CFD ANALYSIS OF MISSILE SHROUD SEPARATION

A THESIS SUBMITTED TO
THE GRADUATE SCHOOL OF NATURAL AND APPLIED SCIENCES
OF
MIDDLE EAST TECHNICAL UNIVERSITY

BY

HASAN ERKAN ÇELİKER

IN PARTIAL FULFILLMENT OF THE REQUIREMENTS

FOR
THE DEGREE OF MASTER OF SCIENCE
IN
AEROSPACE ENGINEERING

FEBRUARY 2015
 Approval of the thesis

CFD ANALYSIS OF MISSILE SHROUD SEPARATION

submitted by HASAN ERKAN ÇELİKER in partial fulfillment of the requirements

for the degree of Master of Science in Aerospace Engineering Department,
Middle East Technical University by,

Prof. Dr. Gülbin Dural Ünver _______________

Dean, Graduate School of Natural and Applied Sciences

Prof. Dr. Ozan Tekinalp _______________

Head of Department, Aerospace Engineering

Assoc. Prof. Dr. Sinan Eyi _______________

Supervisor, Aerospace Engineering Dept., METU

Examining Committee Members:

Prof. Dr. Nafiz Alemdaroğlu _______________

Aerospace Engineering Dept., METU

Assoc. Prof. Dr. Sinan Eyi _______________

Aerospace Engineering Dept., METU

Prof. Dr. Serkan Özgen _______________

Aerospace Engineering Dept., METU

Assoc. Prof. Dr. Oğuz Uzol _______________

Aerospace Engineering Dept., METU

Emrah Gülay, M.Sc. _______________

Senior Engineer, ROKETSAN

Date: 05/02/2015
I hereby declare that all the information in this document has been obtained
and presented in accordance with academic rules and ethical conduct. I also
declare that, as required by these rules and conduct, I have fully cited and
referenced all material and results that are not original to this work.

Name, Last name : Hasan Erkan ÇELİKER

Signature :

iv
 ABSTRACT

CFD ANALYSIS OF MISSILE SHROUD SEPARATION

Çeliker, Hasan Erkan

M.S., Department of Aerospace Engineering
Supervisor: Assoc. Prof. Dr. Sinan Eyi

February 2015, 90 Pages

In this thesis, shroud separation on generic missile is performed by using the

Computational Fluid Dynamics (CFD). This study is divided into two main parts. In
the first part, aerodynamic coefficients of generic missile are calculated and store
separation is performed. Additionally, the numerical analyses results are compared
respectively with the experimental results in order to validate of CFD analyses
method. Three-dimensional steady/unsteady Navier-Stokes equations are solved and
three different turbulence models which are one-equation Spalart-Allmaras, two-
equation Realizable k-ε, and two-equation k-ω SST methods are used. Aerodynamic
coefficients results are validated with a HB generic missile and multi body dynamic
simulation is validated with an EGLIN store separation test case by using Chimera
grid method. Aerodynamic coefficients and store position/angle changes results are
compared with experimental data respectively and all of them are in good agreement.
As a result of all this validation, in the second part, a parametric study for shroud
separations on generic missile is carried out in order to investigate altitude, Mach
number, angle of attack, side-slip angle and ejection moment effects during the
separation, and results of CFD simulation are presented.

v
Keywords: Computational Fluid Dynamics, Shroud Separation, Store Separation,
Aerodynamic Coefficients, Multi Body Dynamic Simulation, Chimera Grid, CFD++

vi
 ÖZ

FÜZE AERODİNAMİK ÖRTÜ AYRILMASINA YÖNELİK YAPILAN HAD

ANALİZLERİ

Çeliker, Hasan Erkan

Yüksek Lisans, Havacılık ve Uzay Mühendisliği Bölümü
Tez Yöneticisi: Doç. Dr. Sinan Eyi

Şubat 2015, 90 Sayfa

Bu tezde, Hesaplamalı Akışkanlar Dinamiği (HAD) kullanılarak jenerik füze

üzerinden aerodinamik örtü ayrılması gerçekleştirilmiştir. Bu çalışma iki ana konuya
ayrılmıştır. Çalışmanın ilk kısmında jenerik füzenin aerodinamik katsayıları
hesaplanmış, harici yük ayrılması yapılmıştır. Çalışmada kullanılan sayısal
yöntemlerin doğrulanması için sırasıyla analiz sonuçları deney verileriyle
karşılaştırılmıştır. Üç boyutlu zamandan bağımsız/bağlı Navier-Stokes denklemleri
çözülmüş ve üç farklı türbülans modeli (bir denklemli Spalart-Allmaras, iki
denklemli Realizable k-ε ve iki denklemli k-ω SST metotları) denenmiştir.
Aerodinamik katsayı sonuçları HB jenerik füze analiziyle doğrulanmıştır. Çoklu
gövde dinamik analizleri ise Chimera çözüm ağı kullanılarak EGLIN harici yük
ayrılması denek taşı ile doğrulanmıştır. Sırasıyla aerodinamik katsayılar ve harici yük
konum/açı değişimleri deney verileriyle karşılaştırılmış, bütün değerler için birbiriyle
örtüşen sonuçlar elde edilmiştir. Bütün bu doğrulamalar sonucunda tezin ikinci
kısmında ise jenerik füze üzerinden aerodinamik örtü ayrılması sırasında, irtifanın,
Mach sayısının, hücum açısının, yanal sapma açısının ve uygulanan ayrılma
momentinin ayrılmaya etkisi parametrik çalışmayla incelenmiş ve analiz sonuçları
gösterilmiştir.

vii
Anahtar Kelimeler: Hesaplamalı Akışkanlar Dinamiği, Aerodinamik Örtü Ayrılması
Harici Yük Ayrılması, Aerodinamik Katsayılar, Çoklu Gövde Dinamik Simülasyonu,
Chimera Çözüm Ağı, CFD++

viii
To My Family

ix
 ACKNOWLEDGEMENTS

I wish my deepest gratitude to my supervisor Assoc. Prof. Dr. Sinan Eyi for his
guidance, advice, criticism and encouragements throughout the thesis.

I am very thankful to my parents Mrs. Zeynep Çeliker, Mr. Muhammed Çeliker for
their love, help and motivation. Without them this work would not be completed.

I want to express my best wishes to Mr. Kıvanç Arslan, Mr. Tuğrul Aydemir, Mr.
Adem Candaş and Mr. Ozan Göker for their friendship and support during this study.

I also wish to thank my manager Mr. Ali Akgül for his guidance and support during
this study. I also would like to thank my colleagues in Aerodynamics Department of
ROKETSAN for all their help and support during the thesis.

I finally would like to thank anyone who has supported my thesis effort in any way.

x
 TABLE OF CONTENTS

ABSTRACT ................................................................................................................ v

ÖZ .............................................................................................................................. vii

ACKNOWLEDGEMENTS ....................................................................................... x

TABLE OF CONTENTS .......................................................................................... xi

LIST OF FIGURES ................................................................................................ [LY

LIST OF TABLES ................................................................................................ [viiL

1. INTRODUCTION .................................................................................................. 1

1.1 Classification of Missiles .............................................................................. 1

1.1.1 Air-to-Air Missiles (AAM) .................................................................... 2

1.1.2 Air-to-Surface Missiles (ASM) .............................................................. 2

1.1.3 Surface-to-Air Missiles (SAM) .............................................................. 2

1.1.4 Surface-to-Surface Missiles (SSM) ....................................................... 2

1.1.5 Underwater-to-Underwater Missiles (UUM) ......................................... 2

1.2 Missile with Shroud....................................................................................... 3

1.3 Aim of Thesis ................................................................................................ 3

1.4 Literature Survey ........................................................................................... 4

2. METHODOLOGY................................................................................................. 9

2.1 Governing Equations ..................................................................................... 9

2.1.1 Fluid Dynamics ...................................................................................... 9

2.2 Numerical Tools and Numerical Simulation Methodology ........................ 13

2.2.1 Computational Fluid Dynamics Solver ................................................ 13

2.2.2 Turbulence Modeling ........................................................................... 14

xi
 2.2.2.1 One-equation Spalart-Allmaras Model ......................................... 15

2.2.2.2 Two-equation Realizable k-ε Model ............................................. 17

2.2.2.3 Two-equation k-ω SST Model ...................................................... 19

2.2.3 Chimera Grid Methodology ................................................................. 22

2.2.3.1 Hole Cutting .................................................................................. 22

2.2.3.2 Intergrid Communication .............................................................. 25

2.2.4 6DOF Methodology ............................................................................. 26

3. VALIDATION OF TEST CASES ...................................................................... 29

3.1 HB-1 Supersonic and Hypersonic Test Case ............................................... 29

3.1.1 Numerical Simulation .......................................................................... 30

3.1.2 Solid Model .......................................................................................... 30

3.1.3 Boundary Conditions............................................................................ 30

3.1.4 CFD Grid Independence Study ............................................................ 31

3.1.5 Turbulence Model Selection ................................................................ 35

3.1.6 Solution Domain Flow Visualization ................................................... 37

3.2 EGLIN Store Separation Test Case ............................................................. 37

3.2.1 Numerical Simulation .......................................................................... 39

3.2.2 Solid Model .......................................................................................... 39

3.2.3 Boundary Conditions............................................................................ 40

3.2.4 Turbulence Model Selection ................................................................ 41

3.2.5 CFD Simulation Results ....................................................................... 42

4. CFD ANALYSIS OF MISSILE SHROUD SEPARATION ............................. 47

4.1 Numerical Simulation .................................................................................. 48

4.2 Solid Model and Grid Generation ............................................................... 48

4.3 Boundary Conditions and Material Properties ............................................ 50

4.4 Shroud Release Procedure ........................................................................... 51

xii
 4.5 Turbulence Model Selection ....................................................................... 52

4.6 CFD Analysis Results ................................................................................. 54

4.6.1 Effects of Altitude ................................................................................ 54

4.6.1.1 Solution Domain Flow Visualization ........................................... 58

4.6.2 Effects of Mach Number ...................................................................... 63

4.6.2.1 Solution Domain Flow Visualization ........................................... 67

4.6.3 Effects of Angle of Attack ................................................................... 68

4.6.3.1 Solution Domain Flow Visualization ........................................... 72

4.6.4 Effects of Ejection Moment ................................................................. 73

4.6.4.1 Solution Domain Flow Visualization ........................................... 77

4.6.5 Effects of Side-Slip Angle ................................................................... 78

4.6.5.1 Solution Domain Flow Visualization ........................................... 82

4.6.6 Effects of Aerodynamic Loads Acting on the Missile During

Separation ........................................................................................................... 83

5. CONCLUSION..................................................................................................... 87

REFERENCES ......................................................................................................... 89

xiii
 LIST OF FIGURES

FIGURES

Figure 1.1 Inflatable Bladder for Shroud Separation [6] ............................................. 5

Figure 1.2 Launch Vehicle with Nose Panels [7] ........................................................ 6
Figure 1.3 Missile with Shroud Covers [10] ................................................................ 7
Figure 2.1 Grid Generation Procedure ....................................................................... 14
Figure 2.2 Illustration of a Chimera Grid System [18] .............................................. 23
Figure 2.3 Assignment of IN/OUT Status to End Points [18] ................................... 24
Figure 2.4 Cut-Cells of the Grid Blanketing all Grid Points [18] .............................. 24
Figure 2.5 A Hole Cut in the Grid With Respect to the Given WB [18] ................... 25
Figure 2.6 Fringe Boundary Cells Adjacent to the Cut Hole [18] ............................. 25
Figure 2.7 Steps of the Flow Solver with 6DOF Calculations ................................... 27
Figure 3.1 Geometry of the HB-1 Test Case.............................................................. 29
Figure 3.2 Solid Model and Computational Grid for HB-1 ....................................... 30
Figure 3.3 Solution Domain and Defined Boundary Condition for HB-1 Test Case 31
Figure 3.4 Coarse, Medium and Fine Grid ................................................................ 32
Figure 3.5 HB-1 Axial Force Coefficient with Respect to Mach for Different Grids 33
Figure 3.6 HB-1 Normal Force Coefficient with Respect to Mach for Different Grids
.................................................................................................................................... 34
Figure 3.7 HB-1 Pitch Moment Coefficient with Respect to Mach for Different Grids
.................................................................................................................................... 34
Figure 3.8 Normal Force Coefficient with Respect to Angle of Attack for Different
Turbulence Models (M=2.0) ...................................................................................... 35
Figure 3.9 Pitch Moment Coefficient with Respect to Angle of Attack for Different
Turbulence Models (M=2.0) ...................................................................................... 36
Figure 3.10 Mach number and Static Pressure Contour for HB-1 Test Case (M=2.0,
α=12°) ........................................................................................................................ 37

[iY
Figure 3.11 Geometry of the EGLIN Test Case......................................................... 38
Figure 3.12 Sketch of the EGLIN Test Case Wind Tunnel Test................................ 38
Figure 3.13 Solid Model of EGLIN Test Case........................................................... 39
Figure 3.14 Solution Domain and Boundary Conditions for EGLIN Test Model ..... 40
Figure 3.15 Store Position Change Results and Experimental Data with respect to
Time for Different Turbulence Models [3] ................................................................ 41
Figure 3.16 Store Angle Change Results and Experimental Data with respect to Time
for Different Turbulence Models [3] .......................................................................... 42
Figure 3.17 Store Position Change and Experimental Data With Respect to Time... 43
Figure 3.18 Store Angle Change and Experimental Data With Respect to Time ...... 43
Figure 3.19 Visual Presentation of CFD and Experimental Results .......................... 44
Figure 3.20 Pressure Distribution with Respect to Time for Store Separation .......... 45
Figure 4.1 Geometry Specification of Generic Missile.............................................. 47
Figure 4.2 Geometry Specification of Shroud Model ................................................ 48
Figure 4.3 Solid Model of the Generic Missile with Shroud (Side View) ................. 48
Figure 4.4 Surface Grid for Generic Missile with Shroud (Top View and Side View)
.................................................................................................................................... 49
Figure 4.5 Surface and Volume Grids for Generic Missile with Shroud ................... 49
Figure 4.6 System of Chimera Grid ........................................................................... 50
Figure 4.7 Solution Domain Boundary Condition .................................................... 50
Figure 4.8 Schematic drawing of the Shroud Release Procedure .............................. 51
Figure 4.9 The Flow Chart of Shroud Release Procedure .......................................... 52
Figure 4.10 Shroud Trajectories for Different Turbulence Models ........................... 53
Figure 4.11 Shroud Trajectories for Different Altitudes ............................................ 55
Figure 4.12 Z-Force Acting on the Shroud with Respect to Time for Different
Altitudes ..................................................................................................................... 56
Figure 4.13 X-Force Acting on the Shroud with Respect to Time for Different
Altitudes ..................................................................................................................... 57
Figure 4.14 Top Views of Mach Number Contour for Shroud Separation
(Altitude=5000 m, M=2.0, AOA=0° and Moment=1000 Nm) .................................. 58
Figure 4.15 Top Views of Static Pressure Contour for Shroud Separation
(Altitude=5000m, M=2.0, AOA=0° and Moment=1000 Nm) ................................... 59
[v
Figure 4.16 Top Views of Mach Number Contour for Shroud Separation
(Altitude=7500 m, M=2.0, AOA=0° and Moment=1000 Nm) .................................. 60
Figure 4.17 Top Views of Static Pressure Contour for Shroud Separation
(Altitude=7500m, M=2.0, AOA=0° and Moment=1000 Nm)................................... 61
Figure 4.18 Top Views of Mach Number Contour for Shroud Separation
(Altitude=10000 m, M=2.0, AOA=0° and Moment=1000 Nm) ................................ 62
Figure 4.19 Top Views of Static Pressure Contour for Shroud Separation
(Altitude=10000m, M=2.0, AOA=0° and Moment=1000 Nm) ................................. 63
Figure 4.20 Shroud Trajectories for Different Mach Numbers.................................. 64
Figure 4.21 Z-Force Acting on the Shroud with Respect to Time for Different Mach
Numbers ..................................................................................................................... 65
Figure 4.22 X-Force Acting on the Shroud with Respect to Time for Different Mach
Numbers ..................................................................................................................... 66
Figure 4.23 Top Views of Mach Number Contour for Shroud Separation
(Altitude=5000m, M=1.2, AOA=0° and Moment=1000 Nm) ................................... 67
Figure 4.24 Top Views of Static Pressure Contour for Shroud Separation
(Altitude=5000m, M=1.2, AOA=0° and Moment=1000 Nm) ................................... 68
Figure 4.25 Shroud Trajectories for Different Angle of Attacks ............................... 69
Figure 4.26 Z-Force Acting on the Shroud with Respect to Time for Different Angle
of Attacks ................................................................................................................... 70
Figure 4.27 X-Force Acting on the Shroud with Respect to Time for Different Angle
of Attacks ................................................................................................................... 70
Figure 4.28 Perspective Views of Mach Number Contour for Shroud Separation
(Altitude=5000 m, M=2.0, AOA=20° and Moment=1000 Nm) ................................ 72
Figure 4.29 Perspective Views of Static Pressure Contour for Shroud Separation
(Altitude=5000 m, M=2.0, AOA=20° and Moment=1000 Nm) ................................ 73
Figure 4.30 Shroud Trajectories for Different Ejection Moments ............................. 74
Figure 4.31 Shroud Pitch Angle Change with Respect to Time for Different Ejection
Moments..................................................................................................................... 75
Figure 4.32 Z Force Acting on the Shroud with Respect to Time for Different
Ejection Moments ...................................................................................................... 76

[vL
Figure 4.33 Top Views of Mach Number Contour for Shroud Separation
(Altitude=5000 m, M=2.0, AOA=0° and Moment=1500 Nm) .................................. 77
Figure 4.34 Top Views of Static Pressure Contour for Shroud Separation
(Altitude=5000 m, M=2.0, AOA=0° and Moment=1500 Nm) .................................. 78
Figure 4.35 Shroud Trajectories for Different Side-Slip Angles ............................... 79
Figure 4.36 Z-Force Acting on the Shroud with Respect to Time for Different Side-
Slip Angle ................................................................................................................... 80
Figure 4.37 X-Force Acting on the Shroud with Respect to Time for Different Side-
Slip Angle ................................................................................................................... 81
Figure 4.38 Top Views of Mach Number Contour for Shroud Separation
(Altitude=5000 m, M=2.0, Side-Slip Angle=5° and Moment=1000 Nm) ................. 82
Figure 4.39 Top Views of Static Pressure Contour for Shroud Separation
(Altitude=5000 m, M=2.0, Side-Slip Angle=5° and Moment=1000 Nm) ................. 83
Figure 4.40 X-Force Acting on the Missile Part During Separation with Respect to
Time ........................................................................................................................... 84
Figure 4.41 X-Force Acting on the Missile with Respect to Time ............................ 85

[viL
 LIST OF TABLES

TABLES

Table 3.1 Test Conditions .......................................................................................... 30

Table 3.2 Reynolds Number With Respect to Mach Number [4] .............................. 31
Table 3.3 Number of Elements for Different CFD Grids .......................................... 33
Table 3.4 Body Axial Force Coefficient Error for Different Turbulence Models
(M=2.0, α=0°) ............................................................................................................ 36
Table 3.5 Store and Ejector Information .................................................................... 39
Table 3.6 Flight Condition for Store Separation ........................................................ 42
Table 4.1 Material Properties For Shroud Model ...................................................... 51
Table 4.2 Flight conditions and Ejection Moment for Turbulence Model Selection. 52
Table 4.3 Flight conditions and Ejection Moment for Effects of Altitude ................ 54
Table 4.4 Summary of the Shroud Separation for Effect of Altitude ........................ 57
Table 4.5 Flight Conditions and Ejection Moment for Effect of Mach number ........ 63
Table 4.6 Summary of the Shroud Separation for Effect of Mach number ............... 66
Table 4.7 Flight Conditions and Ejection Moment for Effect of Angle of Attack .... 68
Table 4.8 Summary of the Shroud Separation for Effect of Angle of Attack ............ 71
Table 4.9 Flight Condition and Ejection Moments for Effect of Ejection Moment .. 74
Table 4.10 Summary of the Shroud Separation for Effect of Ejection Moment ....... 76
Table 4.11 Flight Conditions and Ejection Moment for Effects of Side-Slip Angle . 78
Table 4.12 Summary of the Shroud Separation for Effect of Side-Slip Angle .......... 81
Table 4.13 Flight Conditions and Ejection Moment for Effects of Aerodynamic
Loads Acting on the Missile During Separation ........................................................ 84

[viiL
 CHAPTER 1

INTRODUCTION

Missiles have higher speed, altitude and maneuvering capability compared to the
airplanes. These increased performance parameters may introduce new aerodynamic
problems which are aerodynamic heating, higher dynamic pressures, higher
maneuvering accelerations, operating in the non-linear region of high angle of
attacks, etc. Due to the requirements of different mission, various types of missiles
are designed to compensate arising aerodynamic related problems. Especially air-to-
air missiles are designed to operate at high Mach numbers. At these flow regimes,
the missile’s nose is exposed to significant aerodynamic heating, and it can cause
ablation phenomena and damage the electronics inside the body. Therefore, using an
additional front cover to protect the missile’s nose and its separation during the flight
are important topics particularly for air-to-air and /surface-to-air (due to high flight
velocity) type missiles. Missile classification is given briefly in the following part.

1.1 Classification of Missiles

Type of guidance system, propulsion system, launching and impact, trajectory and,
trim and control device can be used for classifying the missiles. The points of
launching and impact classify the missiles in five groups:

AAM Air-to-air missile

ASM Air-to-surface missile
SAM Surface-to-air missile
SSM Surface-to-surface missile
UUM Underwater-to-underwater missile

1
1.1.1 Air-to-Air Missiles (AAM)
Air-to-Air Missiles (AAM) are fired from an aircraft to another one. Small payload
carrying capacity, relatively short range and high maneuverability requirement are
the general characteristics of the AAM. Long and thin cylinders structure reduces the
cross sectional area of the AAM, so drag force can be decreased. The flight time is
relatively short because the missile can reach high supersonic speeds.

1.1.2 Air-to-Surface Missiles (ASM)

Air-to-Surface Missiles (ASM) are mounted on bombers, attack aircraft, fighter
aircraft etc. for destroying land or sea targets which are near or hundreds kilometers
away. Maneuverability requirement is low. ASMs operate at subsonic or transonic
speeds depending on the requirements.

1.1.3 Surface-to-Air Missiles (SAM)

Surface-to-Air Missiles (SAM) are used for destroying aircraft, helicopters or, other
missiles from the ground. It can be used in both short and long ranges. Ramjet or
other air breathing power plants are used in long range missiles to reduce the weight.
SAMs are suitable for air defense against the high speed threats because of their
maneuverability and high speed requirements.

1.1.4 Surface-to-Surface Missiles (SSM)

Surface-to-Surface Missiles (SSM) are fired from a sea or land based platform to
attack another sea or land units. For example ballistic missiles and artillery batteries
are called as SSM. Hand-held, vehicle mounted systems or fixed installations options
are available for firing the SSMs. Maneuverability requirement is relatively weak.
While they are effective in long range as offensive purpose, at short range they are
used for supporting the ground units. The fin and/or wing are usually used lifting and
providing stability.

1.1.5 Underwater-to-Underwater Missiles (UUM)

Underwater-to-Underwater Missiles (UUM) are fired from one submerged
submarine, they travel at high speeds to the area of the distant submarine, and then

2
delivers an antisubmarine payload usually a torpedo. It creates an underwater bubble
in which it could travel under rocket power, at speeds greatly faster than a torpedo.

1.2 Missile with Shroud

Many flight vehicles and in particular air-to-air missiles are designed to operate at
supersonic velocities. At these flow regimes, the missile’s nose is exposed to
significant aerodynamic heating, and it is common to use an additional front cover to
protect the payload from these hostile conditions. Moreover, a typical seeker
configuration requires a blunt nose if expensive transparent materials required for the
seeker lens are avoided. The wave drag on the nose makes a significant effect on the
range of missile (for a given thrust and fuel weight), the nose design of medium to
long range air-to-air missiles is traditionally optimized for low drag [1].

To further emphasize this point; until the missile seeker is locked on target,
removable shrouds can be used to cover the sensitive parts of the seeker system.
Avoiding aerodynamic drag force and protecting of the seeker against aerodynamic
heating effects at midcourse phase are the essential advantages of the shroud concept
[2].

1.3 Aim of Thesis

Accurate prediction of missile shroud separation trajectory is important for shroud
design since an unsuccessful separation may damage the missile and lead to a
catastrophic failure. Originally these trajectory predictions are made by using wind
tunnel and flight tests. However, high cost and complexity seen in free flight and
wind tunnel testing lead the missile designers to find faster and reliable methods. For
this aim, with the increasing computer technology, researches have been
concentrated on numerical methods of multi body dynamic simulations where the
motivation of the thesis comes from. In order to solve multi body dynamic problem
CFD++ solver is used. Computational analysis is performed in reasonable time by
using parallel computer clusters. The applied engineering approach was validated
using EGLIN Store Separation test case [3] and HB test case [4]. The results were
compared with reference studies. After validating the method, shroud separation
analyses for a generic missile configuration was performed.
3
The aim the thesis is to investigate the effects of different parameters such as
altitude, Mach number, angle of attack, side-slip angle and ejection moment on the
safe separation of shroud from a generic missile.

1.4 Literature Survey

Various missiles in service and next-generation missile systems use the advantages
of modern infra-red seekers. One drawback of IR-systems is their limitation due to
the aerodynamic heating of missiles at high supersonic speed. Constant heat transfer
and high recovery temperatures can be dangerous for the seeker. According to
Oswald et al. [2] this issue can be handled, when the missile seeker is protected with
shrouds. The shrouds cover the seeker while providing lower aerodynamic drag. The
shroud leaves from the missiles at the homing phase. Defining the release vector to
eject the shrouds is the significant problem for engineer to make safer the shroud
separation process. For this purpose, they performed three dimensional CFD
simulations under free-flight flow conditions. Predefined force vector is applied on
the center of gravity of the shroud before the releasing time. Separating of the shroud
from the missile comes after the releasing, so CFD grid has to be adapted because of
the free motion of the shroud.

A shroud is used to protect the payload on the ground and during launch from
external environments. The original Polaris shroud, used for STARS I flights, uses a
single circumferential base separation system, which requires complex, energy
absorbing maneuvers to accomplish separation. Whalley [5] developed a bi-sector
shroud in which the two halves are effected perpendicular to the flight path in order
to increase STARS II launch vehicle capability. The shroud consists of a base ring
machined from aluminum alloy, two bi-sectors that use Kevlar/epoxy and
graphite/epoxy laminate construction, and a Teflon nose cap. The two bi-sectors are
joined longitudinally by two thrusting separation rails and the external surface of the
shroud is coated with an ablative heat shield to maintain the structure at an
acceptable temperature.

4
Hypersonic interceptors require a removable shroud to protect the vital seeker
components from the extreme heating loads of the fly-out environment and to
provide a more aerodynamic shape for the vehicle in the lower atmosphere. Panetta
et al. [6] designed self-contained shroud jettison devices in order to remove a shroud
prior to the commencement of end-game activities by applying forces to the shroud
in a passive or active manner. These forces cause the shroud petals to commence
separation and move away from the vehicle. They have demonstrated a number of
jettison techniques. Each design is dependent on the aerodynamic forces acting on
the shroud and the physical constraints of the vehicle. For example,
pneumatic/hydraulic systems are generally reserved for exo-atmospheric
deployments. Pyrotechnics are used to separate a portion of the shroud (nose-tip,
hinge, split-line, etc.), and either contain the explosion in an inflatable bladder or
utilize ram air to separate the shroud and relevant figure can be seen in Figure 1.1.

Figure 1.1 Inflatable Bladder for Shroud Separation [6]

Also according to Anandhanarayanan [7] shroud panels are attached to the vehicle
through a hinge at the end of panels and they are opened using pyro mechanism to a
small angle against the external aerodynamic forces. As the flow rushes through, the
panels open up further due to the aerodynamic load. The panels should be detached

5
at an optimum angle for safe separation from the launch vehicle and relevant figure
can be seen in Figure 1.2.

Figure 1.2 Launch Vehicle with Nose Panels [7]

It is necessary to estimate the minimum opening angle required to initiate the
opening and the maximum angle at which the panels can be separated safely without
hitting the launch vehicle. It is very difficult and expensive to simulate such
separation studies through experiments. Therefore, he has carried out a CFD study in
order to estimate the aerodynamic loads and moments acting on the nose panels at
different opening angles.

The HEDI shroud design employs four petals, hinged at their base, and a two-piece
nose cap. An ordnance device provides the energy to separate the two halves of the
nose cap. As soon as the moment about the hinge due to the internal loads exceeds
the moment about the hinge due to external loads, the petals separate from the
interceptor. According to Lumb [8] shroud configurations tested included three and
four petals design, 15 and 30 degree hinge release angles. Using the results the
shroud design had become four petals with a split nose cap. The petals were oriented
that the seam was above the sensor window. The hinges released after 30 degree of
rotation in order to ensure safe separation.

The shroud system for the HEDI vehicle consisted of an aerodynamic shroud which
covered the HEDI forebody. According to Resch et al. [9] the shroud was designed
to impart no asymmetric loads to the vehicle during the separation process. The
shroud petals were mounted to the forebody via a hinge ring assembly. The nose cap
contained an explosive thruster which separated the nose cap into two sections and

6
radially propelled each section away from the vehicle. The exposed cavity between
the shroud petals and the forebody was then pressurized due to the dynamic pressures
of the flow. The pressure differential across the petals caused the petals to pivot
about the hinge ring and away from the forebody. After pivoting approximately 15
degrees, the four petals were released from the hinge ring and moved radially away
from the forebody.

According to Cavallo and Dash [10] CFD predictions were performed to support test
of the shroud deployment at hypersonic velocities in wind tunnel. While overset
mesh Chimera methods have been the traditional approach for simulating such
separation problems; they suggest that unstructured methods are more readily able to
treat complex geometries and possess the additional advantage of being able to adapt
to evolving flow features. Missile and separated shroud covers can be seen in Figure
1.3 that was used for CFD analysis.

Figure 1.3 Missile with Shroud Covers [10]

7
8
 CHAPTER 2

METHODOLOGY

Methodology used in the calculations of the numerical simulation is explained in this

chapter. Firstly, governing equations of the fluid flow is given. After that, numerical
tools and simulation procedures are discussed.

2.1 Governing Equations

In this study, compressible and steady/unsteady form of Reynolds-Averaged Navier-
Stokes (RANS) equations are used. Several turbulence models are also examined.

2.1.1 Fluid Dynamics

In this problem, the solution domain consists of three domains; two inners and one
outer. The inner domains are the moving domains and the outer domain is the
stationary domain. In the non-inertial references frame, the integral forms of the
equations are defined for all domains. These governing equations are given below
[11].
Continuity:

∫ ∮ (⃗ ⃗ ) (2.1)

Here,

⃗ ⃗ (2.2)

⃗ ⃗ (2.3)

9
⃗ is the fluid velocity, ⃗ is the moving volume velocity, Ω and S are the surface
enclosed to this finite volume.

Conservation of momentum:

∫ ⃗ ∮( ⃗ ⃗ ̿ ⃗ ⃗ ) ∮ (2.4)

Here, ⃗ ⃗ is the tensor product, P is the pressure.

Conservation of energy:

∫ ∮( ⃗ ⃗ ⃗ ) ∮( ⃗ ̇) (2.5)

In Eq. (2.5) E is the total energy, ̇ is the heat transfer rate is the viscous shear
stress tensor defined as:

[( ) (⃗ ⃗) ] (2.6)

Integral Compact Form of Governing Equations

The suitable form of governing equations for finite volume method is given below.

∫ ⃗ ∮ ∮⃗ (2.7)

Where, ⃗ is the conservation variables, is the convective flux and ⃗ is the viscous
diffusive flux vectors. These vectors are as follows:

10
 ⃗

[ ]
(⃗ ⃗ )
(2.8)
[( ⃗ ⃗ ̿ ⃗ ⃗ )]
( (⃗ ⃗ ) ⃗)

⃗ [ ]
( ⃗ ̇)

The convective flux vector can be defined as follows:

⃗ (2.9)

Where;

( )
( )
( )
( )
[( )( ) ]
( )
( )
( ) (2.10)
( )
[( )( ) ]
( )
( )
( )
( )
[( )( ) ]

The diffusive flux vector ⃗ can be defined as follows:

11
 ⃗ ( ⃗) (2.11)

Here;

[ ̇ ]

(2.12)

[ ̇ ]

[ ̇ ]

The viscous shear stress tensors are given below.

( ) ( )( )

( ) ( )( )

( ) ( )( )
(2.13)
( )( )

( )( )

( )( )

The heat conductions are given below.

12
 ̇ ( )
( )

̇ ( ) (2.14)
( )

̇ ( )
( )

Finally, the pressure is defined as perfect gas.

( )[ ( )] (2.15)

Free stream temperature( ), density( ), speed of sound( ), viscosity ( ) and

reference length( ) are used non-dimensionalized of the equations.

2.2 Numerical Tools and Numerical Simulation Methodology

In this part flow solvers, chimera grid technique and Six-Degree of Freedom (6DOF)
methodology used in this study will be covered.

2.2.1 Computational Fluid Dynamics Solver

When the governing equations for fluid motion are examined, it is obvious that these
equations are highly nonlinear and for very limited problems exact solutions can be
acquired [12].

In this study CFD++, a commercial Navier-Stokes solver, was used for the
calculations in three dimensional, density-based and compressible flow conditions
with unstructured-mesh. CFD++ is based on unified grid, unified physics and unified
computing methodology in an advanced numerical discretization and solution
framework [13]. The unified grid framework unifies the treatment of different cell
shapes and grid topologies so that the flow solver becomes equally applicable to all
cell topologies. The unified physics treatment provides the flexibility to solve many
combinations of governing equation sets from incompressible flow to hypersonic

13
flow, with additional equation sets for the modeling reactions, turbulence acoustics.
The unified computing capability allows total portability between different
computing platforms, including multi-CPU machines [13].

Commercial tools which are GAMBIT and TGRID are used to generate solid models
and solution domains. These grid generation programs have different capabilities.
According to the capabilities, GAMBIT is used for generation of the surface grids
and volume grids also boundary layer grids are created with TGRID. Surface grids of
the models are comprised of triangular elements. Boundary layer grids are type of
wedge elements and volume grids are type of tetrahedral elements. Grid generation
procedure is illustrated in Figure 2.1.

Figure 2.1 Grid Generation Procedure

2.2.2 Turbulence Modeling

Turbulence, by virtue of the range of manifested scales, remains one of the greatest
challenges in computational fluid dynamics. In most flows at even modest Reynolds
numbers, non-linear inertial effects in the Navier-Stokes equations tend to
overwhelm the damping effects of viscosity. This causes a growth of (apparently)
random disturbances, leading to a chaotic flow pattern which is usually described
only in terms of its statistics, since the finest scales of motion are typically much too
small to be directly resolved. The foundation for the subject of modern turbulence
modeling is the contention that this fine-scale turbulence obeys certain universal
rules, the effects of which can, to a large extent, be reproduced by mathematical
models [13].

14
CFD++ provides various different options of turbulent models. In this thesis only
RANS based turbulence models are considered in turbulence modeling since they are
more proper for practical industrial applications.

2.2.2.1 One-equation Spalart-Allmaras Model

The Reynolds stresses are given as

̅̅̅̅̅̅ (2.18)

With

( ) (2.19)

The eddy viscosity is formulated as follows:

(2.20)
̃

Transport equation for ̃

̃
( )̃̃ * ,( ̃) ̃- ( ̃) +
(2.21)
̃
( )( )

Where the rotation function accounting for rotation and curvature effects is

( ) , ( ̃ )- (2.22)

15
Where

̂ ̂
̃ * ( ̂ ̂ ) +

̂ ( )
(2.23)
( )

̂ ̂ ̂

(̂ )

It should be noted that in CFD++, the Lagrangian derivative of the strain rate tensor,
̂
, does not include the time derivatives, nevertheless changes of velocity

derivatives with time will be present, albeit indirectly.

The modified strain rate is

̃
( )
(2.24)

The sink term involves

⁄
( )
(2.25)
, ( )-

16
 ̃
̃( )

To allow for laminar regions, the sink term includes

( ) (2.26)

In the above

Boundary conditions: ̃
Model constants:
⁄ ⁄ ( )

2.2.2.2 Two-equation Realizable k-ε Model

In CFD++’s realizable k-ε model, the Boussinesq relation is used to obtain
Reynolds-stresses (algebraically) from the modeled eddy viscosity ( ) and the
available mean-strain tensor:

̅̅̅̅̅ (2.27)

Where

17
 ( ) (2.28)

The model consist of the following transport equations for k and ε:

( )
( ) *( ) +

( )
( ) (2.29)

*( ) + ( )

In which the rate of production of turbulence energy,

̅̅̅̅̅ (2.30)

Realizable estimate of the turbulence timescale,

* +
(2.31)
√

And the turbulence Reynolds number,

(2.32)

The additional term, E, in the dissipation-rate equation is designed to improve the

model respond to adverse pressure-gradient flows. This term has the form:

√ { ( ) } (2.33)

18
 , -

The model constants are given by:

, , , , ,

The eddy viscosity is obtained from:

, - (2.34)

Where √ is the dimensional strain magnitude and is a low-Reynolds

number function, designed to account for viscous and inviscid damping of turbulent
fluctuations in the proximity of solid surfaces:

{ ( ) } (2.35)
√

2.2.2.3 Two-equation k-ω SST Model

The Reynolds stresses are given by

̅̅̅̅̅ (2.36)

In which

19
 ( ) (2.37)

In the above is the mean strain and eddy viscosity is given as

(2.38)
* +

The ingredients of this formula are as follows:

Turbulence kinetic energy transport equation

( )
̃ ,( ) - (2.39)

Turbulence inverse time-scale transport

( )
,( ) -
̂
(2.40)
( )

First blending function

√
{, * ( ) +- } (2.41)

Where

( ) (2.42)

Second blending function:

20
 √
{* ( )+ } (2.43)

Turbulence generation term

* ( ) + (2.44)

Limited generation used in equation:

̃ ( ) (2.45)

Blending of constants:

( ) (2.46)

where stands for the various model constants.

In the above
, , ,
⁄ ⁄√ ,

, ⁄√
, ,
Smooth wall boundary conditions with

(2.47)
( )

Where is the distance to the first centroid away from the wall. Also, d is the
distance to the nearest wall; S is the invariant measure of strain rate, ⁄
The following limited eddy viscosity is used in the equation:

21
 ̂ ( ) (2.48)

2.2.3 Chimera Grid Methodology

Chimera overset grid scheme is a domain decomposition approach where a full
configuration is meshed using a collection of independent overset grids. This allows
each component of the configuration to be gridded separately and overset into a main
grid [15]. In Chimera, overset grids can move relative to each other without
disturbing the grids in other zones. Since different zones are not required to align
with each other, zonal grids can then be produced completely independently [16].
In this study, the Chimera methodology is used for allowing the moving body
calculations. The computational domain of the body itself is constructed using
patched grid. The methodology behind Chimera approach is described in the
following sections.
In the following sections, process of Chimera methodology, hole-cutting process and
intergrid communication are discussed.

2.2.3.1 Hole Cutting

In the Chimera grid system, intergrid communication is accomplished through data
interpolation. Therefore a hole needs to be cut in each grid in the regions that overlap
with solid bodies or any other non-flow regions which belong to the other grid of the
overset grid system. Afterwards interpolation stencil have to be formed for cells
which lie along the hole boundaries and also for other interpolation boundaries in
each grid. In the present method, in order to allow flexible grid interactions, the
overset grids are allowed to overlap each other in an arbitrary manner [17]. This
procedure is a complex procedure and several steps are required and outlined as
follows with help of an illustrative case shown in Figure 2.2.

22
 Figure 2.2 Illustration of a Chimera Grid System [18]
1. Determine the bounding boxes of all component grids and wall boundary
(WB) faces;
2. If the bounding box of a component grid, let’s call it A, overlaps with that of
a WB face in another component grid, let’s call it B, then prepare an
Alternating Digital Tree (ADT) so that it will provide the coordinates of A’s
cells which are located on the bounding box of the overlapped WB face; as
repeating this step for each face in the WB face list in grid B and identify all
such cells.
3. Find out if one or more edges of the cells (identified step 2) intersect with any
WB face in grid B; then decide the intersection points. The cell that contain
such intersecting-edge(s) are named as cut-cells;
4. Review each edge of the cut-cells and classify the status of its two end points
are either IN (if this points inside WB) or OUT (if this point is outside WB).
5. Once the status of the end points of all edges that intersect the WB faces is
checked, as shown in Figure 2.3, propagate the status checking to other edges
of the cut-cells and then to the adjacent cells until all the grid points that are
IN are identified, as shown Figure 2.4.

23
 Figure 2.3 Assignment of IN/OUT Status to End Points [18]

Figure 2.4 Cut-Cells of the Grid Blanketing all Grid Points [18]
6. Blank out the cells that contain IN points, as shown in Figure 2.5, therefore
the hole is generated. The limiting boundary of the hole is established by the
cut-cells. The cells which lie outside the WB, adjacent to the hole are
identified as fringe boundary cells, as shown in Figure 2.6.

24
 Figure 2.5 A Hole Cut in the Grid With Respect to the Given WB [18]

Figure 2.6 Fringe Boundary Cells Adjacent to the Cut Hole [18]
7. Repeat the steps 2-6 until all the component grids in the Chimera grid system
are processed.

2.2.3.2 Intergrid Communication

In the Chimera grid method it is required to transfer information between the
overlapping grids by interpolation. For each computational node that is a component
grid, we need to determine if there exist some grid elements in different component

25
grids from which the flow variables should be interpolated. In the Chimera grid
system, each component grid may comprise interpolation boundary cells and/or
fringe boundary cells. They represent the linkage between the component grids in the
Chimera grid system. All of these are named intergrid boundary cells (IGBCs).
Intergrid communication happens by interpolating data to IGBCs in each grid from
the respective donors in another grid. A donor is a cell which overlaps with the
interpolation boundary cells or fringe boundary cells and provides and interpolation
stencil [17]. The intergrid communication is guided as follows:
1. Search all the other grids and identify the cells whose bounding boxes
overlap with that of the interpolation boundary or fringe boundary.
2. Using these candidates, determine the cell that overlaps with the centroid of
the interpolation boundary cell or fringe boundary cell. This cell is a donor. If
more than one valid donor is found, then choose the donor with the smallest
control volume.
3. Once donor cells are identified, values of the cell at the interpolation or fringe
boundaries are determined from its donor cells as well as their neighboring
cells by a weighted average method either inverse-distance or least-squares.

2.2.4 6DOF Methodology

One of the capabilities of CFD++ solver is moving body analysis. CFD++ is used for
unsteady and time accurate problems such as shroud separation. This capability
requires the rigidity of the body and the grids of the moving body. The motion of the
body can be determined by a 6DOF calculation or it can be prescribed by the user.
6DOF is based on the fluid flow solution in CFD++. In order to determine forces and
moments acting on the body, pressure and shear stresses are used. To calculate
translational and rotational displacements of the body, forces and moments which are
acting on the body are used in the general equations of motion.
The equations of motion for a rigid body with constant mass and mass moments of
inertia are solved in order to obtain the linear and angular velocities and the
displacements of the body in a delta time step size. These equations are given below.

(2.49)

26
 ⃗
⃗⃗ ⃗ ⃗ (2.50)

Where m is the body mass, is the linear velocity of the center of gravity, ⃗ is the
angular momentum, ⃗ is angular velocity about the body’s center of gravity. The
force Eq. 2.49 is in the inertial frame of reference. The momentum Eq. 2.50 is in the
body fixed frame of reference. The moments of inertia are completely based on the
body rotating about an axis passing through its center of gravity. The main steps that
include these computations are presented in Figure 2.7. This figure also summarizes
the steps of the flow solver with the 6DOF calculations.

Figure 2.7 Steps of the Flow Solver with 6DOF Calculations

27
28
 CHAPTER 3

VALIDATION OF TEST CASES

In this chapter, HB-1 test case is used to validate an engineering approach which is
used for aerodynamic coefficient calculation [4]. Furthermore, EGLIN test case is
used to validate an engineering approach which is used for the multi body CFD
analysis to simulate store separation at transonic and supersonic speeds [3]. Details
of these test cases are given in the following sections.

3.1 HB-1 Supersonic and Hypersonic Test Case

HB-1 Test Case model is hypersonic ballistic correlation model which is used for
data evaluation and calibration in test facilities. The main target of this test case is
used to examine the capabilities of CFD analysis tool by comparing results with the
wind tunnel test data.

The HB-1 test case model has conical nose with cylindrical body. Basic dimensions
of the configurations with respect to body diameter are presented in Figure 3.1.

Figure 3.1 Geometry of the HB-1 Test Case

29
The test model was used in wind tunnel experiments which was conducted at the
Arnold Engineering Development Center in Aerodynamic Wind Tunnel in 1964 [4].
The test conditions, which are used for comparisons are given in the Table 3.1.

Table 3.1 Test Conditions

Mach Number 1.5-5

Angle of Attack -2°<α<12°

3.1.1 Numerical Simulation

The corresponding numerical simulation of the experiment is performed for a steady-
state, compressible and turbulent flow using commercial software which is CFD++.

3.1.2 Solid Model

Solid model and computational domain are created using GAMBIT commercial
software. The generated solid model and computational grid are shown in Figure 3.2.

Figure 3.2 Solid Model and Computational Grid for HB-1

3.1.3 Boundary Conditions

The computational domain is large enough to minimize flow effects between model
and boundaries. Outer boundaries of the computational domain are set as far-field,
with sea level temperature and pressure free stream conditions are calculated with
respect to Reynolds number. Table 3.2 shows Reynolds number with respect to Mach
30
number. Solid surfaces were modeled as a no-slip, adiabatic wall boundary
condition. Solution domain and defined boundary conditions are shown in the Figure
3.3.

Table 3.2 Reynolds Number With Respect to Mach Number [4]

Mach number Reynolds number

1.5 990000
2 820000
3 1210000
4 600000
5 440000

Figure 3.3 Solution Domain and Defined Boundary Condition for HB-1 Test Case

3.1.4 CFD Grid Independence Study

Grid sensitivity analyses conduct to get results independent of grid size before using
the results of the numerical simulation. Three different fluid domain grids are
examined to decide the minimum grid size with acceptable accuracy for this test
case.

31
In order to capture the geometry better, unstructured grids are preferred in this study.
Unstructured hybrid grids are generated by the GAMBIT. Boundary layer grid is
created by TGRID. Unstructured triangular elements are used for surface grids and
30 prismatic elements are created to resolve boundary layer flow. First point of the
surface is chosen to give y+ value of about 1.0 for boundary layer grid modeling.
Tetrahedral elements are used for the volume grids and the mesh growth rate is
preserved below 1.15.

Three different grids are used for the grid independence study. Generated coarse,
medium and fine, surface and volume grids are shown in Figure 3.4.

Figure 3.4 Coarse, Medium and Fine Grid

32
Element numbers of coarse, medium and fine grids are given in Table 3.3.

Table 3.3 Number of Elements for Different CFD Grids

Number of Elements on Total Number of

Grid
the Surface Elements
Coarse 8520 671928

Medium 25850 1496934

Fine 47692 2399480

For grid independence study, realizable k- turbulence model is used. Turbulence
model study is carried out in the following section. CFD analyses are performed for
several Mach numbers at angle of attack of 8°. CFD analysis results of grid
independence study are shown in Figure 3.5, Figure 3.6 and Figure 3.7.

Figure 3.5 HB-1 Axial Force Coefficient with Respect to Mach for Different Grids

33
Figure 3.6 HB-1 Normal Force Coefficient with Respect to Mach for Different Grids

Figure 3.7 HB-1 Pitch Moment Coefficient with Respect to Mach for Different Grids

Fine and medium grids give similar results for the aerodynamic coefficient which are
CAc, CN, and Cm. However coarse gird gives different results. In order to save both
time and computational power, medium gird is chosen for the rest of the thesis work.

34
3.1.5 Turbulence Model Selection
Results obtained using three different turbulence models; Spalart-Allmaras,
Realizable k- and k- SST, are compared with experimental data in order to show
the effect of turbulence model on the aerodynamic coefficients. The results for
normal force and pitching moment coefficients are presented in Figure 3.8 and
Figure 3.9 respectively. For body axial force coefficient, error values are calculated
with respect to experiment value for each turbulence model considered in analyses
and shown in Table 3.4.

Figure 3.8 Normal Force Coefficient with Respect to Angle of Attack for Different
Turbulence Models (M=2.0)

35
 Figure 3.9 Pitch Moment Coefficient with Respect to Angle of Attack for Different
Turbulence Models (M=2.0)

Table 3.4 Body Axial Force Coefficient Error for Different Turbulence Models
(M=2.0, α=0°)

Error Percentage (%)

Coefficient Spalart-Allmaras Realizable k-ε k-w SST

CAc 6.7 6.8 7.5

According to CFD results, these three turbulence models give similar results for the
aerodynamic coefficients with the experimental data. However, for the CAc
coefficient k-ω SST turbulence model gives poor result when compared to other
turbulence models. It can be said that Spalart-Allmaras and Realizable k-ε give
satisfactory results.

36
3.1.6 Solution Domain Flow Visualization
Mach number and static pressure contour for HB-1 Test Case are given in Figure
3.10.

Figure 3.10 Mach number and Static Pressure Contour for HB-1 Test Case (M=2.0,
α=12°)

3.2 EGLIN Store Separation Test Case

The test case model has three parts; wing, pylon and finned body. The wing
consisted of a clipped delta wing with 45° sweep and a constant NACA 64a010
airfoil section. The pylon has an ogive-flat plate-ogive cross section shape. The store
body was an ogive-cylinder-ogive with an aft cylindrical sting. The fins on the store
consisted of a clipped delta wing with a 45° sweep and a constant NACA 0008 airfoil
section. Gap between pylon and finned body is 0.07 in. The wind tunnel model was
37
assumed to be 1:20 scale. The geometric specifications of body are presented in
Figure 3.11. All dimensions are given in inches [19].

Figure 3.11 Geometry of the EGLIN Test Case

This test model is used in wind tunnel experiments conducted at the Arnold
Engineering Development Center in Aerodynamic Wind Tunnel in 1990. Sketch of
the EGLIN test case model and sting used in the wind tunnel test are represented in
Figure 3.12 [19].

Figure 3.12 Sketch of the EGLIN Test Case Wind Tunnel Test
Experiment conducted in transonic regime (M=0.95) and supersonic regime (M=1.2).
During experiments position and angle change of store is obtained for both flow
regimes.
In the test, aft and forward ejector forces are applied to store to provide safe
separation. The ejector characteristics and other full-scale store parameters are
presented in Table 3.5 [3].

38
 Table 3.5 Store and Ejector Information

Weight 907 kg

Center of Gravity 1417mm (aft of store nose)

Roll moment of inertia 27 kg.m2

Pitch moment of inertia 488 kg.m2

Yaw moment of inertia 488 kg.m2

Forward ejector location 1237.5mm (aft of store nose)

Aft ejector location 1746.5mm (aft of store nose)

Forward ejector force 10.7kN

Aft ejector force 42.7kN

3.2.1 Numerical Simulation

The related numerical simulation of the experiment is performed for a transient,
compressible and turbulent flow using CFD++ with chimera grid methodology.

3.2.2 Solid Model

Solid model and computational domain are created using GAMBIT commercial
software. The generated solid models are shown in Figure 3.13. Sting geometry was
not used in CFD modeling because the effect of sting was corrected for wind tunnel
measurement.

Figure 3.13 Solid Model of EGLIN Test Case

39
3.2.3 Boundary Conditions
The computational domain is large enough to minimize flow effects between model
and boundaries. The computational domain inlet was located 17 model wing length
upstream from center of body, outlet was located 25 wing length downstream from
the center of body, above boundary was located 17 model length from the center of
the body, below boundary was located 25 model length from the center of the body
and side boundary was located 17 model length from the center of the body.
Downstream, upstream, and all side boundaries except right side were set as pressure
far-field, using standard atmosphere model at 26000 ft altitude for temperature and
pressure free stream conditions. Right side boundary was defined as symmetric.
Solid surfaces were modeled as a no-slip, adiabatic wall boundary condition.
Solution domain and defined boundary conditions are shown in the Figure 3.14.
Numerical simulation of the experiment model is conducted at Mach number of 1.2.

Figure 3.14 Solution Domain and Boundary Conditions for EGLIN Test Model

40
3.2.4 Turbulence Model Selection
CFD analyses are performed for three different turbulence models to investigate
effect of turbulence model for store separation on EGLIN Test Model. CFD results
(distance and angle) are compared with experimental data and given in Figure 3.15
and Figure 3.16.

Figure 3.15 Store Position Change Results and Experimental Data with respect to
Time for Different Turbulence Models [3]

41
Figure 3.16 Store Angle Change Results and Experimental Data with respect to Time
for Different Turbulence Models [3]

As presented in figures, solutions obtained by k-ε turbulence model are more

accurate than solutions of the other two models compared to the experimental data.

3.2.5 CFD Simulation Results

Store separation is simulated for EGLIN test case. Transient CFD analysis is
performed for given flight condition in Table 3.6.

Table 3.6 Flight Condition for Store Separation

Static Pressure [Pa] 20646

Temperature [K] 217

Mach Number 1.2

Time dependent position and angle changes are obtained from CFD simulation and
results are compared with experimental data [3]. Results are given in Figure 3.17 and
Figure 3.18.

42
 Figure 3.17 Store Position Change and Experimental Data With Respect to Time

Figure 3.18 Store Angle Change and Experimental Data With Respect to Time
As shown in the Figure 3.17 the store position changes are very similar with the
experimental data. Also in the Figure 3.18 store pitch and yaw angle changes are in
good agreement with experimental data. On the other hand the store rolls
continuously outboard throughout the first 0.8 seconds of the separation. This trend
is under-predicted by the CFD, and the curve tends to diverge from the experiments

43
after approximately 0.3 seconds. The roll angle is especially difficult to model
because the moment of inertia about the roll axis is much smaller than that of the
pitch and yaw axes. Consequently, roll is very sensitive to errors in the aerodynamic
force prediction.

In addition CFD analysis is initialized as steady analysis. After convergence of

steady run, transient CFD analysis is continued on steady CFD analysis data file to
achieve easier convergence. Time step is selected as 0.0001 s and 25 sub-iterations
are made for transient CFD analysis.

Also, visual presentation of CFD and experimental results is given in Figure 3.19,
and pressure distribution on store and wing with respect to time is given in Figure
3.20.

Figure 3.19 Visual Presentation of CFD and Experimental Results

44
Figure 3.20 Pressure Distribution with Respect to Time for Store Separation

45
46
 CHAPTER 4

CFD ANALYSIS OF MISSILE SHROUD SEPARATION

In this part of the thesis, CFD analysis of shroud separation on generic missile is
performed to investigate safe separation of the shroud covers from missile for
different parameters which are altitude, Mach number, angle of attack, side-slip
angle and ejection moment. Selected generic missile geometry has conical nose with
cylindrical body. Basic dimensions of the generic missile with respect to body
diameter are presented in Figure 4.1. Selected shroud geometry has tangent-ogive
nose appropriate for conical generic missile nose. Basic dimensions of the shroud
geometry with respect to body diameter are given in Figure 4.2.

Figure 4.1 Geometry Specification of Generic Missile

47
 Figure 4.2 Geometry Specification of Shroud Model

4.1 Numerical Simulation

The corresponding numerical simulation of the shroud separation is performed for a
transient, compressible and turbulent flow using CFD++ with chimera grid
methodology.

4.2 Solid Model and Grid Generation

Solid model of the generic missile geometry with shroud is created by using
GAMBIT commercial software. The generated solid model is shown in Figure 4.3.

Figure 4.3 Solid Model of the Generic Missile with Shroud (Side View)

48
The surface, boundary layer and domain grid generation of missile with shroud is
similar with HB-1 test case. Surface, domain grid and Chimera grid system are given
in Figure 4.4, Figure 4.5 and Figure 4.6.

Figure 4.4 Surface Grid for Generic Missile with Shroud (Top View and Side View)

Figure 4.5 Surface and Volume Grids for Generic Missile with Shroud

49
 Figure 4.6 System of Chimera Grid

4.3 Boundary Conditions and Material Properties

The solid surfaces were defined as no-slip wall boundary condition. The outer
surfaces of fixed fluid domain were defined as pressure far field boundary condition
with constant freestream static pressure, static temperature and velocity. Solution
domain and defined boundary conditions are shown in the Figure 4.7.

Figure 4.7 Solution Domain Boundary Condition

50
The material of shroud model was chosen as steel. Therefore, material properties and
moment of inertia of shroud models can be found in Table 4.1. The value of moment
of inertia for shroud models were given with respect to both center of gravity (CG)
and hinge point (HP).
Table 4.1 Material Properties For Shroud Model
Shroud Model (Half)
Mass 3.917913 [kg]
Ix [kg.m2] Iy [kg.m2] Iz [kg.m2]
Moment Of Inertia (CG)
0.023513 0.031706 0.046233
Moment Of Inertia (HP) 0.037201 0.109247 0.110086

4.4 Shroud Release Procedure

The shroud release procedure consists of two stages. At the first stage, the shroud is
hinged to the main body and can rotate only about the hinge axis up to 30°. At the
second stage, the shroud is released from the hinge and can move in space 6DOF.
Schematic drawing of the shroud release procedure is presented in Figure 4.8.

Figure 4.8 Schematic drawing of the Shroud Release Procedure

51
In addition, the flow chart of shroud release procedure can be seen in Figure 4.9.

Figure 4.9 The Flow Chart of Shroud Release Procedure

Also, steady CFD analysis was performed for missile with shroud to determine
aerodynamic forces acting on the shroud at flight condition of 5000 m altitude, 2.0
Mach number and 0° angle of attack. Therefore, an ejection moment was assumed to
overcome aerodynamic forces acting on shroud and to initialize the separation.
In addition all ejection moment values applied on shroud covers along 10
milliseconds to initialize the shroud separation.

4.5 Turbulence Model Selection

CFD analyses are performed for three different turbulence models which are Spalart-
Allmaras, Realizable k-ε and k- SST to investigate effect of turbulence model for
missile shroud separation. Flight condition and applied ejection moment is given in
Table 4.2 for this study.

Table 4.2 Flight conditions and Ejection Moment for Turbulence Model Selection

Altitude [m] 5000

Mach Number 2

Angle of Attack [°] 0

Ejection Moment [Nm] 1000

52
CFD results (shroud trajectories) are compared with each other and given in Figure
4.10.

Figure 4.10 Shroud Trajectories for Different Turbulence Models

As presented in figure, all turbulence models give similar results for the missile
shroud separation. This study was performed in supersonic velocity so turbulence
model did not affect shroud trajectories dramatically. On the other hand, selection of
turbulence model is important in transient region and this effect was shown in
previous study which is store separation. Therefore for the rest of the analyses k-ε
turbulence model is used.

53
4.6 CFD Analysis Results
CFD analysis is performed in order to investigate safe separation of the shroud
covers from missile for different parameters which are altitude, Mach number, angle
of attack, side-slip angle, ejection moment and analysis results presented in this part.
In addition CFD analysis is initialized as a steady analysis and after convergence of
steady run transient CFD analysis is continued on steady CFD analysis data file to
make convergence easy. Time step is selected as 0.0001 and 25 sub-iterations are
made for transient CFD analysis. Approximately overall clock time of only one
analysis is about 6 days for 40 CPU.

4.6.1 Effects of Altitude

In this part, effects of altitude are investigated. Shroud separation analyses are
performed at five different altitudes by keeping the Mach number, angle of attack
and ejection moment constant. Flight conditions and applied ejection moment are
given in Table 4.3 for this part.

Table 4.3 Flight conditions and Ejection Moment for Effects of Altitude

Altitude [m] 0 3500 5000 7500 10000

Mach Number 2

Angle of Attack [°] 0

Ejection Moment [Nm] 1000

Shroud trajectories belonging to four different altitudes are given in Figure 4.11.
These trajectories reflect position change of center of gravity of shroud.

54
 Figure 4.11 Shroud Trajectories for Different Altitudes

As it is shown in above figure, shroud separation is performed safely for the

specified flight conditions and ejection moment at four different altitudes which are
3.5 km, 5 km, 7.5 km and 10 km. Also, 3500 m altitude is the minimum altitude, for
which shroud separation can be initialized with 1000 Nm ejection moment. The
shroud does not separate at sea level because the ejection moment is not enough for
the initialization of the separation at this flight condition. Besides, the closest shroud
trajectory is followed at maximum altitude level and the furthest shroud trajectory is
followed at minimum altitude level where shroud can be separated. As the reason for
this, it can be shown that Z-force with respect to time for different altitudes in Figure
4.12.

55
 Figure 4.12 Z-Force Acting on the Shroud with Respect to Time for Different
Altitudes
As the altitude increases, density of the air decreases leading to lower aerodynamic
forces acting on the shroud. In the same manner, it can be seen from Figure 4.12 that
an increased Z-Force which is acting on the Z direction of the shroud covers
determines with decreasing altitude. This phenomenon affects the trajectory of the
shroud covers during separation. When the Z-Force change of the shroud is higher,
force acting on the Z direction that separates the shroud from the missile becomes
larger. Therefore, maximum separation distance for the shroud is obtained when Z-
force change of the shroud is the highest. Besides, decreasing of the altitude leads to
increasing of the X-Force which is acting on the X direction in Figure 4.13.
Therefore, shroud covers reach the same position along the X direction in the shorter
time.

56
 Figure 4.13 X-Force Acting on the Shroud with Respect to Time for Different
Altitudes
These analyses show that under these conditions minimum altitude for the safe
separation of shroud is determined. In addition to that the analysis shows that larger
ejection moment are desired for safe separation of shroud as seen with decreasing
altitude. Also, summary of the shroud separation for effect of altitude can be seen in
Table 4.4.

Table 4.4 Summary of the Shroud Separation for Effect of Altitude

Altitude=0 m Shroud Separation Failed

Altitude=3500 m Safe Separation

Altitude=5000 m Safe Separation

Altitude=7500 m Safe Separation

Altitude=10000 m Safe Separation

57
4.6.1.1 Solution Domain Flow Visualization
Mach number and static pressure contours for shroud separation are given in below
figures.

Figure 4.14 Top Views of Mach Number Contour for Shroud Separation
(Altitude=5000 m, M=2.0, AOA=0° and Moment=1000 Nm)

58
 Figure 4.15 Top Views of Static Pressure Contour for Shroud Separation
(Altitude=5000m, M=2.0, AOA=0° and Moment=1000 Nm)
After the shroud is released, it is strongly influenced by the flow field that is induced
by the missile body. Pressure gradient distributions clearly show the complex
shockwave interactions on the shroud and missile body and it can be seen in Figure
4.15. The influence is considerable at the area close to the missile’s front. This
implies that the influence is mostly due to shock waves created by the missile nose.
In the near body case, sharp increase in pressure is observed at the missile nose area
as opposed to the free-body case and a pressure decrease is observed at the same
region. Consequently, a forward pushing force is acting on the shroud near the
missile body, whereas the flow pressure at free flight produces significant drag.
Also similar shock pattern is observed in all other analyses results.

59
Figure 4.16 Top Views of Mach Number Contour for Shroud Separation
(Altitude=7500 m, M=2.0, AOA=0° and Moment=1000 Nm)

60
Figure 4.17 Top Views of Static Pressure Contour for Shroud Separation
(Altitude=7500m, M=2.0, AOA=0° and Moment=1000 Nm)

61
Figure 4.18 Top Views of Mach Number Contour for Shroud Separation
(Altitude=10000 m, M=2.0, AOA=0° and Moment=1000 Nm)

62
 Figure 4.19 Top Views of Static Pressure Contour for Shroud Separation
(Altitude=10000m, M=2.0, AOA=0° and Moment=1000 Nm)

4.6.2 Effects of Mach Number

In this part, effects of Mach number are investigated for the shroud separation.
Shroud separation analyses are carried out for four different Mach numbers by
keeping the altitude, angle of attack and ejection moment constant. Flight conditions
and applied ejection moment are given in Table 4.5.

Table 4.5 Flight Conditions and Ejection Moment for Effect of Mach number

Altitude [m] 5000

Mach Number 1.2 2 2.2 2.5

Angle of Attack [°] 0

Ejection Moment [Nm] 1000

63
Shroud trajectories belonging to three different Mach number are given in Figure
4.20. These trajectories show position change of center of gravity of shroud.

Figure 4.20 Shroud Trajectories for Different Mach Numbers

As it is shown in Figure 4.20, shroud separation is performed safely for the specified
flight conditions and ejection moment at three different Mach numbers which are
1.2, 2.0 and 2.2. Also, 2.2 Mach number is the maximum Mach number, for which
shroud separation can be initialized with 1000 Nm ejection moment. The shroud does
not separate 2.5 Mach number because the ejection moment is not enough for the
initialization of the separation. Besides, the closest shroud trajectory is followed for
minimum Mach number and the furthest shroud trajectory is followed for maximum
Mach number at which shroud can be separated. As the reason for this, it can be
shown that Z force with respect to time for different Mach numbers in Figure 4.21.

64
Figure 4.21 Z-Force Acting on the Shroud with Respect to Time for Different Mach
Numbers
It is known that, as the Mach number decreases aerodynamic forces acting on the
shroud covers decreases. In other words, it can be seen from Figure 4.21 that an
increased Z force acting on the Z direction of the shroud covers determines with
increasing Mach number. This phenomenon affects the trajectory of the shroud
during separation. When the Z-Force of the shroud is higher, force acting on the Z
direction that separates the shroud from missile becomes larger. Therefore,
maximum separation distance for the shroud is obtained when Z-force of the shroud
is the highest. Besides, increasing of the Mach number leads to increasing of the X-
force acting on the X direction in Figure 4.22. Therefore, shroud covers reach the
same position along the X direction in the shorter time.

65
Figure 4.22 X-Force Acting on the Shroud with Respect to Time for Different Mach
Numbers
These analyses show that under these conditions maximum Mach number for the
safe separation of shroud is determined. In addition to that analyses show that larger
ejection moment is desired for safe separation of shroud as seen with increasing
Mach number. Also, summary of the shroud separation for effect of Mach number
can be seen in Table 4.6.

Table 4.6 Summary of the Shroud Separation for Effect of Mach number

Mach Number=1.2 Safe Separation

Mach Number=2.0 Safe Separation

Mach Number=2.2 Safe Separation

Mach Number=2.5 Shroud Separation Failed

66
4.6.2.1 Solution Domain Flow Visualization
Mach number and static pressure contours for shroud separation are given in below
figures.

Figure 4.23 Top Views of Mach Number Contour for Shroud Separation
(Altitude=5000m, M=1.2, AOA=0° and Moment=1000 Nm)

67
 Figure 4.24 Top Views of Static Pressure Contour for Shroud Separation
(Altitude=5000m, M=1.2, AOA=0° and Moment=1000 Nm)

4.6.3 Effects of Angle of Attack

In this part, effects of angle of attack are investigated for the shroud separation.
Shroud separation analyses are carried out for three different angle of attacks by
keeping the altitude, Mach number and ejection moment constant. Flight conditions
and applied ejection moment are given in Table 4.7 for this part.

Table 4.7 Flight Conditions and Ejection Moment for Effect of Angle of Attack

Altitude [m] 5000

Mach Number 2
Angle of Attack [°] 0 10 20
Ejection Moment [Nm] 1000

68
Shroud trajectories belonging to three different angle of attacks are given in Figure
4.25. These trajectories indicate position change of center of gravity of shroud.

Figure 4.25 Shroud Trajectories for Different Angle of Attacks

As it is shown in above figure, in all angle of attacks shroud separation is performed
safely for the specified flight conditions and ejection moment. Since angle of attack
changes depending on the Y direction, it does not affect the shroud trajectories on Z
direction dramatically. Therefore, shroud trajectories for different angle of attacks
are similar up to 2 missile diameter in Z direction. In addition, as it is seen in Figure
4.26 and Figure 4.27, Z force and X force values with respect to time for different
angle of attack are the similar during the separation.

69
Figure 4.26 Z-Force Acting on the Shroud with Respect to Time for Different Angle
of Attacks

Figure 4.27 X-Force Acting on the Shroud with Respect to Time for Different Angle
of Attacks

70
Also, summary of the shroud separation for effect of angle of attack can be seen in
Table 4.8.

Table 4.8 Summary of the Shroud Separation for Effect of Angle of Attack

Angle of Attack=0° Safe Separation

Angle of Attack=10° Safe Separation

Angle of Attack=20° Safe Separation

71
4.6.3.1 Solution Domain Flow Visualization
Mach number and static pressure contours for shroud separation are given in below
figures.

Figure 4.28 Perspective Views of Mach Number Contour for Shroud Separation
(Altitude=5000 m, M=2.0, AOA=20° and Moment=1000 Nm)

72
 Figure 4.29 Perspective Views of Static Pressure Contour for Shroud Separation
(Altitude=5000 m, M=2.0, AOA=20° and Moment=1000 Nm)

4.6.4 Effects of Ejection Moment

In this part, effects of ejection moment are investigated for the shroud separation.
Shroud separation analyses are performed for four different ejection moment by
keeping the flight conditions constant. Flight conditions and applied ejection moment
are given in Table 4.9 for this part.

73
 Table 4.9 Flight Condition and Ejection Moments for Effect of Ejection Moment

Altitude [m] 5000

Mach Number 2
Angle of Attack [°] 0
Ejection Moment [Nm] 500 800 1000 1500

Shroud trajectories belonging to three different ejection moments are given in Figure
4.30. These trajectories display position change of center of gravity of shroud.

Figure 4.30 Shroud Trajectories for Different Ejection Moments

As it is shown in Figure 4.30, shroud separation is performed safely for the specified
flight conditions and three different ejection moments which are 800 Nm, 1000 Nm,
1500 Nm. Also, 800 Nm ejection moment is the minimum ejection moment, for
74
which shroud separation can be initialized at the respective flight condition. The
shroud does not separate with 500 Nm ejection moment because the ejection moment
is not enough for the initialization of the separation at this flight condition. Besides,
the closest shroud trajectory is followed for maximum ejection moment and the
furthest shroud trajectory is followed for minimum ejection moment which shroud
can be separated. As the reason for this, it can be shown that pitch angle changes
with respect to time for different ejection moments in Figure 4.31.

Figure 4.31 Shroud Pitch Angle Change with Respect to Time for Different Ejection
Moments
As it is shown in above figure, higher pitch rate creates higher rotational inertia on
the shroud covers. Therefore shroud covers which were separated with greater
ejection moment, rotate faster. In this way, aerodynamic forces acting on the shroud
on the Z-direction decreases faster and it can be seen in Figure 4.32. This
phenomenon affects the trajectory of the shroud during separation.

75
 Figure 4.32 Z Force Acting on the Shroud with Respect to Time for Different
Ejection Moments
As a result increasing the ejection moment leads to shroud covers following a close
trajectory. Also, summary of the shroud separation for effect of angle of attack can
be seen in Table 4.10.

Table 4.10 Summary of the Shroud Separation for Effect of Ejection Moment

Ejection Moment=500 Nm Shroud Separation Failed

Ejection Moment=800 Nm Safe Separation

Ejection Moment=1000 Nm Safe Separation

Ejection Moment=1500 Nm Safe Separation

76
4.6.4.1 Solution Domain Flow Visualization
Mach number and static pressure contours for shroud separation are given in below
figures.

Figure 4.33 Top Views of Mach Number Contour for Shroud Separation
(Altitude=5000 m, M=2.0, AOA=0° and Moment=1500 Nm)

77
 Figure 4.34 Top Views of Static Pressure Contour for Shroud Separation
(Altitude=5000 m, M=2.0, AOA=0° and Moment=1500 Nm)

4.6.5 Effects of Side-Slip Angle

In this part, effects of side-slip angle are investigated for the shroud separation.
Shroud separation analyses are carried out for two different side-slip angles by
keeping the altitude, Mach number and ejection moment constant. Flight conditions
and applied ejection moment are given in Table 4.11 for this part.

Table 4.11 Flight Conditions and Ejection Moment for Effects of Side-Slip Angle

Altitude [m] 5000

Mach Number 2

Side-Slip Angle [°] 0 5

Ejection Moment [Nm] 1000

78
Shroud trajectories belonging to two different side-slip angles are given in Figure
4.35. These trajectories reflect position change of center of gravity of shroud.

Figure 4.35 Shroud Trajectories for Different Side-Slip Angles

79
As it is shown in Figure 4.35, in all side-slip angles shroud separation is performed
safely for the specified flight conditions and ejection moment. Since side-slip angle
changes depending on the Z direction, it affects the shroud trajectories on Z direction
dramatically. When the missile has side-slip angle, shroud trajectories are not
symmetrical. In addition, as it is seen in Figure 4.36 and Figure 4.37, Z force and X
force values with respect to time are asymmetrical for each shroud component
(left/right) at non-zero side-slip angle whereas symmetrical Z-force and same X-
force values are obtained at zero side-slip angle.

Figure 4.36 Z-Force Acting on the Shroud with Respect to Time for Different Side-
Slip Angle

80
Figure 4.37 X-Force Acting on the Shroud with Respect to Time for Different Side-
Slip Angle
Also, summary of the shroud separation for effect of side-slip angle can be seen in
Table 4.12.

Table 4.12 Summary of the Shroud Separation for Effect of Side-Slip Angle

Side-Slip Angle=0° Safe Separation

Side-Slip Angle =5° Safe Separation

81
4.6.5.1 Solution Domain Flow Visualization
Mach number and static pressure contours for shroud separation are given in below
figures.

Figure 4.38 Top Views of Mach Number Contour for Shroud Separation
(Altitude=5000 m, M=2.0, Side-Slip Angle=5° and Moment=1000 Nm)

82
 Figure 4.39 Top Views of Static Pressure Contour for Shroud Separation
(Altitude=5000 m, M=2.0, Side-Slip Angle=5° and Moment=1000 Nm)

4.6.6 Effects of Aerodynamic Loads Acting on the Missile During Separation

The effects of time dependent variation of aerodynamic forces and moments on
missile during shroud separation are investigated. CFD analysis results were used to
investigate these changing effects on the missile for a definite flight condition and
ejection moment. Applied ejection moment and flight condition can be given in
Table 4.13 for this study.

83
 Table 4.13 Flight Conditions and Ejection Moment for Effects of Aerodynamic
Loads Acting on the Missile During Separation

Altitude [m] 5000

Mach Number 2

Angle of Attack [°] 0

Ejection Moment [Nm] 1000

X-force changing acting on each component of missile and total X-force in time was
given below figure.

Figure 4.40 X-Force Acting on the Missile Part During Separation with Respect to
Time
As seen in Figure 4.40, missile nose is most affected compared to missile body and
base during separation. Therefore X-force acting on missile nose becomes more
important for structural analysis during separation. These results should be taken into
account in the structural analysis of the missile nose.
In addition it is mentioned that shroud covers decrease axial force acting on missile
and shroud covers protect the missile nose from significant aerodynamic heating and
high recovery temperature. Also X-force acting on missile with/without shroud is
given in Figure 4.41.
84
 Figure 4.41 X-Force Acting on the Missile with Respect to Time
As seen in above figure, missile with shroud has lower axial force compared to
missile without shroud. Therefore missile with shroud is useful for range of missile
for same missile system. Variation of axial force acting on missile during separation
is also given in Figure 4.41.

85
86
 CHAPTER 5

CONCLUSION

In this thesis, CFD analysis of shroud separation on a generic missile was performed
to investigate safe separation of the shroud covers from missile for different
parameters which are altitude, Mach number, angle of attack, side-slip angle and
ejection moment by using commercial code CFD++. Firstly the subject was
introduced secondly methodology and governing equations used in the study were
clarified, thirdly analysis for test cases was carried on and compared with experiment
and finally a multi body CFD analysis approach was applied to shroud separation on
generic missile.

EGLIN store separation and HB test cases were used for CFD solver validation
analyses. Prior to shroud separation analyses, grid independence and turbulence
model selection studies were performed to get accurate results for multi body CFD
analyses. Aerodynamic coefficients were compared with experimental data for the
HB test case. Normal force coefficient, pitch moment coefficient and axial force
coefficient were in good agreement with the experimental data. Also the store
separation analysis results were compared with experimental data. Store position and
store angle were in very good agreement with experiment. Solution of these two test
cases and the compared results showed that the commercial CFD++ solver could be
applicable for multi body CFD analysis problems.

As a result of all this validation, the parametric study for shroud separations from
generic missile was studied and results of CFD simulation were presented. Shroud
trajectories were examined for different flight conditions and ejection moments.

87
Flight conditions and ejection moments for safe separation were determined for the
model used in analyses. Analyses results show that as the altitude increases, shroud
follows a trajectory close to missile body because of the decreasing aerodynamic
force. As the Mach number increases, shroud follows a trajectory that is further away
from missile body because of increasing the aerodynamic force. Angle of attack does
not affect shroud separation trajectory dramatically because angle of attack and
separation axis are different. Side-slip angle affects shroud separation trajectory
because side-slip angle and separation axis are same. Therefore, side-slip angle
effects should be taken into account while defining the ejection moment for safe
separation. As the ejection moment value increases, shroud follows a trajectory close
to missile body for same flight condition. Finally, effects of aerodynamic loads
acting on the missile during separation are important for structural analysis of the
missile nose. Therefore, these effects should be taken into account in the design
phases.

In conclusion, by using modeling techniques applied in this study, it is possible to

determine necessary ejection moment and separation trajectory of an arbitrary shroud
design with different shapes and materials at given flight conditions. In the same
manner for a given ejection moment, the flight envelop that enables safe separation
of shroud from missile can be determined.

88
 REFERENCES

[1] Y. Dagan and E. Arad, "Analysis of Shroud Release Applied for High-Velocity
Missiles," Journal of Spacecraft and Rockets, 2014.
[2] M. Oswald, D. S. Eder, K. Weinand, D. Stern and W. Zeiss, "Shroud Separation
Study Using The Six-Degree-of-Freedom Approach For High Mach Numbers,"
AIAA Multinational BMD Conference and Exhibition, Honolulu, 8-11
September, 2008.
[3] L. E. Lijewski and N. E. Suhs, "Time-Accurate Computational Fluid Dynamics
Approach to Transonic Store Separation Trajectory Prediction," Journal of
Aircraft, vol. 31, pp. 886-891, July 1994.
[4] J. D. Garay, "Summary Report On Aerodynamic Characteristics Of Standart
Models HB-1 and HB-2," July 1964.
[5] I. Whalley, "Development of the STARS II Shroud Separation System," 37th
AIAA/ASME/SAE/ASEE Joint Propulsion Conference, Utah, 8-11 July, 2001.
[6] A. D. Panetta, H. T. Nagamatsu, L. N. Myrabo, M. A. Minucci, D. G. Messitt
and T. E. Dominick, "Drag of Shroud Deployment Bladder at Mach Numbers of
8 to 20," AIAA, 1999.
[7] K. Anandhanarayanan, "Grid-Free Kinetic Upwind Solver on Chimera Cloud of
Points," West-East Speed Flow Field Conference, Moscow, 19-22 November,
2007.
[8] S. B. Lumb, "Transient Aerodynamics of a High Dynamic Pressure Shroud
Separation for a Ground-Based Interceptor Missile," Alabama, 19-21 May 1992.
[9] L. Resch, C. Decesaris and E. Hedlund, "The Naval Surface Warface Center
Hypervelocity Wind Tunnel 9 Hypersonic Shroud Separation Testing
Capability," AIAA 30th Aerospace Sciences Meeting and Exhibit, Nevada, 6-9
January, 1992.

89
[10] P. A. Cavallo and S. M. Dash, "Aerodynamics of Multi-Body Separation Using
Adaptive Unstructured Grids," AIAA 18th Applied Aerodynmamics
Conference, Denver, 14-17 August, 2000.
[11] M. Hughson, "A 3-D Unstructured CFD Method for Maneuvering Vehicles,"
Mississippi State University, December, 1998.
[12] F. M. White, "Viscous Fluid Flow," McGraw and Hill, 1991.
[13] "CFD++ User Manual-Version 10.1".
[14] J. Blazek, "Computational Fluid Dynamics: Principles and Applications," 2001.
[15] E. N. Ferry, J. Sahu and K. R. Heavey, "Navier-Stokes Computations of Sabot
Discard Using a Chimera Scheme," AIAA, 1998.
[16] Ö. H. Demir, "Computational fluid Dynamics Analysis of Store Separation,"
Ms. Thesis, METU, 2004.
[17] S. J. Zhang and S. F. Owens, "A Parallel Unstructured Chimera Grid Method,"
AIAA 17th Computational Fluid Dynamics Conference, Toronto, 6-9 June,
2005.
[18] S. J. Zhang, J. Liu, Y. S. Chen and X. Zhao, "Numerical Simulation of Stage
Separation with an Unstructured Chimera Grid Method," AIAA 22nd Applied
Aerodynamics Conference and Exhibit, Providence, 16-19 August 2004.
[19] E. R. Heim, "CFD Wing/Pylon/Finned Store Mutual Interference Wind Tunnel
Experiment," September 10-17 1990.

90