Matrix Method of

Structural analysis
 Impact of the digital computer
Easy availability of high speed digital computers:
revolutionary growth in the field of numerical techniques

Which made it possible to investigate and analyze

large complicated structures with even thousands of
unknown
An outstanding example of this development is the matrix method that has
gained popularity for analyzing skeletal i.e. framed structure in recent years.
The elegant finite element method is primarily the extension of matrix method.

✓ the principles of equilibrium and

compatibility,
✓ the relationship between forces and
displacement

remain basic in all the methods, their interpretations within

the various numerical processes have received much attention.
At early developments compatibility is enforced at discrete points
throughout the structure, yielding a set of equations in which the forces
are unknown and the coefficients in the equations are flexibility quantities.
Hence, the method is termed as force method or flexibility method.
Almost, simultaneously, work was being carried out on a parallel method,
in which equilibrium is enforced at discrete points, giving a set of
equations with displacements as the unknowns and the coefficient as the
stiffness quantities. The method is known as displacement method or the
stiffness method.
In both the methods, the governing equations are constituted in matrix
form.

Matrix notation is a compact and convenient method for generating

systems of equations, and the arrays represent the efficient format for
writing systems of equations and storing the information in computer.
Most structural engineering offices now have access to computers that
permits the use of the stiffness method for analyzing traditional Civil
Engineering structures e.g. various skeletal structures.
 Classification of Framed / Skeletal structure
Framed structures are structures that can be satisfactorily idealized using line
elements. Such structures are often referred to as skeletal structures. Usually the
members of the structures are assumed to be connected either by friction-less
pins or by rigid joints.
Degree of Freedom:
number of independent displacement components to be found during the analysis.

▪ Finding displacements by the stiffness method involves the solution of a

system of linear equation relating the known applied forces to the degree
of freedom of the structure.
▪ In general elastic structures have an infinite number of independent
displacement components as displacement vary continuously through out
the structure.
▪ Obviously, to solve an infinite number of simultaneous equations is not
feasible.
▪ The matrix method renders the analysis tractable by considering
displacements and forces at selected points, called 'nodes’.
▪ When the matrix method is applied to a structure. a node is usually
located at each joint of the structure.
▪ The analysis determines the nodal displacements caused by applied
loading. The number of possible displacement components at each joint is
known as the 'degree of freedom'
Structure Type Nodal d.o.f. Nodal force vector

Plane Truss u,v Fx , Fy

Space Truss u,v,w Fx Fy Fz
Beam w, z F y , Mz
Plane Frame u,v, z F x F y , Mz
Grillage w, x, y Fz Mx, My
Space Frame u,v,w,x, y,z Fx Fy Fz Mx, My, Mz
 Boundary conditions
For stability, a structure is always restrained at one or more nodes.
These restraints are known as boundary condition and they have the effect
of reducing the degree of freedom of the restraint node.
 Force Displacement relationships

{F} =[K]{d}
M1 M2
1 2
F={W} and d={u}

u [K]=AE/L

6EI
M 1= [2 +  2 ]
L
M1 M2
1 2 6EI
M 2= [2 2 +  1 ]
L

M1  1  2 1
  6 EI
 =
 2 1  
1 2   
k  = 6 EI 1 2
M  L     L  
 2  2
 d1(P1) d4(P4)
d2 (P2) d5(P5) d 1 =  1( P 1, P 2 ,..., P 6 ) d 1 = f 11 P 1 + ......+ f 16 P 6
d3(P3) d6(P6) d 2 =  2( P 1, P 2 ,..., P 6 ) ...
d 6 =  6( P 1, P 2 ,..., P 6 ) d 6 = f 61 P 1 + ......+ f 66 P 6

 f11 f12 f 13 f14 f15 f16 

 
 f22 f 23 f 24 f 25 f 26 
P1 
P1 =  1(d 1 ,d 2 ,...d 6 ) d 1   f21
d 2  
 P2 
P2 =  2 (d 1 ,d 2 ,...d 6 ) d 3   f31 f32 f 33 f 34 f 35 f 36 

P3 
d  = P  {d}=[f]{P}
d 4   f41 f42 f 43 f 44 f 45 f 46 
P4 
P6 =  6(d 1 ,d 2 ,...d 6 ) d 5   
 6  f f52 f53 f54 f55 f56  P5 
 35   6
f f62 f 63 f 64 f 65 f 66 
 61
P 1 = k 11d 1 + .....+ k 16 d 6 Flexibility Approach
.
.
 P1   k111 k12 k 13 k14 k15 k16  d 1 
P 6 = k 61d 1 + .....+ k 66 d 6  P2  k 21 k 22 k 23 k 24 k 25 k 26  d 2 
 P3   k 31 k 32 k 33 k 34 k 35 k 36  d 3  {f}=[k]{d}
 P  = k k 42 k 43 k 44 k 45 k 46  d 
 P4   k 41 d 4 
k 52 k 53 k 54 k 55 k 56  d 5  Stiffness Approach
 P5   k 35   6
 6   61 k 62 k 63 k 64 k 65 k 66
 TRUSSES

A truss is a structure composed of straight bars connected at their points on

intersection by means of momentless joints called pinned or hinged.
All loading are assumed to be applied at those points of intersection.
Thus, each straight bar is subjected to axial load only without any
bending/twist.

The independent displacements at a joint of such structure are

translations along a convenient set of orthogonal coordinate axes.

Plane truss is the most important of all structural forms.

Examples of application can be found in bridge trusses, roof
trusses, truss beams and girders, truss walls etc.
Three-dimensional or space trusses are also an important
structural form. The examples include roof trusses for
antenna, tower for supporting power line, truss structure for
space satellite etc.
 A, E
Axial Bar ElementF1, u1 u2 , F2
1 L 2

Displacement Case 1: Displacement Case 2:

If node 2 is fixed and force F1is node 1 is fixed, force F2 is applied at node
applied at node 1 i.e., u1=u1 and u2=0 2 i.e. u1=0 and u2=u2, then the force
force deformation relation deformation relation may be written as
F1 L AE
u1 = or F11 = [ ]u1 F2 L AE
AE L u2 = or F22 = [ ]u2
AE L
Since node 2 is fixed, F2 becomes a reaction.
Since node 1 is fixed, F1 becomes a
Imposing equilibrium on the forces reaction. Imposing equilibrium on the
F21 = − F21 i.e. F21 = −[
AE
]u1
forces we can write
L
AE
F12 = − F22 i.e. F12 = −[ ]u2
General Displacement Case: L
If both displacement components are simultaneously given, the force conforming to
Hooke’s law and maintaining overall equilibrium of the member can be written as
AE AE
F1 = F1 for case1 + F1 for case 2 = u1 −
F1  AE 1 − 1 u 1 
u2
L L
F  =  
 2  L − 1 1 u 2 
F2 = F2 for case1 + F2 for case 2 =−
AE
u1 +
AE
u2
L L
 System Stiffness Matrix
Node 1 Node 2 Node 3 Node 4
 Fx1   k1 − k1  u1   k1 − k 1 
F  =    , thus , [ k ]1− 2 = − k k 
 x 2  − k1 k1  u 2   1 1 

u1 u2 u3 u4
Fx1   k 2 − k 2  u 2   k2 − k2 
  =    , thus [ k ] 2 −3 = − k k 
L1 L2 L3 Fx 2  2 − k 2 k 2  u3   2 2
A1 , E 1 A2 , E 2 A3 , E 3

, k2 = 2 2 , k3 =Px13 3 Px2  k 3 −Pkx33  u3  Px4 k − k 

A1 E1 AE AE
k1 = Fx1 
  =  , thus [k ]3−4 = 
L1 L2 L3 3 3
 − k 3 k 3 
Fx1 Fx2 Fx 2  3 − kF3x1 k 3  u 4  Fx2 
Px1 Px2 Px3 Px4
Px1 Px2 Px3 Px4
Fx1 Fx2 Fx1 Fx1 Fx2
Fx1 Fx2 Fx2 Fx1 Fx2

Fx1 Fx1
Px1 =Fx1 of element 1 elem 2+ Fx1 of
Fx2 elem 3 Px3 =Fx2
Fx2of
Px2 =Fx2 of elem 1+ Fx1 of elem 2 Px4 =Fx2 of element 3
 Px1   k1 − k1 0 0  u1 
P x1 = k1 u1 - k1 u2 P    
 x 2   −k1 − k1 + k2 − k2 0  u2 
Px 2 = − k1 u1 + k1 u2 + k 2 u 2 + k 2 u 3   =   
Px 3 = − k 2 u2 + k 21 u 3 + k 3 u 3 + k 3 u 4 P
 x3   0 − k k + k − k 3  3
u
2 2 3

Px 2 = − k 3 u 3 + k 3 u 4  Px 4   0 0− − k3 k3  u4 

  Px1   k 1 − k 1 0 0  u1 
 Px 2  − k 1 − k 1 + k 2 − k 2 0  u2 
P  =  0 − k k2 + k3 − k 3  u3 
 Px 3   0 0−
2
− k3 k 3  u4 
 x4  

u1 u2 u1 u2 u3 u4
 Fx1   k1 − k1   u1  k1 -k1 0 u1
 =  , 0
 Fx 2   −k1 k1  u2 
-k1 k1 +k2 -k2 0 u2
 k − k1 
[k ]1− 2 =  1
 −k1 k1  0 -k2 k2 +k3 -k3 u3
0 0 -k3 k3 u4

u2 u3 u3 u4
 Fx1   k2 − k2  u2 
  =     Fx1   k3 − k3  u3 
 Fx 2 2  −k2 k2  u3    =  
 Fx 2 3  −k3 k3  u4 
 k − k2 
[k ]2−3 =  2   k3 − k3 
 − k2 k2  [k ]3− 4 = 
 − k3 k3 
 Actual Truss: Stiffness Matrix in Global Coordinate System
v2
v2 Force displacement relation in
v2 v2 uu22 local/member axis system
Y u2  −  −
F
 x1  u1 

 −   1 0 −1 0   − 
v1  Fy1  AE  0 0 0 0  v1 
 − =   − 
vv1 1
 Fx 2  L  −1 0 1 0  u2 
u1  −    
 0 0 0 0  −
uu 11  Fy 2  v2 
    -   
x F  = K  d
 
−  −   
F x1 = F x1c os +F y1s in   Fx1 
   Fx1 
 −     0 0  
Denoting =cos  Fy1    
−
  −  0 0   Fy1 
Fy1 = - F x1 sin +F y1 cos   and =sin   −  = 
 Fx 2   0 0   
 
    Fx 2 
−  
 −  
0 0
F 
 Fy 2   y2 
Similarly, for displacements vectors    

 - 


− Transformation matrix relating i.e. F = T F 
d =  T d quantities in member axis system with  
global/reference axis system
Stiffness matrix in reference coordinate system

 -     
[T] F  = K  [T]d [T] [T] F  = [T]-1 K  [T]d
-1
F  = K  d    
 

   
[T] [T] F  = [T]T K  [T]d
T
i.e. F  = [K]d i.e. [K] = [T]T K  [T]
   

  − 0 0  1 0 −1 0    0 0  2  − 2 −  
  0 0  0  2
− AE 0 0 0 −   0 0    2
−  −  
[ K ] = [T ]T [k ][T ] =  
 0 0  −  L − 1 0 1 0  0 0    = AE − 2 −  2  
     L  
 0 0    0 0 0 0  0 0 −    −  −    
2 2

 
 
Thus the equilibrium equation at element level can be written as
Fy2 (v2)  2  − 2 −  
 Fx1    u 
   −  −  2  1 
2
F 
 y1  AE  2 v 
Fy1 (v1) Fx2 (u2)
 = 
−  −  2    1 
 Fx 2  L −  −  2  u2
 2   
 Fy 2    v 2 
Fx1 (u1)  
 
  2  − 2 −  
 Fx1    u  Obs. 1: the coefficient in the first row are the
   −  −  2  1 
2
F  same but opposite in sign as the coefficient in
 y1  AE  2  v1 
 = −  −  2     the third row.

 Fx 2  L −  −  2  u2
 2   
 Fy 2    v 2 
Fx1 = 2 u1 + v1 − 2 u2 − v 2
 
  Fx 2 = − 2 u1 − v1 + 2 u2 + v
Fx2=-Fx1
The same relation holds between the second and fourth rows. Fy2=-Fy2

So the equilibrium in both x and y directions for the free body is satisfied.
Fy2 (v2)

Fy1 (v1) Fx2 (u2)

Obs 2: Now we take the moment about nodal point 1

Fx1 (u1)

AE
M = L
[− 2u1 −  v1 +  2u2 +  v2 ]L −  − u1 −  2v1 +  u2 +  2v2  L

=0

satisfy the moment equilibrium conditions.

▪ In the expression of [k] we note that the  2  − 2 −  
first and third rows are same but  Fx1    u 
   −  −  2  1 
2
F 
opposite in sign, as are the second &  y1  AE  2  v1 
 = −  −  2    
fourth rows. 
 Fx 2  L −  −  2  u2
 2   
▪ Thus [k] is singular matrix.  Fy 2    v 2 
 
 
▪ It can not be inverted and solution to
stiffness equation is not possible.

Physically, singular stiffness matrix means that the element without

any support is an unstable free body.
The element becomes stable and the stiffness matrix can be
nonsingular if it is properly supported.

Also we see that the diagonal element kii for all i is positive.
This implies that a force Fi directed towards right will not produce
displacement towards left.

For example if all degree of freedom is zero except

Fi=kiiui
ui, then from stiffness equilibrium equation
which shows that Fi and ui have the same sign as kii is positive.
 Assembly of Element Matrix in Global coordinate System

u3 P6 v3 P5 All bars have the same length L & axial rigidity AE

3
Element    2 2 
v1 P2 v2 P4 1-2 0 1 0 1 0 0
2-3 120 -1/2 1/4 ¾
1 2 1-3 60 1/2 1/4 ¾
P1 u 1 u2 P3
For Element 2-3 ,
For Element 1-2
Fx 2   1 − 3 − 1 3  u 
 Fx1   1 0 − 1 0  u1  F    2 
      y 2  AE − 3 3 3 − 3  v2 
 Fy1  AE  0 0 0 0  v1   =   u 
 =   F
 x3  4 L  − 1 3 1 − 3  3 
F
 x2  L  −1 0 1 0  u2 
    Fy 3   3 − 3 − 3 3  v3 
 Fy 2     
   0 0 0 0  v2 

 Fx1   1 3 −1 − 3  u1 
 Fy1  AE  3 3 − 3 − 3  v1 
For Element 1-3 , F  =   
 Fx 3  4 L − 1 − 3 1 3  u3 
 y3  − 3 − 3 3 3  v 3 
 Joint Equilibrium u3 P6 3 v3 P5
P1=Fx1of element 1-2+Fx1 of element 1-3
P2=Fy1of element 1-2+Fy1 of element 1-3
v1 P2 v2 P4
P3=Fx2of element 1-2+Fx1 of element 2-3 1 2
P4=Fy2of element 1-2+Fy1 of element 2-3 P1 u 1 u2 P3
element 1-2
P5=Fx2of element 1-3+Fx2 of element 2-3  Fx1  1 0 − 1 0  u1 
    
P6=Fy2of element 1-3+Fy2 of element 2-3  Fy1  AE  0 0 0 0  v1 
 =  
 Fx 2  L  −1 0 1 0  u2 
 Fy 2   
AE    0 0 0 0  v2 
P1 = [( 4 + 1 )u1 + 3v1 − 4u2 − u3 − 3v3 ]
4L element 2-3
Fx 2   1 − 3 − 1 3  u 
AE F    2 
i.e.P1= = [ 3u1 + 3v1 − 3u3 − 3v3 ]  y 2  AE − 3 3 3 − 3  v2 
L  =   
Fx 3  4 L − 1 3 1 − 3  u 3 
 Fy 3   3 − 3 − 3 3  v3 
   
AE
P2 = [ 0 + 3u1 + 3v1 − 3u3 − 3v3 ]
4L element 1-3
 Fx1   1 3 −1 − 3  u1 
 Fy1  AE  3 3 − 3 − 3  v1 
F  =   
 Fx 3  4 L − 1 − 3 1 3  u3 
 y3  − 3 − 3 3 3  v 3 
 AE AE
P1 = [( 4 + 1 )u1 + 3v1 − 4u2 − u3 − 3v3 ] i.e.P1= = [ 3u1 + 3v1 − 3u3 − 3v3 ]
4L 4L
AE
P2 = [ 0 + 3u1 + 3v1 − 3u3 − 3v3 ]
4L
AE
P3 = [ −4u1 + ( 4 + 1 )u2 − 3v2 − u3 + 3v3 ]
4L
AE
P4 = [ − 3 u2 + 3v2 + 3u3 − 3v3 ]
4L
AE
P5 = [ −u1 + 3v1 − u2 + 3v2 + ( 1 + 1 )u3 + ( 3 − 3 )v3 ]
4L
AE
P6 = [ − 3u1 − 3v1 + 3u2 − 3v2 + ( 3 − 3 )u3 + ( 3 + 3 )v3 ]
4L

4 + 1 3 − 4 0 −1 − 3  u
 P1    1 
P   3 3 0 0 − 3 −3  v1 
 2   
 P3  AE − 4 0 4 +1 − 3 −1 3  u 2 
  =  
 0 − 3 3 
− 3  v 2 
 P4  4 L 0 3
 P5  − 1 − 3 − 1  
  3 1+1 3 − 3  u 3 
 P6    v 
− 3 − 3 3 − 3 3− 3 3 + 3   3 
  Fx1  0 − 1 0  u1   Fx1   1 −1 − 3  u1 
1 
3
 
      
 Fy1  AE  0 0 0 0  v1   Fy1  AE  3
 =
3 − 3 − 3  v1 
 =     
L  −1 0 1 0  u2   Fx 3  4 L  −1 − 3 1 3  u3 
 Fx 2   Fy 3    
 Fy 2      − 3 − 3 3  v3 
   0 0 0 0  v2  3

Member 1 − 3
Member 1 − 2
𝑢1 𝑣1 𝑢2 𝑣2 𝑢3 𝑣3
4 +1 0+ 3 -4 0 -1 − 3 𝑢1
0+ 3 0 +3 0 0 − 3 -3 𝑣1
𝐴𝐸/4𝐿 -4 4 +1 0− 3 -1 3 𝑢2
0
0 0 0− 3 0 +3 3 -3 𝑣2

-1 − 3 -1 3 +1 +1 − 3 + 3 𝑢3
− 3 -3 3 -3 − 3 + 3 3+3 𝑣3

 Fx 2   1 − 3 −1 3  u2 
    
 
Fy 2 AE  − 3 3 3 − 3  v2 
Member 2 − 3  =   
F
 x 3 4 L  −1 3 1 − 3  u3 
 Fy 3    
   3 −3 − 3 3  v3 
 Application of Boundary Conditions

stiffness equation
[K]{u}={F}
▪ Matrix [K] is singular and an unique solution for displacement
vector {u} is possible if the structure is supported.
▪ Some d.o.f. in {u} must be prescribed to enable a solution.

✓ It should be emphasized that improper

specification of boundary conditions can lead to
erroneous results.
✓ Boundary conditions eliminate the possibilities of
the structures moving as a rigid body.
✓ Boundary conditions should properly model the
physical system.
 Elimination Approach
Stiffness Eqn. [K]{u}={F} is basically a set of n simultaneous linear equations
k 11 u1 + k 12 u 2 + k 13 u 3 + .....+ k 1N u N = F 1
k 21 u1 + k 22 u 2 + k 23 u 3 + .....+ k 2N u N = F 2
Consider the single boundary
k 31 u1 + k 32 u 2 + k 33 u 3 + .....+ k 3N u N = F 3
condition u1=a1.
.
k N 1 u1 + k N 2 u 2 + k N 3 u 3 + .....+ k NN u N = F N

▪ To determine N-1 unknown u value, as u1is already known. Thus it is sufficient to

solve N-1 equations for obtaining N-1 unknown d.o.f.
▪ Excluding the first of the above set of equations, we can rewrite by substituting the
known value of u1=a1,Then the above equation can be written in matrix form as
k 22 u 2 + k 23 u 3 + .....+ k 2n u N = F 2 - k 21 a1  k22 k23 . k2 N   u2   F2 − k21a1 
k . k3 N   u3   F3 − k31a1 
k 32 u 2 + k 33 u 3 + .....+ k 3n u N = F 3 - k 31 a1  32 k33
 = 
.  . . . . .   
   
k N 2 u 2 + k N 3 u 3 + .....+ k NN u N = F N - k N 11 a1  k2 N k3 N . k NN  u N   FN − k N 1a1 

Now, [K] is the reduced stiffness matrix ((N-1 x N-1) ) obtained by eliminating the
specified or support d.o.f.
Note that the external load vector is also modified for non zero support conditions.
 Generalization for multiple boundary conditions.

Consider the boundary condition up1=a1, up2=a2, ..., upr=ar

Delete the p1th row and column, the p2th row and column, ..., prth row and column
from the K matrix. The resulting [K] matrix of dimensions (N-r X N-r).

The force vector {F} is of dimension (N-r x1). Modify the each load component as

Fi = Fi - ( ki,p1 a1 + ki,p2 a2 +...+ ki,pr a r )

Now solve [K] {u}={F} for the displacement vector {u}.

 4 + 1 3 − 4 0 −1 − 3  u
3  P1    1 
P   3 3 0 0 − 3 −3  v1 
 2   
 P3  AE − 4 0 4 +1 − 3 −1 3  u 2 
  =  
 0 − 3 3 
− 3  v 2 
 P4  4 L 0 3
2  P5   
− 1 − 3 − 1 3 1+1 3 − 3  u 3 
    v 
 P6 
− 3 − 3 3 − 3 3− 3 3 + 3   3 

the truss is supported as shown. the reduced stiffness matrix is obtained by deleting
The supporting conditioned are the rows & columns of corresponding dof to be
v1= u2=v2=u3=0. Only u2 and v1 fixed. i.e. 2nd, 3rd, 4th and 5th d.o.f. is restrained
remain to be solved. due to support conditions.

Thus those rows and columns need to be deleted  P2   5 − 3  u1 

and the reduced stiffness matrix becomes as:  =  
 P5   − 3 6  v3 
 Penalty Method easy to implement in a computer program.

Consider the single boundary condition u1=a1.

A spring with very large stiffness ks is added to model the corresponding support.
Now add this large stiffness to the corresponding diagonal i.e. with k11 and apply the
large force ks x a1 and discard the load F1.
( k 11 + k s )u 1 + k 12u 2 + k 13u 3 + .....+ k 1N u N = F 1 + k s a 1
The corresponding matrix
equation becomes k 21u 1 + k 22u 2 + k 23u 3 + .....+ k 2N u N = F 2
Now the first of above set of k 31u 1 + k 32u 2 + k 33u 3 + .....+ k 3N u N = F 3
equation can be written as .

F 1 + k s a 1 - k 12u 2 + k 13u 3 + .....+ k 1 k N 1u 1 + k N 2u 2 + k N 2u 3 + .....+ k NN u N = F N

u1 = N
 a1
k11 + k s a 1
the only modifications to handle u1=a1 are that a large number ks is added to the first
diagonal element of [K], and ksa1 is added to F.
Generalization: Consider the boundary condition up1=a1, up2=a2, ...,upr=ar.
Modify the structural stiffness matrix [K] by adding a large number ks to each
p1th,p2th,...,prth diagonal elements of K. Also, modify the global load vector {F} by
adding ksa1 to Fp1, ksa2 to Fp2, ksar to Fpr.
Now solve [K] {u}={F} for the displacement vector {u}.
Internal Forces 2  − 2
−  
Fxi   2  u i 
Fyi  AE  2  2
−  −  v i 
F  = −  −   2
  u j 
F  L  2
−  −     v 
xj 2
Fyj2 (v2j)  yj   j
 

Fyi (vi) Fxj (uj)

AE
Fxi = ( ) i− j [2 u i + v i − 2 u j − v j ]
L
Fxi (ui)

axial forces Si-j can be obtained in terms of Fi & Fj

S i− j = Fxi  + Fyi  = Fxj  + Fyj 

 Modulus of Elasticity
E=2x1011 N/m2,
Cross-section areas
Member 1-2: 4x10-4m2, 2
Member 2-3: 4x10-4m2,
Member 2-4: 6x10-4m2
1 4 3
Element 1(1-2):
ng the element as 1-2 it means that the element points from 1 to 2, as shown follo
2 (x2,y2 i.e. 1,1)
[k1− 2 ]e =
Figure 16.2.12a Element 1 0.5 0.5 − 0.5 − 0.5
 
4 x10−4 x 2 x1011 0.5 0.5 − 0.5 − 0.5
1 (x1,y1 i.e. 0,0) 1.414  −0.5 − 0.5 0.5 0.5 
 
2 2 2 2  −0.5 − 0.5 0.5 0.5 
L 1-2 = ( x 2 - x 1 ) +( y 2 - y 1 ) = (1- 0 ) +(1- 0 )
= 1.414 1 1 −1 −1 
- 1- 0  1 − 1 − 1
 = cos  = x 2 x 1 = = 0.707, = 28.284 x10 6  1

L 1-2 1.414  −1 −1 1 1 
 
y 2 - y 1 1- 0  −1 −1 1 1 
 = sin  = = = 0.707
L 1-2 1.414
 0 0 0 0 
11 
Element 2-3: 4 x10 −4
x 2 x10 0 1 0 − 1
[k2−3 ]e =
2 2 2 2 1.0 0 0 0 0 
L 2-3 = ( x 3 - x 2 ) +( y 3 - y 2 ) = (1- 1 ) +(0 - 1 ) = 1.0  
0 −1 0 1
2 (x2,y2 i.e.1,1)
- 1- 1  0 0 0 0
 = cos  = x 3 x 2 =  1 0 − 1 
= 0.0,
L 2-3 1.0 6  0
= 80.0 x10
y3 - y2 0 -1  0 0 0 0
 = sin  = = = -1.0  
3 (x3,y3 i.e.1,) L 2-3 1.0  0 −1 0 1 

Element 2-4:
e. 1,1) 2(x2,y2 i.e. 1,1)
0.8 − 0.4 − 0.8 0.4 
 
6 x10−4 x 2 x1011  −0.4 0.2 0.4 − 0.2 
[k2− 4 ] =
e

2.236  −0.8 0.4 0.8 − 0.4 

 
 0.4 − 0.2 − 0.4 0.2 
4(x4,y4 i.e. 3,0)
 4 −2 −4 2 
 1 2 −1
6  −2
2 2 2 2
L4-2 = (x4 - x2 ) + ( y4 - y2 ) = (1- 3 ) + (1- 0 ) = 2.236
= 10.733 x10
 −4 2 4 −2 
-
 = cos  = x 4 x 2 =
3-1  
 2 −1 −2 1 
= 0.894,
L 4-2 2.236
y4 - y2 0 -1
 = sin  = = = -0.447
L 4-2 2.236
Assembly of Element Stiffness matrix size of the global stiffness matrix is 8 by 8.

u1 V1 u2 v2 u3 v3 U4 v4
28.29 28.29 -28.29 -28.29 u1
28.29 28.29 -28.29 -28.29 v1
-28.29 -28.29 28.29+0 28.29+0 0 0 -49.26 21.46 u2
+49.26 -21.46
-28.29 -28.29 28.29+0 28.29+80 0 -80 21.46 -10.73 v2
-21.46 +10.73

0 0 0 0 u3
0 -80 0 80 v3
-49.26 21.46 49.26 -21.46 u4
21.46 -10.73 -21.46 10.73 v4

[k1− 2 ]e == 28.284 x106 x [k2−3 ]e == 80.0 x106 x [k2− 4 ]e == 10.733x106 x

1 1 − 1 − 1   u1   0 0 0 0  u2   4 −2 −4 2  u2 
 0 1 0 − 1  v   −2 1 2 −1  v 
1
 1 − 1 − 1  v1     2     2 
   0 0 0 0  u3   −4 2 4 −2  u4 
 −1 − 1 1 1  u2 
     
 −1 − 1 1 1   v2   0 − 1 0 1   v3   2 −1 −2 1   v4 
 28.29 * 28.29 − 28.29 − 28.29 0 0 0 0 
28.29 28.29 * − 28.29 − 28.29 0 0 0 0 
− 28.29 − 28.29 77.55 6.83 0 0 − 49.26 2146. 
 − 10.73 
6 − 28.29 − 28.29 6.83 39.02 0 0 2146
.
K ] = 10 0 0 0 0* 0 0 0 0 
 
0 0 0 − 80 0 80 * 0 0 
0 0 − 49.26 2146
. 0 0 49.26 * − 2146 . 
0 0 2196
. − 10.73 0 0 − 2146
. 10.73* 

Boundary Conditions: The supporting conditions are

ux1=0,uy1=0, ux3=0, uy3=0, ux4=0, uy4=0.
77.56 6.83  u2  3 50 
106    =  
119.02  v2 
10
6.83 −100 

4
u2  
−6 0.012 − 7.43x10−4  3 50  −3 0.674 
  = 10   10   = 10  
v2   −7.43 x10 −4
0.008426   −100   −0.881
 Internal Forces
 Fx11  1 1− 1− 1  0   2.784 
 1   0   
Member 1(1 -2):  Fy1  6  1 1 − 1− 1   −3  2.784 
 2  = 28 .29 x10  10 =  
F
 x2   − 1 − 1 1 1  0.674   − 2.784 
 Fy 2   
− 1 − 1 1 1 − 0.881 − 2.284 
2
 

 Fx21  0 0 0 0   .674   0 
 2  0 1 0 − 1 − .881  − 70 
Member 2 (2-3):  Fy1  6  −3    
 2 = 80 .0 x10 10  =
  
F
 x2   0 0 0 0   0  − 2.784 
 Fy 2     0  − 2.784 
 0 − 1 0 1
2
 
 Fx31   4 −2 −4 2   0.674   52 .784 
 3 − 2 1  − 0.881 − 26 .393 
 Fy1  2 − 1 10    
 3  = 10 .73 
−
=
3
 
Member 3 (2-4): F
 x2   − 4 2 4 − 2   0  − 52 .784 
Fy32    
   2 −1 − 2 1   0   
 26 .394 
Matrix Method of Analysis of Beam Structures
 12 EI 6 EI 12 EI 6 EI 
 L3 − 
L2 L3 L2
 Fy1   6 EI 
v1 
  2 
4 EI 6 EI 2 EI
 − 2
 M z1   L L L L  1 
 =  
 Fy 2   − 12 EI 6 EI
− 2
12 EI 6 EI  v2 
− 2
M   L3 L L3
L   2 
 z2   
 6 EI 2 EI 6 EI
− 2
4 EI 
 L2 L L L 
 Matrix Method of Analysis of Frame Structures
▪ Two dimensional plane frames are extensively used in civil engineering.
Examples of most common such structures are roof trusses, offshore platforms,
all types of building frame etc.
▪ Plain frames constitute one of the most important structural entity. In most of
the situations, a real structural system may be viewed as an assembly of beam
grids and plane frames, wherein, the former consists load transferring and the
later a load bearing entity.
▪ The main feature of these structures is that the members meeting at a joint are
rigidly connected together by welding or bolt as in steel structure or by
monolithic construction as in reinforced and prestressed concrete structures.
▪ Because of the rigid joint large moments are transferred at the joints and such
members are subjected to axial force, shear and bending moment
 1 M1 2 M2

u1,Fx1 u2,Fx2

v1,Fy1 v2,Fy2
A plane rigid frame can be envisioned as
▪ being composed of individual beam members with axial effect i.e.
axial forces and deformations are present,
▪ oriented in various directions and connected together at their ends.
▪ To cope with the plane frame problem, an element must possess three d.o.f. at each
nodal point; two displacement component u and v in the x and y directions
respectively and one angle of rotation .
▪ Corresponding to the d.o.f. u,v and  there are two forces Fx and Fy and a bending
moment M at each nodal point.
The element stiffness matrix for above element will be of size 6x6 relating the six
d.o.f. with the corresponding 6 internal force quantities.

Sstiffness matrix:combination of an axial 2x2 truss bar & 4x4 stiffness of a beam
element in member axis system;
then transform the above combined 6x6 stiffness matrix to the global coordinate system
 Axial-Flexural Beam Element M1 M2
M1
in Local Coordinate System u1, FMx12 A,E,I,L Fx2, u2
u1, Fx1 A,E,I,L Fx2, u2

Fy1, v1 Fy2, v2 Fy1, v1 Fy2, v2

= =
u1, Fx1 Fx2, u2 u1, Fx1 F ,u
x2 2
 AE AE 
M2  L − L 0 0 0 0 
Fy1, v1  F ,
y2 2v 
Fx2, u2
Axial Beam

Fx1  −
AE AE F , v
y1  1 u
 1 
0 0 0 0
 F Axial  u 2 Beam
L L

Fy2, v2   0 0
x 2 12 EI 6 EI 12
− 3
EI 6 EI  
 F y1   L 3
L 2
L L 2
 v1 
F  =    
  0 0 6 EI 4 EI 6 EI 2 EI
− 2  1 
y2

M 1   L 2
L L L  v 2 
   12 EI 6 EI 12 EI 6 EI   
 2  0 0 − 3 − 2
M − 2  2 
Fy1, v1 Fy2, v2  L L L 3
L 
Beam  6 EI 2 EI 6 EI 4 EI
− 2

0 0 L2 L L L 
For the convenience of assemblage, it is desirable to number all the d.o.f. at each
nodal point in a certain sequence. For this purpose, we rearrange the above equation

 AE AE 
 L 0 0 − 0 0 
L
 
0 12 EI 6 EI
0 −
12 EI 6 EI 
 
L3 L2 L3 L2 
 Fx1   
6 EI 4 EI 6 EI 2 EI  u1  − − −
  0
 Fy1   L2 L
0 −
L2
 
L  v1  { f } = [ k ]{d }
M   
 1  AE AE  1 
  = − 0 0 − 0 0  
F
 x2   L L  u2 
F   12 EI 6 EI 12 EI 6 EI  v 
 y 2   0 − − 0 −  2
 M 2   L3 L2 L3 2
L   2 
 6 EI 2 EI 6 EI 4 EI 
0 0 − 
 L2 L L2 L 
 
 
 
 Coordinate Transformation:
M2 F x1 = Fx1 cos  + F y1 sin 
u2, Fx2
F y1  − F x1 sin  + Fy1 cos 
E,A,I, L Fy1, v1 M 1= M 1
M1 In matrix form
− 
 F x1 
u1, Fx1     cos  sin  0   Fx1 
−    
Fy1, v1 x  F y1  =  − sin  cos  0   Fy1 
   0 1   M 
−  
0
 1
−  M 1 
 F x1   
 
 −     0 0 0 0 F
 x1  − 
 F y1   −   0   F   F x 2   cos 
−  
0 0 0
 y1  sin  0  Fx 2 
M   0 0  −    
0  Fy 2 
0 1 0 0
 1     M1   F y 2  = − sin  cos 
 = 0 0   0   −   0 1 M 2 
0
−   0 0 0 −  0  Fx 2  M 2  
0
 F x2    F   
−   0 0 0 0 0 1  y 2 
F     M 2 
 y2   
− 
 M 2 
−
{F } = [T ]{F } {d} = [T]{d}
  -     
[T] F  = K  [T]d [T] [T] F  = [T]-1 K  [T]d
-1
F  = K  d    
 

   
[T] [T] F  = [T]T K  [T]d
T
i.e. F  = [K]d i.e. [K] = [T]T K  [T]
   
 AE AE 
 L 0 0 − 0 0 
L
 
 0 12 EI 6 EI 12 EI 6 EI 
0 − 3  c1 
 L2 
 −  0 0 0 0 
L3 L2 L
   0 0 0 0  c c Symm . 
 
 0 0 0 0   0
6 EI 4 EI 6
0 − 2
EI 2 EI    2 3 
 L2 L L L  −   0 0 0 0  −c4 c5 c6 
0 0 1 0 0 0   AE AE  0 0 1 0 0 0 
= 
   − −  
0 0 0  −  0  L
0 0
L
0 0
0 0 0   0  −
 1 c − c 2 c 4 c 1 
0 0 −   0  0 − 12 EI − 6 EI 0 12 EI − 6 EI   0 0 0 −   0  −c − c − c c3 
0
  
L2  0 0 0 0 0 1 
c2
0 0 0 0 0 1   L3 L2 L3
   2 3 5

 6 EI 2 EI 6 EI 4 EI  −
 4c c c c − c c 6

 0 L2 L
0
L2 L  5 7 4 5
 
 
 
2 AE 2 12 EI AE 12 EI AE 12 EI 6 EI 6 EI
c1 =  + , c2 =  ( − ), c3 =  2 +  2 3 , c4 = 2 , c5 = 2  ,
L L3 L L3 L L L L
4 EI 2 EI
c6 = , c7 =
L L
 Element 1:
150KN
2
150KN 3 =900, then =cos900=0, =sin 900=1.0, L=3000 mm.
2
3M 4M
c1=0xAE/L+1x12x(50x109)/(3000)3=22.22,
1 c2=0, c3= 60,
1
c4=33.33x103, c5=0, c6=66.66x106, c7=33.33x106
100KN-M
7M

 22.2 0 - 33.33e + 3 - 22.2 0 - 33.33e + 3

 
 0 60 0 0 - 60 0 
- 33.33e + 3 0 66.7e + 6 33.33e + 3 0 33.3e + 6 
[k ] 1- 2 =  
 - 22.2 0 33.33e + 3 22.2 0 33.33e + 3
 
 0 - 60 0 0 60 0
 
 - 33.3e + 3 0 33.3e + 6 33.3e + 3 0 66.6e + 6 

 249.5 35.4 - 848.5 - 249.5 - 35.4 - 848.5

 
 35.4 6.754 5940.0 - 35.4 - 6.754 5940
 - 848.5 5940 28.3e + 6 848.5 - 5940 14.1e + 6 
[k ] 2-3 = 
 - 249.5 - 35.4 + 848.5 249.5 35.4 848.5
 
 - 35.4 - 6.754 - 5940.0 35.4 6.754 - 5940
 
− 848.5 5940 14 .1e + 6 848.5 − 5940 28 .28e + 6 
Now as the total d.o.f is 9, then the size of global stiffness matrix will be 9x9.

 22.2 0 3.33x103 −22.2 0 −3.33x103 0 0 0  d1   P1  0 

      
 0 60 0 0 −60 0 0 0 0  d 2   P2  0 
3.33x103 6
0 66.76 x10 3.33x10 3
0 3.33x10 6
0 0 0   d3   P3   0 
      
 −22.2 0 3.33x106 271.7 35.4 32485 −249.4 −35.4 −848.5  d 4   P4  −1.0 x10 
5

 − − − −  d  =  P  = 150 
 0 60 35 .4 35 .4 60 .68 5940 35 .4 6 .75 5940 
 5   5   
 −3.33x106 0 3.33x106 32485 5940 95.0 x106 848.5 −5940 14.1x106  d 6   P6  −1500 
      
 0 0 0 −249 .4 −35 .4 848 .5 249 .5 35 .4 848 .5 
 d7   P7   0 
      
0 0 0 − 35.4 −6.75 −5940 35.4 6.754 −5940  d8   P8  0 
 
 0 0 0 848.5 6 6

5940 14.1x10 848.5 −5940 28.3 x10  d9   P9  0     
 

d1 d 2 d 3 d 4 d 5 d 6 d 7 d8 d 9  = [0 0 − 2.16x10−3 0.792 − 0.297 0.532x10−3 0 0 − 0.179x10−3 ]T

T
 Internal Forces:  
 
 
Member 1-2  FX 1 
1
-33.33e+ 3  0   
 FY 1   22.2 0 -33.33e+ 3 -22.2 0
  
 1   0 60 0 0 -60 0  −2.16 x10−3   
 M1      178 
  -33.33e+ 3 0 66.7e+ 3 33.33e+ 3 0 
33.3e+6 0.79   
 FX 1  =    =  −36.67 
 2   -22.2 0 33.33e+ 3 22.2 0 33.33e+ 3  2 − 0.297  
−9987 
   0 -60 0 0 60  
0 0.532 x10−3   
 FY 1      −178 
 2   -33.3e+ 3 0 33.3e+6 33.3e+ 3 0 66.6e+6  0   
M 1    36.67 
 2  −101130 
 
 
Member 2-3:  2 
FX 1
  -848.5  0.792  189 
 FY 2   249.5 35.4 -848.5 -249.5 -35.4
−0.297  
 1  35.4 6.754 5940.0 -35.4 -6.754 5940    1.43 
 2      
−3
 -848.5 5940 28.3e+6 848.5 -5940 14.1e+6  0.532 x10  10097 
M 1 
 =   = 
 FX 22   -249.5 -35.4 +848.5 249.5 35.4 848.5  0   −189 
   -35.4 -6.754 -5940.0 35.4 6.754 -5940  0  −1.43 
 FY 2 
2     
   −848.5 5940 14.1e+6 848.5 −5940 28.28 e+6 −
 −0.179 x10  
3 24.59 
M 2 
 2 
 Grillage Structures
▪ Plane grid i.e. grillage structure is basically consists of beam in two or more directions
and rigidly connected to each other at joints.
▪ The main difference between the grillage and two dimensional rigid jointed frame
structures are only in which they are loaded.
▪ Plane frames are loaded in the plane of the structures whereas grillages are loaded only
perpendicular to the plane of the structure.
▪ Various reinforced or steel skeletons with their members confined to horizontal plane
(x,y) of the global (x,y,z) space constitute beam grids.
▪ It plays a very important role in offering facilities to support a large floor or roof space
without having any internal supports.
▪ They could be employed in the structural constitution of wide bridge decks.
▪ Furthermore, often a complex structural system can be transformed into equivalent
beam grid for the purpose of at least preliminary calculations
 y

z 2 4 6 8

x
1 3 5 7

z w2 Fz2
My2y2
 Mx2
x x2
x
z w, Fz1
y
My2 y1

x1 Mx1
 z w2 Fz2
My2y2 w2 Fz2 x2, Mx2
 Mx2
x x2
x

z w, Fz1 = w1
y Fz1 +
My2 y1

x1 Mx1 My1 y1 x1, Mx1

w2 Fz2 x2, Mx2

Now to derive the stiffness relationships y2, My2
involving the torques and=twist at the endwof
1 an
element. Fz1 +

If x1 and x2 are the twists at ends, theMy1 y1

net x1, Mx1
twist =x2-x1 and rate of twist =(x2-x2)/L

Then by considering the equilibrium i.e. GJ

T1+T2=0 as T2 = ( −  x1 )
L x2
GJ
T1 = − ( x 2 −  x1 )
T1  GJ 1 − 1  x1  L
 = − 1 1  
T2  L    x2 
The force displacement relation for the combined system

 12 EI 6 EI 12 EI 6 EI 
 L3 − 0 0 
L2 L3 L2
 
 6 EI 4 EI 6 EI
− 2
2 EI
0 0 
 L2 L L L 
 Fz1    w1 
M  −12 EI 6 EI 12 EI
− 2
6 EI
− 2 0 0   
 y1   L3 L L 3
L   y1 
 Fz 2   6 EI 2 EI 6 EI 4 EI  w2 
  = 2 − 2 0 0  
 y2   L
M L L L   y 2 
M   GJ GJ   
 x1   0 0 0 0 −   x1 
M x 2   L L   
 x2 
 GJ GJ 
 0 0 0 0 −
L L 
 
 
 
 z w2 Fz2
My2y2
 Mx2
x x2
x
z w, Fz1
y
My2 y1

x1 Mx1
Now rearranging the above equation (as done in case of plane frame),

 12 EI 6 EI 12 EI 6 EI 
 L3 0 − 0 
L2 L3 L2
 
 0
GJ
0 0 −
GJ
0
 L L 
 Fz1    w 
 M   6 EI 4 EI 6 EI 2 EI   1 
0 − 2 0 
 x1   L2 L L L   x1 

 f  = k  d 
 M y1   12 EI 6 EI 12 EI 6 EI   y1 
  = − 3 0 − 2 0 − 2  
 z2   L
F L L3 L   w2 
M x 2    
0   x 2 
GJ GJ
   0 − 0 0
 M y 2   L L   y 2 
 6 EI 2 EI 6 EI 4 EI 
 0 − 0 
 L2 L L2 L 
 
 
 
 Coordinate Transformation
My2 Fz1 = Fz1
− −
M y2 M x2 M x1 = M x1 cos  + M y1 sin 
−
M y1 = - M x1 sin  + M y1 cos 
M y1 My1 Mx2
 − 
Fz1  1 0 0  Fz1 
 −    
Mx1 M x1  = 0    M x1 
−  −  0 −    M y1 
M x1 M y1  
 
− 
 F z1  F − 
 −  1 0 0 0 0 0   Fz1  1 0 0  
   z−2  
 M x1  0   0 0   M x1      Fz 2 
−  
0 M−x 2  = 0    M x 2 
 M y1   0 −   0 0 0   M y1  M  0 −    M y 2 
− =    y 2 
 F z 2  0 0 0 1 0 0   Fz 2 
 −  0 0 0 0    M x2 
M x 2    

− 
M y 2 
 0 0 0 −     M y 2 

−
F = T  F 
Stiffness matrix, K= 12 EI 6 EI 12 EI 6 EI 
 L 0 − 0
3
L2
L3 L2 
 
0 GJ
0 0 −
GJ
0 
 L L 
1 0 0 0 0 0   6 EI 4 EI 6 EI 2 EI  1 0 0 0 0 0 
0 
  − 0 0 0   L2 0
L
−
L2
0
L  0
   0 0 0 
0   0 0 0   12 EI 6 EI 12 EI 6 EI  0 −   0 0 0 
 − 3 0 − 2 0 − 2  
0 0 0 1 0 0  L L L3 L  0 0 0 1 0 0
0 0 0 0  −   0 −
GJ GJ  0 0 0 0   
   0 0 0  
0 0 0 0    L L  0 0 0 0 −  
 6 EI 2 EI 6 EI 4 EI 
 L2 0 − 0
L L2 L 
 
 
 
GJ
c1 − c2 c3 − c1 − c2 c3  where, = and
EI
 c4 − c5 − c7 
 c2 c6  12 6 6
c1 = 2 , c2 = , c3 = ,
EI  c8 − c3 − c9 c10  L L L
=  
L  c1 − c2 − c3  c4 =  2 + 4 2 , c5 =  (4 −  ),
 c4 − c5  c6 = − 2 + 2 2 , c7 = 2 +  ,
 
 c8  c8 = 4 2 +  2 , c9 =  (2 +  ),
c10 = 2 2 −  2
a three member beam grid system as 2 1
(-3,5.196) (6,0)
shown in figure 16.5. . Each member
has a span of 6.0m. E=25491N/mm2, 6m 6m
u=0.15, I=0.54x104mm4 and torsional 4 (0, 0)
constant J=0.372x104mm4,
6m

3 (-3,-5.196)
Element 1:(4-1):

Nodal co-ordinate : x4=0, y4=0 and x1=6000.0,y1=0.0.

Thus L=6.0m, =1.0 and =0.0. GJ
= =
EI
c1 = c2 =, c3 =, c4 =, c5 =, c6 =
c7 =, c8 =, c9 = and c10 =

 .76E + 04 .00E + 00 .23E + 08 -.76E + 04 .00E + 00 .23E + 08

 .00E + 00 .69E + 10 .00E + 00 .00E + 00 -.69E + 10 .00E + 00 

 .23E + 08 .00E + 00 .92E + 11 -.23E + 08 .00E + 00 .46E + 11 
 
k 4 -1 =  -.76E + 04 .00E + 00 -.23E + 08 .76E + 04 .00E + 00 -.23E + 08 
 .00E + 00 -.69E + 10 .00E + 00 .00E + 00 .69E + 10 .00E + 00 
 
 .23E + 08 .00E + 00 .46E + 11 -.23E + 08 .00E + 00 .92E + 11 
 
 
 .76E + 04 - .20E + 08 - .11E + 08 - .76E + 04 - .20E + 08 - 11E + 08 
 - .20E + 08 .71E + 11 .37E + 11 .20E + 08 .33E + 11 23E + 11 

 - .11E + 08 .37E + 11 .28E + 11 .11E + 08 .23E + 11 63E + 10 
k 4-2 =  
 - .76E + 04 .20E + 08 .11E + 08 .76E + 04 .20E + 08 11E + 08 
 - .20E + 08 .33E + 11 .23E + 11 .20E + 08 .71E + 11 37E + 11 
 
 - .11E + 08 .23E + 11 .63E + 10 .11E + 08 .37E + 11 28E + 11 

.76E + 04 .20E + 08 - .11E + 08 - .76E + 04 .20E + 08 - 11E + 08 

 .20E + 08 .71E + 11 - .37E + 11 - .20E + 08 .33E + 11 - 23E + 11 
 
 - .11E + 08 - .37E + 11 .28E + 11 .11E + 08 - .23E + 11 63E + 10 
k 4 -3 =  
 - .76E + 04 - .20E + 08 .11E + 08 .76E + 04 - .20E + 08 11E + 08 
 .20E + 08 .33E + 11 - .23E + 11 - .20E + 08 .71E + 11 - 37E + 11 
 
 - .11E + 08 - .23E + 11 .63E + 10 .11E + 08 - .37E + 11 28E + 11 
 .11E+31 .00E+00 .00E+00 .00E+00 .00E+00-.20E+08 -.76E+04-.20E+08 .00E+00 .00E+00 .00E+00 .00E+00

.00E+00 .00E+00 .00E+00 .00E+00 .00E+00 .10E+38 .00E+00 .33E+11 .00E+00 .00E+00 .00E+00 .00E+00

-.23E+08 .00E+00 .00E+00 .00E+00 .00E+00-.37E+11 .23E+08-.23E+11 .00E+00 .00E+00 .00E+00 .00E+00

.00E+00 .00E+00 .11E+31-.20E+08 .00E+00 .20E+08-.76E+04 .00E+00 .00E+00 .00E+00 .00E+00 .00E+00

.00E+00-.69E+10 .20E+08 .33E+11 .00E+00 .33E+11-.20E+08 .15E+12 .00E+00 .00E+00 .00E+00 .00E+00

.00E+00 .00E+00 .11E+08 .23E+11 .00E+00-.23E+11-.11E+08 .00E+00 .00E+00 .00E+00 .00E+00 .00E+00

.00E+00-.23E+08 .00E+00 .00E+00 .11E+31 .00E+00 -.76E+04-.23E+08 .00E+00 .00E+00 .00E+00 .00E+00

.00E+00 .00E+00 .00E+00 .00E+00-.20E+08 .00E+00 .20E+08 .00E+00 .00E+00 .00E+00 .00E+00 .00E+00

.00E+00 .14E+38 .00E+00 .00E+00 .11E+08 .00E+00-.11E+08 .46E+11 .00E+00 .00E+00 .00E+00 .00E+00

-.76E+04 .00E+00-.76E+04 .11E+08-.76E+04 .00E+00 .23E+05 .11E+08 .00E+00 .00E+00 .00E+00 .00E+00

.00E+00 .00E+00 .20E+08 .37E+11-.20E+08 .00E+00 .00E+00 .23E+11 .00E+00 .00E+00 .00E+00 .00E+00

-.23E+08 .00E+00 .11E+08 .42E+37 .11E+08 .00E+00-.20E+02 .63E+10 .00E+00 .00E+00 .00E+00 .00E+00
Load Vector:
The consistent load vector for each element is obtained from the fixed end action as
discussed for beam problems.
The beam grid is subjected to an udl of 4.5N/mm. As each member is subjected to
same distributed load having equal length, the local load vector for each will be

{f}4-1={f}4-2={f}4-2=[-0.135e+5 0.0 -0.135e+8 -0.135e+5 0.0 0.135e+8]T

Note that these element load vectors are computed in local axis system. However, the
final loads will be in global directions. Thus, each element load vectors need to
transform to global axis system first then add to get final load vector in global axis
system.

− 
 F z1 
 Fz1  1 0 0 0 0 0  − 
The element load vector in global M  
 −  0 0 0  
M x1 
axis system for each element can  x1  0 − 
 M y1  0   0 0 0   M y1 
be easily obtained as   =  
 z 2  0
F 0 0 1 0 0  − 
F z2
 M x 2  0   
F  = T 
−1

−
= [T ]T F   
 M y 2  0
0 0 0  −
0 0  
−
 M x2 
 
 −

M y 2 
For element 1 =1 and =0, so,
{f}4-1=[-0.135e+5 0.0 -0.135e+8 -0.135e+5 0.0 0.135e+8]T

For element 2 =-0.5 and =0.866 thus,

{f}4-2={f}4-2={f}4-2=[-0.135e+5 0.1168e+8 0.0675e+8 -0.135e+5 -0.1169e+8 -0.675e+8]T

For element 3 =-0.5 and =0.866 thus,

{f}4-3={f}4-2={f}4-2=[-0.135e+5 0.1168e+8 0.0675e+8 -0.135e+5 -0.1169e+8 -0.675e+8]T

And the final global load vector is obtained by adding these three element load vectors in
global axis system as below

-0.135e+5 0.0 0.135e+8 -0.135e+5 -0.117e+8 -0.0675e+8 -0.135e+5 0.117e+8

-0.0675e+8 -0.405e+5 0.0 0.0
Internal Forces: The internal force quantities at each member can be obtained by multiplying
the element stiffness matrix of each element with corresponding nodal degree of freedoms and
then by adding the corresponding fixed end force vector.
element number: 1
-13500.0 -4.896164E-28 -4.050000E+07 13500.0 4.896164E-28 -4.049999E+07
element number: 2
-13500.0 -2.573834 -4.049999E+07 13500.0 -1.109734 -4.049999E+07
element number: 3
-13500.0 2.573834 -4.049999E+07 13500.0 1.109734 -4.049999E+07

The final load vector is obtained by adding the corresponding fixed end force to this load vectors:
element number: 1
-13500.0 +0.135e+5 =0.0 -4.896164E-28 +0 =0.0
-4.050000E+07 +.135e+8 =-0.27e+8 13500.0 +.135e+5 =0.27e+5
4.896164E-28 +0 =0 -4.049999E+07 +(-.135e+8)=-0.539e+8
element number: 2
-13500.0 +0.135e+5 =0.0 -2.573834 +0 = -2.5738340
-4.049999E+07 +.135e+8 =-0.27e+8 13500.0 +.135e+5 =0.27e+5
-1.109734 +0 =0 -4.049999E+07 +(-.135e+8) =-0.539e+8
element number: 3
-13500.0 +0.135e+5 =0.0 2.573834 +0 =2.5738340
-4.049999E+07 +.135e+8 =-0.27e+8 13500.000000 +.135e+5 =0.27e+5
1.109734 +0 =0 -4.049999E+07 +(-.135e+8) =-0.539e+8
 Space Truss

▪ The most difficult truss to analyze are those which form a network in space.
▪ In proceeding with space trusses, the first step is to obtain the stiffness matrix for the
member.
▪ The bar element in space has three forces and three displacements at each node( i.e
Fx,Fy, Fz and u,v,w) in the global co-ordinate system; hence the order of element
stiffness matrix will be 6 by 6.
▪ The element in local co-ordinate system has only one axial force F and one axial
d.o.f. u at each nodal pint.

Augment the 2x2 stiffness matrix equation in local co-ordinate

system to 6x6 stiffness matrix equations by creating two more
non-existence forces Fy and Fz and displacements d.o.f. v and w
in the local co-ordinate system y and z.
 − −
−  − 
Fyj Fyj Fxj  F xi  u i 
−   1 0 0 − 1 0 0  − 
y x  F yi   0 0 0 0 0 0  v i 
Fyi −    − 
Fzj Fxj  F zi  AE  0 0 0 0 0 0  w i 
− =   − 
 F xj  L  − 1 0 0 1 0 0  u j 
−
−   0 00 0 0 0 − 
Y Fzj  F yj    v j 
− −    0 0 0 0 0 0   
Fyi Fxi −  − 
Fxi x  F zj  w j 
Fzi
−
Fzi z
−
Fxi = F xi cos(x, x )+ F yi cos(y, x )+ F zi cos(z, x ) = F xi  x x + F yi  y x + F zi  z x
−
Fyi = F xi cos(x, y )+ F yi cos(y, y )+ F zi cos(z, y ) = F xi  x y + F yi  y y + F zi  z y
−
Fzi = F xi cos(x, z )+ F yi cos(y, z )+ F zi cos(z, z ) = F xi  x z + F yi  y z + F zi  z z
 Fxi   xx  yx  zx  Fxi 
  
Fyi  =  xy  yy  zy  Fyi 
Fzi   xz  yz  zz  Fzi  {Fj } = [ t ]{Fj }
i. e.  Fi  = [ t ] Fi 

 −  [t ] [0] F
F i     i −
− =    F j i.e.{F } = [T ]{F }
 F j  [0] [t ]  
 

Note that the direction in which the angle is measured is immaterial as cos(3600-)=cos. Thus,

,
xy =  yx ,  zx =  xz ,  z y =  yz and  x x =  y y =  z z

x j - xi y j − yi z j - zi1
Denoting  x,x = ,  y,x = ,  z,x =  =
Lij
, =
Lij 2
, =
Lij
 2   − 2 −  −  
 2  −  − 2 −  
 
  2 −  −  − 2 
 2 
−  −  −  2   
 − 2 −   2  
−  −  − 2    2 

Example Problem
 Space Frame:
▪ Three dimensional frame also called space frames are frequently encountered in the
analysis of multistoried building.
▪ They are also to be found in the modeling of car body, shhips and bicycle analysis.
▪ These are the most complex of all skeletal structures met in practice.
▪ In fact, most of the structures are three dimensional in their behavior.
▪ Transformation of space frame into assemblage of components such as beam grids,
plane frames, etc. constitutes an attempt to simplify the problem.

A typical 3-D frame: each node has 6 d.o.f. i.e. 3

translation along x,y and z axis and 3 rotations
about x,y and z axis, making a total 12 d.o.f. for
each member.
Element Stiffness in Member Principal Axis:

For three dimensional behavior of this member, at each u, v, w, x, y, z

node we must consider six generalized displacement
− − − − − −
six corresponding force quantities Fx , Fy , Fz , M x ,M y , M z ,

The linear elastic member can be envisioned to have four distinct response modes:
(i) Tension-Compression in x-direction
(ii) Flexure in x-y plane
(iii) Flexure in x-z plane
(iv) Torsion about the x axis

Note that every stiffness coefficient associated with a space frame element has already been
encountered when dealing with truss, beam, frame and grillage.
Thus, in accordance with that, the stiffness equilibrium equation for each four different cases
can be described
  12 6 12 6 
 L3 −
L2 L3 L2 
Y1   6 4 6 2  v 1 


M z1   2 − 2 
Y  = EI z  L12 L L L   z1 , whereI = y 2 da
X  AE 1 − 1 u 
− − 6 12 6 v z
M2  − 3 − 2 − 2   2 
 z2   L L3 L  2 z 
− = −  L
1 x1
   6 2 6 4
A

X 2  L 1 1  u x 2   L2 L
− 2
L L 

 12 6 12 6 
 L3 −
L2 L3 L2 
 Z1   6 4 6 2  w 1 


  2 − 2 
M x1  GJ  1 − 1 
 x1 
− −
  M y1  L L   y1 , where, I =
y  12
Z  = L L x 2 dA
 − = − 1 1   −  M2 
EI
− 3
6
− 2
12 6 w
− 2   2 
y


M x 2 

L   
 x 2 
  y2   L L L3 L  y2  A
 6 2
− 2
6 4
 L2 L L L 
 F− 
 − x1  u 
A 0 0 0 0 0 −A 0 0 0 0 0  v1 
 F− y1   12 I z 6I z 12 I z 6 I z   w1 
 F   − 2 0
L   1 x1 
− z 1 0 2
0 0 0 0 0 0
M − x1   L
12 I y 6I y
L L
12 I y 6I y   
M − y1   − 0 − 2 0 − 0   y1 
 M z 1  E 0 0 2
0 0 0
   z1 
 − = L J L L J L
 F− x 2  L 0 0 0 0 0 0 0 0 − 0 0   u2 
 2 (1 +  ) 6 I y 2(1 +  )  v2 
 F− y 2  6I y
 F− z 2  0 0 − 0 4I y 0 0 0 0 2 I y 0   w2 
M  L L  x 2 
− x2 
6I z 6I z
M  0 0 0 0 4I z 0 − 0 0 0 2 I z   y 2 
− y2  L L  
 M z2   z2 
 
 There are three independent orthogonal
components of force and moments at each node
x
y Each of which must be transferred from the
member axis system to structure axes system.

Let us consider a vector P at a point over the

element The vector could be resolved into
z components Px,Py,Pz in the x,y,z directions,
x respectively; or the components
− − −
P− , P− , P− in the x, y, z directions respectively.
  x y z

 X−  X1 
− 1  Y1  From the geometric details it can be written as
Y− 1  Z1 
Z−1  M   P−   
M − x1 
[ t ] [0] [0] [0]  M y1 
x1
 Px−  xx−  −  −  
Px 
M − y1   y   
[0]  M z1 
xy xz
 M z1 
− =
[0] [t ] [0]   =  −   −   Py 
[0] [0]  X 2  yz 
−

X− 2 
[0] [t ]  P−   yx yy
 
[0] [t ]  Y   −   Pz 
Y− 2   [0] [0]
 z   zx−  −
zz 
Z− 2 
2
Z 2    zy

M − x2   
M   M x 2 
− y2
 M z2   y2 
M
   M z 2 