Lesson Module 1

THE MATRIX METHOD OF STRUCTURAL ANALYSIS

INTRODUCTION
The analysis of indeterminate structures is the major field in structural engineering. There are several
methods of analysis, the best among these being the “Kani’s” method. But this method will not be
convenient for analysis of present day multi-storey buildings. The need for analysis of high degree of
indeterminate structures and development of computers have given rise to the Matrix Method.
Basically, there are two methods of the Matrix Method: Flexibility Matrix Method and Stiffness Matrix
Method.

FLEXIBILITY MATRIX METHOD

The systematic development of consistent deformation in the matrix form has led to the Flexibility
Matrix Method. In this method, the basic unknowns are the “redundant forces”. Hence, the analysis has
to first identify the basic determinate structure and thereby identify redundant forces. The number of
redundant forces is equal to the degree of static indeterminacy. The displacements in the basic
determinate structure due to a given load and redundant forces are found and the consistency
conditions are formed. The resulting simultaneous equations are solved to get redundant forces. The
required quantities at any section are calculated. The equations developed are in matrix form. This
method is also known as Force or Compatibility Method.

STIFFNESS MATRIX METHOD

The systematic development of slope deflection method in the matrix form has given rise to the stiffness
matrix method. Here, the basic unknowns are the displacements of joints. The equations of equilibrium
are formed and solved to get slopes and deflections at the joints. From these, moments and shears are
calculated. This method is also known as Displacement or Equilibrium Method.

TWO APPROACHES TO MATRIX METHOD

1. Direct Approach
2. Transformation Matrix Approach
The first method gives a clear concept of analysis, while the latter is suited for developing general
purpose computer programs for structural analysis.

DEGREE OF STATIC AND KINEMATIC INDETERMINACY

The number of equations required over and above the equations of static equilibrium for the analysis of
a structure is called “static indeterminacy” or “degree of redundancy”.

A structure is kinematically indeterminate if the displacement components of its joints cannot be

determined by compatibility equations alone. Additional equations based on equilibrium conditi ons
must be formulated to obtain the number of equations necessary to determine all unknown
displacement components. The number of equilibrium conditions to find the displacement components
of all joints of the structure is known as “ degree of kinematic indeterminacy” or “ degree of freedom”.

For a beam, the degree of freedom at an end is given as

1. Free end- 3
2. Simply supported/roller end- 2
3. Hinged end- 1
4. Fixed end- 0
Hence, degree of kinematic indeterminacy, Dk for
1. One end hinged and the other roller- 3
2. Propped cantilever- 2
3. Fixed beam- 0

For pin jointed frames

Dk = 2j – e (plane frames)
= 3j – e (space frames)
Where j = number of joints
e = number of compatibility/boundary conditions

For Rigid Frames

Dk = 3j – e (plane frames)
= 6j – e (space frames)

MATRIX FORMULATION OF THE METHOD OF JOINTS

Theory of Trusses

A truss is a structural framework composed of interconnected bars pinned together to form a series of
triangles. The triangular configuration is the basic shape of the truss. There are two basic assumptions to
the theory of trusses:
1. The members are connected at the ends by frictionless mechanisms that prevent relative
translation but allow rotation;
2. The truss is loaded only at the joints.

The assumptions then indicate that truss elements are axial members in compression or tension.

Depending on member-joint arrangement and support conditions, a truss is either determinate

indeterminate. Recall that static determinacy in plane trusses is determined by the criteria 2j = m + r
where j = number of joints, m = no. of members and r = no. of support reaction components.

Notations and Sign Convention

For a start, denote each member and each joint by numbers instead of letters. The members and joints
shall be numbered successively from 1, 2 and so on. Member forces are positive (+) if they are in
tension, and represented as a vector pointing away from an end or joint; negative (-) if they are in
compression and symbolized by a vector pointing towards the joint or end. For loads, vertical loads
upwards are positive, horizontal loads pointing to the right are also positive. Inclined loads are broken
into their components first, each component treated distinctly. The horizontal and vertical components
of a joint’s translation are treated as “distinct freedom of translation”. The sum of all of these
translational components on all joints in a truss is called “degree of freedom”.
Note: A joint load is effective (could cause stress in the truss) only if it is applied along a “freedom”.
Example, any load applied on a hinged support of the truss has no effect since the load is absorbed by
the support. On a roller support, only a horizontal load could be applied to have any effect.

Figure: Freedom and Load Numbering

6

5 10
2
1 9
8
4
11
3 7
Example: Given truss and loads with joint & member numbering

30 KN 30 KN

5 KN 4 4
2

5 6 3m
1 3

1 2 7 5
4m 3 3m

20 KN

Also, the following shows the freedom numbering (vectors representing the freedoms of translation):

2
4
1
3

5 7

Observe that the loads applied opposite or along the lines of reaction forces have no effect on the
members. Thus, we only consider horizontal equilibrium at joint 5 (see above) and none at joint 1. Only
loads acting along a DOF No. will have an effect on the members.

Any inclined load could be resolved into horizontal and vertical components. The number of possible
distinct loads that could be applied to the joints of this truss is 7, which is also the DOF.

Let P1, P2, … P7 = applied loads corresponding in designation to the vectors of the freedom of
translation in the joints shown above.
Let F1, F2… F7 = member forces as designated in the member numbering above.

Next, at joints 2, 3, 4 and 5, draw the FBD showing the effect of Pi and Fi along the joint; for example,
refer to FBD of joint 2:

P2

P1
F4

0.80F3

F3

0.60F3
F1
Formulate the equilibrium conditions of joint 2:

∑Fh = 0 ∑Fv = 0
P1 = -0.80F3 – F4 P2 = F1 + 0.60F3

Do the same for joints 3, 4 and 5. Note that in this case, member forces were assumed tensile in nature.

The seven conditions produced from the equilibrium conditions are in linear system form:

P1 = -0.80F3 –F4 P5 = F2 + 0.80F3 – F7

P2 = F1 + 0.60F3 P6 = -0.60F3 – F5
P3 = F4 – 0.7071F6 P7 = 0.7071F6 + F7
P4 = F5 + 0.7071F6

The Matrix Equation is AF = P

Where: A = coefficient matrix
F = column matrix of variables
P = column matrix of constants

Referring to the original truss with applied loadings and the corresponding truss with the freedom
numbering, substitute the force value with respect to the vector of freedom of translation P1, P2, etc
considering sign conventions:

P1 = 5 KN
P2 = -30 KN
P3 = 0
P4 = -30 KN
P5 = 0
P6 = -20 KN
P7 = 0

5
⎡−30⎤
⎢ 0 ⎥
⎢ ⎥
Thus, P = ⎢−30⎥
⎢ 0 ⎥
⎢−20⎥
⎣ 0 ⎦

Writing in Matrix Format, AF = P

0 0 -0.80 -1 0 0 0 F1 5
1 0 0.60 0 0 0 0 F2 -30
0 0 0 1 0 -0.7071 0 F3 0
0 0 0 0 1 0.7071 0 F4 = -30
0 1 0.80 0 0 0 -1 F5 0
0 0 -0.60 0 -1 0 0 F6 -20
0 0 0 0 0 0.7071 1 F7 0

Let A,P = Augmented Matrix is shown below

 0 0 -0.80 -1 0 0 0 5
1 0 0.60 0 0 0 0 -30
0 0 0 1 0 -0.7071 0 0
A,P = 0 0 0 0 1 0.7071 0 -30
0 1 0.80 0 0 0 -1 0
0 0 -0.60 0 -1 0 0 -20
0 0 0 0 0 0.7071 1 0

Using Gaussian or Gauss-Jordan Elimination on Matrix A,P and obtaining the row reduced echelon form
(rref),

1 0 0 0 0 0 0 -49.2857
0 1 0 0 0 0 0 5.
0 0 1 0 0 0 0 32.14285
0 0 0 1 0 0 0 -30.7142
A,P = 0 0 0 0 1 0 0 .7142857
0 0 0 0 0 1 0 -43.4369
0 0 0 0 0 0 1 30.71428

Thus, F1 = -49.2857 KN
F2 = 5.00 KN
F3 = 32.14285 KN
F4 = -30.7142 KN
F5 = 0.7142857 KN
F6 = -43.4369 KN
F7 = 30.71428 KN