m4l19.pdf
TRANSCRIPT
-
7 Module 4: Analysis of Frame Structures
Lecture 2: Analysis of Truss
4.2.1 Element Stiffness of a 3 Node Truss Member
Fig. 4.2.1 3-node truss member
Here, the displacement function using Pascals triangle can be expressed as:
20 1 2u x x x 0
21
2
1 x x
(4.2.1)
Applying boundary conditions: At x= 0, u(0)= u 1 , x=L/2, u(L/2) = u2 and at x=L, u(L) = u3
And solving for 0 , 1 and 210 u , 1 2 31 3 4u u uL
and 1 2 32 22 4 2u u uL
Therefore,
2 2 21 2 32 2 23 2 4 4 21 x x x x x xu x u u u N uL L L L L L
(4.2.2)
Here, N is the shape function of the element and is expressed as:
2 2 22 2 23 2 4 4 21 x x x x x xN L L L L L L
(4.2.3)
Now, the element stiffness matrix can be written as Tk B D B d
(4.2.4)
Where, 2 2 23 4 4 8 1 4d N x x xB dx L L L L L L So, the stiffness matrix will be:
Tk B D B d
0L TB E B Adx
-
8
3
44
8
1 4
3
4
4
8
1
4
9 16
24 12
40
32 3
16
16
12 40 32 16
64
64 4
24
32
3 16 16 4
24
32 1
8
16
(4.2.5)
After integrating the above equation, the stiffness matrix of the 3-node truss member will become:
7 8 18 16 81 8 7
(4.2.6) 4.2.2 Worked Out Example Analyze the truss shown below by finite element method. Assume the cross sectional area of the inclined member as 1.5 times the area (A) of the horizontal and vertical members. Assume modulus of elasticity is constant for all the members and is E.
Fig. 4.2.2 Plane truss
-
9
Solution The analysis of truss starts with the numbering of members and joints as shown below:
Fig. 4.2.3 Numbering of members and nodes The member information for the truss is shown in Table 4.2.1. The member and node numbers, modulus of elasticity, cross sectional areas are the necessary input data. From the coordinate of the nodes of the respective members, the length of each member is computed. Here, the angle has been calculated considering anticlockwise direction. The signs of the direction cosines depend on the choice of numbering the nodal connectivity. Now, let assume the coordinate of node 1 as (0, 0). The coordinate and restraint joint information are given in Table 4.2.2. The integer 1 in the restraint list indicates the restraint exists and 0 indicates the restraint at that particular direction does not exist. Thus, in node no. 2, the integer 0 in x and y indicates that the joint is free in x and y directions.
Table 4.2.1 Member Information for Truss
Member No.
Starting Node
Ending Node
Value of
Area Modulus of Elasticity
1 1 2 90 A E 2 2 3 315 1.5A E 3 3 1 180 A E
Table 4.2.2 Nodal Information for Plane Truss
Node No. Coordinates Restraint List x y x y
1 0 0 1 1 2 0 L 0 0 3 L 0 1 1
-
10
The stiffness matrices of each individual member can be found out from the stiffness matrix equation as shown below.
22
22
22
22
sinsincossinsincos
sincoscossincoscos
sinsincossinsincos
sincoscossincoscos
LAEk
Thus the local stiffness matrices of each member are calculated based on their individual member properties and orientations and written below.
and
5 6 1 25
6
1
2
0000
0101
1000
0101
3 LAEk
3 4 5 6 3
4
5
6
1111
1111
1111
1111
243
2 LAEk
1 2 3 41
2
3
4
1010
0000
1010
0000
1 LAEk
-
11
Global stiffness matrix can be formed by assembling the local stiffness matrices into globally. Thus the global stiffness matrix are calculated from the above relations and obtained as follows:
243
243
243
24300
2431
243
243
24301
243
243
2431
24310
243
243
243
24300
001010010001
LAEK
The equivalent load vector for the given truss can be written as:
1
1
2
2
3
3
00
2
00
x
y
x
y
x
y
FFF P
FF PFF
Let us assume that u and v are the horizontal and vertical displacements respectively at joints. Thus the displacement vector will be expressed as follows:
00
00
2
2
3
3
2
2
1
1
vu
vuvuvu
d
Therefore, the relationship between the force and the displacement will be:
1 2 3 4 5 61 2 3 4 5 6
-
12
1
1
2
2
3
3
1 0 0 0 1 00 1 0 1 0 0
03 3 3 30 0 04 2 4 2 4 2 4 22 3 3 3 30 1 1
4 2 4 2 4 2 4 23 3 3 3 01 0 1
4 2 4 2 4 2 4 2 03 3 3 30 0
4 2 4 2 4 2 4 2
x
y
x
y
FF
P uAEP vL
FF
From the above relation, the unknown displacements u2 and v2 can be found out through computer programming. However, as numbers of unknown displacements in this case are only two, the solution can be obtained by manual calculations. The above equation may be rearranged with respect to unknown and known displacements in the following form:
F k k dF k k d
Thus the developed matrices for the truss problem can be rearranged as:
243
24300
243
243
2431
24301
243
243
001010010100
243
243
1024
31243
243
24300
243
243
LAE
. The above relation may be condensed into following
2
2
3 32 4 2 4 2
3 314 2 4 2
uP AEvP L
2P
P
Fx1 Fy1
Fx3
Fy3
u2
v2 0 0
0
0
-
13
The unknown displacements can be derived from the relationships expressed in the above equation.
1
2
2
3 3 3 312 24 24 2 4 2 4 2 4 2
3 3 3 3314 2 4 2 4 2 4 2
u P PAE Lv P PL AE
Thus the unknown displacement at node 2 of the truss structure will become:
2
2
8 233
3
u PLv AE
Support Reactions: The support reactions {Ps} can be determined from the following relation:
s csP P K d Where, {Pcs} correspond to equivalent loadings at supports. Thus, the support reaction of the present truss structure will be:
0 00 0 1
8 20 33 330 4 2 4 2 3
0 3 34 2 4 2
sAE PLPL AE
PPP
2230
Member End Actions: Now, the member end actions can be obtained from the corresponding member stiffness and the nodal displacements. The member end forces are derived as shown below. Member 1
1
1
2
2
00 0 0 0 0
00 1 0 1 3
8 20 0 0 0 0330 1 0 1 3
3
mx
my
mx
my
FF PAE PLF L AEF P
-
14
Member 2
2
2
3
3
8 21 1 1 1 2331 1 1 1 23 3
1 1 1 1 24 2 01 1 1 1 2
0
mx
my
mx
my
F PF PAE PLF PAELF P
Member 3
0000
0000
0000010100000101
1
1
3
3
AEPL
LAE
FFFF
my
mx
my
mx
Thus the member forces in all members of the truss will be:
2 23 3
2 2 2 200
m
P P
F P P P
The reaction forces at the supports of the truss structure will be:
032
2
R
PF
PP
Thus the member force diagram will be as shown in Fig. 4.2.4.
Fig. 4.2.4 Member Force Diagram