Exam 3 Sample Questions

9. For the stress matrix below, the two principal stresses are given as σ3 = −3 and σ1 = 2,
respectively. In addition, the two principal stress directions corresponding to the two
principal stresses are also given below.

2 1
1 0 2
5 5
[ ] 0 1 0 , n1 0 and n 3 0
2 0 2 1 2
5 5

(a) What is the normal and shear stress on a plane whose normal vector is parallel to
(2, 1, 2)?
(b) Calculate the missing principal stress σ2 and the principal direction n2.
(c) Write stress matrix in the new coordinates system that is aligned with n1, n2, and
n3 .
Solution:
n 1.4444, n 1.4229
2 1 . n2 n3 n1 {0 1 0}T

2 0 0
T
N N 0 1 0
0 0 3

11. A solid shaft of diameter d = 5 cm, as shown in the figure, is subjected to tensile
force P = 13,000 N and a torque T = 6,000 Ncm. At point A on the surface, what is
the state of stress (write in matrix form), the principal stresses, and the maximum
shear stress? Show the coordinate system you are using.

A P

6.62 2.44 0
[ ]A 2.44 0 0 MPa
0 0 0

1 7.43, 2 0, 3 0.81 MPa

1 2
max 4.11 MPa
2
12. If the displacement field is given by

ux x2 2y 2
uy y2 2x (y z)
2
uz z 2xy

(a) Write down 3×3 strain matrix.

(b) What is the normal strain component in the direction of (1,1,1) at point (1,–3,1)?

2x y z y
y z 2(x y ) 0
y 0 2z

n ny 3
2.

14. The displacement field in a solid is given by

u kx 2
v 2kxy 2
w k (x y )z

where k is a constant.
(a) Write down the strain matrix.
(b) What is the normal strain in the direction of n = {1, 1, 1}T?

2kx ky 2 1
2
kz
[ ] ky 2 4kxy 1
2
kz
1 1
2
kz 2
kz k (x y)

1
n [ ] n 2ky 2 4kxy 3kx ky 2kz
3

26. The figure below illustrates a thin plate of thickness t. An approximate displacement
field, which accounts for displacements due to the weight of the plate, is given by

ux (x , y ) (2bx x2 y2)
2E
uy (x , y ) y(b x)
E

(a) Determine the corresponding plane stress field.

 CHAP 8 Structural Design Using Finite Elements 3

(b) Qualitatively draw the deformed shape of the plate.

A B
a a
x

Solution:

xx (b x)
yy 0
xy 0

Thus, xx (b x ) is the only non-zero stress component.

(b) The deformed geometry is sketched below

29. The state of stress at a point is given by

80 20 40
[ ] 20 60 10 MPa
40 10 20

(a) Determine the strains using Young’s modulus of 100 GPa and Poisson’s ratio of
0.25.
(b) Compute the strain energy density using these stresses and strains.
(c) Calculate the principal stresses.
(d) Calculate the principal strains from the strains calculated in (a).
(e) Show that the principal stresses and principal strains satisfy the constitutive
relations.
 (f) Calculate the strain energy density using the principal stresses and strains.
Solution:

xx 0.6
3
yy 10 0.35
zz 0.15
xy 3
xy 0.5 10
G
yz 3
yz 1.0 10
G
zx 3
zx 0.25 10
G

(b) Strain energy density: U 59.25kPa

(c) Principal stresses: 1 110, 2 50, 3 0 MPa
3
(d) Principal strains: 1 0.975 10 , 2 0.225 10 3, 3 0.4 10 3

(e) From Eq. (1.55)

1 1 .25 .25 110 0.975

1
2 .25 1 .25 50 106 10 3
0.225
1011
3 .25 .25 1 0 0.4

Thus, the principal stresses and principal strains satisfy the constitutive relations.
(f) Strain energy density

1
U 1 1 2 2 3 3 59.25kPa
2

30. Consider the state of stress in Problem 29 above. The yield strength of the material is
100 MPa. Determine the safety factors according to the following: (a) maximum
principal stress criterion, (b) Tresca Criterion, and (c) von Mises criterion.
Solution:
(a) Maximum principal stress criterion SF 0.91
(b) Tresca criterion SF 0.91
(c) Von Mises criterion SF 1.048

31. A thin-walled tube is subject to a torque T. The only non-zero stress component is
the shear stress xy, which is given by xy = 10,000 T (Pa), where T is the torque in
N.m. When the yield strength Y = 300 MPa and the safety factor N = 2, calculate the
maximum torque that can be applied using
(a) Maximum principal stress criterion (Rankine)
(b) Maximum shear stress criterion (Tresca)
(c) Distortion energy criterion (Von Mises)
 CHAP 8 Structural Design Using Finite Elements 5

Solution:
Since it is a pure shear stress state, the three principal stresses are
(a) Maximum stress criterion, T 15, 000 N m
(b) Maximum shear stress criterion T 7, 500 N m
(c) Von Mises criterion T 8, 660 N m

32. A thin-walled cylindrical pressure vessel with closed ends is subjected to an internal
pressure p = 100 psi and also a torque T around its axis of symmetry. Determine T
that will cause yielding according to von Mises’ yield criterion. The design requires a
safety factor of 2. The nominal diameter D of the pressure vessel = 20 inches, wall
thickness t = 0.1 inch, and yield strength of the material = 30 ksi. (1 ksi = 1000 psi).
Stresses in a thin walled cylinder are: longitudinal stress l, hoop stress h, and shear
stress  due to torsion. They are given by

pD pD 2T
l , h ,
4t 2t D 2t

T 444 103 lb-in

35. The figure below shows a shaft of 1.5 in. diameter loaded by a bending moment Mz =
5,000 lb∙in, a torque T = 8,000 lb∙in, and an axial tensile force N = 6,000 lb. If the
material is ductile with the yielding stress σY = 40,000 psi, determine the safety factor
using: (a) the maximum shear stress theory and (b) the maximum distortion energy
theory.

y
Mz T Mz
N N x

(a) The maximum shear stress theory N 1.315

(b) Maximum distortion energy theory NVM 1.4332

37. A circular shaft of radius r in the figure has a moment of inertia I and polar moment
of inertia J. The shaft is under torsion Tz in the positive z-axis and bending moment
Mx in the positive x-axis. The material is mild steel with yield strength of 2.8 MPa.
Use only the given coordinate system for your calculations.
(a) If Tz and Mx are gradually increased, which point (or points) will fail first among
four points (A, B, C, and D)? Identify all.
(b) Construct stress matrix []A at point A in xyz-coordinates in terms of given
parameters (i.e., Tz, Mx, I, J, and r).
(c) Calculate three principal stresses at point B in terms of given parameters.
(d) When the principal stresses at point C are 1 = 1, 2 = 0, and 3 = 2 MPa,
calculate safety factors (1) from maximum shear stress theory and (2) from
distortion energy theory.
 z x

A
Tz B D x
y
C
Mx y

Solution:
(a) The bending moment will produce maximum stress at points A and C. Thus, A and
C will fail first.
(b) At point A,

Mxr / I 0 Tz r / J
[ ] 0 0 0
Tz r / J 0 0

(c) At point B, the three principal stresses are

Tz r Tz r
1 , 2 0, 3 ,
J J

(d) For maximum shear stress criterion, N 0.933

For von Mises criterion, N 1.06

3. Using two CST elements, solve the simple shear problem described in the figure and
determine whether the CST elements can represent the simple shear condition
accurately or not. Material properties are given as E = 10 GPa,  = 0.25, and
thickness is h = 0.1 m. The distributed force f = 100 kN/m2 is applied at the top edge.

4 f 3

1m

1m
1 2
x

By combining with zero displacements, we have nodal displacements, as

{Qs } 10 5 {0, 0, 0, 0, 0.25, 0, 0.25, 0} m

 CHAP 8 Structural Design Using Finite Elements 7

{ (1)
} [C]{ (1)
} {0, 0, 105 }T Pa
{ (2)
} [C]{ (2)
} {0, 0, 105 }T Pa

Note that only shear stress exists, which satisfy the pure shear condition. Since
distributed force f = 10 kN/m2 is applied at the top edge, the above shear stress is exact.
Thus, the CST element can represent the pure shear condition accurately. The figure
below shows the deformed and undeformed shape of the elements.

5. A structure shown in the figure is approximated with one triangular element. Plane
strain assumption is used.
(a) Calculate the strain–displacement matrix [B].
(b) When nodal displacements are given by {u1, v1, u2, v2, u3, v3} = {0, 0, 2, 0, 0, 1},
calculate the element strain vector.

3 (0,20)

2 (10,10)

1 (0,0)

Solution:

10 0 20 0 10 0
1
B 0 10 0 0 0 10
200
10 10 0 20 10 10

xx = 0.2, yy = 0.05, and xy = −0.05.

6. Calculate the shape function matrix [N] and strain-displacement matrix [B] of the
triangular element shown in the figure

3(0,1)

x
1(0,0) 2(1,0)

1 x y 0 x 0 y 0
[N ]
0 1 x y 0 x 0 y

u1
u
v1
x 1 0 1 0 0 0
v u2
{ } 0 1 0 0 0 1
y v2
u v 1 1 0 1 1 0
u3
y x [B] v3

7. The coordinate of the nodes and corresponding displacements in a triangular element

are given in the table. Calculate the displacement u and v and strains xx, yy, and xy at
the centroid of the element given by the coordinates (1/3, 1/3)
Node x (m) y (m) u (m) v (m)
1 0 0 0 0
2 1 0 0.1 0.2
3 0 1 0 0.1
Solution:

u( 13 , 13) u2N 2 ( 13 , 13) 0.1 1

3
0.033
v( 13 , 13) v2N 2 ( 13 , 13) v3N 3 ( 13 , 13) 0.1

u N2
xx u2 0.1
x x
v N2 N3
yy v2 v3 0.1
y y y
u v
xy 0.2
y x

8. For a rectangular element shown in the figure, displacements at the four nodes are
given by {u1,v1,u2,v2,u3,v3,u4,v4} = {0.0, 0.0, 1.0, 0.0, 2.0, 1.0, 0.0, 2.0}. Calculate
displacement (u, v) and strain εxx at (x, y) = (2, 1).
 CHAP 8 Structural Design Using Finite Elements 9

y
4 (0,2) 3 (3,2)

x
1 (0,0) 2 (3,0)
Solution:

2 1
u 1, v , xx
3 2

9. The quadrilateral element shown in the figure has the nodal displacements of {u1, v1,
u2, v2, u3, v3, u4, v4} = {–1, 0, –1, 0, 0, 1, 0, 1}.
(a) Find the (s, t) reference coordinates of point A (0.5, 0) using iso–parametric
mapping method.
(b) Calculate the displacement at point B whose reference coordinate is (s,t)=(0,−0.5)
(c) Calculate the Jacobian matrix [J] at point B.

y t
4 1 4 1 3

1 1 x -1 1 s
-1 A 3
B

-1 2 1 -1 2

Physical Element Reference Element

Solution:
(a) s = t = 0.5
(b) u 0.75, v 0.25
(c)

1 1
[ J[ 2 2
1 1
2 2

10. A four-node quadrilateral element is defined as shown in the figure.

(a) Find the coordinates of parent element corresponding to (x, y) = (0, 0.5).
(b) Calculate the Jacobian matrix as a function of s and t.
(c) Is the mapping valid? Provide reasons for your answer.
 y
+1 4

1 2 3
-1 +1 x
Solution:
(a) s 1 2, t 2 1
(b) Jacobian matrix

1 1
(1 t )
[ J] 2 4
1 1
(1 s )
2 4

(c) The mapping is not valid at node 2.

 CHAP 8 Structural Design Using Finite Elements 11

11. A 2m2m1mm square plate with E = 70 GPa and = 0.3 is subjected to a uniformly
distributed load as shown in Figure (a). Due to symmetry it is sufficient to model one
quarter of the plate with symmetry boundary conditions as shown in Figure (b). Use
two triangular elements to find the displacements, strains and stresses in the plate.
Check the answers using simple calculations from mechanics of materials.

yy = 100 MPa

yy = 100 MPa

3
41
2m 2
1m
1
2 x
2m 1
1m

yy = 100 MPa

Solution:

{u2 u3 v3 v4 } { 0.428 0.428 1.43 1.43} mm

0 0
1 100 MPa, 2 100 MPa
0 0

The structure is under uniform stress field (x = 0, and y = 100MPa) and the finite
element solution yields the exact stress field. This is possible because the triangular
elements can represent constant stress state exactly.

12. Six rectangular elements are used to model the cantilevered beam shown in the
figure. Draw the graph of xx along the top surface schematically, similar to the one
in the figure. You don’t have to provide accurate values of stress.

xx
?

Solution:
Since displacement is linear on the top surface, the strain and stress are constant along the
top surface within an element. Also, the stress is not continuous at the element boundary.
The actual stress varies linear along x-direction because the bending moment changes
linearly. However, the stress from finite elements will approximate this linearly varying
stress using piece-wise constant stresses, as illustrated in the figure below.

xx

13. A rectangular element as shown in the figure is used to represent a pure bending
problem. Due to the bending moment M, the element is deformed as shown in the
figure with displacement {d} = {u1, v1, u2, v2, u3, v3, u4, v4}T = {−1, 0, 1, 0, −1, 0, 1,
0}T.
(a) Derive expressions of strain component xx, yy, and xy, as functions of x and y
based on the CST element shape functions.
(b) Does the element satisfy pure bending condition? Explain your answer.
(c) If two CST elements are created by connecting nodes 1-2-4 and 4-2-3 of the
rectangular element, what will be xx along line A-B?

y
A
4 (0,2) u4 u3
3 (3,2)

M M

x
1 (0,0) 2 (3,0) u1 B u2

Solution:
1
u(x, y ) N1 N2 N3 N4 ( 4xy 4x 6y 6)
(a) 6
v(x, y ) 0

u 1
xx ( 4y 4)
x 6
v
yy 0
y
u v 1
xy ( 4x 6)
y x 6

(b) Since xx is a linear function of y, it is consistent with the pure bending problem.
However, xy is a linear function of x, which is supposed to be zero for the pure bending
problem. Thus, the rectangular element cannot satisfy the pure bending condition.
(c) For element 1: xx 2 / 3 , Foe element 2: xx 2/3
The figure below illustrates the distribution of xx along B-A line.
 CHAP 8 Structural Design Using Finite Elements 13

xx

B A

14. Five rectangular elements are used to model a plane beam under pure bending. Rigid
body motion is ignored. The element in the middle has nodal displacements as shown
in the figure. Using the bilinear interpolation scheme, calculate the shear strain along
the edge AB and compare it with the exact shear strain.

a a

D C

1m
u(x, y ) 0 1x 2y 3xy
1m
A B
v(x, y ) 0 1x 2y 3xy a a
a = 1 mm
Original element
Deformed element

Solution:
The shear strain along the edge AB becomes

xy (x , 0) 0.001 0.002x

The shear strain decreases linearly along the edge AB. Note that the exact shear strain is
zero for the pure bending problem.

15. A uniform beam is modeled by two rectangular elements with thickness b. Without
performing calculations sketch xx and τxy along the top edge from A to C, as
predicted by FEA. Also, show the exact stresses according to beam theory.

A B C
F
2c x
F
L L

Solution:
Exact solution: M = 2Fc,

b(2c)3 Mc 3F
I , exact , exact 0
12 I bc
 3F exact
bc FE

exact
A B C

FE

18. A quadrilateral element in the figure is mapped into the parent element.
(a) A point P has a coordinate (x, y) = (½, y) in the physical element and (s, t) = (−½,
t) in the parent element. Find y and t coordinates of the point using isoparametric
mapping.
(b) Calculate the Jacobian matrix at the center of the element.
(c) Is the mapping valid? Explain your answer.

y
3 (0,2)
t
4 (-1,1) 3 (1,1)
2 (1,1)

x 1 (-1,-1) 2 (1,-1)
4 (0,0) 1 (1,0)

Solution:
(a) y 3/8
(b)

3
0 4
[ J[ 1 1
2 4

(c) The mapping is valid because the Jacobian is always positive.

1
J 8
(3 t) 0 for 1 t 1

19. Consider the plane stress 4–node element shown below. It is the 27th element in the
mesh and has global node numbers as shown below. The coordinates of the nodes
with respect to a global x-y coordinate systems in shown next to each node.
 CHAP 8 Structural Design Using Finite Elements 15

y
64 63
(0,1) (1,1)

(1,0)
(0,0)
51 52 x
The element connectivity is as follows:
Element # Local node 1 Local node 2 Local node 3 Local node 4
27 51 52 63 64

Nodal displacement vector = {X}T = {u51, v51, u52, v52, u63, v63, u64, v64} = {0, 0, 0.1, 0,
0.1, 0.1, 0, 0}.
(a) Determine the displacement at the point (x, y) = (0.75, 0.75) by interpolating the
nodal displacements.
(b) Compute the Jacobian matrix at the point in (b).
(c) Compute strain yy at the center of the element.
Solution:
(a) u 0.075, v 0.05625
(b)

x y
1
0
[ J[ s s 2
x y 1
0 2
t t

(c)

yy 0.05

20. For the 4–node element shown in the figure, a linearly varying pressure p is applied
along the edge. The finite element method converts the distributed force into an
equivalent set of nodal forces {Fe} such that

uT TdS {q(e ) }T {F(e ) }

S

where T is the traction (force per unit area) on the surface S and u is the vector of
displacements. Since the applied pressure in the figure is normal to the surface (in the
x–direction), the traction can be expressed as T = {p, 0}T where p can be expressed as
p = p0(t+1)/2, where t = -1 at Node 1 and t = +1 at Node 4. Le is the length of the
edge. Integrate the left hand side of the above equation to compute the work–
equivalent nodal forces {F(e)} when {q(e)}T = {u1, v1, u2, v2, u3, v3, u4, v4}.
 t
4 3

1 2
p0L(e ) p0L(e )
{F(e ) }T { 6
, 0, 0, 0, 0, 0, 3
, 0}

21. Determine the Jacobian matrix for the following isoparametric elements. If the
temperature at the nodes of both elements are {T1, T2, T3, T4} = {100, 90, 80, 90},
compute the temperature at the mid point of the element and at the mid point of the
Edge 1-4?

2 (1,12) 1 (2,12)
2 (0,4) 1 (8,4)

3 (0,0) 4 (18,0)
3 (1,1) 4 (2,1)

Solution:
(a) Jacobian:

x y
1
0 11
[ J[ s s 2 , J 0
x y 11
0 2
4
t t

(b) T ( 1, 0) 95

22. Integrate the following function using one–point and two–point numerical integration
(Gauss quadrature). Explain how to integrate it. The exact integral is equal to 2.
Compare the accuracy of the numerical integration with the exact one.

I sin(x ) dx
0

Solution:
Integration Points Exact for polynomial of
NG Weights (wi)
(si) degree
1 0.0 2.0 1
2 .57735 1.0 3
 CHAP 8 Structural Design Using Finite Elements 17

One Point Integration: I 3.1415 , Error = π – I = 1.1415

Two Point Integration: I 1.9358 , Error = 2 – 1.9358 = .0642

23. A six–node finite element as shown in the figure is used for approximating the beam
problem.
(a) Write the expressions of displacements u(x,y) and v(x,y) in terms of polynomials
with unknown coefficients. For example, u(x, y) a0 a1x .

(b) Can this element represent the pure bending problem accurately? Why or why
not? (Bending moment M is applied at the edge 2–3)
(c) Can this element represent the constantly distributed load problem accurately?
Why or why not? (Distributed load q is applied at the edge 4–6–3)

4 6 3

1 5 2 x

Solution:
(a)

u(x, y ) a0 a1x a2y a 3x 2 a 4xy a5y 2

v(x, y ) b0 b1x b2y b3x 2 b4xy b5y 2

(b)

u
xx a1 2a 3x a 4y
x
v
yy b2 b4x 2b5y
y

u v
xy a2 a 4x 2a5y b1 2b3x b4y
y x

Yes, this element is exact for pure bending, as the strain can vary linearly in the y-
direction. Also, since there are three nodes on the top and the bottom, the spurious
shear can be removed.
(c) No, this element cannot accurately represent the constantly distributed load problem
accurately. This is because the strain needs to be a quadratic function of x.