Composite Structures 254 (2020) 112857

Contents lists available at ScienceDirect

Composite Structures
journal homepage: www.elsevier.com/locate/compstruct

Design and thermo-structural analysis of 2D exhaust nozzle with multiple

composite layers
Yongha Kim, Sun Je Kim ⇑
The 4th R&D Institute, Agency for Defense Development, Deajeon 34186, Republic of Korea

A R T I C L E I N F O A B S T R A C T

Keywords: The IR(Infrared) signature generated by hot sections of the aircraft results in severely reducing aircraft surviv-
2D exhaust nozzle ability during the mission because this signature increases potential to be traced by IR detection system. This
Thermo‐structural analysis article deals with the novel 2D exhaust nozzle for reducing IR signature, and design process by thermo‐
Multiple composite layers structural analysis for satisfying thermal and structural requirements. The novel 2D exhaust nozzle is adopted
Specimen tests
as multiple composite layers with variable elliptical cross‐sections for dispersing high temperature core gas
Engine application test
flow and enhancing mixing performance with bypass air flows. Also, this study defines 1D thermo‐structural
model based analytical method for calculating initial performances of the 2D exhaust nozzle. In addition, initial
design parameters are firstly determined by the proposed 1D thermo‐structural model. Furthermore, detailed
thermo‐structural characteristics of the 2D exhaust nozzle are investigated by the finite element method
(ABAQUS). Based on the database, the minimum thickness of the insulator layer can be determined to maintain
structural rigidity and suppress increase of temperature at the outermost surface of the nozzle. Consequently,
we review thermo‐structural stability of the 2D exhaust nozzle by comparison of the finite element analysis and
engine application tests.

1. Introduction The 2D exhaust nozzle can mask internal engine hot parts and lower
the gas temperature by enhancing mixing performance of hot exhaust
The signatures generated by aircrafts are radar, infrared, acoustic, gas with air [4,5]. But IR signature may occur because of the high tem-
visual and etc [1]. Among these, IR(Infrared) signature is passive sig- perature of the outermost surface resulted by the surface of the 2D
nature that occurs at hot sections of the aircraft. Since the 19600 s, the exhaust nozzle exposed to hot gas for a long time. Accordingly, the
development of IR detection systems, such as MAN Portable Air thermo‐structural analysis and design are required for reducing outer-
Defence systems (MANPADS) that defends aircrafts based on IR signa- most surface temperature, improving structural rigidity of the 2D
tures, has led to researches on the reduction of IR signatures in the air- exhaust nozzle. Traditional studies have dealt with thermo‐structural
craft [2]. IR signature in the 3–5 μm wavelength range occurs from the characteristics of the nozzle system by the finite element method.
hot sections (engine case, exhaust plume), and IR signature in the Ahmed R et al. [6] carried out thermo‐structural response of steel‐
8–12 μm wavelength range from relatively low temperature sections composite rocket nozzle using ANSYS. Hao W et al. [7] investigated
(airframe, skin). These IR signatures are mainly used for IR detection the fluid dynamic and infrared radiation characteristics of the 2D‐CD
system due to low absorption rate by the atmosphere. According to vectoring exhaust system at subsonic cruise status with nozzle deflec-
Planck’s law [3], assuming that an object is a black body, the IR emis- tion angles from 0 to 20° by means of computational fluid dynamics,
sion can be expressed in terms of frequency and absolute temperature. and the results are compared with those of the baseline axisymmetric
The higher absolute temperature of the body increases the difference exhaust system. Shiva Shankar K et al. [8] performed thermo‐
in IR radiance between the background and the body, thus it increases structural analyses of a nozzle throat insert composed of a 3D carbon
the probability of aircraft detection. Therefore, a technique for sup- carbon ablative coating on 4D carbon carbon substrate to establish the
pressing IR signature by lowering the temperature of the hot sections thermal/structural integrity of the nozzle. Mahesh Babu P et al. [9]
(engine case, exhaust plume) has been researched. The representative carried out thermo‐structural analyses of the ablative layers for select-
technique is a non‐axisymmetric shape 2D exhaust nozzle technology. ing materials for solid rocket motor casing. Kimihiro K and Hugues J

⇑ Corresponding author.
E-mail address: [email protected] (S.J. Kim).

https://doi.org/10.1016/j.compstruct.2020.112857
Received 7 July 2020; Revised 29 July 2020; Accepted 13 August 2020
Available online 20 August 2020
0263-8223/© 2020 Elsevier Ltd. All rights reserved.
Y. Kim, S.J. Kim Composite Structures 254 (2020) 112857

[10] investigated aerodynamic characteristics; determined design Fig. 1. 1D thermo‐structural model with simplicity and computational
parameters of 2‐D variable exhaust nozzle by heat transfer tests, efficiency should be defined owing to requiring a number of iterative
CFD. Chloé G et al. [11] dealt with experimental/numerical approach processes for determining initial design parameters to satisfy thermal
of heat flux in dual bell nozzles. The finite element method and exper- and structural requirements. In this section, we calculate tempera-
imental method have advantages when investigating the detailed ture–pressure distributions for inner surface shapes by performing
thermo‐structural performances. For the reasons, thermo‐mechanical flow simulations (CFD); define 1D thermo‐structural model based on
characteristics of composite structures in aerospace applications are analytical methods.
generally investigated by the finite element method and experimental
method [12–15]. However these methods are not suitable for prelim- 2.1. Multiple composite layers with variable elliptical cross-sections
inary design, both of which need a number of iterative processes
owing to the difficulty of modifying a model and the amount of time The 2D exhaust nozzle is adopted as the multiple composite lay-
required to reach a solution. Therefore, a recent number of researches ers with variable elliptical cross‐sections for dispersing high temper-
on the thermo‐structural characteristics of composite structures have ature core gas flow resulting in IR signatures. The variable elliptical
been investigated by theoretical approaches due to their relative sim- cross‐sections with high aspect ratios can disperse high temperature
plicity and computational efficiency. Michele F [16] proposed a novel core gas flow; also enhance mixing performance with bypass air
design concept of the wing leading edge of a hypersonic vehicle able to flows. For obtaining temperature–pressure distributions of the inner
sustain the aerodynamic heating caused by the re‐entry trajectory of surface, we perform flow simulations (CFD) using Fluent 14.5 &
the CIRA Unmanned Space Vehicle FTB‐X. Junjie Y [17] developed a ICEM‐CFD. Settings of the flow simulation are presented in Table 1.
microscopic mechanical model to investigate failure of composite By performing flow simulations (CFD), we calculate tempera-
structures subjected to a coupled thermo‐mechanical condition. ture–pressure distributions, as shown in Fig. 2. Because of the vari-
Moleiro F [18] addressed the multi‐objective design optimization for able elliptical cross‐sections with high aspect ratios, the high
metal‐ceramic functionally graded (FG) plates composed of a main temperature is distributed widely on the upper and lower surfaces.
functionally graded material (FGM) layer under thermo‐mechanical
loadings. Saeedeh Q [19] presented the free vibration of composite
plates reinforced with graphene platelets (GPLs) under thermal envi-
ronment based on the higher‐order shear deformation plate theory.
Gyubin K [20] introduced an efficient approach for aero‐thermo‐
mechanical analysis to predict the temperature history of a hypersonic
vehicle during its full mission trajectory. However, these studies did
not perform design and thermo‐structural analysis of multiple compos-
ite layers with variable elliptical cross‐sections for 2D exhaust nozzle.
Also, the multiple composite layers with variable elliptical cross‐
sections are frequently used in aerospace applications. In addition, it
is important to quickly investigate thermo‐structural characteristics
because predicted thermo‐structural performances should be guideli-
nes in conceptual and preliminary designs for aerospace applications.
Thus, we should define 1D thermo‐structural model based analytical a) Isometric view
method with simplicity and computational efficiency.
This article deals with multiple composite layers (C‐SiC, insulator and
CFRP) for the novel 2D exhaust nozzle for reducing IR signature, and
design process by thermo‐structural analysis for satisfying thermal and
structural requirements. By flow simulations, we calculate tempera-
ture–pressure distributions of the multiple composite layers subjected
thermal–mechanical loads. Also, mechanical properties (CFRP) of the
multiple composite layers are obtained by specimen tests. For calculating
initial thermal and structural performances of the multiple composite lay-
ers, 1D thermo‐structural model is defined based the analytical method
with simplicity and computational efficiency. In addition, initial design
parameters are firstly determined by the proposed 1D thermo‐structural
b) Boundary condition
model. Furthermore, detailed thermo‐structural stability/characteristics
of the multiple composite layers are investigated by performing thermo‐
structural analyses using the finite element method (ABAQUS). Based
on the database, the minimum thickness of the insulation can be deter-
mined to maintain structural rigidity and suppress increase of temperature
at the outermost surface (CFRP) of the nozzle. Consequently, we review
thermo‐structural stability of the 2D exhaust nozzle by comparison of
the finite element analysis and engine application tests.

2. 1D problem statement and analysis of the multiple composite

layers

This article carries out the design and thermo‐structural analysis of

the multiple composite layers for the 2D exhaust nozzle subjected to c) Radii distributions of variable elliptical cross-sections
thermal–mechanical loads (high temperature core gas and inner pres-
sures) with fixed boundary condition along an edge, as shown in Fig. 1. Multiple composite layers for the 2D exhaust nozzle.

2
Y. Kim, S.J. Kim Composite Structures 254 (2020) 112857

The pressure gradient is formed in the lower surface by expansion‐ where Tgas, Tair, Troom, T1, T2, T3, T4 are temperatures of core gas flow,
shock structures resulted from expanded supersonic flows. Also, the air flow, ambient radiation, the innermost layer, the insulator layer, the
temperature–pressure distributions are used for 1D/3D thermo‐ outermost layer, and hgas, hair are convective heat transfer coefficients of
structural analyses of the multiple composite layers. By results of core gas flow, air flow, and eext, Sb are emissivity of the outermost sur-
flow simulation, the multiple composite layers consist that a C‐SiC face layer and Stefan‐Boltzmann constant, and k1, k3 are heat conduc-
innermost layer for structural performances under high temperature, tivities of the innermost/outermost layers.
a CFRP outermost layer for structural performances, insulator layer Based on the convection and radiation boundary conditions, we can
between them for preventing heat transfer from C‐SiC layer to CFRP define a governing equation for heat transfer of the multiple composite
layer. layers by the following,

2 3 2 3
Rj hgas þ K 1 K 1 0 0 2 3 Rj hgas T gas
T1
6 K 1 K1 þ K2 K 2 0 7 6 7
6 76 T 7 6 0 7
6 76 2 7 6 7
6 0 K 2 K2 þ K3 K 3 76 7 6 ¼ 0 7 ð3Þ
6   74 T 3 5 6   7
4 3 5 4 3 5
0 0 K 3 Rj þ ∑ t i ðhrad þ hair Þ þ K 4 T4 Rj þ ∑ t i ðhrad T room þ hair T air Þ
i¼1 i¼1

0 j¼i
1
2.2. 1D heat transfer model of multiple composite layers
B Rj þ j¼1
∑ tj C
B C
K i ¼ ki lnB C ð4Þ
1D heat transfer model should be defined for performing the @ j¼i1 A
thermo‐structural analysis. Accordingly, we define firstly schematic Rj þ ∑ t j
j¼1
of a thermo‐structural problem, as shown in Fig. 3. The innermost
layer and outermost layer are subjected to high temperature core gas

hrad ¼ σ b ɛðT 4 þ T room Þ T 4 2 þ T room 2 ð5Þ
flow and air flow. So, convection and radiation boundary conditions
for the multiple composite layers can be written as below [21], where t1, t2, t3 are thicknesses of the innermost layer, the insulator
layer, the outermost layer, and Rj is an inner radius of the innermost
- Convection boundary condition for the innermost layer layer (jth profile). This governing equation is linearized for solving heat
  transfer equations. The convection and radiation boundary conditions

 Rj þ t 1 for the outermost layer are dependent to temperature of the outermost
Rj hgas T 1  T gas ¼ k1 ln ðT 2  T 1 Þ ð1Þ
Rj layer. For improving accurately solution, the convection and radiation
boundary conditions should be redefined by iterative solution proce-
- Convection and radiation boundary conditions for the outermost dure. Accordingly, the heat transfer problem is solved by the iterative
layer solution procedure for redefining the convection and radiation bound-
  ary conditions (recalculating temperature of the outermost surface

 3
layer), as shown in Fig. 4. After solving the heat transfer problem by
σ b ɛ T 4 4  T room 4 þ Rj þ ∑ t i hair ðT 4  T air Þ
i¼1 the iterative solution procedure, we can calculate temperatures of the
0 1
3 innermost layer, the insulator layer, the outermost layer. These temper-
BRj þ ∑ t i C atures are used for calculating thermal stresses in the next section.
B i¼1 C
¼ k3 lnB CðT 3  T 4 Þ ð2Þ
@ 2 A
Rj þ ∑ t i 2.3. 1D thermo-structural model of multiple composite layers
i¼1

In this section, thermal–mechanical stresses are defined for predict-

Table 1 ing simply thermal‐structural stability because the multiple composite
Settings of the flow simulation. layers are subjected to thermal–mechanical loads (high temperature
core gas and inner pressures). Accordingly, schematic of a thermo‐
Solver type Density-based steady simulation
structural problem is presented in the Section 2.2, as shown in Fig. 3.
Numerical Energy Included For the multiple composite layers subjected to thermal–mechanical
model equation loads, the stress–strain relations can be presented in terms of the layer
Viscous STT model with low-Re correction and
model compressibility option
directions (1, 2 coordinates) in 2D as [22]:
Species model Off
@u @2w
Materials Equation of state: ideal gas ɛα ¼ z 2 ð6Þ
(air) Specific heat: piecewise-polynomial @α @α
Thermal conductivity: kinetic energy  
Viscosity: kinetic energy @v w 1 @v @ 2 w
ɛβ ¼ þ þz  2 ð7Þ
Cell zone Operating pressure: ambient pressure @β Rj Rj @β @β
condition
Boundary Inflow: pressure-inlet condition  
condition Outflow pressure/temperature: 0 (operating @v @u 1 @v @2w
ɛαβ ¼ þ þ 2z  ð8Þ
pressure = 101,325 Pa)/288 K @α @β R @α @α@β
Solution Implicit Roe scheme
The orientation of the layers should derive the angle θ with the
Spatial Gradient: least squared cell based method
discretization Transport equations: MUSCL
rectangular coordinates α and β. The transformation of stresses forms
the 1, 2 coordinates to the α and β coordinates can be performed by

3
Y. Kim, S.J. Kim Composite Structures 254 (2020) 112857

a) Pressure distribution

b) Temperature distribution

Fig. 2. Results of flow simulations for the 2D exhaust nozzle.

Fig. 3. Schematic of the 1D thermo-structural problem.

the transformation matrix T. The stress–strain relationship for a typical where σα, σβ, σα,β are normal stress components, a shear stress compo-
kth layer becomes: nent to the global coordinates, and ɛα, ɛβ, ɛαβ are normal strain compo-
nents, the engineering shear strain to the global coordinates, and θ is
2 3 82 3 2 3 9
σα < ɛα
hi > αα >
= the orientation of fibers in a lamina. The stresses and stress couples over
6 7 6 7 6 7
4 σβ 5 ¼ Q 4 ɛ β 5  4 αβ 5ΔT ð9Þ the thickness of the multiple composite layers are integrated to obtain
k>
: >
;
σ αβ k ɛ αβ 0 force and moment resultants.
k 2 3 2 3
ɛ 0α κα
6ɛ 7 6 7
hi ½N M  þ ½N T  ¼ ½A4 0β 5 þ ½B4 κ β 5 ð13Þ
Q ¼ ½T 1 ½Q½T  ð10Þ γ 0αβ κ αβ
2 3 2 3
2 E1 ν12 E 2 3 ɛ0α κα
1ν12 ν21 1ν12 ν21
0 6 7 6 7
6 7 ½M M  þ ½M T  ¼ ½B4 ɛ0β 5 þ ½D4 κβ 5 ð14Þ
½Q ¼ 4 ν12 E2 E2
0 5 ð11Þ
1ν ν 12 21 1ν12 ν21 γ 0αβ κ αβ
0 0 G12
n hi n hi
 2 
2 3 zk  zk1 2
½A ¼ ∑ Q ðzk  zk1 Þ; ½B ¼ ∑ Q ; ½D
cos2 θ sin2 θ 2cosθsinθ k¼1 k k¼1 k 2
6 7 h i  
T¼4 sin2 θ cos2 θ 2cosθsinθ 5 ð12Þ n  zk 3  zk1 3
¼ ∑ Q ð15Þ
cosθsinθ cosθsinθ cos θ  sin θ
2 2
k¼1 k 3

4
Y. Kim, S.J. Kim Composite Structures 254 (2020) 112857

Fig. 4. Flow chart for the heat transfer analysis.

a) Side view

a) Test section of the tensile testing b) High temperature oven with the tensile
machine testing machine

Fig. 6. Specimen test for measurement of stiffness and strength.

2 pgas ΔAj
3
Cj
6 ½ðaj þt j Þðbj þt j Þaj bj π
6   7
7
½N M  ¼ 6 R 7
6 N β ¼ rr 0 þt j 0 gas ð 02 j Þ2 air þ ð
gas air Þ 20 ð 0 j2Þ dR 7
2 2
r 2 p  r þt p p p r 2 r þt ð16Þ
4 0
ðr 0 þt j Þ r 0 ðr 0 þt j Þ r 0 2 R 5
0
b) Front view 2 3
Z h  i αα
6 7
Fig. 5. Area subjected to pressures of core gases in each profile of the multiple ½N T  ¼ Q 4 αβ 5 ΔT k dz ð17Þ
k
composite layers. 0 k
2 3
0
6 7
½M M  ¼ 4 0 5 ð18Þ
0
where ɛ0, κ are mid‐plane strains, curvatures, and N, M are the ther-
mal–mechanical force and moment resultants, and [A], [B], [D] are 2 3
Z h  i αα
membrane stiffness, membrane‐bending stiffness, bending stiffness 6 7
½M T  ¼ Q 4 αβ 5 ΔT k zdz ð19Þ
matrices. According to the thermo‐structural problem, the force resul- k
tant of each profile can be defined as follows, 0 k

5
Y. Kim, S.J. Kim Composite Structures 254 (2020) 112857

Fig. 7. Specimens for the tensile tests under room and elevated temperatures.

Table 2
Test settings under room and elevated temperatures.

Value Unit

Displacement rate 2 mm/min

Strain gage length 5 mm
Gage resistance 120 Ω
Humidity condition 50 %
Temperature conditions 25, 130, 170 °C

where aj, bj, Cj, r0, tj are lengths, circumference, inner radius, thickness
of ellipse (each profile), and ΔAj is area subjected to pgas as shown in
Fig. 5. The mechanical force resultant is defined based on stress formu-
lations of thick cylindrical structures [23]. These parameters (Cj, ΔAj)
can be presented as below [24],
8 9
>
> >
>

 < 3ðaj  bj Þ 2 =
Cj ¼ π aj þ bj 1 þ r ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
h iffi
>
>
2
2
2 >
: 10 aj þ bj þ aj þ bj 4 aj þ bj  ðaj  bj Þ2 > ;

ð20Þ
    
sin2ϕj sin2ϕn
ΔAj ¼ 8 aj bj ϕj þ  an bn ϕn þ ð21Þ
2 2
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi! sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi!
1 an 2 bn 2  an 2 bj 2 1 aj 2 bn 2  aj 2 bj 2
ϕj ¼ sin ; ϕn ¼ sin ð22Þ
aj 2 bn  an 2 bj
2 2
aj 2 bn 2  an 2 bj 2

Based on these thermo‐mechanical force/moment resultants, the

mid‐plane strains, curvatures can be rewritten as follows,

½ɛ 0  ¼ ½A1 ½N M   ½A1 ð½B½κ  þ ½N T Þ ð23Þ

1
½κ  ¼ ½D  ½B½A1 ½B ½M M   ½B½A1 ½N M   ½B½A1 ½N T  þ ½M T 
ð24Þ
Fig. 8. Load–displacement curves of the specimens.
For calculating failure index of each layer, we can formulate the
stresses of each layer by thermal–mechanical loads in terms of the
strain in the global coordinate below, 2 3 2 3
2 3 02 3 2 31 ɛ1 κ1
σ1 ɛ1 κ1 6 7
T 1 6 7
1
6 7 B6 7 6 7C 4 ɛ 2 5 ¼ ½T k ½ɛ0 ; 4 κ 2 5 ¼ ½T k T ½κ 0  ð26Þ
4 σ 2 5 ¼ ½Qk @4 ɛ2 5 þ z4 κ2 5 A ð25Þ
ɛ 12 k κ 12 k
σ 12 k
ɛ 12 k
κ 12 k

6
Y. Kim, S.J. Kim Composite Structures 254 (2020) 112857

Table 3
Summary of specimen test results.

Value Unit

Temperature conditions 25 130 170 °C

Elastic modulus 53,000 51,100 49,400 MPa
Failure stress 544.27 360.54 234.70

Table 4 Table 5
Geometric parameter for parametric analysis. Initial parameters for satisfying thermal and structural requirements.

Value Unit Value Unit

C-SiC innermost layer 2–10 mm C-SiC innermost layer 4.9 mm

Insulator layer 5–25 Insulator layer 21.8
CFRP outermost layer 2–10 CFRP outermost layer 4.9
Profile sections of the multiple composite layers are presented in Section 2.1. Temperature of CFRP outermost layer 99.1 °C
Maximum failure index 0.58 –

σ 1 2 σ 1 σ 2 σ 2 2 σ 12 2
f ¼  þ þ 2 ð27Þ
SL 2 SL ST ST 2 S
where SL, ST, S are longitudinal, transverse, shear strengths. In this
research, the longitudinal, transverse, strengths are the same values
because the CFRP layer is a quasi‐isotropic laminate, the other layers
are isotropic shells. These failure indexes are used for determining ini-
tial design parameters of the multiple composite layers for the 2D
exhaust nozzle in the Chapter 4.

3. Specimen tests for mechanical properties

Carbon fiber reinforced plastic (CFRP) plays an important role in

structural performances of the multiple composite layers. Thus speci-
men tests are conducted according to ASTM for obtaining experimen-
tal mechanical properties (stiffness, strength) of CFRP under room and
elevated temperatures (130 °C, 170 °C) for applying design of the mul-
a) Maximum failure index tiple composite layers.

3.1. Specimen preparation and testing

Specimen tests are conducted by ASTM 3039 (Standard Test

Method for Tensile Properties of Polymer Matrix Composite Materials)
for measurement of stiffness and strength [26]. ASTM 3039 is
designed to produce a simple jig for tensile tests of CFRP specimens.
The test load is normally applied to the specimen via the jig under
room and elevated temperatures; is collected from load cell of material
testing system INSTRON 5900R, as shown in Fig. 6. The test load and
actuator displacement is synchronized in the data acquisition system.
There are eleven specimens for the tensile tests under room and ele-
vated temperatures, as shown in Fig. 7; detailed information on the
test settings, the specimens under room and elevated temperatures is
presented in Table 2.

3.2. Specimen test results

As shown in Fig. 8, the load–displacement curves of the specimens

b) Temperature of CFRP layer
under tensile loading are fairly linear up to final failure. For each spec-
Fig. 9. Results of parametric analyses.
imen, the maximum failure stress and elastic modulus are summarized
in Table 3. The elevated thermally conditioned specimen (170 °C) had
a strength value about 50–60% lower than that under room tempera-
Finally, this article formulates failure index resulted from calcu- ture conditions. For the thermo‐structural analysis, other mechanical
lated the stresses of each layer. In addition, we use Tsai‐Wu failure cri- properties (heat conductivity, thermal expansion) are obtained by
teria for checking material failures as follows [25], open sources.

7
Y. Kim, S.J. Kim Composite Structures 254 (2020) 112857

Fig. 10. Finite element model of the multiple composite layers.

4. Parametric analysis of multiple composite layers by 1D thermo-

Table 6
structural analysis
Geometric parameters of the finite element analysis.

Value Unit In this article, a parametric analysis is performed to investigate the

C-SiC innermost layer 5 mm effects of the design parameters on the thermo‐structural characteris-
Insulator layer 10, 15, 20, 25 tics. Also, initial design parameters of the multiple composite layers
CFRP outermost layer 5 are determined for satisfying mission requirements by the 1D
Profile sections of the multiple composite layers are presented in Section 2.1.
thermo‐structural model, CFD results, specimen test results.

8
Y. Kim, S.J. Kim Composite Structures 254 (2020) 112857

layer, the outermost layer. The responses are temperature of the outer-
most layer, and maximum failure index at each profile section of the
multiple composite layers as shown Fig. 5. Also, thermo‐mechanical
forces are defined based on the results of flow simulations. Ranges
of parameters are presented in Table 4.

4.2. Results of the parametric analyses

Fig. 9 presents the results of the parametric analyses. In addition,

these results indicate that the temperature nonlinearly decreases
according to increase in the thickness of the insulator layer; they are
little changed according to changes in the thicknesses of the C‐SiC/
CFRP layers. Also, the maximum failure index tends to nonlinearly
decrease with the increases in the thicknesses of the C‐SiC/
insulator/CFRP layers. Thus, minimum thicknesses of each layer are
required for thermo‐structural design of the multiple composite layers.
Based on requirements (T4 < 100 °C, fmax < 1), Table 5 presents initial
design parameters for thermo‐structural stability of the multiple com-
posite layers. These initial design parameters are calculated by safety
factor 1.5.

5. 3D thermo-structural analysis and design of the multiple

composite layers

This article performs the thermo‐structural analysis of the multiple

composite layers for reviewing detailed thermo‐structural stability of
the 2D exhaust nozzle, obtaining thermo‐structural characteristics.

5.1. Finite element model

The finite element models of the multiple composite layers for the
2D exhaust nozzle are created by using ABAQUS. These models, as
well as the pre‐processor, are illustrated in Fig. 10. The finite element
model consists of a quadratic solid element (C3D20T) because the 2D
exhaust nozzle is not a thin‐shell structure. Also, the solid element is
suitable for the multiple composite layers subjected to thermal–me-
chanical loads because thermal–mechanical responses in the thickness
direction are necessarily investigated in this simulation. Accordingly,
each ply of the composite layer is modelled in the finite element
model. The number of solid elements is over than 39,000 (element size
is approximately 5) by the convergence analysis for quality of the
finite element analysis, and the thicknesses of layers are presented in
Table 6. The multiple composite layers are subjected to thermal–me-
chanical loads (high temperature core gas and inner pressures); fas-
tened with bolts. The load conditions are imported to
temperature–pressure distributions calculated by the flow simulation
in the Section 2.1. The boundary condition is defined that central
nodes of the bolts are fixed using multiple point constraints (MPC).
Also, cyclic and symmetric boundary conditions (x axis) are applied
to the finite element analysis for computational efficiency.

5.2. 3D thermo-structural analysis and design

The 3D thermo‐structural analyses are carried out by ABAQUS

(Coupled temp‐displacement) used as the solver of the finite element
analysis. Fig. 11 presents results of the 3D thermo‐structural analyses
(stress/temperature distributions). Stress concentrations are occurred
Fig. 11. Results of the 3D thermo-structural analyses. at positions fastened with bolts, and the failure index is less than 1.
Also, overall structural members are investigated to be enough to pro-
4.1. Main parameters and responses vide structural stability at the requirements because their margin of
safety are greater than zero. In addition, maximum failure index of
Based on CFD results, the thermal‐structural requirements are the CFRP layer except positions fastened with bolts nonlinearly
defined as that temperature of the outermost layer is less than 100 ° decreases as thickness of the insulator layer. Furthermore, temperature
C and failure indexes of the innermost layer, the insulator layer, the of the outermost layer should be less than 100 °C for thermal structural
outermost layer are less than 1. The main parameters of the multiple stability of the multiple composite layers. Thus, thickness of the insu-
composite layers are thicknesses of the innermost layer, the insulator

9
Y. Kim, S.J. Kim Composite Structures 254 (2020) 112857

Fig. 12. Engine application test settings.

face, as shown in Fig. 12. Fig. 13 presents temperatures of the outer

surface by the engine application tests. By measuring temperatures
at the four positions, the highest temperatures can be occurred at posi-
tion 1 and 2 attached to the 2D exhaust nozzle. These results are
because the outermost CFRP layer is directly exposed to the core gas
due to the nozzle tip closely located to these gases. Furthermore,
Table 7 presents a difference of less than 15% between temperature
results by the 3D thermo‐structural analyses and engine application
tests. Also, thermo‐structural stability of the 2D exhaust nozzle is ver-
ified because temperatures of the outer surface maintain below 100 °C
and the 2D exhaust nozzle is normally operated without structural fail-
ures during the engine application tests.

Fig. 13. Temperature results of the engine application tests. 6. Conclusion

This article deals with multiple composite layers (C‐SiC, insulator

and CFRP) for the novel 2D exhaust nozzle for reducing IR signature,
and design process by thermo‐structural analysis for satisfying thermal
lator layer should be over than 20 mm by results of the 3D thermo‐ and structural requirements. Accordingly, the followings are carried
structural analyses; is determined to 20 mm as design parameter. out:

5.3. Comparison by engine application tests - 1D heat transfer/thermo‐structural models are defined based the
analytical method with simplicity and computational efficiency.
Engine application tests are performed using the 2D exhaust nozzle In these models, the convection and radiation boundary conditions
under the ground test condition, and we investigate temperatures of (recalculating temperature of the outermost surface layer) can be
the outer surface. K‐type thermocouples are used for measuring high redefined by the iterative solution procedure for improving accu-
temperatures; are attached to a total of four positions of the outer sur- rately solution.

Table 7
Comparison of the 3D thermo-structural analyses and engine application tests.

Value Unit
Position #1 Position #2 Position #3 Position #4

3D thermos-structural analysis 82.14 82.14 80.32 82.34 °C

Engine application test 94.98 87.6 82.66 87.32
Difference 13.52 6.24 2.84 5.70 %

10
Y. Kim, S.J. Kim Composite Structures 254 (2020) 112857

- A parametric analysis is performed to investigate the effects of the [4] Banken G, Cornette W, Gleason K. Investigation of infrared characteristics of three
generic nozzle concepts. Proceedings of AIAA/SAE/ASME 16th Joint Propulsion
design parameters on the thermo‐structural characteristics. Also,
Conference, 1980.
initial design parameters of the multiple composite layers are deter- [5] Capone FJ, Gowadia NS, Wooten WH. Performance characteristics of
mined for satisfying thermal and structural requirements by the 1D nonaxisymmetric nozzles installed on the F-18 aircraft. J Aircraft 1980;17
thermo‐structural model. (6):387–92.
[6] Ahmed R, Gomma HH. Thermo-structural analysis of steel-composite wall nozzle.
- 3D thermo‐structural analyses of the multiple composite layers are Int J Modell Optim 2016;6(5):261–7.
performed by CFD, FEM. Also, detailed thermo‐structural stability [7] Hao W, Honghu J, Haohao L. The influence of nozzle deflection on fluid dynamic
of the 2D exhaust nozzle are investigated by obtaining thermo‐ and infrared characteristics of a two-dimensional convergent–divergent vectoring
exhaust system. Proc Inst Mech Eng Part G J Aerospace Eng 2019;233
structural characteristics. Accordingly, the final design parameters (12):4646–62.
of the multiple composite layers are determined for satisfying ther- [8] Shiva SK, Shailes BS, Sai KA, Rao TBS. Thermo-structure of dual-bell nozzle. Int J
mal and structural requirements. Eng Innov Technol 2015;4(11):292–9.
[9] Mahesh BP, Bala KG, Siva PB. Design & analysis of solid rocket motor casing for
- Engine application tests are performed using the 2D exhaust nozzle aerospace applications. Int J Curr Eng Technol 2015;5(3):1947–54.
under the ground test condition, and we investigate temperatures [10] Kimihiro K, Hugues J. Research of 2-D variable exhaust nozzle. Int J Gas Turbine,
of the outer surface. There is a difference of less than 15% between Propul Power Syst 2013;5(1):17–22.
[11] Chloé G, Andreas G, Ralf S. Experimental and numerical study of heat flux in dual
temperature results by the 3D thermo‐structural analyses and bell nozzles. J Propul Power 2013;29(1):21–6.
engine application tests. finally, thermo‐structural stability of the [12] Michele F, Concetta P, Andrea S, Aniello R. Investigating the Thermo-Mechanical
2D exhaust nozzle is verified because temperatures of the outer sur- Behavior of a Ceramic Matrix Composite Wing Leading Edge by Sub-Modeling
Based Numerical Analyses. Computation; 2020:published online.
face maintain below 100 °C and the 2D exhaust nozzle is normally
[13] Norman FK, Kyongchan S, Ivatury SR. Space shuttle orbiter wing-leading-edge
operated without structural failures during the engine application panel thermo-mechanical analysis for entry conditions. Proceedings of 51st AIAA/
tests. ASME/ASCE/AHS/ASC structures, structural dynamics, and materials conference,
2010.
[14] Kumara RR, Vinod G, Renjith S, Rajeev G, Jana MK, Harikrishnan R. Thermo-
Consequently, we review thermo‐structural stability of the novel structural analysis of composite structures. Mater Sci Eng, A 2005;412:66–70.
2D exhaust nozzle that can reduce outermost surface temperature for [15] Riccio A, Raimondo F, Sellitto A, Carandente V, Scigliano R, Tescione D. Optimum
reducing the IR signature by comparison of the finite element analysis design of ablative thermal protection systems for atmospheric entry vehicles. Appl
Therm Eng 2017;119:541–52.
and engine application tests. [16] Michele F, Roberto S, Aniello R, Emanuele B, Marco R. Thermo-structural design of
a Ceramic Matrix Composite wing leading edge for a re-entry vehicle. Compos
Declaration of Competing Interest Struct 2019;207:264–72.
[17] Junjie Y, Yiwei W, Ziwei L, Mohammed S, Fei J, Bo H, Jianqiao Y. Failure analysis
of fiber-reinforced composites subjected to coupled thermo-mechanical loading.
The authors declare that they have no known competing financial Composite Structures; 2020:published online.
interests or personal relationships that could have appeared to influ- [18] Moleiroa F, Madeira JFA, Carrera E, Reddy JN. Design optimization of functionally
graded plates under thermo-mechanical loadings to minimize stress, deformation
ence the work reported in this paper. and mass. Composite Structures; 2020:published online.
[19] Saeedeh Q, Farzad E, Vinyas M. Free vibration analysis of graphene platelets-
Acknowledgements reinforced composites plates in thermal environment based on higher-order shear
deformation plate theory. Int J Aeronaut Space Sci 2019;20:902–12.
[20] Kim G, Kang YC, Woo J, Kim JH, Cho JY. Efcient prediction of the temperature
This work has been supported by the project “A Study on Core history of a hypersonic vehicle throughout the mission trajectory with an
Technology of Stealth UAV” of Defense Acquisition Program Adminis- aerodynamic thermal load element. Int J Aeronaut Space Sci 2020;21:363–79.
tration and Agency for Defense Development. [21] Frank PI, David PD, Theodre LB, Adrienne SL. Introduction to heat transfer. 5th
ed. John Wiley & Son; 2005.
[22] Ronald FG. Principles of composite material mechanics. 4th ed. CRC Press; 2016.
References [23] Reddy JN. Theory and analysis of elastic plates and shells. CRC Press; 2007.
[24] Michael DH, Mark BV. A refinement of Ramanujan’s factorial approximation.
[1] Robert EB. The fundamentals of aircraft combat survivability: analysis and design. Ramanujan J 2014;34:73–81.
2nd ed. AIAA Education Series; 1985. [25] Tsai SW, Wu EM. A general theory of strength for anisotropic materials. J Compos
[2] Shripad PM, Hemant RS, Arvind GR. Infrared signature studies of aerospace Mater 1971;5:58–80.
vehicles. Prog Aerosp Sci 2007;43(7–8):218–45. [26] ASTM 3039, Standard Test Method for Tensile Properties of Polymer Matrix
[3] Hudson Jr RD. Infrared system engineering. Wiley; 1969. Composite Materials.

11