Chapter 4 Solutions
Chapter 4 Solutions
Exercise
P4.1 A force is gradually applied at the end of an elastoplastic bar such that it is in the
plastic phase. When the total magnitude of strain is e = 0.003, calculated the applied
force, axial stress, elastic strain, and plastic strain. Use the following material properties:
E = 100GPa, H = 10GPa, and sY = 100MPa. The cross-sectional area of the bar is A
= 1.0´10-4m2.
Solution:
If the total strain is pure elastic, then stress will become s = Ee = 300MPa, which is
larger than the yield stress. Thus, the material is in the plastic phase. It is convenient to
separate the initial elastic deformation, e(1), until the yield stress from the elastoplastic
deformation, e(2) such that e = e(1) + e(2). The initial elastic strain can be calculated from
sY
e(1) = ee(1) = eY = = 0.001, s(1) = sY = 100MPa
E
After the initial elastic deformation, the remaining deformation is elastoplastic
deformation. In this phase, the strain increment is De(2) = 0.002. Since the total strain
increment is given, Eq. (4.9) can be used to calculate the plastic strain increment as
De(2)
De(2)
p = = 0.00182
1+H /E
P4.2 A force 12kN is gradually applied and then removed at the end of an elastoplastic
bar. When the yield stress of the material is 100MPa, calculate plastic strains and tip
displacement after removing the applied force. Use the following material properties: E
= 100GPa and H = 10GPa. The cross-sectional area of the bar is A = 1.0´10-4m2 and
the length of the bar is L0 = 1m.
Solution:
4-2
In a one-dimensional bar, it is assumed that the force is uniformly distributed over the
cross section. During the loading process, since the total stress, s = F/A = 120MPa, is
larger than the yield stress, it can be concluded that the material is under plastic
deformation. It is convenient to divide the entire deformation into elastic and plastic
phases. The material is initially elastic until it reaches yield stress. Thus, when stress
reaches yield stress s(1) = sY = 100MPa, strain is purely elastic:
sY
Dee(1) = = 0.001
E
After yielding, the remaining stress increment, Ds = 20MPa, is in the plastic phase. The
elastic and plastic strain increments can be calculated from
Ds(2)
Dee(2) = = 0.0002
E
Ds(2)
Dep(2) = = 0.002
H
Thus, the total elastic and plastic strains become
Now, during the unloading, elastic strain is removed, but the plastic strain remains. Thus,
ep = 0.002 and u = epL = 2mm . ▄
P4.3 A uniaxial bar is under tensile force F = 12kN at load step tn. (a) When the
plastic strain is epn = 0.002 , determine the yield status of the material. (b) If the applied
force is increased to F = 15kN at load step tn+1, calculate plastic strain and tip
displacement. Assume the initial yield stress sY = 100MPa, E = 100GPa and H =
10GPa. The cross-sectional area of the bar is A = 1.0´10-4m2 and the length of the bar
is L0 = 1m. Assume isotropic hardening model.
Solution:
(a) At load step tn, the stress in the bar is
F
sn = = 120MPa
A
The yield stress is
(b) During the loading stage up to tn, the elastic and plastic strains can be calculated by
4-3
sn
een = = 0.0012, epn = 0.002
E
At load step tn+1, the elastic and plastic strain increments can be calculated from stress
increment Dsn +1 = 30MPa as
Dsn +1 Dsn +1
Deen +1 = = 0.0004, Depn +1 = = 0.004
E H
Thus, the total elastic and plastic strains become
P4.4 An elastoplastic bar is under variable load history. At load step tn, the stress and
plastic strain are sn = 200MPa and epn = 1.0 ´ 10-4 , respectively. (a) Is the material in
elastic or plastic state? (b) When strain increment is De = -0.003, calculate stress and
plastic strain. Assume isotropic hardening with E = 200GPa, H = 25GPa, and sY =
250MPa.
Solution:
(a) At a given plastic strain epn = 1.0 ´ 10-4 , the yield stress is
Ds = E De = -600MPa, s tr = s n + Ds = -400MPa
Since f tr =| str | -sY = 147.5 > 0 , the material is yielded in the compression side.
From Eq. (4.26), the plastic strain increment becomes
tr
f
Dep = = 6.5556 ´ 10-4
E +H
sn +1 = tr
s - sgn( tr s)E Dep = -268.9MPa
P4.5 Repeat Problem P4.4 using the kinematic hardening model. For back stress, use
an = 2.5MPa.
Solution:
(a) At a given stress sn = 200MPa and back stress an = 2.5MPa , the trial shifted
stress is
hn = sn - an = 197.5MPa
Ds = E De = -600MPa, s tr = s n + Ds = -400MPa
Since f tr =| str - an | -sY = 152.5 > 0 , the material is yielded in the compression
side. From Eq. (4.41), the plastic strain increment becomes
tr
f
Dep = = 6.7778 ´ 10-4
E +H
Therefore, the stress, back stress and plastic strain are updated as
sn +1 = tr
s - sgn( tr h)E Dep = -264.4MPa
P4.6 Repeat Problem 4.5 using the combined hardening model with b = 0.5.
4-5
Solution:
(a) At a given stress sn = 200MPa, back stress an = 2.5MPa , and plastic strain
epn = 1.0 ´ 10-4 , the trial shifted stress and the yield stress are
hn = sn - an = 197.5MPa
Ds = E De = -600MPa, s tr = s n + Ds = -400MPa
Since f tr =| str - an | -sY = 151.25 > 0 , the material is yielded in the compression
side. From Eq. (4.41), the plastic strain increment becomes
tr
f
Dep = = 6.2222 ´ 10-4
E +H
Therefore, the stress, back stress and plastic strain are updated as
sn +1 = tr
s - sgn( tr h)E Dep = -265.6MPa
P4.7 For the combined isotropic/kinematic hardening model, derive the expression of
plastic strain increment from the plastic consistency condition.
Solution:
The plastic consistency condition means that the yield function at the current time step,
tn+1, remains zero; that is,
n +1 n +1 n +1
f = h - sY = 0
The following stress update formulas and yield stress are used for the plastic consistency
condition:
n +1 tr
s= s - sgn( tr h)E Dep
4-6
n +1
a = n a + sgn( tr h)bH Dep
n +1
sY = n sY + (1 - b )H Dep
Note that the above formula works for both positive and negative trial stress. Therefore,
the plastic strain increment can be obtained as
tr
f
D ep =
E +H
Note that the formula for plastic strain increment is identical to that of isotropic
hardening in Eq. (4.26) and kinematic hardening in Eq. (4.41). ▄
P4.8 An elastoplastic bar is clamped at the left end, and variable loads are applied at
the right end, as shown in the table. Plot the stress-strain curve by changing the applied
forces by 5kN increments. Assume the following material properties with isotropic
hardening: E = 70GPa, H = 10GPa, sY = 250MPa. The length of the bar is L = 1m,
and the cross-sectional area is A = 1.0´10-4m2.
Load step 1 2 3 4
Force (kN) 30 20 35 20
Solution:
Since the given condition is applied force at the tip, the convergence iteration must be
performed to find equilibrium. Below is the MATLAB program that solves for the
problem:
%
% P4.8 Variable loadings
%
E=70000; H=10000;Et=1000; sYield=250;
mp = [E, 0, H, sYield];
Et=E*H/(E+H);
nS=0; nep=0; nA=0;
A = 100; L = 1000;
tol = 1.0E-5; u = 0; iter=0; Res=0;
Force = 1000*[5:5:30 25 20 25 30 35 30 25 20];
N = size(Force',1);
X=zeros(N,1);Y=zeros(N,1);
fprintf('\nstep iter u S ep Residual');
fprintf('\n %3d %3d %7.4f %7.3f %8.6f %10.3e',i,iter,u,nS,nep,Res);
for i=1:N
P = Force(i); iter = 0;
Res = P - nS*A;
conv = Res^2/(1+P^2);
4-7
du=0;
while conv > tol && iter < 20
Eep = E; if epnew>nep; Eep = Et; end
delu = Res / (Eep*A/L);
du = du + delu;
delE = du / L;
[Snew, Anew, epnew]=combHard1D(mp,delE,nS,nA,nep);
Res = P - Snew*A;
conv = Res^2/(1+P^2);
iter = iter + 1;
end
u=u+du;
nS = Snew; nep = epnew;
X(i) = u; Y(i) = nS;
fprintf('\n %3d %3d %7.4f %7.3f %8.6f %10.3e',i,iter,u,nS,nep,Res);
end
X=[0;X];Y=[0;Y];plot(X,Y);
P4.9 An elastoplastic bar is clamped at the left end, and variable displacements are
applied at the right end, as shown in the table. Plot the stress-strain curve by changing the
tip displacement by 1mm increments. Assume the following material properties with
isotropic hardening: E = 70GPa, H = 10GPa, sY = 250MPa. The length of the bar is
L = 1m, and the cross-sectional area is A = 1.0´10-4m2.
Load step 1 2 3 4
Displacement (mm) 5.0 3.0 7.0 6.0
Solution:
Since the tip displacements are given, there is no need to perform convergence iteration.
At each step, displacement increment is calculated by the difference between current and
previous load increments. After that, the strain increment is calculated from the small
deformation assumption. The program combHard1D will calculate stress and plastic
4-8
strain for the given strain increment. Below is the MATLAB program that solves for the
problem:
%
% P4.9 Variable displacement
%
E=70000; H=10000;Et=1000; sYield=250;
mp = [E, 0, H, sYield];
Et=E*H/(E+H);
nS=0; nep=0; nA=0;
A=100; L=1000;
tol = 1.0E-5; u=0; iter=0; Res=0;
disp=[0 1 2 3 4 5 4 3 4 5 6 7 6];
N = size(disp',1);
X=zeros(N,1);Y=zeros(N,1);
fprintf('\nstep u S ep');
fprintf('\n %3d %7.4f %7.3f %8.6f',i,u,nS,nep);
for i=2:N
delu = disp(i) - disp(i-1);
delE = delu / L;
[Snew, Anew, epnew]=combHard1D(mp, delE, nS, nA, nep);
nS = Snew; nep = epnew;
X(i) = disp(i); Y(i) = nS;
fprintf('\n %3d %7.4f %7.3f %8.6f',i,u,nS,nep);
end
plot(X,Y);
P4.10 A force of P = 15 is applied to the two parallel bars in Example 4.2 and then
removed. Using combHard1D program, calculate tip displacement and residual stresses
for the two bars after unloading. Use 15 load increments for each loading and unloading
cycle. Plot stresses vs. tips displacement in the XY graph.
Solution:
Below is the MATLAB program to solve loading/unloading cycle for two parallel bars in
Example 4.2. After unloading, the tip displacement is u = 0.1227, and stresses are s1 =
4-9
-4.091, s2 = 2.455. Since the area of bar2 is twice than that of bar1, the stress s2 is half
of s1.
%
% P4.10 Two-bar in parallel - unloading
%
E1=10000; Et1=1000; sYield1=5;
E2=5000; Et2=500; sYield2=7.5;
mp1 = [E1, 0, E1*Et1/(E1-Et1), sYield1];
mp2 = [E2, 1, E2*Et2/(E2-Et2), sYield2];
nS1 = 0; nA1=0; nep1 = 0; epnew1=0;
nS2 = 0; nA2=0; nep2 = 0; epnew2=0;
A1 = 0.75; L1 = 100;
A2 = 1.25; L2 = 100;
tol = 1.0E-5; u = 0;
Force = [1:15 14:-1:0];
N = size(Force',1);
X=zeros(N,1);Y1=zeros(N,1);Y2=zeros(N,1);
fprintf('\nstep iter u S1 S2 ep1 ep2 Residual');
fprintf('\n %3d %3d %7.4f %7.3f %7.3f %8.6f %8.6f %10.3e',...
0,0,u,nS1,nS2,nep1,nep2,0);
for i=1:N
P = Force(i); iter = 0;
Res = P - nS1*A1 - nS2*A2;
conv = Res^2/(1+P^2);
while conv > tol && iter < 20
Eep1 = E1; if epnew1>nep1; Eep1 = Et1; end
Eep2 = E2; if epnew2>nep2; Eep2 = Et2; end
delu = Res / (Eep1*A1/L1 + Eep2*A2/L2);
u = u + delu;
delE = delu / L1;
[Snew1, Anew1, epnew1]=combHard1D(mp1, delE, nS1, nA1, nep1);
[Snew2, Anew2, epnew2]=combHard1D(mp2, delE, nS2, nA2, nep2);
Res = P - Snew1*A1 - Snew2*A2;
conv = Res^2/(1+P^2);
iter = iter + 1;
nS1 = Snew1; nep1 = epnew1; nA1 = Anew1;
nS2 = Snew2; nep2 = epnew2; nA2 = Anew2;
end
X(i) = u; Y1(i) = nS1; Y2(i) = nS2;
fprintf('\n %3d %3d %7.4f %7.3f %7.3f %8.6f %8.6f %10.3e',...
i,iter,u,nS1,nS2,nep1,nep2,Res);
end
X=[0;X];Y1=[0;Y1];Y2=[0;Y2];plot(X,Y1,X,Y2);
4-10
P4.11 A force 12kN is gradually applied at the end of an elastoplastic bar. When the
yield stress of the material is 100MPa, calculate displacement at the tip. Use the
following material properties: E = 100GPa and H = 10GPa. The cross-sectional areas
of the bars are A(1) = 1.0´10-4m2 and A(2) = 0.5´10-4m2.
1 2 3
E, A(1) E, A(2) F
x
L0=1m L0=1m
Figure P4.11
Solution:
Since the two bars are connected in parallel, the element forces of the two bars are the
same as the applied force at the tip. By considering the cross-sectional areas of the bars,
the stresses of the bars can be calculated by
F F
s(1) = = 120MPa, s(2) = = 240MPa
A(1) A(2)
Since both bars are in the plastic state, it is possible to separate the entire deformation
into the initial yielding stage (at s = 100MPa) and followed by the elastoplastic state.
For bar1, the elastic strain at the initial yielding stage can be calculated from
sY
ee(1) = = 0.001
E
After that, the remaining Ds = 20MPa is in the elastoplastic stage:
4-11
Ds(1) Ds(1)
ee(1) = = 0.0002, e(1)
p = = 0.002, e(1) = ee(1) + ep(1) = 0.0022
E H
Thus the total strain becomes e(1) = 0.001 + 0.0002 + 0.002 = 0.0032 , and the
displacement at Node 2 becomes
u2 = e(1)L0 = 0.32mm
For bar2, the strain at the initial yielding stage can be calculated from
sY
ee(2) = = 0.001
E
After that, the remaining Ds = 140MPa is in the elastoplastic stage:
Ds(2) Ds(2)
ee(2) = = 0.0014, e(2)
p = = 0.014, e(2) = ee(2) + e(2)
p = 0.0154
E H
the total strain becomes e(1) = 0.001 + 0.0014 + 0.014 = 0.0164 , and the displacement
at Node 3 becomes
u3 = u2 + e(2)L0 = 1.96mm
The same problem can be solved using combHard1D program with convergence
iteration. The following MATLAB program solves for the displacement and stress for the
two-bar problem:
%
% P4.11 Two-bar in serial
%
E=100000; H=10000; sYield=100;Et=E*H/(E+H);
mp = [E, 0, H, sYield];
nS1 = 0; nA1=0; nep1=0; epnew1=0; Eep1=E;
nS2 = 0; nA2=0; nep2=0; epnew2=0; Eep2=E;
A1 = 100; L1 = 100; A2 = 50; L2 = 100;
tol = 1.0E-8; iter = 0; u = [0 0]'; F = [0 12000]';
Res = F - [nS1*A1-nS2*A2;nS2*A2];
conv = norm(Res)^2/(1+norm(F)^2);
fprintf('\n iter u1 u2 S1 S2 ep1 ep2
Residual');
fprintf('\n %3d %7.4f %7.4f %7.3f %7.3f %8.6f %8.6f %10.2e %10.2e',...
iter,u(1),u(2),nS1,nS2,nep1,nep2,Res);
while conv > tol && iter < 20
Kt = [Eep1*A1/L1+Eep2*A2/L2,-Eep2*A2/L2;-Eep2*A2/L2,Eep2*A2/L2];
delu = Kt\Res;
u = u + delu;
delE1 = delu(1) / L1;
delE2 = (delu(2)-delu(1)) / L2;
[Snew1, Anew1, epnew1]=combHard1D(mp, delE1, nS1, nA1, nep1);
[Snew2, Anew2, epnew2]=combHard1D(mp, delE2, nS2, nA2, nep2);
Eep1 = E; if epnew1>nep1; Eep1 = Et; end
Eep2 = E; if epnew2>nep2; Eep2 = Et; end
nS1 = Snew1; nep1 = epnew1; nA1=Anew1;
nS2 = Snew2; nep2 = epnew2; nA2=Anew2;
Res = F - [nS1*A1-nS2*A2;nS2*A2];
conv = norm(Res)^2/(1+norm(F)^2);
4-12
iter = iter + 1;
fprintf('\n %3d %7.4f %7.4f %7.3f %7.3f %8.6f %8.6f %10.2e %10.2e',...
iter,u(1),u(2),nS1,nS2,nep1,nep2,Res);
end
As shown in the output below, the Newton-Raphson iteration converges in the second
iteration.
P4.12 Two one-dimensional bars are connected serially as shown in the figure. At load
step n, bar1 was plastic and bar2 was elastic. At load step n+1, the increments of nodal
displacements are given as Du = [Du1, Du2, Du3] = [0.0, -0.01, 0.0]. Calculate
stresses and plastic strains of both bars at load step n+1.
1 2 3
bar1 bar2
L=100 L=100
Figure P4.12
bar1 bar2
Young modulus (E) 10,000 5,000
Tangent modulus (Et) 1,000 500
Previous stress (sn) 6.0 7.4
Initial yield stress (sY) 5.0 7.5
Plastic strain (ep) 9E-4 0.0
Yield status Plastic Elastic
Hardening Isotropic Isotropic
Solution:
Bar1: From the given nodal displacements, the strain of the element can be calculated by
Du2 - Du1
De = = -10-4
L1
First, assume that the strain increment is purely elastic to obtain the following trial state:
Since the material is initially in the plastic state and the stress is positive at load step n,
the bar yielded in tension. However, since the incremental strain is negative, the bar is
under unloading. Thus, the material becomes elastic and
4-13
sn +1 = str = 5.0, epn +1 = epn = 9 ´ 10-4
Bar2: From the given nodal displacements, the strain of the element can be calculated by
Du3 - Du2
De = = 10-4
L1
First, assume that the strain increment is purely elastic to obtain the following trial state:
Since str > sYn , the material becomes plastic in this load increment. Since the element is
elastic at load step n, the ratio R needs to be calculated, which can be written as
str - sYn
R = 1- = 0.2
Ds
1-R
epn +1 = epn + De = 7.2 ´ 10-5
1+H /E
P4.13 Write the expression of the rank-4 unit symmetric tensor and unit deviatoric
tensor in the 6´6 matrix notation.
Solution:
The index of rank-2 tensor can be converted into a vector via [11 22 33 12 23 13]. From
the definition of I ijkl = (dik djl + dil djk ) / 2 , the matrix version of the fourth-order unit
symmetric tensor becomes
é1 0 0 0 0 0 ùú
ê
ê0 1 0 0 0 0 úú
ê
ê0 0 1 0 0 0 úú
ê
I=ê ú
ê0 0 0 21 0 0 ú
ê ú
ê0 0 0 0 21 0 ú
ê ú
ê0 0 0 0 0 21 ú
ë û
In addition, from the definition of rank-4 unit deviatoric tensor Idev = I - 13 1 Ä 1 , its
matrix version becomes
4-14
é 2 -1 -1 0 0 0 ùú
ê 3 3 3
ê- 1 2 -1 0 0 0 úú
ê 3 3 3
ê- 1 - 1 2 0 0 0 úú
ê
IDEV = ê 3 3 3
ú
ê 0 0 0 21 0 0ú
ê ú
ê 0 0 0 0 21 0ú
ê ú
ê 0 0 0 0 0 1ú
ë 2û
P4.14 A solid shaft as shown in the figure is subjected to tensile force P and a torque T.
The force and torque are such that the normal stress sxx = s and shear stress t = s. The
shear stress is along the circumference of the shaft. Using the von Mises criterion,
determine the values of s when the material yields first time. The yield stress from the
uniaxial tension test is sY.
X2
P
X1
T
X3
Figure P4.14
Solution:
In the case of P, the stress component is fixed to be s11. In the case of T, the shear stress
component varies at different location on the boundary of the cross-section. For the
simplicity, let’s consider the case in which t12 = s. Thus, the stress matrix and deviatoric
stress become
és s 0ù é2s s 0 ùú
ê ú ê3
s = êê s 0 0 úú , s = êê s - 13 s 0 úú
ê 0 0 0ú ê 0 0 - 13 s úú
ëê úû êë û
The norm of deviatoric stress can be calculated by
s =s 4 + 19 + 19 + 1 + 1 = s 8
9 3
f = s - 2s = 8s - 2s =0
3 Y 3 3 Y
1
s= s
2 Y
4-15
▄
P4.15 A plane stress plate is under biaxial stress state in which sxx = -syy = s. When
the applied load is proportional, determine s when the material yields first time. The
yield stress from the uniaxial tension test is sY.
Solution:
In the case of biaxial loading, the stress matrix and deviatoric stress become
és 0 0ù
ê ú
s = s = êê 0 -s 0 úú
ê 0 0 0ú
êë úû
s =s 2
f = s - 2s = 2s - 2s =0
3 Y 3 Y
s= 1s
3 Y
P4.16 A square is under proportional loading with shear stress t12 = t21 = t. When the
effective plastic strain is ep = 0.1, calculate the value of shear stress. Consider three
different hardening models: (a) isotropic, (b) kinematic, and (c) combined hardening with
b = 0.5. Assume that the initial yield stress is 400MPa and the plastic modulus is H =
200MPa.
Solution:
Since the applied stress is proportional loading, it is expected that the material is in the
plastic phase, and all three models provide the same stress value. The difference occurs
only when the direction of loading changes. In the case of pure shear loading, the stress
and deviatoric stress become
é 0 t 0ù
ê ú
s = s = êê t 0 0 úú
ê 0 0 0ú
êë úû
s = 2t
s = 171.5MPa
(b) Kinematic hardening: from the definition of yield function,
s-a - 2s 0 = 0
3 Y
Note that Da is parallel to h and the loading direction remains fixed, and thus, a is
parallel to s. In that case, the norm of the shifted stress can be written as
s - a = s - a . Thus, the yield function can be rewritten as
s-a - 2s 0 = s - a - 2s 0 = 2t - 2 He - 2s 0 = 0
3 Y 3 Y 3 p 3 Y
t = 1 2 (s 0 + He p ) = 171.5MPa
2 3 Y
s-a - é 0 2 ù
êë sY + (1 - b )Hep úû
3
= s - a - 23 éê sY0 + (1 - b )Hep ùú
ë û
2 2 0 2
= 2t - 3 bHep - 3sY - 3 (1 - b )Hep
=0
Thus, the applied stress can be solved for
t = 1 2 (s 0 + He p ) = 171.5MPa
2 3 Y
Note that all three models provide the same stress value. ▄
P4.17 A pure shear deformation is applied to the square element as shown in the figure
such that s12 = s21 is only non-zero stress component. At load step n, the stress value
was s12 = 50, and there was no plastic deformation. At load step n+1, incremental strain
De12 = De21 = 0.005 is applied. Calculate stress components and effective plastic strain
at load step n+1. Use the following material properties: shear modulus m = 1,000,
plastic modulus H = 100, initial yield stress sY = 100.
4-17
X2
Figure P4.17
X1
Solution:
Since s12 is only non-zero component, it is convenient to work with it as a scalar rather
n
than stress matrix. With given information of s12 = 50 and De12 = 0.005 , the trial
state can be calculated as
tr n
s12 = s12 + 2mDe12 = 60
f = htr - 2sn
3 Y
= tr
2s12 - 2s 0
3 Y
= 3.2032 > 0
Thus, the material becomes plastic in this load increment. The plastic consistency
parameter can be calculated by
f
Dg = = 0.00155
2m + 23 H
In addition, the deviatoric unit tensor for the trial state becomes
é0 1 0ù
htr 1 êê ú
ú
N= = ê 1 0 0 ú
htr 2ê ú
êë 0 0 0 úû
Thus, only s12 will be updated due to plastic deformation. The updated stress and plastic
strain become
n +1 tr 2mDg
s12 = s12 - = 57.808
2
x1
Figure P4.18
Solution:
From the infinitesimal deformation assumption, the strain at load step n can be calculated
by
é 0 1k ù
en = êê 1 2 ú n
e12 = 0.008, De12 = 0.004
êë 2 k 0 úú
û
Since e12 is only non-zero component, strain will be considered as a scalar in the
following calculations. For the given strain at load step n, the stress becomes
é 0 mk ù é 0 1.6 ù
sn = sn = 2me = 2me = êê ú=ê ú
ê 1.6 0 ú , n
s12 = 1.6
êë mk 0 úú ê ú
û ë û
Again, since s12 is only non-zero component, stress will be considered as a scalar. For a
given increment De12, the trial stress becomes
tr n
s12 = s12 + 2mDe = 2.4
The yield function can be used to check the status of the material, as
4-19
f = str
12 -
2s 0
3 Y
= tr 2
2(S12 ) - 8 = 0.566 > 0
Thus, the material becomes plastic in this increment, yieldStatus = 1. The unit
deviatoric tensor and the plastic consistency parameter can be calculated by
str 1 éê 0 2.4 ùú
N= =
str 3.394 êêë 2.4 0 úúû
f 0.5657
Dg = = = 0.002737
2m + 3 H
2 200 + 6.667
Using the unit deviatoric tensor and the plastic consistency parameter, the stress and the
effective plastic strain can be updated by
é 0 2.4 ù 0.5473 é 0 2.4 ù é 0 2.013 ù
sn +1 = str - 2mDg N = êê ú-
ú
ê
ê
ú=ê
ú ê
ú
ú
êë 2.4 0 úû 3.394 êë 2.4 0 úû êë 2.013 0 úû
epn +1 = epn + 2 Dg
3
= 0.002234
Thus, the updated state of the material is on the surface of the yield function.
Below is the MATLAB program that solves the above problem:
%
% P4.18 shear deformation of a square
%
lambda=100; mu=100;
mp = [lambda, mu, 0, 10, sqrt(12)];
Iden=[1 1 1 0 0 0]';
stressN=[0 0 0 0 0 0]';
alphaN=[0 0 0 0 0 0]';
epN=0;
D=2*mu*eye(6) + lambda*Iden*Iden';
D(4,4) = mu; D(5,5) = mu; D(6,6) = mu;
deps=[0 0 0 0.016 0 0]';
[stress, alpha, ep]=combHard(mp,D,deps,stressN,alphaN,epN);
stressN=stress; alphaN=alpha;epN=ep;
deps=[0 0 0 0.008 0 0]';
[stress, alpha, ep]=combHard(mp,D,deps,stressN,alphaN,epN)
▄
P4.19 At load step tn, a unit cube is under unaxial stress state with s11 = 100MPa, and
all other stress components and plastic variables are zero. At load step tn+1, additional
shear stress is applied such that Dg12 = 0.002. Determine stress, back stress, and
4-20
Solution:
At load step tn, the material reaches the initial yield stress. At load step tn+1, the trial stress
becomes
tr n
s= s + D : De = [100 0 0 200 0 0]T
Since back stress is zero at load step tn, the deviatoric stress is the same with the shifted
stress as
tr tr tr
h= s= s - tr ( tr s)1 = [66.67 -33.33 -33.33 200 0 0]T
From the shifted stress, the unit deviatoric tensor that is normal to the yield surface can
be obtained as
tr
h
N= = [.2265 -.1132 -.1132 .6794 0 0]T
tr
h
Using the norm of the shifted stress, the yield function can be evaluated as
tr 2s 0 2
f = h - 3 Y
= 294.4 - 100 3
= 212.7 > 0
It is clear that the material is in the plastic state. The plastic consistency parameter can be
calculated from
f
g = = 0.001
2m + 23 H
Using and N, the stress, back stress, and effective plastic strain can be updated as
n +1 tr
s= s - 2mg N = [53.4 23.3 23.3 60.1 0 0]T
n +1
a = n a + 23 b H g N = [.777 -.389 -.389 2.331 0 0]T
n +1
e p = ne p + 2g
3
= 8.405 ´ 10-4
Below is the MATLAB program that solves for the above problem:
lambda=100000; mu=100000;
mp = [lambda, mu, 0.5, 10000, 100];
Iden=[1 1 1 0 0 0]';
stressN=[100 0 0 0 0 0]';
alphaN=[0 0 0 0 0 0]';
epN=0;
D=2*mu*eye(6) + lambda*Iden*Iden';
D(4,4) = mu; D(5,5) = mu; D(6,6) = mu;
deps=[0 0 0 0.002 0 0]';
[stress, alpha, ep]=combHard(mp,D,deps,stressN,alphaN,epN)
4-21
▄
P4.20 Using Abaqus perform a uniaxial tension test of a unit cube (C3D8) in x3-
direction. Assume elastoplastic material with linear isotropic hardening (E = 2.0E5, =
0.3, Y = 200, H = 2.0E4). Displace at x3 = 1 surface is controlled as shown in the figure
with three steps. Use 10 increments in each step. Plot stress-strain curve for all 30
increments.
0.006
0.004
0.001
t1 t2 t3 Figure P4.20
Solution:
The following program list shows the ABAQUS input file for the elastoplastic loading:
4-22
*HEADING *BOUNDARY
MISES PLASTICITY/LINEAR ELASTICITY, 7,3,,.001
UNIAXIAL TENSION TEST, C3D8 5,3,,.001
*NODE,NSET=ALLN 6,3,,.001
1,0.,0.,0. 8,3,,.001
2,1.,0.,0. *EL PRINT,FREQ=1
3,1.,1.,0. S,
4,0.,1.,0. E,
5,0.,0.,1. EP,
6,1.,0.,1. *NODE PRINT
7,1.,1.,1. U,RF
8,0.,1.,1. *EL FILE,FREQ=1
*ELEMENT,TYPE=C3D8,ELSET=ALLE S,
1,1,2,3,4,5,6,7,8 E,EP
*SOLID *END STEP
SECTION,ELSET=ALLE,MATERIAL=ALLE *STEP,INC=10
*MATERIAL,NAME=ALLE *STATIC,DIRECT
*ELASTIC 1.,10.
200.E3,.3 *BOUNDARY
*PLASTIC 7,3,,.006
200.,0. 5,3,,.006
400.,.01 6,3,,.006
*BOUNDARY 8,3,,.006
1,PINNED *EL PRINT,FREQ=1
2,2 S,
5,2 E,
6,2 EP,
4,1 *NODE PRINT
5,1 U,RF
8,1 *EL FILE,FREQ=1
2,3 S,
3,3 E,EP
4,3 *END STEP
*STEP,INC=10
*STATIC,DIRECT
▄
4-23
P4.21 Calculate Dep and Dalg for one-dimensional elastoplasticity problem using the von
Mises yield criterion and linear combined isotropic/kinematic hardening. Assume
material properties: (E , H , sY0 , b ) .
Solution:
(a) In one-dimensional problems, the lateral strains are ignored, and stress s and strain e
are scalar. In addition, there is no need to calculate deviatoric stress and strain. The yield
function becomes
Note that the above yield function is slightly different from the multi-dimensional yield
function in Eq. (4.84) because equivalent stress and effective strain are not used. The
constitutive relation is s = E (e - e p ) with e p = ep = g , and the hardening models are
a = bH g
There is no difference between the plastic strain and effective plastic strain. From the
consistency condition:
f = E (e - e p ) - a - (1 - b )Hep = 0
E e
g =
E +H
Thus,
æ E ÷ö÷ e = EH e
s = E e - E e p = E çç 1 - ÷ø
èç E +H E +H
EH
Dep =
E +H
Note that the above Dep is the same as the tangent modulus in Eq. (4.8).
f ( n +1h, n +1ep ) = tr
h - (E + b H )Dg - (sY0 + (1 - b )H ( tre p + Dg )) = 0
¶Dg E
=
¶De E +H
E2 EH
D alg = E - =
E +H E +H
Note that Dalg = Dep because in the case of one-dimensional case, the direction N is
fixed. ▄
P4.22 In the saturated isotropic hardening model, the yield stress starts from initial value
of sY0 and approaches sY¥ as the plastic strain increases.
é æ e öù ÷
ç
- sY0 ) êê 1 - ÷÷ úú
p ÷
k(ep ) = sY0 + (sY¥ ç
exp ç -
êë ççè ep ÷ø úû
¥
Solution:
The return-mapping algorithm will be similar to the linear isotropic hardening model
except for the yield function and the local Newton-Raphson method to find the plastic
consistency parameter.
1. Yield function
æ é æ k öù ö
2ç
çç sY0 ê çç e p ÷÷ ú ÷÷
f k
= tr
h - + (sY¥ - sY0 ) ê 1 - exp ç - ÷ ÷
3ç
çè êë ççè e ¥ ÷ø÷ úú ø÷÷
p û
2. Jacobian relation
4-25
¶f 2 (sY¥ - sY0 ) éê æ e k öù
ç p ÷
= 2m + ê 1 - exp ç
ç - ÷÷÷ úú
¶g 3 ep¥ êë ççè e ¥ ÷ø ú
p û
3. Update the plastic consistency parameter and effective plastic strain
fk
g k +1 = g k + , epk +1 = nep + 2 g k +1
¶f ¶g 3
4. Check convergence
Below is MATLAB program, combHardSat, which can solve for stress and effective
plastic strain from the saturated hardening model:
%
% Saturated isotropic hardening model
%
function [stress, ep]=combHardSat(mp,D,deps,stressN,epN)
% Inputs:
% mp = [lambda, mu, epinf, Y0, Yinf];
% D = elastic stiffness matrix
% stressN = [s11, s22, s33, t12, t23, t13];
%
Iden = [1 1 1 0 0 0]';
two3 = 2/3; stwo3=sqrt(two3); %constants
mu=mp(2);epinf=mp(3);Y0=mp(4);Yinf=mp(5); %material properties
ftol = Y0*1E-6; %tolerance for yield
stresstr = stressN + D*deps; %trial stress
I1 = sum(stresstr(1:3)); %trace(sigmatr)
eta = stresstr - I1*Iden/3; %deviatoric stress
etat = sqrt(eta(1)^2 + eta(2)^2 + eta(3)^2 ...
+ 2*(eta(4)^2 + eta(5)^2 + eta(6)^2));%norm of eta
fyld = etat - stwo3*(Y0+(Yinf-Y0)*(1-exp(-epN/epinf)))
if fyld < ftol %yield test
stress = stresstr; ep = epN; %trial states are final
return;
else
iter =0; gamma = 0; ep = epN;
while fyld > ftol %local N-R iteration
iter=iter+1;
if iter > 40; error('Fail to compute gamma'); end;
dfdg = 2*mu+two3*((Yinf-Y0)*exp(-ep/epinf)/epinf)
gamma = gamma + fyld/dfdg %update plast cons param
ep = epN + stwo3*gamma;
fyld = etat -2*mu*gamma - stwo3*(Y0+(Yinf-Y0)*(1-exp(-ep/epinf)))
end
end
N = eta/etat; %unit vector normal to f
stress = stresstr - 2*mu*gamma*N; %updated stress
Below is the MATLAB program that can solve for P4.15 with the saturated isotropic
hardening model:
4-26
%
% P4.22 shear deformation of a square (saturated isotropic hardening)
%
lambda=1000; mu=1000;epinf=0.05;Y0=100;Yinf=200;
mp = [lambda, mu, epinf, Y0, Yinf];
Iden=[1 1 1 0 0 0]';
deps=[0 0 0 0.01 0 0]';
stressN=[0 0 0 50 0 0]';
alphaN=[0 0 0 0 0 0]';
epN=0;
D=2*mu*eye(6) + lambda*Iden*Iden';
D(4,4) = mu; D(5,5) = mu; D(6,6) = mu;
[stress, ep]=combHardSat(mp,D,deps,stressN,epN)
After convergence, the stress and effective plastic strain are updated to
n +1 n +1
s12 = 58.63, ep = 7.87 ´ 10-4
x 1 = X1 + kX 2 , x 2 = X 2, x3 = X3
Using the linear relationship between principal Kirchhoff stress and logarithmic stretch,
find the Kirchhoff stress tensor when k = 0.02. Use the following material properties: l
= m = 100GPa.
Solution:
Since the cube is elastic, there is no need to separate elastic and plastic part of
deformation. Thus, the superscript ‘e’ will be omitted in the following derivation. For
given deformation, the deformation gradient and left Cauchy-Green deformation tensor
becomes
é 1 0.02 0 ù é 1.0004 0.02 0 ù
ê ú ê ú
F = êê 0 1 0 úú , b = FFT = êê 0.02 1 0 úú
ê0 0 1 úú ê 0 0 1 úú
ëê û ëê û
The three eigenvalues and eigenvectors of b become
ê1 1 3ú ï ï ï
ï ï
ï 0 ïï ï
ëê úû ï
ï 0 ï
î ï
þ îï
ï þï
Y
DY
DX
1m
X
1m Figure P4.24
5 -R 0 Plastic flow
6 -R 0 Plastic flow
7 0 -R Elastic unloading
8 0 -R Plastic flow
Solution:
The problem is modeled using one square element in ANSYS. Since the strain in Z-
direction is constrained, stress in Z-direction. Below is the table of stresses at each load
step.
/COM,ANSYS MEDIA REL. 11.0 (10/27/2006) REF. VERIF. MANUAL: REL. 11.0
/VERIFY,VMR049-PL1A-182
/TITLE, VMR049-PL1A-182, 2D PLANE STRAIN PLASTICITY BENCHMARK
/COM, REFERENCE: NAFEMS REPORT-R0049
/PREP7
R = 2.5E-5
ET,1,182,,,
KEYOPT,1,3,2
N,1,,,,
N,2,0,1,,
N,3,1,0,,
N,4,1,1,,
E, 1,3,4,2
MP,EX,1,250E3,
MP,NUXY,1,0.25,
TB,BISO,1,1, , ,
TBMODIF,2,1,5
TBMODIF,3,1,0.0
TB,HILL,1
TBDATA,1,1.0,1.0,1.0,1.0,1.0,1.0
NSEL,S,LOC,X
D,ALL,UX
NSEL,S,LOC,Y
D,ALL,UY
NSEL,S,LOC,Y,1.0
D,ALL,UY
FINISH
/SOLU
NLGEOM,ON
NSEL,S,LOC,X,1.0
D,ALL,UX,R
4-29
NSEL,ALL
NSUBST,10,10,10,
OUTRES,ALL,5
SOLVE
NSEL,S,LOC,X,1.0
D,ALL,UX,2*R
NSEL,ALL
NSUBST,10,10,10,
OUTRES,ALL,5
SOLVE
NSEL,S,LOC,Y,1.0
D,ALL,UY,R
NSEL,ALL
NSUBST,10,10,10,
OUTRES,ALL,5
SOLVE
NSEL,S,LOC,Y,1.0
D,ALL,UY,2*R
NSEL,ALL
NSUBST,10,10,10,
OUTRES,ALL,5
SOLVE
NSEL,S,LOC,X,1.0
D,ALL,UX,R
NSEL,ALL
NSUBST,10,10,10,
OUTRES,ALL,5
SOLVE
NSEL,S,LOC,X,1.0
D,ALL,UX,0.0
NSEL,ALL
NSUBST,10,10,10,
OUTRES,ALL,5
SOLVE
NSEL,S,LOC,Y,1.0
D,ALL,UY,R
NSEL,ALL
NSUBST,10,10,10,
OUTRES,ALL,5
SOLVE
NSEL,S,LOC,Y,1.0
D,ALL,UY,0.0
NSEL,ALL
NSUBST,10,10,10,
OUTRES,ALL,5
SOLVE
FINISH
/POST26
/GROPT,VIEW,0
/GTHK,CURVE,1
/GROPT,FILL,OFF
/GRID,1
/GTHK,GRID,1
/GROPT,CGRID,1
/AXLAB,X,STEP
/AXLAB,Y,STRESS
/GTHK,AXIS,1
/GRTYP,0
/XRANGE,0,8
/YRANGE,-10,25
ESOL,2,1,4,S,X,
ESOL,3,1,4,S,Y,
4-30
ESOL,4,1,4,S,Z,
ESOL,5,1,4,S,EQV,
PLVAR,2,3,4,5
PRVAR,2,3,4,5
*DIM,VALUEX,ARRAY,8,1
*DO,JJ,1,8,1
*GET,VALUEX(JJ,1),VARI,2,RTIME,JJ
*ENDDO
R1=VALUEX(1,1)/7.500
R2=VALUEX(2,1)/11.666
R3=VALUEX(3,1)/14.166
R4=VALUEX(4,1)/16.418
R5=VALUEX(5,1)/9.927
R6=VALUEX(6,1)/5.134
R7=VALUEX(7,1)/2.635
R8=VALUEX(8,1)/1.218
*DIM,VALUEY,ARRAY,8,1
*DO,JJ,1,8,1
*GET,VALUEY(JJ,1),VARI,3,RTIME,JJ
*ENDDO
R9=VALUEY(1,1)/2.500
R10=VALUEY(2,1)/6.666
R11=VALUEY(3,1)/14.166
R12=VALUEY(4,1)/19.669
R13=VALUEY(5,1)/15.622
R14=VALUEY(6,1)/10.745
R15=VALUEY(7,1)/3.245
R16=VALUEY(8,1)/(-3.715)
*DIM,VALUEZ,ARRAY,8,1
*DO,JJ,1,8,1
*GET,VALUEZ(JJ,1),VARI,4,RTIME,JJ
*ENDDO
R17=VALUEZ(1,1)/2.500
R18=VALUEZ(2,1)/6.666
R19=VALUEZ(3,1)/9.166
R20=VALUEZ(4,1)/13.912
R21=VALUEZ(5,1)/11.951
R22=VALUEZ(6,1)/9.120
R23=VALUEZ(7,1)/6.620
R24=VALUEZ(8,1)/3.521
*DIM,VALUEEF,ARRAY,8,1
*DO,JJ,1,8,1
*GET,VALUEEF(JJ,1),VARI,5,RTIME,JJ
*ENDDO
R25=VALUEEF(1,1)/5.000
R26=VALUEEF(2,1)/5.000
R27=VALUEEF(3,1)/5.000
R28=VALUEEF(4,1)/5.000
R29=VALUEEF(5,1)/5.000
R30=VALUEEF(6,1)/5.000
R31=VALUEEF(7,1)/3.719
R32=VALUEEF(8,1)/5.000
*DIM,STEP,CHAR,10
*DIM,TARGETX,CHAR,10
*DIM,TARGETY,CHAR,10
*DIM,TARGETZ,CHAR,10
*DIM,TARGETEF,CHAR,10
*DIM,RATIOX,,8,1
4-31
*DIM,RATIOY,,8,1
*DIM,RATIOZ,,8,1
*DIM,RATIOEF,,8,1
STEP(1)='1.0','2.0','3.0','4.0','5.0','6.0','7.0','8.0'
TARGETX(1)='7.500','11.666','14.166','16.418','9.927','5.134','2.635','1.218'
TARGETY(1)='2.500','6.666','14.166','19.669','15.622','10.745','3.245','-3.715'
TARGETZ(1)='2.500','6.666','9.166','13.914','11.951','9.120','6.620','3.521'
TARGETEF(1)='5.000','5.000','5.000','5.000','5.000','5.000','3.719','5.000'
*VFILL,RATIOX,DATA,R1,R2,R3,R4,R5,R6,R7,R8
*VFILL,RATIOY,DATA,R9,R10,R11,R12,R13,R14,R15,R16
*VFILL,RATIOZ,DATA,R17,R18,R19,R20,R21,R22,R23,R24
*VFILL,RATIOEF,DATA,R25,R26,R27,R28,R29,R30,R31,R32
/COM,
/COM, ----------- VMR049-PL1A-182 RESULTS COMPARISON --------------
/COM,
/COM, vmr049-pl1a-182.jpeg RESULTS SHOULD MATCH R0049 NAFEMS MANUAL
/COM, GRAPH RESULTS ON PAGE 49, FIGURE 2.14(A). THE RESULTS
/COM, DISPLAYED ARE INCREMENTED FOR THIS PURPOSE.
/COM,
/COM,
/COM, ------ VMR049-PL1A-182 STRESS RESULTS IN X DIRECTION -------
/COM,
/COM, | STEP | TARGET | ANSYS | RATIO
/COM,
*VWRITE,STEP(1),TARGETX(1),VALUEX(1,1),RATIOX(1,1)
(1X,A8,' ',1x,A8,' ',F8.3,' ',F8.3,' ')
/COM,
/COM, ------ VMR049-PL1A-182 STRESS RESULTS IN Y DIRECTION -------
/COM,
/COM, | STEP | TARGET | ANSYS | RATIO
/COM,
*VWRITE,STEP(1),TARGETY(1),VALUEY(1,1),RATIOY(1,1)
(1X,A8,' ',1x,A8,' ',F8.3,' ',F8.3,' ')
/COM,
/COM,
/COM, ------ VMR049-PL1A-182 STRESS RESULTS IN Z DIRECTION -------
/COM,
/COM, | STEP | TARGET | ANSYS | RATIO
/COM,
*VWRITE,STEP(1),TARGETZ(1),VALUEZ(1,1),RATIOZ(1,1)
(1X,A8,' ',1x,A8,' ',F8.3,' ',F8.3,' ')
/COM,
/COM,
/COM, ------ VMR049-PL1A-182 EFFECTIVE STRESS RESULTS -------
/COM,
/COM, | STEP | TARGET | ANSYS | RATIO
/COM,
*VWRITE,STEP(1),TARGETEF(1),VALUEEF(1,1),RATIOEF(1,1)
(1X,A8,' ',1x,A8,' ',F8.3,' ',F8.3,' ')
/COM,
/COM,
FINISH