jiang et al. [28] evaluated the e�ects of basic design features and material
choices on a basic panel for ITPS by developing a FEA tool using von Mises criterion. While the
von Mises criterion provided certain insights, it is not the best choice for the materials considered.
In this paper, a design of ITPS consisting of CMC, insulation core and PMC is studied. The
goal here is to build the framework for, and examine preliminary results of, this design to compare
3
3Distribution A. Approved for public release (PA): distribution unlimited.
the basic characteristics of several possible options. To reveal the potential of various design options,
optimizations of the di�erent design candidates are conducted by using the ANSYS Parametric De-
sign Language (APDL). The bending sti�ness of the optimized plate con�gurations is computed by
using SwiftCompTM, an e�cient yet accurate tool for constitutive modeling of composite materials
and structures. Then the results are compared to conclude a preferred design strategy considering
both thermal and structural loading conditions.
II. Structural Con�gurations of the Integrated Thermal Protection Systems
A. The Integrated Thermal Protection Systems
The structure studied in this paper is a sandwich-like plate consisting of three layers; an outer
layer of CMC, an inner layer of PMC, and an insulation core. The corresponding properties of the
materials composing the ITPS can be found in the Appendix. The area of the plate considered is
0.762 × 0.762 m2 (30 × 30 in2). The core is mainly responsible for thermal insulation, but for an
ITPS structure, it also carries mechanical loads. In this study, there are 5 options for such a core,
as shown in Table 1.
Table 1 Core options of ITPS
Option label CMC Sti�ener Foam
ITPS-1 Blade sti�eners Unbonded
ITPS-2 Hat-section sti�eners Unbonded
ITPS-3 Blade sti�eners Bonded
ITPS-4 Hat-section Sti�eners Bonded
ITPS-5 No Sti�ener Bonded
In ITPS-1, blade-shaped sti�eners of CMC carry primary loads across the core. The foam is
divided into separate cells, as shown in Fig. 1a. The assembly of the blade sti�eners is shown in
Fig. 1b. The sti�eners are assembled so that those running in the X direction are continuous and
those in the Z direction are separated into 3 pieces each. The foams are not bonded to CMC, PMC
or sti�eners.
In ITPS-2, hat-section sti�eners made of CMC carry loads across the core. The foam is in at
4
4 Distribution A. Approved for public release (PA): distribution unlimited.
CMC
PMCFoam
Blade stiffeners Y, y3
X, y2
Z, y1
(a) ITPS-1 (b) Assembly of blade sti�eners
Fig. 1 Foam with blade sti�eners as insulation core.
least 15 separate pieces, as shown in Fig. 2a. The assembly of the hat-section sti�eners is shown in
Fig. 2b. The assembly of hat-sections consists of long ones in the X direction separating segmented
ones in the Z direction. The foams are not bonded to CMC, PMC or sti�eners.
CMC
PMCFoam
Hat-section stiffeners Y, y3
X, y2
Z, y1
(a) ITPS-2 (b) Assembly of hat-section sti�eners
Fig. 2 Foam with hat-section sti�eners as insulation core.
ITPS-3 is the same as the ITPS-1 except that the insulating foam is bonded onto all walls of
the chamber. ITPS-4 is the same as the ITPS-2 except that the insulating foam is bonded onto
all walls of the chamber. In ITPS-5 the insulation layer composes of foam only and it is bonded
directly to the CMC and PMC layer to form the sandwich structure, as shown in Fig. 3.
B. Baseline Parasitic Scheme (BPS)
AETB tiles with TUFI coating protecting a PMC structure is modeled and analyzed as a
reference conventional system. The AETB ceramic tile with TUFI coating was developed at the
NASA Ames Research Center as an improvement to the LI-900 tile. The system is composed of an
8× 8 in2 insulation tiles mounted on a felt Strain Isolation Pad (SIP) by using Room Temperature
Vulcanizing (RTV) adhesive, as shown in Fig. 4. The corresponding material properties can be
5
5 Distribution A. Approved for public release (PA): distribution unlimited.
CMC
PMCFoam
Y, y3
X, y2
Z, y1
Fig. 3 ITPS-5: bonded foam as insulation core.
found in the Appendix. This tile is parasitic as the AETB carries no mechanical loads, which means
only the PMC layer contributes to the load-bearing capability of the system as whole. FoMs of this
parasitic case are used as references to assess the FoMs of ITPS with the �ve core options.
TUFI
X, y2
Y, y3
Z, y1
PMCSIP
(RTV coated) AETB8
Fig. 4 Structure of ITPS with parasitic insulation tiles.
C. Representative Sample Structures
For e�cient optimization, the con�gurations in Fig. 1a, 2a, 3 and 4 are considered as the
geometries of the FE models in ANSYS. In order to eliminate the edge e�ect of the panel, the
elements and nodes in the central area of these con�gurations are selected as the sample structures
to obtain the strain and stress values. The sample structures in ITPS-5 are shown in Fig. 5.
The CMC and foam parts of the sample structures in ITPS-3 and 4 are shown in Fig. 6 and 7,
respectively. Note the geometries of the PMC parts of the sample structures in ITPS-3 and 4 are
the same with Fig. 5c. In addition, for ITPS-1 and 2, the foam material is unbonded and thus will
6
6 Distribution A. Approved for public release (PA): distribution unlimited.
(a) CMC (b) Foam (c) PMC
Fig. 5 Finite element models of the sample structures of ITPS-5.
not contribute to load bearing. For simplicity, in the thermomechanical analysis, we assume them to
be bonded but with very small Young's moduli. As a result, no gaps are considered in the thermal
analysis.
(a) CMC (b) Foam
Fig. 6 Finite element models of the sample structures of ITPS-3.
(a) CMC (b) Foam
Fig. 7 Finite element models of the sample structures of ITPS-4.
7
7 Distribution A. Approved for public release (PA): distribution unlimited.
III. Figures of Merit (FoMs)
A notional temperature cycle with a maximum value of 1324.37 Kelvin (1050 C, 1924 F) is
added on the external surface of CMC for a duration of 2400 seconds, as shown in Fig. 8.
0 240 480 720 960 1200 1440 1680 1920 2160 2400200
400
600
800
1000
1200
1400
Tem
pera
ture
(K)
Time (sec)
Fig. 8 Temperature cycle applied to the external surface of CMC.
The mechanical loads and boundary conditions are also applied to the plates. These loads and
boundary conditions together with the temperature changes result in the stresses and strains. The
values of the resultant stresses and strains in the sample structures are taken into consideration as
the constraints (also called the statement variables) in the optimization. The allowable magnitude
of the stresses and strains will be discussed in Section V. As the mechanical loads can be applied
on either CMC and PMC layers, we consider two types of loading:
(1) Inner-Sheet Loading (ISL): the tensile and bending loads are applied on PMC layer only;
(2) Fully Loading (FL): the tensile and bending loads are applied on both CMC and PMC layers.
To evaluate the performance of the panels with �ve core options and the BPS, we de�ne three
di�erent FoMs as follows:
A. FoM of Tensile Strength
The FoM of tensile strength is de�ned as the max tensile force applied to the ends of the plates
divided by the e�ective areal density, which is expressed as
FoMT =P̄
ρa(1)
8
8 Distribution A. Approved for public release (PA): distribution unlimited.
where P̄ is the max tensile force the plate can sustain before any material point in the sample
structure reaches the any of the failure stresses and strains. ρa is e�ective areal density of the plate
which is equal to the total weight of the panel divided by its area.
B. FoM of Bending Strength
The FoM of bending strength can be easily calculated by
FoMB =M̄
ρa(2)
where M̄ is the max bending moment the plate can sustain before any material point in the sample
structure reaches the any of the failure stresses and strains.
C. FoM of Bending Sti�ness
The most critical feature of the plate structure is its bending sti�ness. Consequently, the FoM of
bending sti�ness is the most important index for evaluating the ITPS. The FoM of bending sti�ness
is de�ned as
FoMS =S̄
ρa(3)
where S̄ is the bending sti�ness of the plate with the con�guration from optimizing the FoMB. The
bending sti�ness is predicted by using SwiftCompTM of which the methodology is brie�y introduced
in the following section.
IV. Mechanics of Structure Genome for Plates
Mechanics of Structure Genome (MSG) is an advanced theory unifying micromechanics and
structural mechanics for composite structural analysis. Yu [29] �rst introduced this concept and
the corresponding methodology in a systematic way. This theory has been integrated into a highly
e�cient commercial FE code, SwiftCompTM, which can compute the e�ective properties for beams,
plates, shells, and 3D structures.
The study of plate homogenization problem using MSG can be found in Ref. [30]. The typical
feature of the plate is that the plate thickness is much smaller than the other two in-plane dimensions.
In light of this fact, the reference in-plane surface can be reasonably modeled as a 2D continuum.
9
9 Distribution A. Approved for public release (PA): distribution unlimited.
The material point of this 2D continuum poses a generalized 2D plate constitutive relation, that is,
a linear mapping from the generalized plate strains to the generalized plate stress resultants through
the plate sti�ness matrix. To obtain this plate sti�ness matrix, one needs constitutive modeling
over the material point, that is, the structure genome (SG). The original 3D heterogeneous structure
can be described by macro coordinates xi, where xα are two orthogonal arc-length coordinates in
the in-plane reference surface and x3 is the thickness coordinate. (Greek indices assume values 1
and 2 while Latin indices assume 1, 2, and 3. Repeated indices are summed over.). In addition,
a set of micro coordinates yi = xi/ε are introduced to denotes the rapid change in the material
characteristics in SG, with ε being a small parameter denoting the order of a term. As shown in
Figs. 1, 2, and 3, one can translate the micro coordinates into the ANSYS coordinates system using
y1 = Z y2 = X y3 = Y (4)
The deformation of the original 3D structure can be formulated by the global displacements
from the 2D in-plane reference surface and the 3D unknown �uctuation w(xα, yj) in SG. Variational
Asymptotic Method (VAM) is applied to construct an asymptotically correct macroscopic plate
model by solving the 3D �uctuation function in SG. By assessing the orders of all the quantities in
the variational statement and neglecting the terms in the order of ε , the �rst approximation of the
variational statement can be obtained as
δU ≡ δ1
2⟨ΓTDΓ⟩ = 0 (5)
where D(y1, y2, y3) is the 3D 6×6 material matrix condensed from the fourth-order elasticity tensor
expressed in the yi coordinate system. By assuming small local rotation [31], the 3D Jauman-Biot-
Cauchy strain can be expressed linearly in terms of the plate strains, that is,
Γ = ⌊Γ11 2Γ12 2Γ13 Γ22 Γ22 2Γ23 Γ33⌋T = Γϵϵ̄+ Γhw (6)
10
10 Distribution A. Approved for public release (PA): distribution unlimited.
with
Γh =
∂∂y1
0 0
∂∂y2
∂∂y1
0
0 ∂∂y2
0
∂∂x3
0 ∂∂y1
0 ∂∂x3
∂∂y2
0 0 ∂∂x3
Γϵ =
1 0 0 x3 0 0
0 1 0 0 x3 0
0 0 1 0 0 x3
0 0 0 0 0 0
0 0 0 0 0 0
0 0 0 0 0 0
(7)
and the �uctuation functions arranged in
w = ⌊w1 w2 w3⌋T (8)
as well as the plate strains arranged in
ϵ̄ = ⌊ϵ11 2ϵ12 ϵ22 κ11 κ12 + κ21 κ22⌋T (9)
with ϵαβ denoting the in-plane strains and καβ denoting the curvature strains of the plate model.
The notation ⟨•⟩ denotes a weighted integration over the domain occupied by the geometry of the
SG.
Finite element method is used to solve Eq. (5) for the �uctuation w which is using shape
functions de�ned over the SG as
w(xα; yj) = S(yj)V (xα) (10)
where S represents the shape functions and V a column matrix of the nodal values of the �uctuation
functions. Substituting Eq. (10) into Eq. (5), we obtain the leading terms for the zeroth-order
approximation in the following discretized form as
U =1
2(V TEV + 2V TDhϵϵ̄+ ϵ̄TDϵϵϵ̄) (11)
where
E = ⟨(ΓhS)TD(ΓhS)⟩ Dhϵ = ⟨(ΓhS)
TDΓϵ⟩ Dϵϵ = ⟨ΓTϵ DΓϵ⟩ (12)
Minimizing U in Eq. (11) subject to the kinematic constraints provides the following linear
system
EV = −Dhϵϵ̄ (13)
11
11 Distribution A. Approved for public release (PA): distribution unlimited.
Substituting the solution of Eq. (13) back into Eq. (11), we can calculate the strain energy storing
in the SG as the �rst approximation as
U =ω
2ϵ̄TD̄ϵ̄ (14)
where D̄ is the e�ective plate sti�ness. For plate model, ω is the area spanned by the y1 and y2
for 3D SGs. Consequently, in this study of ITPS, ω = 0.762 × 0.762 m2. We are interested by the
bending sti�ness with respect to the Z direction. Then the bending sti�ness in Eq. (3) is obtained
by
S̄ = D̄66 (15)
V. Computational Aspects
A. Finite Element Analyses of Stresses and Strains
The FEA code is developed by using ANSYS Parametric Design Language (APDL) to analyze
the stress and strain �elds in the plate structure. The stress and strain �eld can be caused by the
temperature changes, the pure mechanical loadings, or a combination of them two. Considering
of this fact, to apply the transient thermal analysis, the geometries of the plate structures are
�rstly meshed by using 3D solid element, SOLID70, which has a 3-D thermal conduction capability.
After the thermal analysis is �nished the nodal temperatures will be saved into a result �le. Then
command "ETCHG,TTS" is used to change the thermal elements to their corresponding equivalent
structural ones. The �owchart of the stress and strain analyses is shown in Fig. 9. Depending on
the loads and the mechanical boundary conditions applied to the structure, the analyses can be
classi�ed into three categories:
(1) Pure mechanical analysis: mechanical loads and boundary conditions are applied;
(2) Thermomechanical analysis: temperature loads and mechanical boundary conditions are ap-
plied;
(3) One-way coupled thermomechanical analysis: temperature loads together with mechanical
loads and boundary conditions are applied.
12
12 Distribution A. Approved for public release (PA): distribution unlimited.
Finite element mesh with thermal element
Transient heat conduction analysis
Pure mechanical analysisThermomechanical analysis
Obtain maximum stresses and strains & Calculate the Figures of Merits
Change element type to structural
One-way coupled thermomechanical analysis
Change element type to structural
Apply mechanical loadings
Apply time-dependent temperature loadings
Save time history of the nodal temperature distribution
Parametrically model geometries
Apply mechanical loadings
Apply mechanical boundary conditions
Apply mechanical boundary conditions
Apply mechanical boundary conditions
Apply time-dependent temperature loadings
Input thermal and mechanical material properties
Fig. 9 Finite element stress and strain analysis �owchart.
If the entire structure of a plate, including the insulation core and the CMC, has some load-
bearing capability, the structure underneath it (in this simple case, the PMC layer) can be lighter,
thereby compensating for the increased weight of the core and CMC. The optimization procedure
needs to loop in the analysis shown in Fig. 9 to optimize the dimension of the plate and meanwhile
�nd the max applied mechanical loads corresponding to the max FoMs. The optimization to achieve
this goal is introduced in the following.
B. Optimization Parameters
A statement variable (SV) is de�ned as the constrained conditions which the optimized results
must satisfy. During comparison of the �ve insulating core options, the SVs include the interesting
stresses and strains extracted from the sample structures. Another SV is the temperature on the
top surface of the PMC layer which is limited to be not higher than the glass transition temperature
of the PMC, 560.928 Kelvin, with the user de�ned tolerance.
A design variable (DV) is de�ned as the parameter which in�uences the optimized objective. In
order to optimize the FoMs de�ned by Eq. (2) and Eq. (1), the loads applied on the plate should also
be considered as DVs under the ANSYS conventions. To optimize the FoMs of the �ve insulating
13
13 Distribution A. Approved for public release (PA): distribution unlimited.
core options focusing on determining the best structural scheme, all of the material properties are
speci�cally �xed, and the thickness of the core is considered as a DV. In the BPS, the AETB tiles
and their TUFI coatings contribute to the total weight of this system. Therefore, in the optimization
the thickness of the ATEB layer should be set as a DV as well.
A objective function (OBJ) is de�ned as to be minimized in ANSYS and only one OBJ can be
set in one optimization. In this study, we de�ne the OBJs of bending and tensile loading cases using
Eq. (16) and Eq. (17), respectively,
OBJB = AMmax
mmin− FoMB (16)
OBJT = APmax
mmin− FoMT (17)
with A the area, mmin the minimal weight, Mmax and Pmax the maximum bending and tensile
loadings, respectively. mmin is obtained by setting the core thickness to be the minimum value.
Mmax and Pmax is obtained by conducting the pure mechanical analysis with the max core thickness
under the allowable stresses and strains. The parameters used in the optimization are summarized
in Table 2.
C. Optimization Strategy
We use the ANSYS optimization module to maximize the FoMs. The optimization is conducted
in steps with di�erent optimizing methods, as shown in Fig. 10a. The �rst step is to perform
a random search of the design space of the variables. This step is terminated after user de�ned
enough feasible solutions were found, otherwise the user de�ned total times of analyses will be
conducted. The feasible design sets are kept and others are removed before the second step, the
�rst order iteration. This step is accurate but time-consuming because there are usually several
repeats of the analyses in Fig. 9 in each optimization iteration in order to �nd the best trend of
the design variables. For example, the curve of optimized feasible FoMT of ITPS-5 (ISL) is shown
in Fig. 10b. By applying the optimization steps in Fig. 10a, the APDL code �rst scans the whole
design variable space with a random search method, which results in feasible design sets quickly.
After 200 loops of random search, the unfeasible solutions are deleted, and the �rst order iteration
14
14 Distribution A. Approved for public release (PA): distribution unlimited.
Table 2 Parameters used in optimization
Parameter Physical Meaning Type Maximum Minimum
εCMCp Max principal strain in CMC SV 0.0015 0
εFoamp Max principal strain in Foam SV 0.0015 0
εPMCp Max principal strain in PMC SV 0.01 0
σCMCy (MPa) Inter-lamina normal stress in CMC SV 10 0
σFoamy (MPa) Inter-lamina normal stress in Foam SV 5 0
σPMCy (MPa) Inter-lamina normal stress in PMC SV 25 0
σCMCxy , σCMC
yz (MPa) Inter-lamina shear stress in CMC SV 38 0
σFoamxy , σFoam
yz (MPa) Inter-lamina shear stress in Foam SV 7 0
σPMCxy , σPMC
yz (MPa) Inter-lamina shear stress in PMC SV 100 0
T (K) Temperature of PMC top surface SV 560.928 293.15
M (kNm) Bending moment DV User De�ned 0
P (kN) Tensile force DV User De�ned 0
hcore (m) Thickness of insulation core DV 0.2 0.125
hAETB (m) Thickness of ATEB layer of parasitic scheme DV 0.2 0.125
method starts with the best feasible sets and provides an accurate and �ne convergence to the �nal
FoM value.
Select the feasible design sets and remove the others
Use the best of the selected sets as the start-point of next optimizations
First order iteration optimization method
Convergence ?
Final optimized design set
Random search optimization method
Yes
No
(a) Flowchart of optimization steps
0 20 40 60 80 100 120 140 160 180 200 2200
2
4
6
8
10
12
14
16
18
FoM
of t
ensi
le st
reng
th (k
Nm
2 /kg)
Number of design set in optimization
(b) Optimizing evolution of FoMT for ITPS-5 (ISL)
Fig. 10 Optimization strategy
15
15 Distribution A. Approved for public release (PA): distribution unlimited.
VI. Results and Discussions
It turns out that only ITPS-1, 5, and BPS have feasible optimized solutions. In addition, it is
shown by looping the thermomechanical analysis from Fig. 9 into the optimization that ITPS-2, 3,
and 4 can not even survive the temperature change. In other words, option 2, 3, and 4 will fail due
to pure thermal loads.
A. Solution with Tensile Load
Table 3 shows the optimized design variables from the feasible solutions under tensile loads.
The loading type does not a�ect the optimized con�guration (that is, the mass) of the ITPS-1.
In contrast, ITPS-5 (FL) has larger mass than ITPS-5 (ISL). BPS has the minimum weight and
maximum load-bearing capability among the candidates.
Table 3 Optimization data under tensile load
Variables ITPS-1 (FL) ITPS-1 (ISL) ITPS-5 (FL) ITPS-5 (ISL) BPS
m (kg) 40.5522 40.5005 38.6917 43.2708 11.2455
P (kN) 8.9164 473.57 823.91 717.71 3404.0
The comparison of FoMT from feasible optimizations is shown in Fig. 11. The BPS provides
much larger FoMT than ITPSs. Under FL, the FoMT of ITPS-5 is 118 times of the FoMT of ITPS-1.
Under ISL, the FoMT of ITPS-5 is 14 times of the FoMT of ITPS-1.
0.17978 1.1712421.29428 16.58648
302.6999
ITPP-1 (FL) ITPP-1 (ISL) ITPP-5 (FL) ITPP-5 (ISL) BPS0.0E+00
5.0E+01
1.0E+02
1.5E+02
2.0E+02
2.5E+02
3.0E+02
3.5E+02
FoM
s of T
ensil
e St
reng
th (k
Nm
2 /kg)
Fig. 11 Comparison of FoMs of tensile strength from feasible optimizations.
16
16 Distribution A. Approved for public release (PA): distribution unlimited.
B. Solution with Downwards Bending
Table 4 shows the optimized design variables from the feasible solutions under downwards
bending loads. The loading type does not singi�cantly a�ect the optimized con�guration (that
is, the mass) of the two ITPSs. BPS has the minimum weight while ITPS-5 (ISL) has maximum
load-bearing capability among the candidates.
Table 4 Optimization data under downwards bending load
Variables ITPS-1 (FL) ITPS-1 (ISL) ITPS-5 (FL) ITPS-5 (ISL) BPS
m (kg) 40.5142 40.9323 45.6487 45.8773 13.1183
M (kNm) 3.3029 0.53951 51.554 53.171 4.5592
S (kNm) 7872.0006 8074.0202 19584.404 19865.993 76.8721
The comparison of FoMB from feasible optimizations under downwards bending loads is shown
in Fig. 12. ITPS-5 provides much larger FoMB than the others. Under FL, the FoMB of ITPS-5
is 14 times of the FoMB of ITPS-1. Under ISL, the FoMB of ITPS-5 is 88 times of the FoMB of
ITPS-1.
81.5247813.18478
1127.35343 1158.95343
347.49891
ITPP-1 (FL) ITPP-1 (ISL) ITPP-5 (FL) ITPP-5 (ISL) BPS0
250
500
750
1000
1250
1500
FoM
s of D
ownw
ards
Ben
ding
Stre
ngth
(Nm
3 /kg)
Fig. 12 Comparison of FoMs of downwards bending strength from feasible optimizations.
The comparison of FoMS from feasible optimizations under downwards bending loads is shown
in Fig. 13. Again, ITPS-5 provides much larger FoMS than the others. For both of the FL and ISL,
the FoMS of ITPS-5 is 2.3 times of the FoMS of ITPS-1.
17
17 Distribution A. Approved for public release (PA): distribution unlimited.
107.89221 109.47642
249.11042 251.43305
3.40252
ITPP-1 (FL) ITPP-1 (ISL) ITPP-5 (FL) ITPP-5 (ISL) BPS0
50
100
150
200
250
300
FoM
s of D
ownw
ards
Ben
ding
Stif
fnes
s (kN
m3 /k
g)
Fig. 13 Comparison of FoMs of downwards bending sti�ness from feasible optimizations.
C. Solution with Upwards Bending
Table 5 shows the optimized design variables from the feasible solutions under upwards bending
loads. The loading type does not signi�cantly a�ect the optimized con�guration (that is, the mass)
of the two ITPSs. BPS has the minimum weight while ITPS-5 (ISL) has maximum load-bearing
capability among the candidates.
Table 5 Optimization data under upwards bending load
Variables Option 1 (FL) Option 1 (ISL) Option 5 (FL) Option 5 (ISL) Parasitic Scheme
m (kg) 40.6965 40.9804 40.0593 41.1182 12.1451
M (kNm) 7.3186 0.42728 14.718 15.130 4.8145
S (kNm) 7959.4435 8099.7116 13475.184 16332.648 76.8721
The comparison of FoMB from feasible optimizations under upwards bending loads are shown
in Fig. 14. BPS provides larger FoMB than the others. FoMBs of BPS and ITPS-5 are quite close.
In addition, they are much larger than FoMBs of ITPS-1. Under FL, the FoMB of ITPS-5 is 1.9
times of the FoMB of ITPS-1. Under ISL, the FoMB of ITPS-5 is 3.5 times of the FoMB of ITPS-1.
The comparison of FoMS from feasible optimizations under upwards bending loads are shown
in Fig. 15. ITPS-5 provides much larger FoMS than the others do. Under FL, the FoMS of ITPS-5
is 1.8 times of the FoMS of ITPS-1. Under ISL, the FoMS of ITPS-5 is 2.1 times of the FoMS of
18
18 Distribution A. Approved for public release (PA): distribution unlimited.
189.87478
100.42478
367.45343 355.4534
396.3989
ITPP-1 (FL) ITPP-1 (ISL) ITPP-5 (FL) ITPP-5 (ISL) BPS0
50
100
150
200
250
300
350
400
450
FoM
s of U
pwar
ds B
endi
ng S
treng
th (N
m3 /k
g)
Fig. 14 Comparison of FoMs of upwards bending strength from feasible optimizations.
ITPS-1.
108.57864 109.68899
195.31756
230.63884
3.67517
ITPP-1 (FL) ITPP-1 (ISL) ITPP-5 (FL) ITPP-5 (ISL) BPS0
50
100
150
200
250
FoM
s of U
pwar
ds B
endi
ng S
tiffn
ess (
kNm
3 /kg)
Fig. 15 Comparison of FoMs of upwards bending sti�ness from feasible optimizations.
D. Critical Statement Variables in ITPS-5
In this section we check the critical SV which limits the FoMs of the champion among the
ITPSs. Based on the optimization results, ITPS-5 is selected as the best option. Compared with
BPS, although the FoMT of ITPS-5 is much less than that provided by BPS, it has much larger
FoMS and downwards bending FoMB. Its upwards bending FoMB is very close to that of BPS.
From the stress and strain analyses, the critical SV of the optimization is found to be the
maximum principal strains in either CMC, foam, or both of them. Fig. 16 plots these critical
principal strain values with respect to the time duration of 2400 seconds of the temperature cycle.
19
19 Distribution A. Approved for public release (PA): distribution unlimited.
Note that the plotted data is the maximum of the elemental averaged principal strain from the
elements meshing the sample structures. As a result, the data may not represent the value from the
unique location in the sample structures. Fig. 16a shows that under tensile loads the principal strains
0 240 480 720 960 1200 1440 1680 1920 2160 2400-1.6E-03
-1.2E-03
-8.0E-04
-4.0E-04
0.0E+00
4.0E-04
8.0E-04
1.2E-03
1.6E-03
Prin
cipa
l Stra
ins
Time (sec)
1 (FL)
2 (FL)
3 (FL)
1 (ISL)
2 (ISL)
3 (ISL)
(a) Sampled from foam under tensile load
0 240 480 720 960 1200 1440 1680 1920 2160 2400-8.0E-04
-4.0E-04
0.0E+00
4.0E-04
8.0E-04
1.2E-03
1.6E-03
Prin
cipa
l Stra
ins
Time (sec)
1 (FL)
2 (FL)
3 (FL)
1 (ISL)
2 (ISL)
3 (ISL)
(b) Sampled from CMC under downwards bending load
0 240 480 720 960 1200 1440 1680 1920 2160 2400-1.6E-03
-1.2E-03
-8.0E-04
-4.0E-04
0.0E+00
4.0E-04
8.0E-04
1.2E-03
Prin
cipa
l Stra
ins
Time (sec)
1 (FL)
2 (FL)
3 (FL)
1 (ISL)
2 (ISL)
3 (ISL)
(c) Sampled from foam under downwards bending load
0 240 480 720 960 1200 1440 1680 1920 2160 2400-1.6E-03
-1.2E-03
-8.0E-04
-4.0E-04
0.0E+00
4.0E-04
8.0E-04Pr
inci
pal S
train
s
Time (sec)
1 (FL)
2 (FL)
3 (FL)
1 (ISL)
2 (ISL)
3 (ISL)
(d) Sampled from foam under upwards bending load
Fig. 16 Time dependent maximum principal strains in ITPS-5.
in the foam reached the maximum allowable value. In addition, the loading type of tensile load does
a�ect the change of maximum strain with respect to time in the sample structure. Compared with
FL, ISL results in larger ε2 and ε3. Consequently, FL should be applied as the loading condition in
the practical utilization of the ITPS-5. Fig. 16b and 16c show that under downwards bending loads,
principal strains in both of the CMC and the foam reached the maximum allowable values. Fig. 16d
shows that under upwards bending loads, principal strains in the CMC reached the maximum
allowable value. Furthermore, from Fig. 16b, 16c, and 16d it can be seen that the loading type
of bending load does not signi�cantly a�ect the change of maximum strain with respect to time in
20
20 Distribution A. Approved for public release (PA): distribution unlimited.
the sample structures. It can also be concluded that developing sti� but light CMC and SiC foam
materials will contribute the strength of the ITPS.
VII. Conclusion
A framework for the design and analysis of sandwich panels that could form the basis for an
Integrated Thermal Protection System (ITPS) has been developed using the ANSYS parametric
design language (APDL) and SwiftCompTM. Figures of Merit (FoMs) are de�ned to evaluate the
thermomechanical performance of plates with various options for the layout of the insulation core.
By optimizing the FoMs, the core option of simply bonded foam appeared to be the best design
con�guration, as it obtained the best FoMs among the core options. The baseline parasitic scheme
(BPS) containing AETB tiles with representative con�guration of 0.762 × 0.762 m2 area was also
analyzed. By evaluating the FoMs, the ITPS with core option 5 (perfectly bonded foam only) has
better bending potentials but worse tensile performance than BPS. The fully loading type is shown
to be better than the inner-sheet loading type for core option 5 according to the time dependent
principal strains. The data also indicates that the identi�cation of the necessary requirements of a
foam and CMC to satisfy ITPS requirements turns out to be critical study contributing the devel-
opment of the ITPS. Evaluating other parameters, such as the material properties, in the same way
should be a good use of this approach in the future. In addition, integrating SwiftCompTM directly
into ANSYS optimization algorithm would further enhance the computational e�ciency. Lighter
CMC materials and functionally graded sti�ers might increase the feasibility and competitiveness
of those core options with sti�eners including option 1, 2, 3, and 4.
Appendix: Material Properties
For simplicity, the materials are assumed to be isotropic. S200 CMC manufactured by COI
Ceramics, Inc. [32] is used as the external face sheet of the ITPS. The reinforcing �ber is Ceramic
Grade Nicalon (CG-Nicalon, NL-201), and the SiC matrix is made by using a Polymer In�ltration
Pyrolysis (PIP) process. The density is 2000 kg/m3at 293 Kelvin and 1900 kg/m
3at 3000 Kelvin.
The Young's modulus is 96 GPa at 293 Kelvin and 90 GPa at 1900 Kelvin, and the Poisson's
ratio is 0.27, not changing with respect to temperature. The temperature dependent speci�c heat
21
21 Distribution A. Approved for public release (PA): distribution unlimited.
capacity, thermal conductivity, and coe�cient of thermal expansion (CTE) are shown in Figure 17.
The proportional limit and the onset of signi�cant matrix cracking, and hence of environmental
degradation is in the neighborhood of 0.15% strain for most SiC based CMCs. Consequently, that
is taken to be the allowable strain in this work.
250 500 750 1000 1250 1500600
700
800
900
1000
1100
1200
1300
Spec
ific
heat
cap
acity
(J/k
g-K
)
Temperature (K)
(a) Speci�c heat capacity
250 500 750 1000 1250 15001.70
1.75
1.80
1.85
1.90
1.95
2.00
2.05
2.10
Ther
mal
con
duct
ivity
(W/m
-K)
Temperature (K)
(b) Thermal conductivity
250 500 750 1000 1250 15002.5
2.6
2.7
2.8
2.9
3.0
3.1
CTE
(1.0
e-6/
K)
Temperature (K)
(c) CTE
Fig. 17 Material properties of CMC.
SiC foam produced by Ultramet [33] with density of 320 kg/m3is used in the insulation core.
The Young's modulus of this foam is 2.873 GPa, and the Poisson's ratio is 0.22. In addition, the
speci�c heat capacity is 1422.56 kJ/kg-K. The other temperature dependent material properties
are shown in Figure 18.
The PMC is the 4-ply-fabric T650 laminate [20] of which the Young's modulus is 77 GPa at
293.15 Kelvin and 81 GPa at 616.15 Kelvin, and the Poisson's ratio is 0.08. The density is 1900
kg/m3, and the speci�c heat capacity is 1200 kJ/kg-K. The other temperature dependent material
properties of the PMC are shown in Fig. 19.
22
22 Distribution A. Approved for public release (PA): distribution unlimited.
250 500 750 1000 1250 15001.1
1.2
1.3
1.4
1.5
1.6
1.7
1.8
Ther
mal
con
duct
ivity
(W/m
-K)
Temperature (K)
(a) Thermal conductivity
250 500 750 1000 1250 15002.0
2.5
3.0
3.5
4.0
4.5
5.0
CTE
(1.0
e-6/
K)
Temperature (K)
(b) CTE
Fig. 18 Material properties of Foam.
250 300 350 400 450 500 550 600 6500.20
0.25
0.30
0.35
0.40
0.45
0.50
Ther
mal
con
duct
ivity
(W/m
-K)
Temperature (K)
(a) Thermal conductivity
250 300 350 400 450 500 550 600 650 7000.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
CTE
in p
lane
(1.0
e-6/
K)
Temperature (K)
(b) CTE in plane
250 300 350 400 450 500 550 600 6503
4
5
6
7
8
9
10
11
CTE
thro
ugh
thic
knes
s (1.
0e-5
/K)
Temperature (K)
(c) CTE through thickness
Fig. 19 Material properties of PMC.
Acknowledgments
This work is supported by the Air Force Research Laboratory Rapid Development and Insertion
of Hypersonic Materials program. The views and conclusions contained herein are those of the
23
23 Distribution A. Approved for public release (PA): distribution unlimited.
authors and should not be interpreted as necessarily representing the o�cial policies or endorsement,
either expressed or implied, of the funding agency.
References
[1] Henline, W. D., �Thermal Protection Analysis of Mars-Earth Return Vehicles,� Journal of Spacecraft
and Rockets, Vol. 29, No. 2, 1992, pp. 198�207.
[2] Laub, B. and Venkatapathy, E., �Thermal Protection System Technology and Facility Needs for De-
manding Future Planetary Missions,� Planetary Probe Atmospheric Entry and Descent Trajectory Anal-
ysis and Science (ESA SP-544), edited by A. Wilson, European Space Agency, ESA Publications Di-
vision, Noordwijk, Netherlands, February 2004, pp. 239�247.
[3] Kelly, H. N. and Blosser, M. L., �Active Cooling from the Sixties to NASP,� NASA Conference Publi-
cation 3157, Current Technology for Thermal Protection Systems, edited by S. J. Scotti, NASA, O�ce
of Management NASA, NASA Langley Research Center, Hampton, Virginia, USA, February 1992, pp.
189�250.
[4] Bertin, J. J. and Cummings, R. M., �Fifty Years of Hypersonics: Where We've Been, Where We're
Going,� Progress in Aerospace Sciences, Vol. 39, No. 6-7, 2003, pp. 511�536.
[5] Goldstein, H., �Reusable Thermal Protection System Development�a Prospective,� NASA Conference
Publication 3157, Current Technology for Thermal Protection Systems, edited by S. J. Scotti, NASA,
O�ce of Management NASA, NASA Langley Research Center, Hampton, Virginia, USA, February
1992, pp. 1�17.
[6] Baumgartner, R. I., �VenturestarTM Single Stage to Orbit Reusable Launch Vehicle Program Overview,�
AIP Conference Proceedings Space Technology and Applications International Forum, Vol. 387, Amer-
ican Institute of Physics, January 1997, pp. 1033�1040.
[7] Blosser, M. L., Advanced Metallic Thermal Protection Systems for Reusable Launch Vehicles, Ph.D.
thesis, Dept. of Mechanical and Aerospace Engineering, University of Virginia, Charlottesville, VA,
May 2000.
[8] Chase, R. L., �Comments on a Military Transatmospheric Aerospace Plane,� AIP Conference Proceed-
ings, Vol. 387, American Institute of Physics, 1997, pp. 1185�1194.
[9] Penn, J. P. and Lindley, C. A., �Space Tourism Optimized Reusable Spaceplane Design,� AIP Confer-
ence Proceedings, Vol. 387, American Institute of Physics, 1997, pp. 1073�1090.
24
24 Distribution A. Approved for public release (PA): distribution unlimited.
[10] Cook, L. M. and Ball, J., �High-Alpha Space Trucks,� AIP Conference Proceedings, Vol. 387, American
Institute of Physics, 1997, pp. 1119�1124.
[11] Andrews, D., Paris, S., and Rubeck, M., �Suborbital Freight Delivery Concept Exploration,� AIP
Conference Proceedings, Vol. 387, American Institute of Physics, 1997, pp. 1025�1031.
[12] Hunt, J. L., Lockwood, M. K., Petley, D. H., and Pegg, R. J., �Hypersonic Airbreathing Vehicle Visions
and Enhancing Technologies,� AIP Conference Proceedings, Vol. 387, American Institute of Physics,
1997, pp. 1285�1295.
[13] Rasky, D. J., Milos, F. S., and Squire, T. H., �Thermal Protection System Materials and Costs for Future
Reusable Launch Vehicles,� Journal of Spacecraft and Rockets, Vol. 38, No. 2, 2001, pp. 294�296.
[14] Bansal, N. P., Handbook of Ceramic Composites, Chap. 9, Springer Science+Business Media, Inc., 233
Spring Street, New Yotk, NY 10013, USA, 1st ed., 2005, pp. 197�224.
[15] Glass, D. E., �Ceramic Matrix Composite (CMC) Thermal Protection Systems (TPS) and Hot Struc-
tures for Hypersonic Vehicles,� 15th AIAA Space Planes and Hypersonic Systems and Technologies
Conference, AIAA Paper 2008-2682, April 2008.
[16] Myers, D. E., Martin, C. J., and Blosser, M. L., �Parametric Weight Comparison of Advanced Metallic,
Ceramic Tile, and Ceramic Blanket Thermal Protection Systems,� Tech. rep., NASA TM 2000-210289,
June 2000.
[17] Daryabeigi, K., Cunnington, G. R., and Knutson, J. R., �Heat Transfer Modeling for Rigid High-
Temperature Fibrous Insulation,� Journal of Thermophysics and Heat Transfer , Vol. 27, No. 3, July-
September 2013, pp. 414�421.
[18] Gogu, C., Bapanapalli, S. K., Haftka, R. T., and Sankar, B. V., �Comparison of Materials for an
Integrated Thermal Protection System for Spacecraft Reentry,� Journal of Spacecraft and Rockets,
Vol. 46, No. 3, May-June 2009, pp. 501�513.
[19] Zinkle, S. J., �Thermophysical and Mechanical Properties of Sic/Sic Composites,� Tech. Rep. DOE-
ER-0313-24, US. Dept. Energy. O�ce Adm. Serv., Washington, DC, 1998.
[20] Whitley, K. S. and Collins, T. J., �Mechanical Properties of T650-35/AFR-PE-4 at Elevated Tempera-
tures for Lightweight Aeroshell Designs,� 47th AIAA/ASME/ASCE/AHS/ASC Structures, Structural
Dynamics, and Materials Confere, AIAA Paper 2008-2682, May 2006.
[21] Villanueva, D., Le Riche, R., Picard, G., Haftka, R. T., et al., �Dynamic Design Space Partitioning
for Optimization of an Integrated Thermal Protection System,� 54th AIAA/ASME/ASCE/AHS/ASC
Structures, Structural Dynamics, and Materials Conference, AIAA Paper 2013-1534, April 2013.
25
25 Distribution A. Approved for public release (PA): distribution unlimited.
[22] Blosser, M. L., �Fundamental Modeling and Thermal Performance Issues for Metallic Thermal Protec-
tion System Concept,� Journal of Spacecraft and Rockets, Vol. 41, No. 2, March-April 2004, pp. 195�206.
[23] Bapanapalli, S. K., Martinez, O. M., Gogu, C., Sankar, B. V., Haftka, R. T., and Blosser, M. L.,
�Analysis and Design of Corrugated Core Sandwich Panels for Thermal Protection Systems of Space
Vehicles,� 47th AIAA/ASME/ASCE/AHS/ASC Structures, Structural Dynamics, and Materials Con-
ference, AIAA Paper 2006-1942, May 2006.
[24] Milos, F. S. and Squire, T. H., �Thermostructural Analysis of X-34 Wing Leading-Edge Tile Thermal
Protection System,� Journal of Spacecraft and Rockets, Vol. 36, No. 2, March-April 1999, pp. 189�198.
[25] Ravishankar, B., Sankar, B. V., and Haftka, R. T., �Homogenization of Integrated Thermal Protec-
tion System with Rigid Insulation Bars,� 51st AIAA/ASME/ASCE/AHS/ASC Structures, Structural
Dynamics, and Materials Conference, AIAA Paper 2010-2687, April 2010.
[26] Poteet, C. C., Abu-Khajeel, H., and Hsu, S.-Y., �Preliminary Thermal-Mechanical Sizing of a Metallic
Thermal Protection System,� Journal of Spacecraft and Rockets, Vol. 41, No. 2, March-April 2004,
pp. 173�182.
[27] Blosser, M. L., Chen, R. R., Schmidt, I. H., Dorsey, J. T., Poteet, C. C., Bird, R. K., and Wurster,
K. E., �Development of Advanced Metallic-Thermal-Protection System Prototype Hardware,� Journal
of Spacecraft and Rockets, Vol. 41, No. 2, March-April 2004, pp. 183�194.
[28] Jiang, F., Yu, W., Ye, Z., Kerans, R., and Chen, M. Y., �Analysis of Reusable Integrated Thermal Pro-
tection Panel Elements with Various Insulating Core Options,� 55th AIAA/ASME/ASCE/AHS/ASC
Structures, Structural Dynamics, and Materials Conference, AIAA Paper 2014-0351, January 2014.
[29] Yu, W., �Structure Genome: Fill the Gap between Materials Genome and Structural Analysis,� 56th
AIAA/ASCE/AHS/ASC Structures, Structural Dynamics, and Materials Conference, AIAA Paper
2015-0201, January 2015.
[30] Lee, C.-Y. and Yu, W., �Homogenization and Dimensional Reduction of Composite Plates with In-Plane
Heterogeneity,� International Journal of Solids and Structures, Vol. 48, No. 10, 2011, pp. 1474�1484.
[31] Danielson, D. A. and Hodges, D. H., �Nonlinear Beam Kinematics by Decomposition of the Rotation
Tensor,� Journal of Applied Mechanics, Vol. 54, No. 2, June 1987, pp. 258�262.
[32] �Properties of S200 CMC,� http://www.coiceramics.com/nonoxidepg.html.
[33] �Typical Physical Properties of Open-Cell Silicon Carbide,� http://www.ultramet.com/index.html.
26
26 Distribution A. Approved for public release (PA): distribution unlimited.