r. h. plaut thangjuly, s.€¦ · 1t.p~rt.6=zr. h. plaut thangjuly,a s. ia · 1989 blacksburg,...
TRANSCRIPT
STRESS ANALYSIS OF ROCKET MOTORS WITH VISCOELASTIC PROPELLANT BY A
MIXED FINITE ELEMENT MODEL
bv
Yung Tun Lin
Dissertation submitted to the Faculty of the
Virginia Polytechnic Institute and State University
in partial fulfillment of the requirements for the degree of
DOCTOR OF PHILOSOPHY
in
Engineering Mechanics
APPROVED:
R. A. Heller, Chairmané
J. N. Redä, —l. l- M. P. Singh
(ein-)
1t.p~rt.6=Z A IA ·R. H. Plaut S.ThangJuly,
1989
Blacksburg, VirginiaU
STRESS ANALYSIS OF ROCKET MOTORS WITH VISCOELASTIC PROPELLANT BY A
MIXED FINITE ELEMENT MODEL
bv
Yung Tun Lin
R. A. Heller, Chairman
Engineering Mechanics .
(ABSTRACT)
A mixed variational statement and corresponding finite element model are de-
veloped for an axisymmetric solid body under external symmetric Ioads using the
updated Lagrangian formulation. The mixed finite element formulation treats the
nodal displacements and stresses as the variables that can be approximated inde-
pendently. The method of static condensation is used to keep some stresses across
interfaces of a solid of revolution discontinuous. The stiffness matrix is transformed
from semi-positive definite to positive definite.
A rocket motor is composed of (1) case (2) propellant and (3) hollow air core
and is modelled as an axisymmetric solid. The propellant of a rocket motor is
treated as a viscoelastic material.
Static and dynamic analyses are performed under (1) two opposite line Ioads (2)
two opposite patch Ioads and (3) one line and one patch load combination. The
modified Newton-Raphson method is used in the solutions of nonlinear algebric
equations. The analysis of free vibration is executed first and then the Newmark
direct integration method is used in a transient analysis. Results of these analyses
are compared with solutions obtained from different methods that are independent
of the finite element method.
Acknowledgements
My sincere appreciation and respect are given to my advisor, Professor R. A.
Heller, for his guidance and support during the course of this study. Special thanks
are extended to Dr. J. N. Reddy for some suggestions on the finite element methods
and to Dr. S. Thangjitham for his ideas and help. I am grateful to Drs. M. P. Singh·
and R. H. Plaut for their encouragements and suggestions during this investigation.
This work has been supported by the U. S. Army Missile Command, Huntsville,
AL under Contract No. DAH-0186-C·D185. The support is greatly acknowledged.
Acknowlcdgcmcnts iii
Table of Contents
_ 1.0 lntroductlon ...................................................1
1.1 General comments ..............................................1
1.2 Literature review ...............................................3
2.0 Development of govemlng equations ..................................7
2.1 Introduction ...................................................7
2.2 Principle of virtual work ...........................................7
2.3 Hellinger-Reissner variational principle ................................9
2.4 Another formulation and boundary conditions ..........................14
2.5 Axi·symmetric tinite element model .................................19
3.0 Description ol vlscoelastic sollds ...................................32
3.1 Introduction ..................................................32
3.2 Derivation of the constitutive law ...................................33
3.3 Finite element model ............................................45
4.0 Algorithms ol solution ...........................................49
4.1 Static analysis ................................................49
4.1.1 Static condensation ..........................................49
Table of Contents iv
4.1.2 Modified Newton—Raphson method ................................51
4.2 Transient analysis ..............................................56
4.2.1 Free vibration ..............................................56
4.2.2 Newmark direct integration method ...............................59
5.0 Numerlcal examples and discussions ................................62
5.1 Static analysis with elastic propellant ................................66
5.2 Static analysis with viscoelastlc propellant ............................70 A
5.3 Dynamic analysis ..............................................73
5.3.1 Analysis with elastic materials ..................................73
5.3.2 Analysis with viscoelastlc propellant ..............................77
5.4 Summary ....................................................82
Appendlx A. .................................................... 176
Appendlx B. .................................................... 178
References ..................................................... 181
Table of Contents V
List of Illustratlons
Figure 1. A physical model of a multi-layered cylinder ................. 2
Figure 2. A finite element model ................................. 4
Figure 3. A solid of revolution ................................. 21
Figure 4. Nevvton-Raphson method .............................. 53
Figure 5. Modified Nevvton-Raphson method ....................... 54
Figure 6. Some vibration modes of a hollow tube .................... 58
Figure 7. Force distribution of the rocket motors .................... 63
Figure 8. Displacements of the cylinder vs. length ................... 68
Figure 9. Deformed shape due to self-weight ....................... 85
Figure 10. Displacement and stress distribution due to self-weight ........ 86
Figure 11. Deformed shape due to two opposite line Ioads .............. 87
Figure 12. Displacement and stress distribution due to two line loads ...... 88
Figure 13. Deformed shape due to two opposite patch loads ............. 89
Figure 14. Displacement and stress distribution due to two patch loads ..... 90
Figure 15. Deformed shape due to one line and one patch combination ..... 91
Figure 16. Displacement and stress distribution (line and patch) .......... 92
Figure 17. Radial displacement vs. time (two line Ioads) ................ 93
Figure 18. Radial stress vs. time (two line Ioads) ..................... 94
Figure 19. Tangential stress vs. time (two line loads) .................. 95
Figure 20. Radial displacement vs. time (two patch loads) .............. 96
List of lllustrations vi
Figure 21. Radial stress vs. time (two patch Ioads) .................... 97
Figure 22. Tangential stress vs. time (two patch Ioads) ................. 98
Figure 23. Radial displacement vs. time (line/patch Ioads) .............. 99
Figure 24. Radial stress vs. time (line/patch Ioads) ...................100
Figure 25. Tangential stress vs. time (line/patch Ioads) .................101
Figure 26. Force histories .....................................102
Figure 27. Radial displacement vs. time (two line loads: Force (A)) ........103
Figure 28. Radial stress vs. time (two line loads: Force (A)) .............104
Figure 29. Tangential stress vs. time (two line loads: Force (A)) ..........105
Figure 30. Radial displacement vs. time (two patch loads: Force (A)) .......106
Figure 31. Radial stress vs. time (two patch loads: Force (A)) ............107
Figure 32. Tangential stress vs. time (two patch loads: Force (A)) .........108
Figure 33. Radial displacement vs. time (line/patch loads: Force (A)) .......109
Figure 34. Radial stress vs. time (line/patch loads: Force (A)) ............110
Figure 35. Tangential stress vs. time (line/patch loads: Force (A)) .........111
Figure 36. Radial displacement vs. time (two line loads: Force (5)) ........112
Figure 37. Radial stress vs. time (two line loads: Force (5)) .............113
Figure 38. Tangential stress vs. time (two line loads: Force (5)) ..........114
Figure 39. Radial displacement vs. time (two line loads: Force (C)) ........115
Figure 40. Radial stress vs. time (two line loads: Force (C)) .............116
Figure 41. Tangential stress vs. time (two line loads: Force (C)) ..........117
Figure 42. Radial displacement vs. time (two line loads: Force (D)) ........118
Figure 43. Radial stress vs. time (two line loads: Force (D)) .............119
Figure 44. Tangential stress vs. time (two line loads: Force (D)) ..........120
Figure 45. Radial displacement vs. time (two line loads: Force (E)) ........121
Figure 46. Radial stress vs. time (two line loads: Force (E)) .............122
List of lllustrations-
vii1
Figure 47. Tangential stress vs. time (two line Ioads: Force (E)) ..........123
Figure 48. Radial displacement vs. time (two patch Ioads: Force (B)) .......124
Figure 49. Radial stress vs. time (two patch loads: Force (B)) ............125
Figure 50. Tangential stress vs. time (two patch loads: Force (B)) .........126
Figure 51. Radial displacement vs. time (two patch loads: Force (C)) .......127
Figure 52. Radial stress vs. time (two patch loads: Force (C)) ............128
Figure 53. Tangential stress vs. time (two patch Ioads: Force (C)) .........129
Figure 54. Radial displacement vs. time (two patch loads: Force (D)) .......130
Figure 55. Radial stress vs. time (two patch loads: Force (D)) ............131
Figure 56. Tangential stress vs. time (two patch Ioads: Force (D)) .........132
Figure 57. Radial displacement vs. time (two patch loads: Force (E)) .......133
Figure 58. Radial stress vs. time (two patch Ioads: Force (E)) ............134
Figure 59. Tangential stress vs. time (two patch loads: Force (E)) .........135
Figure 60. Radial displacement vs. time (line/patch loads: Force (B)) .......136
Figure 61. Radial stress vs. time (line/patch loads: Force (B)) ............137
Figure 62. Tangential stress vs. time (line/patch loads: Force (B)) .........138
Figure 63. Radial displacement vs. time (line/patch Ioads: Force (C)) .......139
Figure 64. Radial stress vs. time (line/patch loads: Force (C)) ............140
Figure 65. Tangential stress vs. time (line/patch loads: Force (C)) .........141
Figure 66. Radial displacement vs. time (line/patch loads: Force (D)) .......142
Figure 67. Radial stress vs. time (line/patch loads: Force (D)) ............143
Figure 68. Tangential stress vs. time (line/patch loads: Force (D)) .........144
Figure 69. Radial displacement vs. time (line/patch loads: Force (E)) .......145
Figure 70. Radial stress vs. time (line/patch loads: Force (E)) ............146
Figure 71. Tangential stress vs. time (line/patch Ioads: Force (E)) .........147
Figure 72. Force histories for viscoelastic propellant ..................148
· List oflllustmtions viii
Figure 73. Radial displacement vs. time (two line loads: Force (F)) ........149
Figure 74. Radial stress vs. time (two line loads: Force (F)) .............150
Figure 75. Tangential stress vs. time (two line loads: Force (F)) ..........151
Figure 76. Radial displacement vs. time (two line loads: Force (G)) ........152
Figure 77. Radial stress vs. time (two line loads: Force (G)) .............153
Figure 78. Tangential stress vs. time (two line loads: Force (G))...........154
Figure 79. Radial displacement vs. time (two line loads: Force (H)) ........155
Figure 80. Radial stress vs. time (two line loads: Force (H)) .............156
Figure 81. Tangential stress vs. time (two line loads: Force (H)) ..........157
Figure 82. Radial displacement vs. time (two patch loads: Force (F)) .......158
Figure 83. Radial stress vs. time (two patch loads: Force (F)) ............159
Figure 84. Tangential stress vs. time (two patch loads: Force (F)) .........160
Figure 85. Radial displacement vs. time (two patch loads: Force (G)) .......161
Figure 86. Radial stress vs. time (two patch loads: Force (G)) ............162
Figure 87. Tangential stress vs. time (two patch loads: Force (G)) .........163
Figure 88. Radial displacement vs. time (two patch loads: Force (H)) .......164
Figure 89. Radial stress vs. time (two patch loads: Force (H)) ............165
Figure 90. Tangential stress vs. time (two patchloads: Force (H)) .........166
Figure 91. Radial displacement vs. time (line/patch loads: Force (F)) .......167
Figure 92. Radial stress vs. time (line/patch loads: Force (F)) ............168
Figure 93. Tangential stress vs. time (line/patch loads: Force (F)) .........169
Figure 94. Radial displacement vs. time (line/patch loads: Force (G)) .......170
Figure 95. Radial stress vs. time (line/patch loads: Force (G)) ............171
Figure 96. Tangential stress vs. time (line/patch loads: Force (G)) .........172
Figure 97. Radial displacement vs. time (line/patch loads: Force (H)) .......173
Figure 98. Radial stress vs. time (line/patch loads: Force (H)) ............174
List or lllustrations ix
Figure 99. Tangential stress vs. time (line/patch Ioads: Force (H)) .........175
List of lllustrations X
List of Tables
Table 1. Mechanical properties of rocket motors .................... 65
Table 2. Comparison of solutions (FEM and exact solutions) ............ 67
Table 3. Comparison between two methods ........................ 69
Table 4. Static responses ..................................... 71
Table 5. Data for the viscoelastic propellant ........................ 72
Table 6. Frequencies of a rocket motor ........................... 74
Table 7. Comparisons between static and dynamic loads (interface) ....... 76
Table 8. Two line loads under force history (B) ..................... 78
Table 9. Two patch loads under force history (B) .................... 79
Table 10. One line and one patch load under force history (B) ........... 80
Table 11. Magnification factors vs. force histories .................... 81
Table 12. Comparison between static and dynamic analyses ............. 83
use of muc;l
xi
1.0 Introduction
1.1 General comments
Solid propellant rocket motors used for tactical missiles are stored in field
storage. They may be subjected to impact damage during handling or may be ac-
cidentally dropped parallel to (or perpendicular to) their axes during the loading
process. The stress distribution due to such causes can result in termination of the
useful service life of rocket motors. The purpose of this study is to obtain the stress
and displacement responses of the propellant in rocket motors (1) to external loads
and (2) to various temperatures at different environments.
A solid rocket motor considered in the present analysis is modelled as a
multi-layered cylinder consisting of (1) case (2) propellant and (3) hollow air core.
The structural diagram is shown in Fig. 1.
The visco·elastic propellant, of course, is the interesting part of the investi-
gation. The case can be treated as either an elastic or a viscoelastic material. The
mechanical properties of a viscoelastic propellant are obtained from experimental
data. The relaxation modulus is expressed in terms of Prony’s series and the time-
temperature shift factor (a,) is derived from the WLF equation. The constitutive law
Introduction 1
A(1) air core
(2) propellant
(3) case
Flgure 1. A physlcal model of a multl-Iayered cyllnder
Introduction 2
is derived from the hereditary integral for relaxation modulus ( Y(t) ) by lengthy
manipulation.
The finite element method has long been established as a versatile and pow-
erful tool of analysis for solids and structural mechanics. The mixed finite element
method is based on the Hellinger-Reissner variational principle that is a stationary
principle. This principle treats all the dependent variables as independent of each
other. The stationary conditions of the principle are the strain-displacement
equations, stress-strain equations, stress-equilibrium equations and both natural
and essential boundary conditions, in short, all of the governing equations of
elasticity.
A muIti·layered cylinder can be modelled as a solid of revolution. An
axisymmetric solid finite element model is used in the present analysis; the finite
element is a ring of constant cross section (Fig.2). Centers of nodal circles lie on
the z axis (axis of revolution).
The external loads can be expressed in terms of Fourier series. By recognizing
the orthogonality properties of trigonometric functions, the coefficients involved in
the calculation of 6 terms can be dropped because of their constancy. The problem
can be reduced to a series of two-dimensional problems. According to the super-
position principle, the original problem is performed by adding the solutions of each
component problem (corresponding to each coefficient of the Fourier series).
1.2 Literature review
The problem has been solved for cylindrical single layer shells [1-4]; the stress
analysis due to thermal loads or external loads for multi-Iayered cylinders is done
Introduction
U3
U
/ \ II \1 \ 11 x 1 \
1 x 1 \I1 J l I1 J \‘I \ I
· ll§lY% 'ix Z /1/\
Z 1/xx 1 V \ 1/
x / \ 1x / \ 1\ / x 1
Flgure 2. A flnite element model“
Introduction 4
by closed form solutions [4-8]. In the above papers [1-8], all material properties
were assumed to be elastic and homogenous.
The textbooks by Flugge [9] and Christensen [10] contain the definitions of the
heredltary integral and relaxation modulus for viscoelastic materials. For
viscoelastic materials the mechanical properties can be obtained from experimental
data [11-12]. The constitutive law of viscoelastic materials has been derived in
[13-19] by the assumption of linear strains between time intervals. Zak [20] solved
the structural analysis of viscoelastic propellant of rocket motors by integral trans-
form techniques.
Chung [21] solved the dynamic stress analysis of viscoelastic shells by a finite
element method. Wilson and Vinson [22] considered the viscoelastic analysis of the
plates by the Rayleigh-Ritz method. The dynamic buckling of viscoelastic cyllndrical
shells by analytical solutions and experimental techniques has been solved by
Bukowskin [23] and FIorence.[24]. They all assumed that the shell is a thin and
single layer.
Wilson [25] and Crose [26] have presented the finite element model for axially
symmetric solids; for more detail, refer to the textbooks by Cook [27], Yang [28] and
Zienkiewicz [29]. Jones [14] has developed displacement finite element methods for
solid rocket motors. This method only considers the nodal displacements as vari-
ables; the stresses are calculated from those known displacements by basic
equations of elasticity [30]. For multi-layered cylinders the tangential stresses
across layers are not necessarily continuous. Apparently the displacement finite el-
ement method can not circumvent this impasse.
Although the mathematical properties of mixed finite element approximations
have seen extensive development in recent years [31-33], the applications of mixed
finite element models to actual physical problems have been rare. In linear, elastic
introduction 5
problems the mixed finite element methods can be developed using the Hellinger-
Reissner variational principle [34-36]. This method treats nodal displacements and
stresses as independent variables. Not all stresses across interfaces of multi-
layered cylinders are continuous; the static condensation technique [28,36-38] is in-
troduced.
The relaxation modulus of viscoelastlc materials is dependent on time and
temperature, as is the constitutive matrix. ln static analysis the modified Newton-
Raphson method is used [27,39]. The equation of motion has been developed for
mixed finite models [40] by using the concept of variations [35]. The Newmark direct
integration method has been described in books [27,29,35] explicitly.
Introduction 6
2.0 Development of govemlng equations
_ 2.1 Introduction
ln this chapter we begin with the statement of virtual work for an arbitrary solid
body under external load and derive the variational functional that will be applied
to the mixed flnite element model of the problem.
When describing the motion of a solid body under external Ioads, a Lagrangian
frame of reference is typically adopted. Referring the variables to the initial,
undeformed configuration is known as the "total Lagrangian description”, and refer-
ring the variables to the current configuration is known as the updated Lagrangian
description. The updated Lagrangian description will be used to descrlbe the mo-
tion in this chapter.
2.2 Prlnclple of vlrtual work
The governing equations can be derived from the principles of virtual work
(i.e., virtual displacement, virtual force or mixed virtual displacements and forces).
Considering the principle of virtual displacements, the principle requires that the
sum of the external virtual work done on a body and the internal virtual work stored
Development of governing equations 7
in the body should be zero, i.e., a continuous body, V , should be in equilibrium
under the action of body forces, X, , and surfacetraction(s),Suppose
that on the portion, S, , of the boundary ,S. the displacements are E
and on the portion, S2, the tractions are The boundary portions, S, and S, , are
disjoint, and their sum is the total boundary, S .
Let the displacement vector of the body in equlibrium be u= (u,, u2, ua), and let
6,, and 6,, be the associated stress and strain components, respectively.
The principle of virtual work states that
öW,+öWE=0 (2.1)
öW, = virtual work coming from internal forces
ÖWE = virtual work coming from external forces
For a linear elastic solid, Eq.(2.1) can be written as .
=J.
X,öu,dV
*1**.*.-V
V S2
where 6 denotes the variational symbol ( öu, means the virtual displacement of u, )
and dV and dS2 are the volume and surface elements in configurations over which
the integrations are executed.
The linear strain-stress relations of a (visco·) elastic solid are
{8} = ¤{{8} — {80}) + {G} (23)
Dcvclopmcnt of governing cquations 8
where {a}, {6} are the stress and strain vectors. Assume U, the strain energy, ex-
ists such that
V
and a quantity U', called complementary enel'QY„ such that
ÖU· =IÖ0Ü£ÜdV (2.5)V
With the potential energy of the applied load, V , as a function of the displace·
ments, u,, given by
V = —f XiuidV+‘I· tiuidV (2.6)
the virtual work statement is equivalent to that of minimizing the total potential en-
ergy, H:
ÖH = Ö(U + V) = O (2.7)
2.3 Hellinger-Reissner varlational prlnclple
In deriving the variation functional (Eq. (2.7), we have assumed that strains were
related to the displacements (strain-displacement relationship),
Development of goveming equations 9
1. Bü =
? (Ui,} + Ui,
i)andon the boundary S, (displacement boundary conditions)
U; = E (2-9)
lf we want to relax these constraints (Eqs. (2.8) and (2.9) in the variational
statement Eq. (2.7)), we lntroduce Lagrangian multipliers, 1,,, defined in the domain
V (associated with the strain-displacement equations) and v,, defined only on the
boundary S, (associated with the displacement boundary conditions). Then a new
variational principle is obtained:
1 -H1 = H '* " *5 (Ui,} +—·v[ vi(ui —•ui)dSV
S1
After substltutlng the total energy into the above equation, Eq. (2.10) can be written
as
n1= U
-12
XiüidvVVS2 (2.11)
—( w(¤; — EWSS1
Performing the variation of Eq. (2.11), l.e.,
ÖH, = 0 (2.12)
gives
Dcvclopmcnt of govcming cquations l0I
ÖH1 ölÜ(sÜ —é- (ULL + UL L))dV —
jl}lÜ(6gÜ - öuLj)dV
V V V
vLÖuLdSS2 $1 S1 ·
jöimlmj—-I,
@öuLdS—].
vLÖULdS -16,1„(g„-é (UL} + UL i))dV
S2 S1 V
övi(uiS1
Using integration by parts,
V V S
Substituting Eq. (2.14) into Eq. (2.13), then the Eq. (2.13) becomes
O = fVösU(oÜ — lÜ)dV —- XL)dV + öUL(njÖLmlmj — vL)dS1
— 1+1‘
öUi(njöim'1·mj— GWS * 1;/Ö4y(@y * 5011,;+ E, i)WV (2.15)S2
*1 öV1{E · EWSS1
As the above is true for any variations, we immediately observe that the last
two Euler conditions give us the constraint conditions.- Since the above equation
(2.15) is true for any variations, the coefficients of the variations are zero:
· Development of goveming equations ll
öui Z [öim,lmi]_i + Xi = O in V (2.16a)
öcü : aii = ÄÜ in V (2.16b)
öui : niöimlmi — vi = 0 in Si (2.16c)
6Ui I Djöimztmj— 6= 0 in S2
I Bü = (Ui.] + Uji i) in
Vövi: ui = E in Si (2.160
We can observe that the Lagrangian multiplier can be identified as followsz
Äii=oÜvi
= {Ü (2.17b)
Apparently lii and vi are stress components and boundary tractions, respectively.
Hence Eq. (2.10) can be rewritten in explicit form,
Hi = H -1;/¤Ü(sÜ— é- (ui_i -+- ui_ i))dV — ti(ui — E)dS (2.18)
This expression is called the”Hu
- Washizu" variational principie; however this
formulation is not a practical principle. We recall that the sum of the complementary
and strain energies is written as
Development of goveming equations I2
V
Another variational form can be obtained by replacing U by U' in Eq. (2.18),i.e.,
H2 =
jl
GÜEUÖV — U·(¢7ij)-~|·
uiXidV*j~V
V S2 (2.20)1 -— '[ —- u;)dS
A special form is obtained by assuming that the strains are related to the dis-
placements by _1
1EU = *5 (Ui,} + U},
i)Wecan write Eq. (2.20) as the "HelIinger - Reissner variational princigle" that can
be stated as
H2 = GÜSUÖV — U.(cÜ)—‘I‘
uiXßV —1*
u;@dS—j·
Mu; — @)dS (2.21)V V S2 S1
wherein 6,) and u, are independent variables.
For linear solids we assume zero initial strains ({60} = 0); then Eq. (2.3) can be
written as
{S} = ¤({¤} — {6}) (2-22)
Development of goveming equations 13
The complementary energy of the system is
· D 2U =g—({¤}-{6}) (2-23)
and the strain — displacement relationship is 6,, =%(u,_,+ ul_,).
Now we differentiate Eq. (2.21) with respect to displacement and stress; this
yields the two approximate equations.
lt is remarked that the prescribed displacements, u, , on the boundary, 8,, are
_ removed (it means no prescribed displacements on boundary) in deriving these two
equations. They are given by
V v
J-öu,(Xi — @)d8 = 0 (2.24b)
V S2
The above two equations are the mixed formulations; stresses are no longer
dependent functions of the displacements.
2.4 Another formulatlon and boundary conditions
Although we derived the mixed finite element model, no explicit statement has
explained the specificatlon of the boundary conditions. In order to determine the
nature and characteristics of the boundary conditions, we use the variational form
Development of governing equations 14
of basic equations : equilibrium equation and stress - displacement relationship.
This will give us the classification of the boundary conditions of the mixed finite el-
ement model into natural and essential boundary conditions, and yield explicit for-
mulation of this method.
ln this section we want to derive the explicit equations in cylindrical coordinates
from the equilibrium equations and stress - displacement relationship since a rocket
motor is modelled as a multilayered cylinder.
The equilibrium equations in cylindrical coordinates are
öarr 1 öorß öorz (°rr—"08)——- ——— —— -—l = .25ör+r ö0+öz+ r +X,.O (2a)
öarß 1 ÖG66 ö°z6 2°z8—— ————— ——— —- = 2.25b()
öarz 1 öazß öazz 020——- ··i· ·—·· X = O 2.25
and stress - displacement relationships are
öu,"ET=D11Urr+D12088+D130zzM
1 1 MT + T 6 + Dzadzz (2-26b)
öuzjf = Disdrr + 023*70 6 + Dasdz z (2-26c)
öuü öu
DCVCIOPIDBHÜ of g0V€l’IIiIIg BQIIRÜOIIS
1 öu ÜU6 U6
737 +7— T = D6:662 6 (2-266)
öu, 1 öu,7+TE = DÜGGZÜ (2-260
We will construct the variatlonal form of these nine equations by taking all
terms to one side and multiplying each equation by a different test function
w, ,i= 1, 2, 3,...9 and then integrating the resulting equations over the domain of a
typical finite element.
The test functions are arbitrary. However, in order to achieve some physical
meaning for these equations, the test functions are chosen as the first variations of
the displacements and stresses, i.e.,
— w1 = 6u,—· W2 = 6u0— w3 = 6u,
W4 = ÖGN-
W6 = 6666 (2-27)W6 = ödzzW7 = öarzW8 = öaröwg = 6626
ln this case the signs of the test functions are arbitrary and these result in a
symmetric matrix. ln order to arrive at the physical meaning of boundary conditions,
we perform the calculations by multiplying Eqs. (2.25) with 6u,, 6u0 and 6u, and
multiplying Eqs. (2.26) with 66,,, 6600 , 66,,, 66,,, ÖÖ,6 and ÖÖZO,
Development of g0V£I'I’IiI‘Ig BQUQÜOIIS
86 8 86
V v V
= I x,66,6v 6,,66,61661 + '[ 6, ,66,666-10 (2·288)v s s
v
8 86 86
v v v
X9öu0dV 6, Oöußrdüdz 6, Üöugrdrdz (2-288)v s s
= IVXÜÖUÜÖV6öu
1 66 0 66u
v
=jl
x,66,6v 6, ,66,r666z + I 6, ,66,66616 (2-288)v s s
u,dV + Lt,6, ,dS
6u,
v
DCVCIOPIIIEIK of g0VCI’Ili|'lg CQIISÜOIIS
ÖU6 1 1 ÖU6( 80 T
_T 60
_D12°rr '“ D22"0 0 " D23°z z)ö°00dV = 6 (2286)
V
öuz( 82
— D‘|3Urr_
D23U0 6 " D336z z)ö6zzdV = 0 (2-280V
ÖU8 öuz(-57 + E- — D44a, z)ö0,zdV = 0 (2,28g)
V
1 öu, öuo ug(Y36- + —
Y— D55d,. 8)Öd,.8dV = O (2.2Bh)
V
öuz 1 öuz .(E-· + Y 717— DGSUZ 8)Ö0‘z8dV = 0 (2.28I)
V
The above three Eqs. (2.28a, b, c) are obtained by integrating by parts, and
noting that dV = r dr d8 dz with
t, = 6, rn, + 6, ang + cr, znz (2.29a)
t9 = U,. 8/7,- + G8 Gr/6 + dz8/12·
Development of governing equations 18
fz = 0, zh, + 0z gng + 02 znz (2.29c)
where n = (n,, ng, nz) is a unit normal to the boundary, S.
We also note the nature of boundary conditions in the first three equations.
Since only the variations of the displacement components u,, ug and uz appear in
the boundary integral,U
ür = ur(76 =U6Üz
= UZ
constitute the essential boundary conditions of the problem. Prescribing the coeffl-
cients of the variations of the displacements u,, ug and uz, the natural boundary
conditions are constructed:
ir = Ur rnr + Ur znz(6 = O', 6/7, + 0'Z6/7ztz
= Ur znr + dz znz
Apparently t= (t,, tg, tz) is a stress vector on the prescribed boundary, S, in cy-
lindricalcoordinates.2.5
Axl-symmetric flnlte element model
A solid of revolution is axially symmetric if its geometry and material properties
are independent of the circumferential coordinate, 6.
DOVCIOPITICIIY of gOV¢|’l’l(IIg €q\l8(iOI'IS
A rocket motor can be modelled as a multi-layered cylinder, and its material
properties are independent of 6. Hence it can be treated as a solid of revolution
(Fig. 3).
The centers of all nodal circles lie on the z — axis (axis of revolution). Each el-
ement is a solid of revolution about the z- axis. Fig. 3 depicts a solid of revolution
modelled by rectangular flnlte elements, and each finite element is a ring of con-
stant cross section.
If the loading is axially symmetric, there is no displacement in the 6 — direction.
The displacement vector has only u, and u, components.
lf the loading is not axially symmetric, it can be expanded into a Fourier series.
The given series is expressed as the sum of several component Ioadings, so the
analysis can be performed for each component. Applying the principle of superpo- ‘
sition, the original problem is solved by adding the solutions of the component
problems.
Let the loading be expressed as a Fourier series
X, = ZX, „ cos(n 6)
X, = ZX, „ cos(n 6)
F, = EQ, cos(n 6)_ _ (2.32)tz = Zt, ,, cos(n 6)
Xg = 2Xg ,, sin(n 6)
tg = Zig ,, sin(n 6)
where X's and Ps are, respectively, body forces per unit volume and prescribed
surface tractions on the boundary, S, ln the r, 6 and z directions.
Development of goveming equations 20
Z
Ä"-/il ‘
I°‘
U 11
Z8
Q ‘0000000000000000
' I'
Flgure 3. A Solid of Revolutlon
DCVCIOPITICIÜ of g0VCI’|'Ilhg CQLIQÜOIIS
Similarly, for displacements, we assume that these can also be expressed as
u,. = EIUM cos(n 6)
ug = ZUG „ sin(n 6) (2.33)
uz = ZUZ „ cos(n 6)
where u's are radial, circumferential and axial displacements,respectively.Eqs.
(2.33) represent a state of symmetry with respect to 6 about the plane
6 = 0°. The strain-displacement relationship in cylindrlcal coordinates is expressed
asU
88,, ·ä· 0 0
1 1 896 6 T TE)? O
6 u'szz OO=
• U9 (2.34)
E .Q. 0 L’zöz ör
UzLE. .ä._l. O*9 r 86 8r r
E 0 L LLZ9 8z r 86
From this relationship the strains can be written as
Development of g0VBf|liIIg CQIIBÜOIIS
6, , = ZE, ,6 cos(n6)
66 6 = 266 ,6 cos(n6)
62 2 = 262 ,6 cos(n6)(2.35)
S6: = Em,. ¤¤S(¤9)
6,6 = 26,6 ,, 6:6066)626 = Z6'26 ,6 sln(n6)
where
E =öü,,,’"ör
- _ urn JL -86 n "° r_
r u6n
_ öü2,66= r· = E"
- @¤„„ öüzn (2·36)@6 6 „ =7;+-
_ _Q_ - öu6n Ü6n8r6n“’ r urn+ Ö, " ;-
E ü'z 6 n öz r zn
Furthermore, from the constitutive law (Eq. (2.3)) the stress components should
have the following formulatlonsc
Development ol' goveming equations 23
6, , = Z5, ,, cos(nö)
60 0 = Z50 ,, cos(nö)
62 2 = 252 ,, cos(nö)(2.37)
¤„ = 2822 „ ¤¤S(¤9)
6,0 = 25,0 ,, sin(nö)
UZ9 = 2529 ,, $ll'l(l'IÜ)
where the barred quantities are functions of r and z but not of ö.I
Substituting Eqs. (2.32), (2.33) and (2.37) into the Eq. (2.28) and noting the coef· .
ficients of Eqs. (2.28) (D,,,..., DSG) for isotropic materials
- 1D11 = D22 = Das = ‘g‘ (2-338)
D12 = D1:1 = D23 = —é (2-33b)
2(1 -l— v)D44 = D55 =-' DGS = "
Tthefollowing equations can be obtained:
-ööürn 2 Fön — 2 Ll_- —
- 2{6,,, cos (nö) + -7 öu,,, cos (nö)- , 6,0,,öu,,, sin (nö) +
_ (2.396)
- Döurn 2 _ — — 2 ‘ — 26,2,, -5- cos (nö)}dV - X, ,,öu,,, cos (nö)dV + t,öu,,, cos (nö)dS
s
Development of goveming equations Z4
66- 6-fg {L 58,,6Ü8,, cos2(nö) + 5,8,,( — lßl) sin2(nö) +
ööü (2.39b)5 —QQ- sin2(nö)}dV = Y 65 sin2(nö)dV + [65 sin2 nö dSzön 82 ö n ön ö ön
S
66- 66- .<[‘{52,, sin2(nö) + 5,2,, sin2(nö) — 528,,652,, sin2(nö)}dV
(2.39c)
= ,,652,, cos2(nö)dV + j‘?26ü2,, cos2(nö)dS ‘
s
öürn - 2 1 - - 2 v - - -{T66,,, cos (nö) — —E 6,,,66,,, cos (nö)
(68,,cos2(nö)}dV= 0
Ü - - 1 .. - 269nöUÜf7 °°S (ng)
(2.396)
+ iv (5,,, + 52,,)658,, cos2(nö)}dV = O
65 _ _ _662,, cos2(nö) 52,,652,, (66,, + V;-n)öUzn (2.39,)
cos2(nö)}dV = O
Development of governing equations 25
öü öü 2 1 ++ %)öarzn C°S2("6) " Erznöarzn Sin2(n0)}dV (2.3gg)
= O
- öü F _ _ 2 1 +(2 22,2
sin2(n9)}dV = 0
ÖÖÜ _ _ _ 2(1 -l- v) _ _ _uzn)öown own s1n2(n0)öown}dV (2239,)
= O
where dV = rdrdüdz for a typical finite element with a finite volume dV. If
5,,, ,50,,,..., Un, can be interpolated by the following form
5,n = (2.406)
özn = Zöznilvi (2*40c)
Development of goveming equatlons 26
aröln = Zarünilvi (2-40d)
E-f'ZI‘l =EEFZITÜNÜam
= 2öz6„1^'1 (2-400
Ürn = Zlürni/Vi (2-409) V
Üßn = ZÜ6„1^': (2-40h)
üzn = 2ÜzniNi (2-40])
substituting Eq. (2.40) into Eq. (2.39) a typical equation
E1<J„ · (X)., = (1=)„ (2-41)
is obtained where [Kl, is the element stiffness matrix, {X},, is the element nodal ,
variable vector and {F}„ is the the element force vector; these correspond to the
n - th term of the Fourier seriesl
---.... - - - T{X} = {urn
-Ußn ·
uzn ·am
·°r8n • Grzn ·
°”9n • azn • Uzßn} (2-428)
‘For convenience, we drop the subscript n in the subsequent analysis
Development of governing equations 27
{F} = {Y1 J2 · fa » °· °- °·O- 0- 0}T (2-42b)
[K] is a 9 by 9 symmetric matrix.Its coefficients are ([K] is a 6 by 6 symmetric
matrix for axisymmetric load, see Appendix A.)
ÜN-14 1 36
K55 = I- % 1v,1~g6A = 1<5° (2.43b)
ÖN-16 1 38 29KU dA = KU = KU (2.43C)
N.dA (2.43d)
ÜN. N·K55 )dA (2.436)
K57 = Nil\QdA (2.43f)
Development of goveming equations 28
44 -1Kr = ITNlNfdA = K6? = Kgs (2·439)
Kf - I-E- ~,:~g6A = Kf - ;<,§° (2.43h)
1 +:<§5 = -24%~,~j6A
= :<,§?6 - :<,§?° (2.43i)
Other coefficients of [K] are zero, and
:, Y, „~,dA + IZ, „~,ds (2.448)
:2 - 79 „~,dA + fi), „~,6:s (2.44b)
:2 272 „:v,dA + fil „~,ds (2.446)
From Eq. (2.39), we find, for each coefficient n , integrands of the element
stiffness and load vector containing cos2(n6) and sin2(n6) in every term. Integrating
with respect to 6 leaves the problem two - dimensional (r, z).
Development of governing equations 29
p Furthermore, from Eq. (2.39) we can observe that the n—th circumferential
component of loading is associated with the n—th circumferential component of
stresses and displacements. This means that the coefficients of the Fourier series
are decoupled.
The resulting constant of TE (and 21: for n = 0 ) can be canceled from every term
of the equations. The final element stiffness matrix is of two - dimensional case
(r, z) . The element stiffness matrix depends upon n (the n — th term of the Fourier
series); the different values of n represent different problems that do not interact
with one another. Therefore the solution is the superposition of n different prob-
lems.The
axisymmetric model can reduce a three - dimensional problem physically
to a two - dimensional problem mathematically. lt saves computational storage and
decreases the CPU time so that it adds to the availability of the program.
The preceeding discussion involved only loadings and displacements that have
6 = 0° as a plane of symmetry. ln fact, anti - symmetric loadings can be considered.
Eqs. (2.32) can be augmented:
X, = ZX,„ cos(n0) + X,„ sin(n6)
Xz = ZX, „ cos(n6) + Xlzn sin(n6)
- = 2- ~ _t, t, ,, cos(n0) + t, ,, sm(n0)
(2.45)tz = Zt, „ cos(n6) + tz „ sin(n6)
X6 = 2)% ,, sin(nl9) — X9 ,7 cos(n8)
tg = Zig „ sin(n6) -— Z, „ cos(n9)I
Development of governing equations 30
where the anti - symmetric modes are represented by ~ . Displacements, stresses,
and strains can be modified similarlyß
The n=0 term in Eqs. (2.45) represent the axially symmetric problems, and
even terms n = 2, 4, 6... correspond to loadings and deformations that have 0 = O°
as a plane of symmetry with respect to 0. Similarly, odd terms correspond to anti -
symmetric loadings.
2 the tilde quantitiles are functions of r and z but not of 9
DQVCIOPIDCDI of g0VC|‘I'liIIg BQUIIÜOIIS
3.0 Description of viscoelastic solids
3.1 lntroductlon
The formulation of the isothermal viscoelastic stress - strain relation states that -
the current values of the stress tensor depend upon the complete past history of the
components of the strain tensor. Usually this relation is described by a ”hereditary
integral" for relaxation, Y(t), that ls a characteristic of viscoelastic materials.
The basic assumptions of the viscoelastic solids are :
(1). isotropic thermal expansion
(2). linear, small deformation
(3). thermorheologically simple temperature response
(4). Poisson’s ratio is a constant
The viscoelastic material accepts (1) material property data (Prony series) ob-
tainable from standard laboratory stress relaxation tests from which the shear and
bulk relaxation moduli are derived and (2) a temperature - time shift factor : a,, a
function of temperature. The latter is the key element in a thermorheologically sim-
ple concept in which a reference temperature (usually room temperature) relaxation
function can be utilized for elevated temperature responses by elongation of the
time scale. This is done by replacing the real time with a reduced time with the
Description of viscoelastic solids 32
magnitude determined by the WLF (Wi//iams, Lande! and Ferry) time- temperature
shift function.
The following sections deal with the construction of the constitutive law of
viscoelastic materials and derivation of the flnite element model.
3.2 Derlvatlon of the constitutive law
Following a procedure parallel to those in elasticity whereby the deviatoric
components of stress Sy and strain ey are introduced, then,
t Ö HQ
ISylt)= G1(C(t) — C (t)) -:9;- dt (3-1)*00
and
' I öe(r)vkklt) = G2(€(t) " C (Ü) 98 (3-2)*00
where
1SÜ= ÜÜ‘°§öÜÜkk1
9/7 = EU — T3'Öüökk9
= (811 + 822 + 862) (35)
Description of viscoelastic solids 33
and 6,, is the Kronecker symbol.
To be consistent with common notations in elasticity, the symbols for the
isotropic relaxation function in simple shear and dilatation are frequently taken as
G(t) and K(t), respectively, where —
G lG(l)=3.6@20) * 1
KU)
=Theviscoelastic stress functions in terms of G(t) and K(t) relaxation functions
are now given by
t
2 G(C — f )-jajdr
f (3.7)
+ {K(€ — — €')}Ö··@d‘l'3 V ÖT0
where { and {' are the reduced times at t and 1 and are defined as
g: (6.63)a¢(V)0
5- = 1-*1%- (6.6:6)ar(T)0
Description of viscoelastic solids 34
Here a, is the time- temperature shift factor based on an assumption of
thermorheological simplicity for converting stress relaxation data gathered at a se-
ries of temperatures to a single curve. It is assumed that stress curves obtained
from relaxation tests performed at different temperatures are appropriately shifted
along the Log time axis.
An empirical function relating temperature to the shift factorTa,
, which is appli-
cable to many viscoelastic materials, was formulated by Williams, Landell and
Ferry, and is known as the WLF equation. It has the form
C1(T* TR)l =i— 3.9¤¤ a„(T) C2+(T_ TR) ( )
where the constants C, and C2 are determined from experimental data, and the ref-
erence temperature, TR, is the absolute temperature for the base curve.
The stress relaxation test generates a curve which can be approximated nu-
merically by a Prony series. lt has the form
8 t/1,,,,E(t)=EO+ ZETT, exp(—-T) (3.10)
m=1
where EO, E,,,, and 1,,, are moduli and relaxation times of parallel Maxwell elements;
EO is the long-term modulus (t—> oo).
For the special case where Poisson's ratio is a constant with time, the shear
modulus can be written as
E(t)G t
=_.l3.11O 2(1 + v)
( )
and the bulk modulus as
Description of viscoelastic solids 35
2 EU)2 t "' i 'l"K(t) A()+ 3 G(t)
3(1_2v)(3.12)
Denoting the first term on the right hand side of Eq. (3.7) as I,,
2'/1=2 G({ - { )Tdt (3.13)
0
and combining the Eqs. (3.10) and (3.11), then G(t) has the following exponential
form:
8 {/rm‘
G(t) = go + gm exp( — -T-) (3.14a)m=1
where
2g<> = im;(3.14b)
2gm (1 + V)
Substituting Eq. (3.14a) into Eq. (3.13),
'8
I 62„(V)/1 =2 [220+ 29,,, @><¤—C„„(é—€)]—g;··dr (3-15)
0 lT7=1
where é=é and§’
= äare reduced time parameters, and §,„ =%- .
Description of viscoelastic solids 36
Simplifying the Eq. (3.15),
I 8 gö€1j(T) ö81j(T)/1 = 2 911Tdr + 2 9m sxpl — C„,(C — C')1Tdr (3-16)
0 m=‘I 0
The second integration term on the right hand side of Eq. (3.16) is now sepa-
_ rated into two parts : the first part has limits from "zero” to”t- At " and the second
has limits from " t- At”
to”
t (current time step). Hence,
' I Ö8Ü(T)9„, @><i¤[ — C,,,(C — C )1Td1 =0 (3.17)"^' I 6,;,,(1) ' I ösÜ(·r)
9,,, @><p[ — C,,,(C — C )1Tdr + 9„, 6><p[ - C„,(C — C )1Tdr0 1-81
The first term on the right hand side of Eq. (3.17) can be written as
'—^'I9,,,6><p[ — C„,(C — C )1Tdr0
*·^* I 6,„(-)= 9,,16><P{—·C„,EC—(€—AC)+(C—AC)—C]}Td¢
O (3.18)‘·^'
I ösÜ(·c)= 9,,, 6><¤( - C,,,AC) @><r>{ — C„,[(C — AC) — C 1}Tdr
0
= 9,,1 6><¤( — C,„AC)pi,Z°°
Description of viscoelastic solids 37
where Aé = %-, ande
2-Ar_ Ö6··(t)Pfnijét
=.['@><P{ — Cm[(C — AC)- C']}# dr (3-19)
0
The second part of Eq. (3.17) now is integrated by parts:
Qm @><p[ — Cm(C — C')] -5;- dr C¢—Ar
=‘—- GX -— _ÜEÜÜ)
[ C (C C’)1 a'T 6 20Ö? gm p m “
Cm 2-Ar( )
— ————exp — — 1
t özsü 8* L C rc mdöt2 Cm mr—A1
ln the current time step [t—At, t], we have assumed 6,] to be a linear function
of time:
ÖT At
Hence its second derivative is zero.
Substituting Eq. (3.21) into Eq. (3.20) yields,
Description of viscoelastic solids 38 .
' Ö8Ü{T)9m expt — Cm(C -
C’)] —g— dr:—A:
8 -·(t) ö ;·(t — At)Cmté —(C
A:)A
(3-22)— —————A, 9m Cm t — ¤><p( — Cm C)]
m
= (81j(t) " 8U(t ' Atngmßin
where —
, _ 1 — @><1>(—CmAC)
Then substituting Eqs. (3.18) and (3.22) back into Eq. (3.15) gives
6:-A:/1 = 290@g(*) + 2 gm @><P( — Cm^§)Pmy
m=16
+ 22am(¤„)(¢) — @q(¢ — A1))ßi„m=1 (3.24)
6
= (290 + 2 9m://::1)¢y(()m=1
8:-A: :+ 2 amt @><¤( — CmAC):¤m) — ¤,y(¢ — A1)/fm]
m=1
Description of viscoclastic solids 39
I We can write Eq. (3.24) in an abbreviated form,
I1 = G)sU(l) + GU (3.25)
where
8
GI = 2QOm=1
and
8
GU = 2 2 9mi: @><r>( — C„„Aé)i¤Ä„ÜU°‘ — ¤U(¢ — A0ßi„] (3-27)m=1
The coefficient, G,, is called the instantaneous shear modulus, while G,) is the
result of hereditary shear stresses. It is noted that the term p§,;f' in Eq. (3.27) is the
m — th component of the hereditary stress 6,) at the last time step (t- At) and it can
be derived from a recurrence formula as shown below. By definition (Eq. (3.19))
Description of viscoelastic solids 40
Ipmü = <6><p[ — C,„(é — é')]Tdr0"^‘
6¤,~<1>'”
6 -1 >=j‘
@><1¤[ — C,„(C — §’)]Tdr+I[ 6><p[ — C„,(~§ - é')] äldr0 r-Ar
=f GXPE-C,„(€-C+€-C’)]·%;Ldr0 I ÖSÜITI (3.28)
+ <6><¤[ — C,„(~f — é’)]Tdr1-A:
"^' _I ösÜ(r)
= exp( — C„,A€) <6><p[ — Cmlé — é )]Tdr ·0
+ (81y(Ü — SUU — Amßin= exp( — c„.Ac>p£,;;" + <¤„<0 -
6„(1where{= { —A{ =%
Consldering the second term of Eq. (3.7) and noting that from Eq. (3.12)
l(t) = K(t) —%G(t), the l(t) will have an exponentlal form:
8 t/·r,l(t) = ho + hm exp( — 7%) (3.29a)
m=1
h_ Eov
°‘
(1+ v)(1—2v) 3 2%_ Emv ( - )hm —
(1 + 0)(1 -20)
and substituting Eq. (3.29a) into Eq. (3.7), it gives
{
Description of viscoelastic solids 4l
2 8
/2 = 2 [//0 + hm exp — C„„(é — é )]7; de (330)0 m=1
Using the same procedure derived in the above context yields,
I2 = Hle + H,, + H22 + H33 (3.31)
where
8
m=1
and
8A— t IHkk exp( (3-33)
m=1
Combining the results of I, and I2, we have the following formuiation in matrix
form:
Description of viscoelastic solids 42
6, ,(t) G, + H, H, G, 0 0 0 6,,(1)06 6(t) H, G, + H, H, 0 0 0 66 6(t)0ZZ(t)
=H, H, H, + G, O 0 0 • 6zz(t)
0, 6(I) 0 0 0 G, 0 0 6, 6(t)0,z(t) 0 0 0 0 G, O 6,z(t)6z 6(t) 0 0 O 0 0 G, 6Z6(t)
(3.34)G11 +H11+H22+H33G22 + H11 + H22 + Has‘
+ Gaa+H11+H22+HaaG12G1aG2a
Then Eq. (3.34) can reduced to a compact form _4
— 1611)) = E,N16<1)) + ~{8<1>}A (3.36)
= c{6(1)} + N{0(t)}
where C=E,N and {0}, {6} and {8} are the vectors of stress, strain and hereditary
stress, respectively. Hence
{6} = {6, ,(t), 66 6(t), 6,Z(t), 6, 6(t), 6, z(t), 6z 6(t)}T (3.37)
{6}= {6,,(t), 66 6(t), 6„(t), 6,6(t), 6,Z(t), 6z6(t)}T (3.38)
{G} = {$1 r(* — A!) 86 6(F — N)» GZ ZU — N)· (3 39)8,6(1— A1), 6,,(1 — A1), 8Z6(1— At)}T
Description of viscoclsstic solids 43
8
E, = Eo + 2:*5,,,/*1,,1 (3-40)m=1
8
8,%* — 4*) = 2E„,[ @><P( — C,„A€)PÄ,]8t — 8%* — A*)ß‘„,] (3-41)m=1
uand
(1 — v) v v 0 O O .
v (1 — v) v 0 O 0
1 v v (1 — v) 0 0 ON = (1 + ")(1 -2V) Q
‘Q Q Q Q
(3.42)
20 0 0 0 00 0 0 0 0 lf!
Eq. (3.36) gives a general viscoelastic constitutive relation that applies to the
problem. Rewriting Eq. (3.36) yields,
(,;(:)} = ¤(¤(¢)} - {Su — 4:)} (3.43)
where D = C", and {€(t)} are the hereditary strains that depend upon the last time
step, and
A t- At{€(t — At)} = (3.44)
Description of viscoelastic solids 44
3.3 Flnlte element model
The procedures used to derive a finite element model of viscoelastic materials
are the same as those in Chap. 2 for elastic materials. The only difference is that
the element force, right hand side of Eq. (2.43), must include the force due to the
effect of hereditary stresses.
To derive this finite element model, we assume that surface traction(s), heredi-
tary - stresses, stresses,...and displacements can be expanded into Fourier series:
622(0 = 262 „(¢) ¤¤S(¤9)
66 6(0 = 266 „(¢) ¤¤S(~9)
6220) = 262 „(#) <=¤S(¤9)
Ur z(t) = Zar: n(t) CO$(n9)
ar 9(t) = 2520 „(t) sin(n0)(3.45)
62 9(t) = 252 9 „(t) sin(n8)
(......)
¤„(¢) = Z6„„(t) ¤¤S(¤9)
u0(t) = ZU9„(t) sin(n8)
u2(t) = 2Ü2„(t) cos(n9) 2
Then we assume again that 5,„(t), 5„„(t),....17„,(t) can be interpolated by shape
functions, N,, i =1,...r in a typical finite element with a finite volume:
Description of viscoelastic solids 45
Fm = Zörnilvi
(.............) (3.46)
Üzn = Züzni/Vi
Invoking the Hellinger-Reissner principle and substituting Eqs. (3.46) into Eq.
(2.24), the following compact formulation is obtained:
[Kto]?. · tx}: = im — Ao}: (3-4v>
It is remarked that superscript indicates "visco-" and subscript indicates”
n—th" _
coefficient of the Fourier series.
[K(t)]; depends upon the time; it is the same as the [K]„ of Eq. (2.41) except
E is replaced by E,, the element force {F(t—At)} is the force of the last time step,
and
{F} = {f1, f2,....fg} (3.48)
where
f1 = „N,dA + „N,dS (3.49a)
f2 = fi-9 „N;dA + fte „N,dS (3.49b)
Description of viscoelastic solids 46
A
{6 = „N;dA + [{2 „NidS (3.49c)
{4 = — fg, „(t — At)NidA (3.49d)
{5 = — [gr 6 „(t — At)NidA (3.496)6
{6 = — I2, Z „(t — At)NidA (3.49{)
{7 = — [S6 „(t — At)N,dA (3.49g)
{6 = —([*22
„(t — At)NidA (3.49h)
fg = —4[QQZ
6 „(t — At)N;dA (3.49i)
Description of viscoelastic solids
l
47
I
From Eq. (3.49) the element force ls calculated from the last time step and it ls
affected by the past time history.
Description of viscoelastic solids 48
4.0 Algorithms ofsolutlon4.1
Statlc analysis
4.1.1 Static condensatlon .
ln the mixed finite element model, the stress components are given as nodal
variables. Hence, the assembly of the element stiffness matrices into the global
stiffness matrix results in stresses that are continuous across each interface be-
tween elements. For the multi-layered cylinder, the stress components 6,,,,6,,, and
6,,,, are discontinuous across the interface}
To circumvent this difficulty, these stresses are condensed out at the element
level so that the discontinuity of the stresses across the interface no longer exists.
To do this, Eqs. (2.41) or (3.46) are rewritten in partitioned form as
K K X F
li 11 12il_[{ 1}]=|¥{1}] (4.1)K21 K22 {X2} {F2}
where {X,} is the vector of continuous variables across the interface,
3 For the general loading (symmetric), as for axisymmetric loading 6,,, and 6,,, are the discontinuousvariables.
Algorithms ol' solution 49
{X1} = {Um· Ü6,, - Üzn - Fm - örßn -F,2r,}T (4-28)
and {X2} is the vector of discontinuous variables across the interface,
{X2} = {?6,, - azn ·<7z6,,}T (4-2b)
Hence the nodal stress vector for a given element becomes
K11{X1} + K12{X2} = {F1} (4-38)
K21{X1} + K22{X2} = {F2} (4-3b)'
From Eq.(4.3b) .{
{X2} = K2;({F2} ‘ K21{X1}) (4-4)
Substituting Eq.(4.4) into Eq.(4.3a),
{F1} (4-5)
results, or
[R]{X1} = {F} (4-6)
where
[X] = K11 K21 (4-78)
{F} = {F1} — K22 {F2}{
(4-7b)
Algorithms of solution S0
Eq. (4.7) is assembled as usual and is solved for the nodal variables after ap-
plying the boundary conditions. The discontinuous nodal stresses that were con-
densed out are then computed by using Eq. (4.4) at the element level. Since these
stresses are no longer nodal variables, they will be discontinuous between inter-
faces.
4.1.2 Modified Newton-Raphson method
If a rocket motor is composed of (1) elastic case and (2) viscoelastlc propellant,
then E, is a function of reduced time, and depends upon stress and temperature at
the current time step. A direct solution cannot be used in this analysis, hence anI
iterative solution technique is sought.
We can rewrite Eq. (4.6)as‘
{R} = [k]{x} — {F} (4.6)
Suppose that we know the solutions of Eq. (4.6) at the i-th iteration and are
interested in the solutions at the (i + 1)- th iteration in the t- th time step.
We expand Eq. (4.8) about the i- th solutions in TayIor’s series:
ö{R} 2 1 ö2{R} 2 2{RI+1}= {Ri} +——;lö{X}l+ ———.l6 {xi}+IÖ{X} 2 ö{x}2 (4.9)
= O
Assuming the second and higher—order terms in ö{X} are negligible, we can
write Eq. (4.9) as
* for convenience, the bar above K and F and subscript 1 are dropped
Algorithms of solution 5l
6{x}§ = — [i<$]"{Rf} (4.10)
where [K;] is the slope (tangent) of the curve {R} at the i—th iteration in the
t- th time step and
1 1 1{X};+; = {X}; + ö{X}; (4.11)
We iterate through Eqs. (4.10) and (4.11) until the ratio of the norms of incre-
mental variables and total variables satisfies
I IA(x)¥ I I< ro:.(4.12)II1x)1+1I I
where TOL is a convergence tolerance, a small value.
The above statements (from Eq. (4.8) to Eq. (4.12)) are called the ”Newton-
Raphson method" [41]. A geometric interpretation of the Newton-Raphson method
is shown in Fig. 4.
The Newton-Raphson method requires that the tangent matrix [KT] be calcu-
Iated at each iteration. This will result in massive expenses when a lot of degrees
of freedom are included; therefore occasional updates of the tangent are desirable.
The modified Newton-Raphson method updates the imbalance force {R} at each
iteration for a fixed time step while it keeps the tangent matrix [KT] fixed
(Fig. 5).
The [K;] is updated only at the beginning of each time step; the modified
Newton-Raphson method may need more iterations to reach an equilibrium point.
The solution is started by assuming a free state. iterations continue until the
convergence tolerance is achieved. The problem is restarted for the next time step,
Algorithms of solution 52
I= i..____II
I II II I
·I II I I .I I II I II I II I I
1 2u u uc
Flgure 4. Newton-Raphson method
Algorithms of solution 53
I=
' I. I I
I II |I 'II
"II 'II
u' uz Uc U
Flgure 5. Modlfled Newton-Raphson method
Algorithms of solution S4
assuming a zero state at the beginning of that time step. Thus a series of problems
is solved.
We may note that the main advantage of this method ls that it is not necessary
to update the stiffness matrix (tangent), [KT] , in a time step but requires more it-
erations.
The solution procedures are shown below for each time step.
1. At the beginning of each time step, the stress vector {6} from the
previous time step is given. For the initial time step (starting time
step), the stress vector {6(t—At)} denotes the initial stress state at
t= 0, indicated by {6°} . lt is customary to assume a stress-free state
at the start of the analysis; {6°} is usually set to be zero. ·
2. Assume the time-temperature shift factor a, is an average value,
and ls given by
af = (af + a{_At)/2.0 (4.13)
3. Calculate hereditary integral, using the recursive formula, Eq.
(3.27) and hereditary strains, {€(t— At)}.
4. Evaluate element stiffness matrices [K] and element forces
{F}„,=p{F}„p_, where p is the load factor that corresponds to the
time step under conslderation, and compute [KT] and {R}.
5. Assemble element tangent matrices [KT] and {R}, save [KT] and
its lnverse matrix for use in this time step.
6. Apply the boundary conditions on the assembled equations.
7. Solve the assembled equations.
8. Update the solution vector using Eq. (4.11).
9. Check for convergence.
Algorithms of solution S5
10. lf the convergence tolerance is satisfied, go to the next time step
and repeat steps 1 - 9 until the final time step is reached; otherwise,
compute {R} in step 4 and go on.
4.2 Translent analysis
The Hellinger-Reissner variational form is rewritten as
I-ID =
I-T2—TwhereT is the kinetic energy, ITD ls the H-R variational form for the dynamic prob-
lems, and H2 is the H-R variational form for the static problems.”
By using the same technique as for the static analysis, the typical equation be-
comes
EM]1S()+E1<J{x)=(1=) (4-15)
where [M] is called the mass matrix, and
mf = '·.p1v,~jdA = M§2 = M,j?° (4.16)
with p the density; other coefficients of [M ] are zero.
4.2.1 Free vibratlon
If the system is subject to zero external loads, assume that the solution vector
to Eq. (4.15) can be expressed as
Algorithms of solution 56
{X} = {X}6i“°' (4.17)
Substituting Eq. (4.17) into Eq. (4.15) yields
[K]{x)—w2l[M]{x}=o (4.18)
Rearranging
([1<] — A[M]){x} = o (4.19)
where A=w2 is the eigenvalue of Eq. (4.19) and w is called "circular frequency”
(radians per second).
In determining the frequencies (or periods) of the system, the displacement fl- ·
nite element method is usually used.
Using the same technique in the displacement finite element model, the
equation of motion is the same as Eq. (4.19) but {X} ls the displacement vector
{X} = {ur, ug, uz} (4.20)
In free vibration analysis of a solid of revolution, we are interested in the first
three terms, n = 0, n =1 and n = 2 (with m =1,
2,...).5 Fig. 6 depicts the geometric
interpretations.
Expanding Eq. (4.19) for n=0, n=1 and n=2 into a more explicit form, for
n = O
11 12 11 0 Ur( 21 22
_'l22 ){ }=0 (421)
KU 0 UZ
5 n’s correspond to circumferential modes and m's correspond to longitudinal modas
Algorithms of solution S7
I@ ü ir ‘·'
Ü * 0 Ü Q •2
·„ CIRCUMFERENTIAL
{ Ah moocsfl • 3 n rg • L
LONGITUOINAL_____, __ _ __ MOCESm • 1 m • 2 rn • 3
Figure 6. Some vlbratlon modes ol a hollow tube
Algorithms of solution 53
for n=1 and n=2
11 12 13 11Kü Kü MU Ü Ü Ur
21 22 23 22( KU KU KU — Ä. 0 MU 0 ){u6 }= 0 (4.22)
31 32 33 33
The coefficients of the stiffness and mass matrices of a displacement finite element
model are presented in Appendix B.
Even though the displacement finite element method is used, the big difference
between moduli of two materials in the rocket motor still incurs difficulty in solving
Eq. (4.19).‘
To circumvent this difficulty, an alternate method is suggested. Rewriting Eq.
(4.19) into another form to avoid the singularity of the stiffness matrix,
[(K + FM) - (A + I”)M]{X} = 0 (4.23)
wherein F is chosen so large that K + FM is not a singular matrix.
From Eq. (4.23) eigenvalues, A+ I", can be found; finally A can be determined
for different n .
By knowing the period of the system, a time increment for transient analyses
can easily be set (usually 2-]% period).
4.2.2 Newmark direct Integration method
There are several approximation schemes available for solving Eq. (4.15). The
most commonly used one is the Newmark direct integration method.
Algorithms of solution S9
In the Newmark direct integration method, the displacements and velocitles at
the (n +1) -th time step ( At1 = At, = .... = At ) are approximated by
{x}n+1 = {X}-, + [(1 — ¤)lY}„ + °l{X}n+1]At (4-24)
{x}n+1 = {X}-, + {><}„4¢ + [( g — ß){X}„ + ß{X}„+1J(At)2 (4-25)
where a and ß are the parameters that control the accuracy and the stability of the
scheme, and the subscript, n, indicates that the solution is evaiuated at the n-th
time step (i.e., time t=t,,). The choices oz = 1/2 and ß= 1/4 are known to result in
an unconditionally stable scheme, Substituting Eqs. (4.24) and (4.25) into Eq. (4.15), _
A A[K]{x}n+1 = {F}n+1 (4-25)
is obtained, where
[R] = [K] + a0[M] (4.27)
A . „.
{F}n+1 = {F}n+1 + [M](8o{X}„ + 81{X)„ + 82{X}„) (4-28)
and ao =1/(ßAt2), a, = aoAt, and az =1/2ß -1.
Combining Eqs. (4.24) and (4.25), the acceleration can be found:
81 {X}- — 82{Y}„ (4-28)
lf the solution {X} is known at the (n + 1) -th time step, then the first and sec-
ond derivatives of {X} can be calculated from Eqs. (4.24) and (4.29).
The solution procedures are summarized below for the (n + 1) -th time step.
Algorithms of solution 60
1. Compute the element stiffness matrices [K] and element mass
matrices [M] .2. lnitialize the starting conditions, i.e., {X},, {X}, and {X},.
3. Select the time increment At, parameters ot and ß.
4. Calculate [K] and {lg} by using Eqs. (4.27) and (4.28).
5. Assemble element stiffness matrices and mass matrices and apply
the boundary conditions.
6. Solve the assembled equations.
7. Calculate velocities and accelerations at (n+1)—th time step by
using Eqs. (4.24) and (4.29).
8. Repeat steps 1-7 until the final time step. ·
Algorithms of solution 61
5.0 Numerlcal examples and discussions
Responses of rocket motors are calculated under the following four loading
conditions : (Fig.7)
(1). Self weight -
(2). Two patch loads centered at 0 = 0° and 0 =180°' on the midspan of the rocket
motors.
(3). One line load ( centered at 6=0°) and one patch load (centered at 0=180°,
on the midspan of the rocket motors ).
(4). Two line loads centered at 0 = 0° and 8 =180°.
The loads can be expressed in terms of Fourier series (symmetric), i.e.,
f(x) = pA/1: + Z(2p sin(nA) cos(n8)/mr) cos(nx) (5.1)
where p= 1 psi. and A=1°. The following boundary conditions are applied : Mo-
tors are supported on fixed end diaphragms, i.e., at z = 0, z=l
'The area of patch loads is viewed as square with dimension of 0.15 in. ( 2° ) by 0.15 in. in currentexamples
Numerical examples and discussions 62
III. IpsiVü
,/ \ Z
I /@ P= 9#
P=9#f 2
#: —— ·W=
äI
Flgure 7. Force dlstrlbutlon ot the rocket motors A
Numericd Examples end discussions 63
U, = OUG = 0 (5-2)UZ = O
at the inner and outer surfaces (r= r, and r= ro ) (except prescribed surface(s))
6,(r„) = the applied load divided by the area of application
Ur(ri) = 0- Ur(ro) = 0
Ur 0(ri) = 0· Ur 6(ro) = 0
Ur z(ri) = 0* of z(ro) = 0
at the interface
u,(prope//ant) = u,(stee/) ·
u9(prope//ant) = u8(stee/) '
uZ(prope//ant) = uZ(stee/)5
6,(prope//ant) = 6,(stee/)( '3)
6, 8(propel/ant) = 6, 0(stee/)6, Z(prope/lant) = 6, Z(stee/)
Initial conditions :
{X}0 = O. (5.4)
{X)0 = 0
where subscript 0 indicates initial time ( t=t„ ). ln practice, one does not know
{X}; it must be calculated from Eq.(4.15):
(Si) = [M]”1({F} — [K]{X}) (5-5)
with the length of a rocket motor = 60 inches. Its mechanical and geometric prop-
erties are listed in Table 1.
Numcrical cxamplcs and discussions 64
Table 1. Mechanical properties of rocket motors
Material Air Propellant Steellayer 1 layer 2 layer 3Outer Radius (in) 1.875 4.3 4.4
Elastic Modulus (psi) 281.84 30,000,000Poisson's Ratio 0.49 0.25Density (lb/cubic in.) .0622. 0.3Weight/unit length (lb/ln.) _ 2.926 .820
Numcrical Examplcs and discussions 65
5.1 Statlc analysls wlth elastlc propcllant
In order to verify the program, comparison is made with the following three
cases.
(1).A hollow cylinder subjected to a uniform pressure (1. psi.) on the outer surface
(see Table 2). The exact solution (closed form) can be found on pages 70-71 [23].
The results of the finite element method converge to the exact solution when using -
a fine mesh.
(2). Isotropic single layer (steel) cylinder under two opposite line loads [3] (Fig.8).
Cederbaum and Heller [3] worked on the responses of a finite length, moderately
thick composite laminated cylinder due to dynamic loads by using the theory ofu
shells. They obtained solutions (displacement at mid-length as a function of cylinder
length) represented by ”triangIes" in Fig. 8; the solutions obtained from FEM are
presented by "stars” in the same figure. The two solutions are almost the same in
a shorter length range, but for the longer length range there exists some discrep-
ancy. The possible reason is that the ratio of width to length of a finite element is
very large.
(3). Responses of finite length cylinder (length = 60 inches) with those of plane-
strain problem (infinite cylinder) [8] under two opposite line loads (Table 3). Results
of the plane-strain problem [8] were obtained by using the semi—series solution (the
solution of stress function represented in a series form [23]). The displacement
solutions of FEM are in agreement with the solutions of Ref. [8], but not stresses.
Apparently the ends of a finite cylinder have a major effect on the stress solutions
in the mixed FEM.7
7 Radial and tangential displacements (inch), radial, tangential and shear stresses (psi.)
Nnnieiieel exeinples and discussions 66
Table 2. Comparlson ot solutlons (FEM and exact solutions)
i' ‘ arr 090
r=4.3 in. Exact solution -.9896 psi. -1.454 psi.FEM solution -1.031 psi. -1.465 psi.
r=4.4 in. Exact solution -1.0 psi. -1.444 psi.FEM solution -1.041 psi. -1.457 psi.
Numerical examples and discussions 67
0.00038A A _.,..--A—-—-·—·
0.00030
··s , .- *—· "” *‘ f I-
Z */ /•·—• Ü.ÜOÜ24 /‘·’ /l-< / .Z /L;] /
ä 0.00018/’ ·
u ^ 11/¤•<an17)
0.00012 » 1 .
E 1/
/
0.00008 //
¢
0.000000 30 80 120 180 300 800
LENGTH (INCH)LEGEND : A A A uam vom s0w.(s1u·:u. mr:.)=•= =•= ¤•= FINITEEIEIDIT SOLU.
Flgure 8. Dlsplacements of the cyllnder vs. length: at the mldspan ot the cyllnder
Numerical Examples and discussions 68
Table 3. Comparison between two methods
Ur Us *7n Gro Goa
F.E.M. .00025 in. .00012 in. .00552 psi .00567 psi .0040 psisolution
Closedform .00026 in. .00012 in. .00597 psi .00512 psi .0053 psisolution
Numerical Examples and discussions 59
Figs.(9), (10), (11), (12), (13), (14), (15) and (16) show the deformed shape and
response distributions of the rocket motors subjected to (1) self-weight (2) two op-
posite line Ioads (3) two opposite patch Ioads and (4) one line load and one patch
load.°
The responses (absolute maximum) under four different Ioadings are listed in
the Table 4.
5.2 Static analysis wlth viscoelastic propellant
lf the rocket propellant is treated as a viscoelastic material, with its mechanical .
properties listed in Table 5, the hereditary integral solution (Sect. 3.2) is used.
Figs. (17), (18), (19), (20), (21), (22), (23), (24) and (25) are the radial displace-
ments, radial stresses and tangential stresses on the interface at midspan (0 =0°)
due to (1) two opposite line loads, (2) two opposite patch Ioads and (3) one line load
and one patch load at different temperatures ( 35°, 45°, 55°, 65°, 75° and 85° F).
From the above Figs. (17-25), it is seen that the long term radial displacements
(t—>oc) will attain the elastic responses (Figs. 12, 14 and 16) no matter what the
temperature is. Further, it can be observed that the radial displacements increase
as the temperature increases; conversely, the stresses decrease.
' L'} ll?}$*2.3planeperpendicular to the longitudinal axis at midspan. .One line load - 1 psi. (9 lbs / (.15 in. x 60 in.)) one patch load
-400 psi. (9 lbs / (.15 in. x 15 in.))
and total weight of a cylinder (60 inches) is 224.77 pounds
Numeiical examples and discussions g 70
Table 4. Statlc responses
Ur U6 arr drä GOG _
Two patch .00043 in. .00014 in. .04467 psi .01077 psi .03758 psiIoads
One line .00040 in. .00012 in. .04613 psi .01055 psi .03908 psiand one patch load
Two line .00014 in. .00006 in. .00346 psi .00214 psi .00351 psi _Ioads
ISelf-weight .00025 in. .00025 in. .18830 psi .02705 psi .18030 psi
Numcrical Examples and discussions 7|
Table 5. Data for the vlscoelastic propellanf
81 1EU) = Eo +,7gEm @><¤( —Q;)
EO = 281.4E, = .19789 X 10+5E2 = .79896 X 10++EO = .25217 X 10++
O‘
E4=.11526><1O+4 _ ‘EO = .73456 X 10+5 (Ps")EO = .20414 X 10+5E, = .27178 X 10+5
EO = .86427 X 10+*5
T, = .333 X 10+**:2 = .333 X 10+5TO = .333 X 10-7
TO = .333 X 10+5·
TO = .333 X 10+*1, = .333 X 10+*TO = .333 X 10+5 1 ·
Numerical Examplcs and discussions 75
5.3 Dynamlc analysls
When the time duration of the applied load is very short, the propellant may be
considered to be an elastic material whose modulus is calculated using the appro-
priate terms of the Prony’s series of Table 5.
For comparison, the natural frequencies of a motor were calculated using (a)
the rest modulus (E = 281.4 psi.) and (b) the modulus corresponding to a loading
time of zero seconds (E = 32945.2 psi,). _
The calculations were carried out with the aid of the IMSL library routine
"DG2LRG" and the elastic computer program ”AXIF". Results are presented in
Table 6.‘
Based on Table 6, the time increment is chosen to be 0.001 seconds for the
steel case and 0.0018 seconds for the viscoelastic propellant.
5.3.1 Analysis with elastic materials
In order to investigate the responses of rocket motors, different kinds of force
histories are applied (Fig.26), i.e.,
(A) ramp-rectangular load ( with the time interval of ramp
3 -times the basic period )
(B) rectangular pulse with application time greater than the
basic period
(C) rectangular pulse with time 2 0.75 period
(D) triangular pulse with time 2 0.75 period
(E) sine-wave pulse with time 2 0.75 periodA
Force histories (C), (D) and (E) simulate the influence of impact.
Numeiical examples and discussions 73
Table 6. Frequencles of a rocket motor
Frequencles (Hz) (r = ät-)m = 1 m = 2 m = 3
> n = 0 98 108 124 _n = 1 105 110 120n = 2 134 136 145
(A). A cyllnder wlth steel case and propellant (E=281.4 psl) ‘
coFrequen_cles (HZ) (f = E;-)
m = 1 m = 2 m = 3n = 0 678 1015 1240n = 1 244 562 933n = 2 480 588 839
(B). A cyllnder wlth steel case and propellant (E=32945.2 psl)
Numcrical Examplcs and discussions 74
ln order to investigate the changes between static Ioads and dynamic Ioads, the
force history (A) is applied. Figs. (27), (28), (29), (30), (31), (32) and (33), (34), (35) are
the radial displacements, radial stresses and tangential stresses on the interface at
midspan ( 0 = 0° ) due to two opposite line Ioads, two opposite patch Ioads and one
line load and one patch load, respectively.
From Figs. (27), (30) and (33) the radial displacements° are increased to the
maxima ( i.e., the results due to static Ioads ) in proportion to the gradually in-
creasing ramp force. Beyond these values displacements oscillate sinusoidally with
the constant applied force ( 1 psi. ). Correspondingly, the radial and tangential
stresses‘° due to two opposite patch Ioads or one line load and one patch load (
reach their maxima when the applied forces become constant
( 1 psi. ), see Table 7."
Figs.(36) - (47) are the radial displacements, radial stresses and tangential
stresses on the interface at midspan (0=0°) due to two opposite line Ioads with
four different force histories (B), (C), (D) and (E).
Figs.(48) - (59) are the radial displacements, radial stresses and tangential
stresses on the interface at midspan (0 = 0°) due to two opposite patch Ioads with
four different force histories (B), (C), (D) and (E).
Figs.(60) · (71) are the radial displacements, radial stresses and tangential
stresses on the interface at midspan (0=0°) due to one line load and one patch
load with four different force histories (B), (C), (D) and (E).
°The radial displacements at the interface are almost the same as those at the outer surfaces.
1° The maximum radial and tangential stresses are not located at the loading points, but somewhereelse when two opposite line loads are applied (see Fig.12) .
ll The values in parentheses are the results due to static Ioads
Numerieal examples and discussions 75
A
Table 7. Comparlsons between static and dynamic loads (interface)
Two line .00014 in. - ·loads (.00014 in.)
Two patch .00043 in. .045 psi. .045 psi.loads (.00043 in.) (.04467 psi.) (.04613 psi.)
One line .00040 in. .036 psi. .040 psi.and one patch load (.00040 in.) (.03758 psi.) (.03908 psi.)
Numcrical Examplcs und discussions 76
For application of two opposite line Ioads, Figs.(36) - (47) indicate that the basic
period of the rocket motors is around 0.01 seconds. The "snap" phenomena in
Figs.(37), (40), (43) and (46) occur due to the effect of higher modes (second mode
mainly ). As for (1) two opposite patch Ioads and (2) one line load and one patch
load the basic period can be estimated to be 0.01 seconds from the displacement
vs. time plots ( Figs.(48) - (71) ).u
More interestingly, at the interface the stress responses are opposite to the
displacement responses (tension stress, negative displacement; compression
stress, positive displacement, i.e., the strains 6:,, at the outer surface are opposite
to those at the interface (see Tables 8, 9 and 10))."
From Figs.(36) - (71) the magnification factors ( ratio of maximum dynamic dis-A
placement to static displacement ) are listed in Table 11.
5.3.2 Analysis with viscoelastic propellant
In analyzing the dynamic problem of a rocket motor with viscoelastic propellant,
it is necessary to calculate the modulus of viscoelastic materials from a Prony’s
series at each time (step). Further, the forces due to the hereditary effect must be
updated at each iteration.
Three different kinds of force histories are used in this analysis (Fig. 72). The
total time Iapse of force histories (F), (G) and (H) is 0.18 seconds (100 steps), and
(F) rectangular pulse with a complete application time
(G) rectangular pulse with a shorter time (40 time steps)
(H) triangular pulse with a shorter time (40 time steps)
12 Strains at midspan (0 = 0°)
Numerical examples and discussions 77
Table 8. Two llne Ioads under force history (B)
radial strain 6,,
Time (sec.) Outer surface Mid - Interface0.001 .22465-6 -.20005-6 -.81965-50.002 .46905-6 -.35055-6 -.90375-50.003 .63955-6 -.31915-6 -.53485-50.004 .83205-6 -.30915-6 -.30805-50.005 .95225-6 -.36555-6 -.47385-60.006 .92125-6 -.41 145-6 -.23365-50.007 .66985-6 -.29875-6 -.24545-50.008 .43905-6 -.24425-6 -.79165-50.009 .32275-6 -.30545-6 -.10555-4 ·
0.010 .25635-6 -.26415-6 -.10065-4
Numcrical Examplcs and discussions 78
Table 9. Two patch Ioads under force hlstory (B)
radial strain 6,,
Time (sec.) Outer surface Mid - Interface0.001 .5161 E-5 -.2320E-5 -.7335E-40.002 .6524E—5 -.2813E-5 -.2608E-40.003 .5913E-5 -.2460E-5 -.4882E·40.004 .6724E-5 -.2662E-5 -.4188E-40.005 .66629-5 -.2659E-5 -.4738E-40.006 .7139E-5 -.2864E-5 -.4370E-40.007 .5993E-5 -.2431 E-5 -.2550E-40.008 .5977E-5 -.2546E-5 ·.7066E·40.009 .5725E-5 -.2609E-5 -.3534E-40.010 .5554E-5-.24635-5Numcrical
Examplcs and discussions 79
Table 10. One Ilne and one patch load under force history (B)
radial strain 6,,
Time (sec.) Outer surface Mid - Interface
0.001 .1078E-3 -.4758E-4 -.1538E-20.002 .1372E-3 -.5790E-4 -.5239E-30.003 .1247E-3 -.5165E-4 -.9505E-30.004 .1375E-3 -.5643E-4 -.7901 E-30.005 .12679-3 -.5268E-4 -.1014E·2
0.006 .1351E·3 -.5431 E-4 -.11445-20.007 .1229E-3 -.5038E·4 -.7307E-30.008 .1278E-3 -.5452E-4 -.1367E·20.009 .1207E-3 ·.5305E-4 -.5255E-30.010 .1226E·3 -.5385E-4 -.1430E-2
Numerical examples and discussions 80
Table 11. Magnlflcatlon factors vs. force hlstories
(B) (C) (D) (E)
Two patch 1.53 1.53 0.93 1.12loads
One line 1.38 1.38 0.95i
1.15and one patch load
Two line 1.86 1.86 1.29 1.57loads
Numerical Examplcs and discussions 8l
Figs.(73) - (81) are the radial dispiacements, radial stresses and tangential
stresses on the interface at midspan (8=0°) due to two opposite line loads with
three different force histories (F), (H) and (G).
Figs.(82) - (90) are the radial dispiacements, radial stresses and tangential
stresses on the interface at midspan (6 =O°) due to two opposite patch loads with
three different force histories (F), (G) and (H).
Figs.(91) - (99) are the radial dispiacements, radial stresses and tangential
stresses on the interface at midspan (0=0°) due to one line load and one patch
load combinations with three different force histories (F), (G) and (H).
It is noted that the responses under force history (F) (radial dispiacements,
stresses and tangential stresses) will converge to stable values which are the same .
as those in the static analysis with viscoelastic propellant at a specified time (t =
0.18 sec., see Table 12). Under force histories (G) and (H) the responses (radial
displacement, radial stresses and tangential stresses) of the viscoelastic propellant
will decay to "zero state" after the applied forces are removed at a specified time
step from the cylinder.
5.4 Summary
The analysis of a finite length, multilayered cylinder with viscoelastic propellant
due to static and dynamic Ioads by a mixed finite element model was presented.
This approach was used because the external loads induced boundary condi-
tions on stresses which could not be satisfied by the more frequently applied dis-
placement model. Because loads were expanded into Fourier series, the problem
could be treated as a symmetric one and this simplified the solution. A large
number of Fourier series’ terms were required.
Numerical examples and discussions 82
Table 12. Comparison between static and dynamic analyses
two-line Ioads
ur _ arr Ur0 090
dynamic -.00081 inch -.011 psi. .0147 psi. -.0142 psi.responses
static -.00082 inch -.015 psi. .015 psi. -.012 psi.responses
two-patch Ioads
ur arr ar0 G90
dynamic -.00030 inch -.236 psi. .108 psi. -.2175 psi.responses
static -.00031 inch -.25 psi. .11 psi. -.20 psi.responses
Iinelpatch Ioads
urdynamic-.00029 inch -.238 psi. .105 psi. -.2205 psi.responses
static -.00029 inch -.24 psi. .11 psi. -.22 psi.responses
Numerical Examples and discussions 83
Motors were supported on end diaphragms that permitted no displacements.
While this assumption is a simplification, it represents, simply supported motors
with end caps and nozzles that prevent radial displacement, reasonabiy well.
Numerical examples and discussions 84
.;._ DEFORHED SHRPEUNDEFORMED SHQPE
HRX. • .25E-B HCH
¤III“{Egg!}
1;;:::::}*** FääääääälläIIIIEEEHIIII!!!§”.!EEKIII\!.!!!.!!.!.}.jII
!.!g!.'-==}|II1!.!!.!g==«JIIt!‘!.!=„-}}}IFgääätnl[Ill}
Flgure 9. Deformed shape due to self·wetghtI
Numcrical Examplcs and discussions Y85
I. HDI!. 0l9L. t®'-‘G!£DI ä. MX. • .5ß-3 INZH2. Clälf. OI9\.. IECIVEDI ä. MX. • .52-3 It¤·t3. XIH. STRESB IOISTRIBJTIOJI E. MX. • .I䣕0 PSI4. SIM STES IDISTRIBJTICNI %. MX. • ZE-I PSI5. TRCETIR STRESS IOISTRIBJ &. MX. • .I&J5•O PSI
Figure 10. Dlsplacement and stress dlstrlbutlon due to self-weight :
along Interface at the mldspan of the cyllnder
Numcrical Exemples and discussions 86
‘——‘—..._.._ BääääääägäffäpgIAX. u . l4l§·3 IPO-!
Il!!!*
I!§=-'!!![/IIII!!.§¤'=!!/III
iII nlEäE§·'=äiIälIIÜIÜÜIHIII
lll_!!!Ll1I|II!!!!!.!„ä§I||I!!.!!!=ägI||l!.!.!!l=§iI||IllllligilllI!.!=‘|='i_!_\i||I!!!l=il{f!|tlgzgztägil__!‘!.
-“-laFigure 11. Deformed shape due to two opposlte Ilne Ioads
Numerical Examplcs and discussions 87
I. ÄIR, Ulf,. fE®‘EDI MX. • .l4l§—3 IPO!2. Cl$.H·'. UI?. IEGUEII ä. PRX. • .§l4E·4 IPO-!
3. HDIGI, Sfüä IUISTRIBJTIGII MX. « .34Sä-2 PSI4. S-EM STESS IUISTRIBHIUGI ä. MX. ¤ .2I37E·2 PSI
6. Thiülfl. STRESS fUlS‘|’RlB.l E. MX. n .§U’7E-2 PSI
Flgure 12. Dlsplacement and stress dlstrlbutlon due to two llne loads :
along Interface at the mldspan of the cyllnder
Numerical Examplcs and discussions 88
_.........——- OEFORHED SHRPE_____._.,. UNDEFORMED SHQPE
IQX. ¤ Ai-3 l©·|
-l!!|!;!¤*!.!r!;=„llnm-„-.,;I1§§§:=:.!n|!!!?=!.!Ha|
lllnggél/IllIuuuamuI!Il!!!äMI!!!.".:g;\\I|l· I!!!!Eg_g_\\\||I!!!‘I-•g\Y|I§‘¤‘•%ä¥'="·‘="“%4.-nuilil
____¤\
Flgure 13. Deformed shape due to two opposlte patch Ioads
Numcrical Examples and discussions 89
l. ßßll Ol§.. tEFU!€Dl ä. MX. • .43§·3 HCH2. CIE)!. Ol?. lßüüfbl &. MX. • .t377E·3 HCH3. ill NESS IOISTRIBJTIGII ä. MAX. • .4467E-l PSI4. 8·£M Nüä IOISTRIBJTIQU ä. MAX. • .lO77E-I PSIG. TF!~®¢Tll NESS tDlSTRl8.l &. MAX. • .375ä% PSI
C) Ö\4/
Flgure 14. Dlsplacement and stress distrlbutlon due to two patch Ioads
Numcrical Examples and discussions t 90
.........-..-. CEFORHED SHAPE-_-. UNCJEFORMED SHAPE
MAX. • .404§·3 IPO!
ll---- -l--111-!!I !!§§„i11-g= 1;-.2,*.11!!!- f!1;1¤¤·=:=·
— e _!I|- s '
1'ällliäI‘*5ä""'
A IIIIÜHII|!I!!!l¥N||
_ W-- =
I!!!.‘:.䧧1\i||-1 1 -!!¤•-\\\||-Ü-I Il|!'I=i'i\\”||
I-!!i=iE1|II-Ü--1 1 ¤§1—¤¤¤-E11"-——-—~11---
!.!.!=.*1si--- --;;.1
---
----11
Figure 15. Deformed shape due to one line and one patch combination
Numerical Examples and discussions 9l
I. EINE DI$'L. IEGOEDI &. Mx. • .404I·3 lr¤·I2. CIC)?. DISK. (EIDE) E. MX. • .I2£~3 HCH3. XIII. ST55 IOISTRIHJTIDU &. MX. • .48II-I PSI4. SEM STESS IOISTRIBJTIOII fü. MX. u JGEE-I PSI6. Thßül ST§ IOISTRIBJ ä. MX. • BSE-I PSI
® ®·Flgure16. Dlsplacement and stress dlstrlbutlon (llne and patch)
Numerical Examples and discussions 92
T•nve•·aIu·6lFl: 6-5-5 86 :1-E1--13 75 Q-9-G 55B-B--9 66
«••-+•—·••46 ·|—l——·l- 35
0.00000
-0.00001-·‘·l-0.00002
:‘Xx-0.0000:1.} Q-0.00004 ¤ \\\
.-0.00006 —‘\ \\\\\\
-̂0.00006.:1 Xca-0.00007 - Öxö
•" -0.00008 \gz" äxN\apnl \
\Q -0.00000s\
\-0.00010 Ö‘*>¥i\-0.00011
‘—}Q\
-0.00012 Qä .-0.0001:1>
F
-0.00014‘-0.00015
3.61-10 3.61-9 3.61-8 3.61-7 3.81-5 3.61-3 3.61-1 36. 3.61+3 3.51+5 3.61+10
Time (sec.)
Flgure 17. Radlal dlsplacement vs. tlme (two Ilne Ioads): vlscoelastlc propellant
Numerical examples and discussions 93
Temperal1¤·e<Fl: a-¤—¢ 86 cr-a-13 T6 6-6-G 669-9--9 56 ·••—•••-••· 46 +—•—•- 36
0.00
. ?
-0.01 //
. 6 /, / >'/
-0.02 _ //,6,;//,/.·// .
Ö -0.03 0///
m -0.04 r Ü/m/.2
-0.06 ,' /.6 .. //’M
·"-0.06 "i,;p/
-0.07 ,Ü,’/
lr-0.06_ 6. ' I
-0.00BJI-10 BJI-9 BJI-! BJB—7 3.6I—5 3JI-B BJI-1 36. BJI+B BJI+5 3JI+10
Time (sec.)
Flgure 18. Radlal stress vs. tlme (two llne loads): vlscoelastlc propellant
Numerical examples and discussions 94
85 D.a.G 75 €__G_G66
9--9--B 56 «••—+—••• 45 +-1-1- 36
0.00
— 1 VE jr/I
./// i
-0.01 { f /
lg?/
I,
„,5 -0.02 ,7/ 0/gg I/1 7¤- · /7 .\•/
I/1:3 -0.03 , IWU] '1',I[f
gv 7/¤ 1 I¤¤ ,· 1 //
*0-04I,1,
[ /”
I,I
-0.06 _ ,7/. '; {;/
-0.063-II-10 3.6]-9 3.6]-8 3.6K—7 3.6]-5 3.6]-3 3.6]-1 36. 8.6]+3 $$+5 3-6]+10
Time (sec.)
Flgure 19. Tangentlal stress vs. time (two Ilne Ioads): vlscoolastic propellant
Numerical examples and discussions 95
T•rm•r•tu•(Fl: 16-a—¢ 5 D-EI--I3 76 G-G-G Ü9-·B--9 56 «••-•••-••• 46 -•-•—•- 5-0.00006-0.00000-0.00010
ö-0.00011 . ~?§\-0.00012 =.\ _-0.00016
‘\-0.00014 -,x-0.00015*0}
-0.00016 *.(\\-0.00017 \\-0.00018 X}
\-0.00019 ¤k\-0.00020 11 \\\-0.00021
‘
,.1 -0.00022 9}.¤ -0.00028 ‘,\Ää -0.00024 *.1
-.-. -0.00025*0Y-0.00026
‘—lS\\ _·-• -0.00027 =v$* -0.00026 ‘
-0.00029 *0}-0.00030 ‘.}\-0.00031 QD-0.00032-0.00033T-0.00064‘-0.00035 }\§ ·-0.00086 gu °-0.00037 . _ Ä _-0.00088T-0.00039
‘~~1 L-0.00040 ·-0.00041 · _-0.00042 · _ 1-0.00046 . .-0.000446.4:-10 6.4:-• 6.4:-6 6.4:-7 6;-6 64:-: 6;-1 64. 64:+6 6;+4 I.I+l|Time (sec.)
Flgure 20. Radlal dlsplacement vs. tlme (two patch Ioads): vlscoolastlcpnopellant
Numerical examples and discussions 96
Tenpcrpturs F): 6-6-6 86 ¤·-E1--13 76 G-G-G 65·- 9-9--9 66 «••-•••-••» 46 ·•—•—+ 36
’ ‘/1 2/
;;7/°VI1// / '
rv " [ //nä ·}' /
2 · //Ü
1,//2. ,-1
'P 1 IIl I /r-1 ' 1 /
E'lrl /)‘( /’° Bi 1 /
$2 ·' '//-2 I.;·’;’ I
-3
l
3.8]-10 3.6]-9 3.6l-6 3.6]-7 3.8]—5 3.6]-3 3.8]-1 38. 3.8]+3 3.8]+5 3.8]+lI)
Time (sec.)
Flgure 21. Radlal stress vs. tlme (tvvo patch Ioads): vlscostastic prop•IIant
Numerical examples and discussions 97
T•mperaIu·•<F): *-1-* 85 g.-E;..¤ 75 g-g.G 669--B--9 66 —•••—-•••-••~ 46 +-1-1- 36
0.0
-0.1 J, %-0.2 [y 'Ä
- //-0.3
·”6 /
,é’-0.4_*10.6 I 7/-o_g
I/I? 4
ZT -0.7,/,0/ '
U) 1/jl/-• -0.8 [lll //Q -0.0 7/ 7cu '1 7
-1.0 ,1]m -1 1
[ [
C: -1.2 /'/1 pÄ -1.6 ,7// /*/
I I /-1.4 P1 //-]__5
-
"„;’}
E
-1.6
-1.7-1.8-1.9
-2.0 —lßl-10 3.6l-9 66:-6 6.6:-1 6.6:-6 66:-6 6.6:-1 66. 66:+6 6.6:+6 6.6:+10
Time (sec.)
Flgure 22. Tangentlal stress vs. tlme (two patch Ioads): vlscoelasttc propellant
Numerical examplu and discussions 98
T•¤\P•f¤¤¤¤¢Fl= ü-—a—& 86 5-5--43 75 g-g_G 66
-0.00000 -——-0.00010
““—- 2 x-0.00011-0.00012 ”‘Ö§\\-0.00013 E \-0.00014·-0.00015 ‘,x
\-0.00018‘.‘\-0.00017
‘·‘Ö0
-0.00018 Q } .-0.00010
‘\ \\\
-0.00020-0.00021 \‘}\
^ -0.00022‘—x\
rd \ \\0 -0.00023 ~„\-0.00024 MQ
Y -0.00025‘ \\
TL -0.00026 Q}E -0.00027 ·5}\Q -0.00026
‘.}
-0.00029‘~§&
-0.00030‘\-0.00031 T Q
-0.00032-0.00033Q}
-0.00064-0.00035‘
-0.00066 · _-0.00037
“‘.j
-0.00038 .-0.00039_-0.00040 * _-0.00041
3.8]-10 3.8]-9 3.8]-8 3.8]-7 3.8]-5 3.8]-3 3.8]-1 38. 3.8]+3 3.8]+5 3.8]+1.0
Time (sec.)
Flgure 23. Radlal dlsplacement vs. tlme (llnelpatch Ioads): vlscoelastlcpropellant
Numerical examples and discussions 99
T•m¤••‘¤u•u=1:6-e-e 86 a-a-6 75 6-9-6 669--9-9 66 •••—•••-••• 46 -•—•—+ 5
O e $
r .,;’ L /
427/°(, /
11[ /
Ö -1 ,1 j/ ·/ /B. . 1 [ jsa l/ .
I
gg1,]s-.
,'/4-)1U1•-1‘ l
d/Iy//'U 1B?.-2 ,1 ß
1 I Q ~.
-3$6]-10 $6]-9 3.6]-8 3.6B-7 3.6]-5 $6]-8 3.6]-1 36. $6]+3 $6]+5 $$+10
Time (sec.)
Flgure 24. Radlal stress vs. tlme (llnelpatch l¤ads): vlscoelastic propellant
Numerical examples end discussions IW
T•tm¢&lu‘•(F): g-g-5 85 5.-5.63 76 6-6.6, 659**9*9 56 ••-*—••• 46 +—•—+ 36
0.0 _•O.1 ‘·-0.2 A:-0.3 I,}1/-0.4 I4}?-0.6 Ié-0.6Ii'? _
T5 —<>·7 :I/1/m 1 I¤-I -0.8
-0.9 "[ //G) H I
5 -1.0 I I //ui -1.1 ,"¢/Q 1l
l / »an -1.2 ,//1 p
-1.4 xn I //
-1.5-1.6
-1.7
-1.6
-1.9
-2.06.6:-10 6.6:-0 6.6:-6 6.6:-1 6.6:-6 6.6:-6 6.6:-1 66. 6.6:+6 6.6:+6 6.6:+16
Time (sec.)
Flgure 25. Tangentlal stress vs. tlme (llnelpatch Ioads): vlscoelastlc pmpollant
Numerical examples and discussions l0l
5cg 43 sQ 2ß 1
00.00 0.01 0.02 0.08 0.04 0.05 0.08
Porc• history (A): time (see.)
A 5-* 4
8•
2 22 1
Ö
0.000 0.005 0.010 0.015 0.020 0.025 0.080
Force history (B): time (sec.)A 5
E- 4**
3
Q 2Q 1
00.000 0.005 0.010 0.015 0.020 0.05 0.080
Force history (C): time (sec.)
A 5
K 4v 8Q 2Q 1 .
00.000 0.005 0.010 0.015 0.020 0.025 0.080
Force history (D): time (sec.)
A 5° 4
Ä 8Q 2£_ 1
° ,0.000 0.005 0.010 0.015 0.020 0.025 0.030
”.
Force history (E): time (sec.)
Figure 26. Force hlstorles
Numerical Example: und discussions mz
0.00000I · A
-0.00001
-0.00002
-0.00003
-0.00004
Q -0.00006 _
-0.00006 ·
4-9cl -0.00007G)
-0.00008O
-0.00009 .~
U)-•-•Q -0.00010
-0.00011
-0.00012
-0.00013“°°l‘ AAA-0.00015
0.000 0.005 0.010 0.015 0.020 0.025 0.030 0.035 0.040 0.045 0.050 0.055 0.060
Time (sec.)
Flgure 27. Radlal dlsplacement vs. tlme (two llne Ioads: Force (A)) .
Numerical examples and discussions l03
0.0014
0.0012
0.0010
0.0008
0.0006
0.0004
0.0002
0.0000
"T -0.0002‘5Sgt. -0.0004
‘ _
-0.0006U)Q) -0.0008isU) -0.0010
E -0.0012-.-«U -0.0014Q
Q:} -0.0016
-0.0018
-0.0020
-0.0022
-0.0024
-0.0026
-0.0028
-0.0030
-0.00320.000 0.005 0.010 0.015 0.020 0.025 0.030 0.035 0.040 0.045 0.050 0.055 0.060
Time (sec.)
Flgure 28. Radlal stress vs. tlme (two llne Ioads: Force (A))
Numerical examples and discussions [O4
0.00160.00150.00140.00130.00120.00110.00100.00090.00080.00070.00060.00050.00040.0003
"T 0.0002·5; 0.0001gg, 0.0000-# -0.0001
·
cn -0.0002Q -0.0003gg •'Ü.ÖOO4Ü; -0.0005
-0.0006
QS-0.0007
un -0.0008gg -0.0009Q -0.0010E -0.0011
-0.0012-0.0013-0.0014-0.0015-0.0016-0.0017-0.0018-0.0019-0.0020-0.0021-0.0022-0.0023-0.0024-0.0025
0.000 0.005 0.010 0.015 0.020 0.025 0.030 0.035 0.040 0.045 0.050 0.055 0.060
T1me (sec.)
Figure 29. Tangentlal stress vs. time (two line Ioads: Force (A))
Numerical examples and discussions105
0.00001I
-0.00001
-0.00003
-0.00005
-0.00007
-0.00009
-0.00011
-0.00013
-0.00015g -0.00017 _
I -0.00019-0.00021
E -0.00023
ä -0.00025E -0.00027
ä -0.00029-0.00031
-0.00039
-0.00035
-0.00037
-0.00039-0.00041 A-0.00045 '
'0.0000.005 0.010 0.015 0.020 0.025 0.030 0.035 0.040 0.045 0.050 0.055 0.060
Time (sec.)
Figure 30. Radlal dlsplacement vs. tlme (two patch Ioads: Force (A))‘
Numerical examples and discussions |()6
0.00I '
-0.01 °
rx -0.02.,.1 .03Q, .
~.«0)03
2U -0.03U) .•—•G-•-«'Ußns -0.04
-0.05
-0.08
0.000 0.005 0.010 0.015 0.020 0.025 0.030 0.035 0.040 0.045 0.050 0.055 0.080
Time (sec.)
Figure 31. Radial stress vs. tlme (two patch Ioads: Force (A))
Numerical examples and discussions l07
0.000-0.001-0.002 ,-0.003-0.004-0.005-0.008-0.007-0.008-0.009-0.010-0.011-0.012-0.013
A -0.014. -0.015
-0.016_
gl -0.017ar -0.018m -0.019m -0.020$5 “8·3ä$gf; -0.02:;_ -0.024Q -0.025¤0 -0.026:1 -0.0276 -0.028E- -0.029
-0.030-0.031 r-0.032-0.033-0.034-0.035
’
-0.038-0.037-0.038-0.039-0.040-0.041-0.042-0.043-0.044
0.000 0.005 0.010 0.015 0.020 0.025 0.030 0.035 0.040 0.045 0.050 0.055 0.080
Txme (sec.)
Flgure 32. Tangentlal stress vs. tlme (two patch Ioads: Force (A))
Numerical examples and discussionsl08
Eäoo:3:0000:8:30300
;g¢gg22;
,3
zgiäoääi
gg
"0'oooov
Ü
:3
:g'oooo6
·,;
·<>Ü333iä
q;
-oÜ00011
E
"0'oomz”0‘ooo1s
°’
"0‘ooo
4
äéäläääää
äjäsxsia
’0°ooo2o
-0.00021'0'ooozz
-0.00023-0.00024-0.00025jg‘00g26'0:ggoä;'0'oooag"0'oooso:363061”0Ü8oggä
-0.00034-0.00035"0'ooose0'ooosv
‘„3„i¤
843
41
Ü
‘o
F
00
Ü9U
or
.0
N
e
0
U
3
5
me
3
O
r
-
ÜO
ic
1
a
O
¤R
Cxa
adl
0.0
mplc
al d
15
5 an
Is
O·O
d
pl
20
d8
is
C
O
cussgo
emen
0;*25
ns
t
iulo
s
Q
30
tim
(560-0
e (|
C
js
In
·
0
‘o
BIP
40
atc
o
.
n
04
Io
5
a
o
d
.0
s
6
Z F
0
00 0])
°°
loo
0.00
-0.01
rx -0.02-.-E .U)¤«Q
m .
3IH -0.0E/J
'U .Q .
nc -0.04
-0.05-
-0.06
0.000 0.005 0.010 0.015 0.020 0.025 0.030 0.085 0.040 0.045 0.050 0.055 0.060
Time (sec.)
Flgure 34. Radial stress vs. tlme (line/patch Ioads: Force (A))
Numerical examples and discussions H0
0.001i
-0.001
-0.003
-0.005
-0.007 _
-0.009
-0.011
-0.013
-0.015g, -0.017
gz .m -0.019 ·U}O -0.021$4ä -0.023 · _
qj -0.025
EF -0.027
g -0.029
-0.031
-0.033
-0.035
-0.037
-0.039
-0.041
-0.043
-0.045
0.000 0.005 0.010 0.015 0.020 0.025 0.030 0.035 0.040 0.045 0.050 0.055 0.060
Time (sec.)
Flgure 35. Tangentlal stress vs. tlme (Ilnelpatch Ioads: Force (A))
Numerical examples and discussions lll
0.00000-0.00001-0.00002-0.00008-0.00004-0.00005-0.00006-0.00007-0.00006-0.00000
Q -0.00010_
•-•
" -0.00011+9Q -0.00012
_
-0.000184; -0.00014
2 -0.00015.-4$.1 -0.00016
E -0.00017-0.00016-0.00019-0.00020-0.00021-0.00022-0.00028-0.000%-0.000%-0.000%
0.000 0.005 0.010 0.015 0.020 0.0% 0.080
Time (sec.)
Flgure 36. Radlal dlsplacement vs. tlme (two line Ioads: Force (8))
Numcrical Examplcs und discussions H2
0.04
0.03
0.02
"TE 0.01Q-«äf
IDU)
2 .Q 0.00
· U)•—•6••¤|
E -0.01Q
-0.02
-0.08
-0.040.000 0.005 0.010 0.015 0.020 0.025 0.080
Time (sec.)
Flgure 37. Radlal stress vs. tlme (two llne Ioads: Force (8))‘
Numcrical Examplcs und discussions U3
0.04
0.03
0.02
A
'G 0.01Ö1~«WW0 .g 0.ooV U)
aiE°G -0.01 l[·•
-0.02
-0.03
-°o•°‘I
0.000 0.005 0.010 0.015 0.020 0.025 0.030
Time (sec.)
Flgure 38. Tangentlal stress vs. tlme (two llne Ioads: Force (8))
Numcrical Examplcs und discussions II4
0.00022
0.00020
0.00018
0.00016
0.000140.00012
0.00010
0.00008
fg 0.00006ä 0.00004 _
C'- 0.00002
Ei 0.00000 _E -0.000023 -0.000042 -0.00006
E -0.00008Q -0.00010
-0.00012
-0.00014-0.00016
-0.00018
-0.00020
-0.00022
-0.00024
-0.00026 '
0.000 0.005 0.010 0.015 0.020 0.025 0.030
Time (sec.)
Flgure 39. Radlal dlsplacement vs. time (two llne Ioads: Force (C))
Numerical Examples und discussions IIS
0.05
0.04
0.03
0.02
"T-5 0.01¤— .\/
0.00
E.
m -0.01PIG!•!¤|
E -0.02N
-0.03
-0.04
-0.05
-0.000.000 0.005 0.010 0.015 0.020 0.025 0.030
Time (sec.)
Flgure 40. Radlal stress vs. tlme (two Iine Ioads: Force (C))
Numcrical Examples and discussions II6
0.05
0.04
0.08
0.02
A'E 0.01 IßaQ
So.ooQDI-«
ü_ -0.01Q3¤0GQ -0.02E··•
-0.08
-0.04
-0.05
-0.080.000 0.005 0.010 0.015 0.020 0.025 0.080
Time (sec.)
Flgure 41. Tangentlal stress vs. time (two llne loads: Force (C))
Numerical Examples uid discussions ll7
0.000170.000160.000150.000140.000180.000120.000110.000100.000090.000080.00007
A 0,00006A 0110005
ä 0@0004•—• 0.00003" 0.00002E 0.00001 _0 0.00000E -0.00001QD -0.00002
. Q -0.0000:;E -0.00004m —o§oo00sE -01)0006
-0.00007-0.00008-0.00009-0.00010-0.00011-0.00012-0.00018-0.00014-0.00015-0.00016-0.00017-0.00018
0.000 0.005 0.010 0.015 0.020 0.0% 0.080
Time (sec.)
Flgurg 42. Radlal dlsplacement vs. tlme (tvvo·|Ine Ioads: Force (D))
Numcrical Exampics and discussions N8
0.05
0.04
0.03
0.026
-•-iID¤. 0.01\J .tb
2Q 0.00(IJ•—•
.9.QM
-0.02
-0.03
-0.04
-0.050.000 0.005 0.010 0.015 0.020 0.025 0.080
'I‘ime (sec.)
Flgure 43. Radlal stress vs. tlme (two llne Ioads: Force (D))
Numcrical Examples and discussions H9
0.05
0.04
0.08
0.02ex~•-IEfg 0.01U)ID
2 .Q 0.00U)
gl, .¤ -0.01GSE•¤
-0.02
-0.03
-0.04
-0.050.000 0.005 0.010 0.015 0.020 0.025 0.080
Time (sec.)
Flgure 44. Tangentlal stress vs. tlme (two Ilne Ioads: Force (0))
Numerical Example: and discussions l20
0.000200.000190.000180.000170.000160.000150.000140.000180.00012
. 018*%%0.000093.00003
Q 0:00006U 0.00005g 0.00004•-• 0.00003‘·’
0.00002E 0.00001 -
- 0.00000
ä -0.00001Q -0.00002
-0.00008Q -0.00004«-• -0.00005¤« -0.00006.2 -0.00007Q -0.00008
-0.00009-0.00010-0.00011-0.00012-0.00018-0.00014-0.00015-0.00016
- . 11*%%-0.00019-0.00020
I3:%§é0.000 0.005 0.010 0.015 0.020 0.0% 0.080
Time (sec.)
Flgure 45. Radlal dlsplacement vs. time (two line Ioads: Force (E))
Numcrical Examples und discussions l2l
0.05
0.04
0.03
0.02A-.4lbBe 0.01~.«U1In .GJ‘3 0.00U).-4.9,.¤ -0.01G
Dt'.-0.02
-0.03
-0.04
-0.050.000 0.005 0.010 0.015 0.020 0.0% 0.030
Time (sec.)
Flgure 46. Radlal stress vs. time (two Ilne Ioads: Force (E))
Numcrical Examples und discussions 122
0.05
0.04
0.08
0.02»·~-.-Z3_/
0.01
wIn '
0.00U)
6an -0.01dl
[··•-0.02
' -0.08
-0.04
-0.050.000 0.005 0.010 0.015 0.020 0.025 0.030
Time (sec.)
Flgure 47. Tangentlal stress vs. time (two llne loads: Force (E))
Numericsl Examples and discussions l23
0.0000
-0.0001
-0.0002/\•¤Q .GE 00086 °° .G
ä8G -0.0004
•-I
¤«rn••¤l
Q
-0.0005
-0.0008
-0.00070.000 0.005 0.010 0.015 0.020 0.025 0.030
Time (sec.)
Flgure 48. Radlal dlsplacement vs. tlme (two patch loadsz Force (B))
Numerical Examplcs and discussions l24
0.170.180.150.140.180.120.110.100.090.080.070.080.05
8*82.Q 0.02
py 0.01 ,Q 0.00‘-'
-0.012,2 ‘3·3§ V°’
-0.04 V:3 -0.05UJ -0.08 .•-n -0.07
-0.08•¤ -0.09 _q -0.10Q -0.11
-0.12-0.18 .-0.14
"
-0.15-0.18-0.17-0.18-0.19-0.20-0.21 .
-0.22-0.28-0.24-0.25-0.28
0.000 0.005 0.010 0.015 0.020 0.025 0.080
Txme (sec.)
Flgure 49. Radlal stress vs. time (two~ patch Ioads: Force (8))
Numcrical Examples and discussions 125
0.160.150.140.130.120.110.100.090.080.070.060.050.040.03
Q 0.02m 0.01
0.00-0.01 .
$ -0.02-0.04
ä -0.06, -0.06 r0:-0.07¤¤
-0.00 .Q -0.001E- -0.10 ‘
-0.11-0.12 .-0.13-0.14-0.15-0.16-0.17-0.18-0.19-0.20-0.21-0.22-0.23-0.24
0.000 0.005 0.010 0.015 0.020 0.025 0.030
T1me (sec.)
Flgure 50. Tangentlal stress vs. time (two patch loads: Force (B))
Numerical Examples and discussions IZ6
0.0004
0.0008
0.0002
0.0001
6
fa 0.0000¤•-4
W.?
4 ¤0.0001
äCD ..Q
0.0002
NE.m -0.0008•-•G
-0.0004
-0.0005
-0.0008
-_
700000.000 0.005 0.010 0.015 0.020 0.025 0-030
'1‘ime (sec.)
Flgure 51. Radlal dlsplacement vs. tlme (two patch loads: Force (C))
Numerlcal Example; and discussion;U
|Z7
0.8
0.2
0.1
0.0—
-0.1A.é _ _
-0.2~..«IDU9 -0.8
r E'
VJ -0.4•—•G••¤I
'¤ -0.6GM
-0.8
-0.7
-0.8
-0.0
-1.0
0.000 0.005 0.010 0.015 0.020 0.025 0.080
Time (sec.)
Flgure 52. Radlal stress vs. time (two patch loads: Force (C))'
Numerical Examples and discussions 128
0.3
0.2
0.1
0.0‘
-0.1A
I
~..¢
3Q -0.3 U;„I••-JV) -0.4Q)¤D¤ -0.5Q .r Ä-0.8
-0.7
-0.8
-0.9
-1.0
0.000 0.005 0.010 0.015 0.020 0.025 0.030
'I‘ime (sec.)
Figure 53. Tangentlal stress vs. time (two patch Ioads: Force (C))
Numcrical Examples and discussions I29
0.0004
0.0008
0.0002
E 0.0001U
.
¤•-«Q0 0.0000 ·G
äS -0.0001
E
6•-IG4ID¤•d
Q -0.0002
-0.0008
-0.0004
-0.00050.000 0.005 0.010 0.015 0.020 0.025 0.080
Time (sec.)
Flgure 54. Radlal dlsplacement vs. tlme (two patch loads: Force (D))
Numerical Example: and discussion: Z I30[
0.140.130.120.110.100.090.080.070.06
A 0.050.04 ·
g, 0.08S-!
2 SIZ?Q; .3 0.00 _V U3 -0.01
*3 -0.02$8 -0.0s ’Q -0.04
-0.05-0.06-0.07-0.08 .-0.09-0.10-0.11-0.12-0.13
-0.14
0.000 0.005 0.010 0.015 0.020 0.025 0.030
Time (sec.)
Flgure 55. Radlal stress vs. time (two patch loads: Force (0))
Numcrica! Examplcs and discussions I3!
0.130.12
0.11
0.100.09
0.080.07
0.080.05
fg 0.04‘•’
0.02. 3 0.01
0 00i
t
V? -0.01ga -0.02
-0.03·‘ ‘
.2 -„..„l-0.05-0.08-0.07
-0.08-0.09
-0.10-0.11
-0.12-0.13
0.000 0.005 0.010 0.015 0.020 0.025 0.030
Time (sec.)
Flgure 56. Tangentlal stress vs. time (two patch Ioads: Force (D))
Numcricsl Examples and discussions Ü:
0.0004
0.0008
0.0002
E o.ooo1OG•-xr
— -•-P 0.0000G
ä .-0.0001
6•·-•G-U3••-Q -0.0002 *
-0.0008
-0.0004
-0.00050.000 0.005 0.010 0.015 0.020 0.025 0.080
Time (sec.)
Flgure 57. Radlal dlsplacement vs. time (two— patch Ioads: Force (E))
Numerical Example; and discussion; Y l33
0.140.130.12
0.110.10
0.090.080.070.06
"? 0.060.04
"’0.03
IDcn 0.02
0 01V3 0.00•-•G -0.01
0 O2-0.03
-0.04-0.05-0.06-0.07-0.08-0.09-0.10-0.11-0.12
0.000 0.005 0.010 0.015 0.020 0.025 0.030
Time (sec.)
Flgure 58. Radlal stress vs. tlme (two patch Ioads: Force (E))
Numcrical Examplcs und discussions I34
0.18
0.12
0.11
0.100.090.080.070.06
A 0.05
ä Ziä;szm 0.02 ~g 0.013 0 00. U) ·:;.:3*2;ä
-0:08E-
-0.04-0.05-0.06 '
-0.07-0.08-0.09-0.10-0.11 ~ .
-0.120.000 0.005 0.010 0.015 0.020 0.025 0.080
Time (sec.)
Figure 59. Tangentlal stress vs. time (two patch Ioads: Force (E))
Numcrical Examplcs and discussions ÜS
0.0000
-0.0001
Q -0.0002UQI-l~.«-6->
ElE -0.0005GJOGl-IG-•tn~•-l
Q ~·0.0004
-0.0005
-0.00060.000 0.005 0.010 0.015 0.020 0.025 0.080
Time (sec.)
Flgure 60. Radlal dlsplacement vs. tlme (Ilnelpatch loads: Force (8))”
Numerical Example: end discussion: I36
0.170.180.150.140.130.120.110.100.090.080.070.080.050.04
Q 0.03m 0.02¤« 0.01T; 0.00
-0.01-0.02· N -0.03
ä -0.04 ~Q -0.05
9 ‘8‘8°E -0.08 ·Q -0.09
-0.10 -]\-0.11 j-0.12-0.13 „
-0.14-0.15-0.16-0.17-0.18-0.19-0.20-0.21-0.22-0.23
0.000 0.005 0.010 0.015 0.020 0.025 0.030
Txme (sec.)
Flgure 61. Radlal stress vs. tlme (llne/patch l0ads: Force (8))
Numerical Examplcs and discussions l37
0.170.160.150.14i 0.130.120.110.100.090.080.070.060.05
*7 0.04*5; 0.02C4 0.02‘;)’
0.010.00
ä -0.01· -0.02
-0.06 „Q; -0.04 ' -
-0.05‘
gf -0.06 Q . eQ -0.07E·• -0.08 .
-0.09-0.10-0.11-0.12-0.13
·‘
-0.14-0.15-0.16 ~-0.17-0.18-0.19-0.20-0.21
i
0.000 0.005 0.010 0.015 0.020 0.0% 0.030
Time (sec.)
Flgure 62. Tangentlal stress vs. time (llnelpatch Ioads: Force (8))
Numerical Examplcs and discussions |38
0.0008‘
0.0002
0.0001
E ogooooO ;
_
tl 1h-I .-# .8 -0.0001Cl ;
E E8 -0i0002G :
pä .
¤« :ID I••-• .Q -010008
-0.0004
-0.0005
-0.00060.000 0.005 0.010 0.015 0.020 0.0% 0.080
Time (sec.)
Flgure 63. Radlal dlsplacement vs. tlme (llnelpatch loads: Force (C))
Numerical Examples und discussions [39
0.8
0.2
0.1
0.0
-0.1»·«.$ .nl -0.2
W -0.32.HW -0.4•—•6za 0 5,
Cd-0.8
-0.7
-0.8
-0.9
-1.0
0.000 0.005 0.010 0.015 0.020 0.025 0.030
Time (sec.)
Flgure 64. Radlal stress vs. time (Ilnelpatch Ioads: Force (C))
Numcrical Examples und discussions l40
0.8
0.2
0.1
0.0W
-0.12
al -0.2\•/
ä -0.8 l·•->VJ -0.4¤$¤D .G -0.5GE-•
-0.8
-0.7
-0.8
-0.9
-1.0
0.000 0.005 5 0.010 0.015 0.020 0.025 0.080
Time (sec.)
Figure 65. Tangential stress vs. time (linelpatch Ioads: Force (C))I
Numcrical Examples and discussions Ni
0.0008
0.0002
0.0001AEÜ .Cl
CL8 0.0000Cl
EGJQ -0.0001
I-1¤-•cn••¤l
Q
-0.0002
-0.0008
-0.00040.000 0.005 0.010 0.015 0.020 0.025 0.080
Time (sec.)
Flgure 66. Radlal dlsplacement vs. tlme (llnelpatch Ioads: Force (D))
Numericai Exampies uid discussions 142
0.13
0.12
0.11
0.10
0.090.08
0.07
0.08
Ö 0.050.04
V 0.08g 0.02
0.01·
-•-> 0.00::3 -0.01 iG -0.02 _N -0.02. ‘
-0.04
-0.05
-0.08
-0.07
-0.08 '
-0.09
-0.10
-0.11
0.000 0.005 0.010 0.015 0.020 0.025 0.030
Time (sec.)
Flgure 67. Radlal stress vs. tlme (Ilne/patch Ioads: Force (D))
Numcrical Examplcs and discussions _ N3
0.12
0.11
0.10
0.09
0.08
0.07
0.08
0.05
fg 0.040.03
g) 0.02g 0.01
· $-45*; 0.00
es -0.01ED -0.02 5 .
[2 -0.03
-0.04
-0.05
-0.08
-0.07
-0.08
-0.09
-0.10
-0.11
0.000 0.005 0.010 0.015 0.020 0.025 0.030
Time (sec.)
Figure 68. Tangentlal stress vs. tlme (llnelpatch loads: Force (0))
Numcrical Examples and discussions I44
0.0003
0.0002
0.0001
A€¤
0.0000~••|\-/
QC1
E -0.0001- GJ
OGS•-•
Q -0.0002-•-«Q
-0.0003
-0.0004
-0.0005
0.000 0.005 0.010 0.015 0.020 0.025 0.030
Time (sec.)
Flgure 69. Radial displacement vs. time (Iinelpatch Ioads: Force (E))
Numcrical Examples and discussions NS
0.11
0.10
0.09
0.08
0.07
0.08
0.05
0.042
-5}-· 0.03·E 0.02
Z} 0.01EU 0.00m -0.01•—(
E -0.02E -0.03E -0.04 ‘ °
-0.05 ··»
-0.08
-0.07
-0.08 ·]-0.09
-0.10
-0.11
0.000 0.005 0.010 0.015 0.020 0.025 0.030
Time (sec.)
Flgure 70. Radlal stress vs. tlme (Ilnelpatch Ioads: Force (E))
Numerical Example: und discussions N6
0.11
0.10
0.09
0.08
0.07
0.08
0.05
A 0.04
0.03 .\-/ .
m 0.02mQ) 0.01
0.00
ci -0.01Cm¤ -0.02 1.2 x \-0.03 |
-0.04
-0.05 _
-0.06
-0.07
-0.08
-0.09
-0.10
0.000 0.005 0.010 0.015 0.020 0.025 0.030
Time (sec.)
Flgure 71. Tangentlal stress vs. tlme (line/patch Ioads: Force (E))
Numcrical Examplcs and discussions M7
5
6 4'S3 3u2 2c
"· 100.000 0.018 0.036 0.054 0.072 0.090 0.108 0.126 0.144 0.162 0.180
Force history (F): time (sec.)
5
6 4
3 3
8H 2·
"· 100.000 0.018 0.036 0.054 0.0720.090 0.108 0.126 0.144 0.162 0.180
Force history (G): time (sec.)
5
6 4'53 3gg
-no"· 1
00.000 0.018 0.036 0.054 0.072 0.090 0.108 0.126 0.144 0.162 0.180
Force history (H): time (sec.)
Figure 72. Force hlstorles for vlscoelastlc propellant
Numerical Example; and discussions _ 148
U
0.00000
-0.00001
-0.00002
-0.00003
{Ä„¤O -0.00004C7••-I
§-/*6*.m -0.00005
-•¤l¤ 1-0.00006]-0.00007 -·
-0.00008
-0.00009
0.000 0.018 0.036 0.054 0.072 0.090 0.108 0.126 0.144 0.162 0.180
Time (sec.)
Flgure 73. Radlal dlsplacement vs. tlme (two Ilne Ioads: Force (F))
Numcrical Example: and discussion: 149
-0.01 A
rvvvv-_
•
-0.02
-0.03
ax'ES -0.04gz
CD0)Q)Xn -0.05
4-*U}
•—•QEtg -0.08
QS
-0.07
-0.08
-0.09
0.000 0.018 0.036 0.054 0.072 0.090 0.108 0.126 0.144 0.162 0.180
Time (sec.)
Flgure 74. Radlal stress vs. time (two Ilne Ioads: Force (F))
Numcrical Examplcs und discussions l50
-0.005
-0.007
-0.009
-0.011
-0.013 _
-0.015
-0.017
-0.019
-0.021 _*7 -0.023-•¤l
gf -0.025V -0.027Ä -0.029
. -0.031«•-J
°'* -0.0:1:1gl) -0.0:161G -0.0:17
-0.039
-0.041
-0.043
-0.045
-0.047
-0.049-0.051
-0.053
-0.055
-0.057
0.000 0.018 0.036 0.054 0.072 0.090 0.108 0.126 0.14-4 0.162 0.180
Time (sec.)
Flgure 75. Tangentlal stress vs. time (two Ilne Ioads: Force (F))
Numerical Examples and discussions ISI
0.00002
l
0.00001
0.00000
-0.00001
-0.00002zs
|
¤ -0.00003•¤1
ga
iim -0.00004•¤¤I
Q -0.00005 ·-0.00006
-0.00007
-0.00008
-0.00009
0.000 0.018 0.038 0.054 0.072 0.090 0.108 0.128 0.144 0.182 0.180
Time (sec.)
Flgure 76. Radlal dlsplacement vs. tlme (two llne Ioads: Force (G))
Numerical Examplcs und discussions 152
0.03
0.02
0.01
0.00
}-.-2
E-/ -0.02CD03Q)S-4 -0.03+9Ui
•-I
E -0.04'UCU
-0.05
-0.06
-0.07
-0.08
-0.09
0.000 0.018 0.036 0.054 0.072 0.090 0.108 0.126 0.14-4 0.162 0.180
Time (sec.)
Flgure 77. Radial stress vs. time (two Ilne Ioads: Force (G))
Numerical Examples and discussions . iss
0.02
0.01
0.00
2
'5 -0.01 .¤«\-/
mcn2__, -0.02m¤$ä"Q -0.03
E-•
-0.04
-0.05
-0.080.000 0.018 0.038 0.054 0.072 0.090 0.108 0.128 0.144 0.182 0.180
Time (sec.)
Figure 78. Tangentlal stress vs. tlme (two Ilne Ioads: Force (G))
Numerical Examples und discussions [S4
0.00000
-0.00001
-0.00002
-0.00003
.¤OC1
-0.00004· •-n
¤«U)-•-«Q
-0.00005
-0.00006
-0.00007
-0.00008
0.000 0.018 0.036 0.054 0.072 0.090 0.108 0.126 0.144 0.182 0.180
Time (sec.)
Figure 79. Radlal dlsplacement vs. tlme (two ilne Ioads: Force (H))
Numerical Examples end discussions 155
0.004
0.003
0.002
0.001
0.000
-0.001
-0.002
-0.003
'°> -0.004EQ! ·0.005 .\-/gg) -0.006
Q; -0.007.5m -0.008
gg -0.009••-4
'O -0.010as
0:: -0.011-0.012
-0.013
-0.014
-0.015
-0.016
-0.017
-0.018
-0.019
0.000 0.018 0.036 0.054 0.072 0.090 0.108 0.126 0.144 0.162 0.180
Time (sec.)
Figure 80. Radlal stress vs. tlme (two line loads: Force (H))
Numerical Examples and discussions *56
0.002
0.001
0.000
-0.001
-0.002
-0.003
-0.004"?
-0.005O-4"“
-0.006mä -0.0075m -0.008
¤5 -EB 0.009Q -0.010 I ·E- I
-0.011 ·+
-0.012
-0.013 ·
-0.014
-0.015
-0.016
-0.017
0.000 0.018 0.036 0.054 0.072 0.090 0.106 0.126 0.144 0.162 0.160
.Time (sec.)
Figure 81. Tangentlal stress vs. tlme (two Ilne Ioads: Force (H))
Numerical Example: und discussions IS7
-0.00008
-0.00009
-0.00010
-0.00011
-0.00012
-0.00013
-0.00014
-0.00015
-0.00016
-0.00017
E -0.00016ä -0.00019 .:7 -0.00020
& -0.00021 ·E -0.00022
-0.00023
-0.00024
-0.00025
-0.00026
-0.00027
-0.00028
-0.00029
-0.00030
-0.00031
-0.00032
0.000 0.018 0.038 0.054 0.072 0.090 0.108 0.128 0.144 0.182 0.180
Time (sec.)
Flgurc 82. Radlal dlsplacement vs. tlme (two patch loads: Force (F))
Numcrical Examplcs and discussions _ [S8
-
*1.,.2
_
U)Q2~./U)U')CDS4
4-JU)
T6-•-«'UQ
Q: -2
-8 U
0.000 0.018 0.038 0.054 0.072 0.090 0.108 0.128 0.144 0.182 0.180
Time (sec.)
Flgure 83. Radlal stress vs. tlme (two patch Ioads: Force (F))
Numcrical Examplcs and discussions |59
"°·2 "{UÄVYVVI!1!1VVVV""*‘
-0.3‘ -
‘
-0.4
-0.5
-0.6
-0.7
xx -0.8.„-E
Q -0.9x/
VJ -1.0- m
E0 -1.1m· -1.2
GJ¤I>Q -1.3:5E" -1.4
-1.5
-1.6
-1.7
-1.8
-1.9
-2.00.000 0.018 0.036 0.054 0.072 0.090 0.108 0.126 0.144 0.182 0.180
Time (sec.)
Figure 84. Tangentlal stress vs. tlme (two patch Ioads: Force (F))
Numerical Example: and discussions 160
0.000030.00002
‘
0.000010.00000
-0.00001-0.00002-0.00003-0.00004-0.00005-0.00008 (
-0.0000*7-0.00008-0.00009 ·-0.00010
A -0.00011·g -0.00012¤ -0.00013
L'; -0.00014. .-2 -0.00015
gk) -0.00018-•-• -0.00017Q -0.00018
-0.00019-0.00020-0.00021-0.00022-0.00023-0.00024-0.00025-0.00028-0.00027-0.00028-0.00029-0.00030-0.00031-0.00032
0.000 0.018 0.038 0.054 0.072 0.090 0.108 0.128 0.144 0.182 0.180
Time (sec.)
Flgure 85. Radlal dlsplacement vs. time (tvvo·patch loads: Force (6))
Numcrical Exampies and discussions I6!
1
0
2 1-•-ZO'}
D-4\-/
U)
3· s-. -1-‘
0m
•-4
C5n•-4’¤ .GS
QS
-2
-3
0.000 0.018 0.036 0.054 0.072 0.090 0.108 0.126 0.144 0.162 0.180
Time (sec.)
Flgure 86. Radial stress vs. time (two patch Ioads: Force (G))
Numcrical Examples and discussions l62
0.2
0.1
0.0
-0.1
-0.2
-0.3
-0.4
-0.5A-0.6
V EL -0.7$ -0.82Q -0.9CD_ -1.0QJCD -1.1CIQ -1.2E-
-1.3
-1.4
-1.5
-1.8
-1.7
-1.8
-1.9
-2.00.000 0.018 0.038 0.054 0.072 0.090 0.108 0.128 0.144 0.162 0.180
Time (sec.)
Figure 87. Tangentiai stress vs. time (two patch Ioads: Force (G))
Numcrical Examplcs und discussions I63
0.00000-0.00001-0.00002-0.00003-0.00004-0.00005-0.00006-0.00007-0.00008-0.00009-0.00010
,„ -0.00011
-5 -0.00012_ :1 -0.00013
C? -0.00014E -0.00015ä -0.00018
-0.00017-0.00018-0.00019-0.00020
-0.00021-0.00022-0.00023-0.00024-0.00025-0.00026-0.00027-0.00028
0.000 0.018 0.038 0.054 0.072 0.090 0.108 0.128 0.144 0.182 0.180
Time (sec.)
Flgure 88. Radial dlsplacement vs. tlme (two patch Ioads: Force (H))
Numcrical Examplcs end discussions 164
0.080.050.040.030.020.010.00
-0.01-0.02-0.03-0.04-0.05-0.08-0.07
Q -0.08In -0.09
C-4 -0.10”
-0.11-0.12
Q} -0.133 -0.14m -0.15__‘
-0.18‘
-.1-0.19
Q: -0.20-0.21-0.221-0.23-0.24-0.25-0.28-0.27-0.28-0.29-0.30-0.31-0.32-0.33-0.34
0.000 0.018 0.038 0.054 0.072 0.090 0.108 0.128 0.144 0.182 0.180
T1me (sec.)
Figure 89. Radlal stress vs. time (two· patch Ioads: Force (H))
Numerical Examples and discussions 165
0.030.020.010.00
-0.01-0.02-0.03-0.04-0.05-0.08
A -0.07-0.08
¤_‘ -0.09‘·’-0.10
g -0.11Q3
-1 1: KIS, -0.14
¤> -tw 0.15c: -0.18 ·QE -0.17
-0.18-0.19-0.20-0.21-0.22-0.23-0.24-0.25-0.28-0.27-0.28
0.000 0.018 0.036 0.054 0.072 0.090 0.108 0.126 0.144 0.182 0.180
Time (sec.)
Flgure 90. Tangentlal stress vs. time (two patch Ioads: Force (H))
Numcrical Examplcs und discussions [66
-0.00008
-0.00009
-0.00010
-0.00011
-0.00012
-0.00013
-0.00014
-0.00015
-0.00016
-0.00017-0.00018
·‘f-0.00019
-0.00020Q -0.00021
-0.00022
-0.00023
-0.00024
-0.00025
-0.00026
-0.00027
-0.00028
-0.00029
-0.00030
0.000 0.018 0.036 0.054 0.072 0.090 0.108 0.126 0.144 0.162 0.180
Time (sec.)
Figure 91. Radlal dlsplacement vs. time (linelpatch loads: Force (F))
Numcrical Examples and discussions f 167
-0.2
-0.3-0.4
-0.5-0.8
-0.7
-0.8-0.9
-1.0-1.1
/'\-•-Z -1.2 _{ -1.3“"
-1.4IDm -1.5CD$4 -1.8•->m -1.7T; -1.8••¤l
'¤ -1.9‘
G5 -2.0-2.1-2.2
-2.3-2.4
-2.5-2.8-2.7
-2.8-2.9
-3.0
0.000 0.018 0.038 0.054 0.072 0.090 0.108 0.128 0.144 0.182 0.180
Time (sec.)
Flgure 92. Radlal stress vs. time (line/patch Ioads: Force (F))
Numerical Exsmplcs und discussions 168
-0*2-0.3
‘
-0.4
-0.5
-0.8
-0.7
A -0.8..-5{ -0.99
g -1.01 EU -1.1U1
Q3 -1.2¤0gi -1.3CUE" -1.4
-1.5
-1.8
-1.7
-1.8
-1.9
-2.00.000 0.018 0.038 0.054 0.072 0.090 0.108 0.128 0.144 0.182 0.180
Time (sec.)
Flgure 93. Tangentlal stress vs. tlme (llne/patch Ioads: Force (F))
Numcrical Example: und discussions 169
0.000030.000020.000010.00000 _
-0.00001u
-0.00002-0.00003-0.00004-0.00005 ·-0.00006-0.00007-0.00008-0.00009 '
,.„ -0.00010.¤ -0.00011Q -0.00012
Q; -0.00013..: -0.00014
Q -0.00015 J-0.00016 _-0.00017 j-0.00018-0.00019 J
-0.00020 J
-0.00021-0.00022-0.00023-0.00024-0.00025-0.00026-0.00027-0.00028-0.00029-0.00030
0.000 0.018 0.036 0.054 0.072 0.090 0.108 0.126 0.144 0.162 0.180
Time (sec.)
Figure 94. Radial displacement vs. time (llnelpatch Ioads: Force (G))
Numerical Examples and discussions 170
1
0
rs-.-ZU)
Cu
U1CDQ)La -1+9U1
•-<N-•-4'UNM
-2
-3
0.000 0.018 0.036 0.054 0.072 0.090 0.108 0.126 0.144 0.162 0.180
Time (sec.)
/ _„~ 7
Flgure 95. Radlal stress vs. time (llnelpatch Ioads: Force (G))
Numcrical Example: and discussions 17I
0.2
0.1
0.0
-0.1
-0.2
-0.3‘
-0.4
A -0.5
E-0.8
„, -0.7
3 -0.8‘
-0.9 -_ -1.0
-1.1 äL5 “1„2AI-1.3
-1.4
-1.5
-1.8
-1.7
-1.8
-1.9
-2.0 I0.000 0.018 0.036 0.054 0.072 0.090 0.108 0.128 0.144 0.162 0.180
Time (sec.)
Flgure 96. Tangenttal stress vs. time (Ilne/patch Ioads: Force (G))
Numerical Examplcs and discussians . 172
0.00000-0.00001
-0.00002-0.00003
-0.00004-0.00005
-0.00008
-0.00007
-0.00008
-0.00009-0.00010
E -0.00011
ä -0.00012-0.00013
.51* -0.000141
.52 -0.00015Q -0.00018
-0.00017
-0.00018
-0.00019
-0.00020
-0.00021-0.00022
-0.00023
-0.00024
-0.00025
-0.00028
-0.00027
0.000 0.018 0.038 0.054 0.072 0.090 0.108 0.126 0.144 0.182 0.180
Time (sec.)
Figure 97. Radlal dlsplacement vs. tlme (llnelpatch loads: Force (H))
Numcrical Examples and discussions l73
0.050.040.03 .0.020.010.00
-0.01-0.02-0.03-0.04-0.05-0.06-0.07-0.08
Q -0.09m -0.10
CL. -0.11T; -0.12
-0.13Z} -0.14S-• -0.15
*5*, -0.18,_, -0.17
E -0.18' -0.19-0.2031Z -0.21
·-0.22 :1-0.23-0.241-0.25-0.28-0.27-0.28-0.29-0.30-0.31-0.32 _-0.33-0.34-0.35
0.000 0.018 0.038 0.054 0.072 0.090 0.108 0.128 0.144 0.182 0.180
T1me (sec.)
Figure 98. Radial stress vs. time (llnelpatch Ioads: Force (H))
Numcrical Examplcs and discussions U4
0.03
0.010.00
-0.01-0.02-0.03-0.04-0.05-0.06
A -0.07.,.1 -0.08
E -0.09‘-“
-0.10
Q -0.11Q -0.12ja -0.13
, -0.14
QB -0.15Cl -0.16[2 -0.17
-0.18-0.19-0.20-0.21-0.22-0.23-0.24-0.25-0.26-0.27-0.26
0.000 0.018 0.036 0.054 0.072 0.090 .0.108 0.126 0.144 0.162 0.180
Time (sec.)
Flgure 99. Tangential stress vs. tlme (llnelpatch Ioads: Force (H))
Numcrical Examplcs and discussions ns
Appendlx A.{X} = {Um- Um- Um- Erzm ö6n· Um} (7-*)Kf = ‘[~j%¥ dA = K§" (1-2) '
K,}" = fr-/j% dA = )<,§" (1-3)
Kf = j·~,~jdA (1-4)
;<,§° = I- é ~,~jdA = Kj?5 = r<§6 (1-5)
K§" = ~,~jdA (1 -6)
Appendix A. *76
35 _ _g_ _ 36 _ 56Kü —E NiNjdA — Kü — Kü (1.7)
The other coefficients of [K] are zero.
Appcndix A. _ I77
Appendlx B.
Forn=O
V ör ör ör V ör V(21)
N; ^’;1-2v ÜN; Ö";
+(‘””)7"7+T' öl 62 WA
K;2_ E ÜN; WAÜ_
(1+v)(1—2v)V ör öz V öz 2 öz ör (2.2)
=K§‘
z2_ E _ ÜN; ö"Ü 1-2;; ÜN; ÜN; dA 23Ku ' (1+v)(1—2v) fw V) öz öz + 2 ör ör } (°)
Appcndix B. H8
For n=1, 2
11 _ E ÜN; ÜN! ÜN; N! ÜN! N;KU ‘
(1+1)(1-21) fw"! öf ÜF +" ÜF7“+”
61 '?'
N. N- _ ÜN. ÖN-2v 1 1 (2.5)V V 2 Öz Öz
_ N. N-
12_ E ÜN; N! N; N!K'! _
(1 +v)(1 —2v) ör
V+·————2n’.(l,-ö,_)}dA
21
ÜN. ÜN. N. ÖN- _ ÜN. ÖN-E 1 /+1 2v 1 /}dA( +v)(1—2v) Ör Öz Ö; 2 Öz Ör (2.7)=K$‘
E N. N- _2 ÜN. N. ÖN- N-
(28)
- 61v- öN·'
Appcndix B. [79
N. öN· _ 5N. N-“?°=vm%m){””+·ä··l%~6ä+}d^(2.9)
32
K;s= E ÜN!+ 1-2v ÜN! ÜN!
U (1 + v)(1 -2v) öz öz 2 ör ör (210)z 1-2v N· N!
I+f7 T*+—F·}dA
mf = mf = mf = jp~,~/6A (2.11)
Appcndix B. ISOI
References
1. Vlasov, N.Z., General Theory of Shells and lts Applications in Engineering,NASA TT-99, 1964
2. Kraus, H., The Elastic Shells, John Wiley, New York, 1967
3. Cederbaun, G. and Heller, R.A., “Dynamic Deformation of Orthotropic Cyl-inders", J. of Pressure Vessel Tech. Vol. 111, 97-101, 1989
4. Gatewood, B.E., "Note on the Thermal Stresses in a Long Circular Cylindersof (m+ 1) Concentric Material", Q. Appl. Math., Vol. 53, 84-86 1948
5. Singh, M.P., Heller, R.A., and Thangjitham, S., “ThermaI Stresses in Con-centric Cylinders due to Asymmetric and Time Dependent Temperatures"J. of Thermal Stresses, Vol. 7,183-195, 1984
6. Thangjitham, S., Heller, R.A., and Singh, M.P., “Frequency Response forThermal Stresses in Multilayered Cylindrical Structures", J. of ThermalStresses, Vol. 9, 133-150,1986
7. Strub, R.A., "Distribution of Mechanical and Thermal Stresses in MultilayerCylinder", Trans. Journal ASME, Vol. 75, 73-82, 1953
8. Heller, R.A., Thangjitham S. and Lin, Y.T., “Stress and Displacements inLine·Loaded Solid Propellant Rocket Motors", Contract No.DAAH—01-86-C-D185, U.S.Army Missile Command, Huntsville, AL., Oct. 1987
9. Flugge, W., Viscoelasticity, 2nd Edition, Springer-Verlag, New York, 1975
10. Christensen, R. M., Theory of Viscoelasticity, 2nd Edition, Academic Press,New York, 1982
11. Singh, M.P. and Heller, R.A., "Random Probability Techniques for RocketMotor Service Life Predictions", Technical Report RK-CR-82-7, U.S.ArmyMissile Command, Huntsville, AL., Sept. 1981
12. Taylor, S. et al., "Thermomechanical Analysis of Viscoelastic Solids", Uni-versity of California, Berkeley, June, 1968
”References l8l
13. Henriksen, M., “Nonlinear Viscoelastic Stress Analysis - A Finite ElementApproach", lnt. J. Compu. & Struc. Vol. 18, 133-139, 1984
14. Jones, l.W. and Pierre, L.E.,“A
Linear Thermoelastic Material Model forSolid Rocket Motor Structural Analysis", Compu. & Struc. Vol. 21, 235-243,1985
15. Yang, T-W et al., “Endochronic Viscoelastic Creep and Analysis of 2-DStructure", Compu. & Struc. Vol. 25, 425-429, 1987
16. Roy, S. and Reddy, J.N.,“A
Finite Element Analysis of Adhesively BondedComposite Joints including Geometric Nonlinearity, GeometricViscoelasticity, Moisture Diffusion, and Delay FaiIure", VPI-E-82-28, VirginiaPolytechnic Institute and State University, Blacksburg, Dec. 1987
17. Roy, S. and Reddy, J.N. "Finite Element Methods of Viscoelasticity and Dif-fusion in Adhesively Bonded Joints", Int. J. of Num. Meth. in Eng. Vol. 26,2531-2546, 1988
18. Roy, S. and Reddy, J.N., "Nonlinear Viscoelastic Finite Element Analysis ofAdhesive Joints", Tire Science and Technology, Vol. 16, No. 3, 146-170, 1988
19. Reddy, J.M. and Roy, S. "Nonlinear° Analysis of Adhesively BondedJoints", Int. J. of Non-Linear Mechanics, Vol.23, No.2, 97-112, 1988
20. Zak, A. R., “Structural Analysis of Realistic Solid Propellant Materials", J.
of Spacecraft, Vol. 5, 270-275, 1968
21. Chung, T.J. and Eidson, R.L. "Dynamic Analysis of ViscoelastoplasticAnisotropic Shells" Compu. & Struc. Vol. 3, 483-496, 1973
22. Wilson, D.W. and Vinson, J.R. "Viscoelastic Analysis of Lamlnated Plate
BuckIing'°, AIAA J. Vol. 22, No. 2, 982-988, 1984
23. Bukowski, R. and Wojewodzki, W. "Dynamic Buckling of ViscoplasticSpherical ShelI" Int. J. Solids & Struc. Vol. 20, No. 8, 761-776, 1984
24. Florence, A.L. and Abrahamson, G.R."Critical Velocity for Coppapse ofViscoplastic Cylindrical Shells without BuckIing", J. of Applied Mechanics,89-94, March, 1977
25. Wilson, E.L., "Structural Analysis of Axisymmetric Solids", AIAA J., Vol. 3,
No. 12, 2269-2274. 1965
26. Crose, J.G., “Stress Analysis of Axisymmetric Solids with AsymmetricPropertles", AIAA J., Vol. 10, No. 7, 866-871,1972
27. Cook, R.D., Concepts and Applications of Finite Element Methods, JohnWiley, New York,1981
References . 182
28. Yang, T.Y., Finite Element Structural Analysis, Prentice-Hall, New Jersey,1986
29. Zienkiewicz O.C., The Finite Element Method , 3rd Edition, McGraw-Hill,London, 1977
30. Timoshenko, S.P., Theory of Elasticity, 3rd Edition, McGraw-Hill, NewYork,1970
31. Atluri, S.N. et al., (eds.) Hybrid and Mixed Finite Element Methods, JohnWiley, New York,1982
32. Reddy, J.N. and Oden, J.T., “Mixed Finite Element Approximations of Linear
Boundary-Value ProbIems", Q. Appl. Math. Vol. 33, 255-280, 1975
33. Oden, J.T. and Reddy J.N., “On Mixed Finite Element Approximations",SIAM J. Num. Ana. Vol. 13, 393-404, 1976
34. Washizu, K., Variational Methods in Elasticity and Plasticity, 2nd Edition,
Pergamon Press Inc., New York, 1975
35. Reddy, J.N., Energy and Variational Methods in Applied Mechanics, JohnWiley, New York,1984
36. Heyliger, P.R., A Mixed Computational Algorithm based on UpdatedLagrangian Description for Plane Elastic Contact Problems , Ph.D Thesis,Virginia Polytechnic Institute and State University, Blacksburg, VA, 1986
37. Heyliger, P.R. and Reddy, J.N. "A Mixed Computational Algorithm for PlaneElastic Contact Problems - I. Formulation", Compu. & Struc. Vol. 26, No. 4,
621-634, 1987
38. Heyliger, P.R. and Reddy, J.N.“A
Mixed Computational Algorithm for PlaneElastic Contact Problems - I. Numerical Examples", Compu. & Struc. Vol.26, No. 4, 635-653, 1987
39. Reddy, J.N., An Introduction to the Finite Element Method, McGraw-Hill,
New York, 1984
40. Akay, H.U., “Dynamic Large Deflection Analysis of Plates using Mixed FiniteElements", Compu. & Struc. Vol.11, 1-11, 1980
41. Reddy, J.N., “CIass Notes on Finite Element Methods II", Virginia
Polytechnic Institute and State University, Blacksburg, VA, 1988
References l83