University of Benghazi
 
Faculty of Engineering
 

 
Civil Engineering Department
 

Numerical Analysis CE 501

Program No.3
Finite difference analysis of plate structures
 

Submitted to:
Prof. Salem Alsanusi
10, January, 2023

Prepared by:
Bashir A. El-Saadawi
 CE 501 Numerical Analysis
 
Find the Deflection at all nodes of thin plate 10×10 using Finite Difference Method
10 division in each direction (simply support)

Solution:
The stresses on the plates can be expressed as a function of the deflection of the plate w, which is
a function of the two coordinates (x,y) of the plate
𝜕 𝑤 2𝜕 𝑤 𝜕 𝑤 𝑞
𝜕𝑥 𝜕𝑥 𝜕𝑦 𝜕𝑦 𝐷
𝐸𝑡
𝐷
12 1 𝜇
Where:

w Small deflection
q Applied load
E Modules of elasticity
D Flexural rigidity of plate
t Thickness of plate
𝜇 Poison’s ratio

Support Conditions
For simply supported plate, we know that the deflection at those points is equal to zero

 
1 
 
 CE 501 Numerical Analysis
 
we can evaluate f(x,y) numerically by laying a
rectangular grid over the domain, and evaluate f(x,y) at
the nodes - the points where the lines of the x-axis and
y-axis intersect These expressions are often represented
as patterns or computational molecules for the easy
computation of the numerical solutions. The
biharmonic equation pattern for computing the
displacement at the interior nodes are as shown
(Computational Molecule when Δx = Δy)

The plate has been represented with gridlines that have 10 divisions on all the planar dimensions.
Therefore, Δx = Δy = 1.0. This shows that the gridline aspect ratio (α) = 1.0. The gridlines have
been extended by fictitious nodes, in order to capture the boundary conditions. The nodes have
been numbered, and all the points where the deflections are expected to be the same due to the
symmetry of the plate and the loading condition have been given the same numbering.
By implication, we have eleven distinguished nodes, hence, we are going to solve a 11 x 11 matrix,
which can be readily solved using MATLAB. The computational molecule (biharmonic
operator/pattern) used is the one shown below. We place the computational molecule node by node
and generate the appropriate equation for each node. After the equations are generated, they are
arranged as appropriate and solved. The results obtained from this method have been compared
with the results from the classical solution and finite element analysis

Number of Nodes

 
2 
 
 CE 501 Numerical Analysis
 
After obtaining a set of equations for each node. It can be arranged in matrix form
[A].[w] = [B]

20 -32 4 0 0 8 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 w1 0.002184
-8 25 -8 1 0 -16 4 0 0 2 0 0 0 0 0 0 0 0 0 0 0 w2 0.002184
1 -8 20 -8 1 4 -16 4 0 0 2 0 0 0 0 0 0 0 0 0 0 w3 0.002184
0 1 -8 20 -8 0 4 -16 4 0 0 2 0 0 0 0 0 0 0 0 0 w4 0.002184
0 0 1 -8 19 0 0 4 -16 0 0 0 2 0 0 0 0 0 0 0 0 w5 0.002184
2 -16 4 0 0 22 -8 1 0 -8 2 0 0 1 0 0 0 0 0 0 0 w6 0.002184
0 3 -8 2 0 -8 21 -8 1 2 -8 2 0 0 1 0 0 0 0 0 0 w7 0.002184
0 0 2 -8 2 1 -8 21 -8 0 2 -8 2 0 0 1 0 0 0 0 0 w8 0.002184
0 0 0 2 -8 0 1 -8 20 0 0 2 -8 0 0 0 1 0 0 0 0 w9 0.002184
0 3 -8 2 0 -8 2 0 0 21 -8 1 0 -8 2 0 0 1 0 0 0 w10 0.002184
0 0 2 0 0 2 -8 2 0 -8 20 -8 1 2 -8 2 0 0 1 0 0 w11 = 0.002184
0 0 0 1 0 0 2 -8 2 1 -8 20 -8 0 2 -8 2 0 0 1 0 w12 0.002184
0 0 0 0 1 0 0 2 -8 0 1 -8 19 0 0 2 -8 0 0 0 1 w13 0.002184
0 0 2 -8 2 1 0 0 0 -8 2 0 0 21 -8 1 0 -8 2 0 0 w14 0.002184
0 0 0 1 0 0 1 0 0 2 -8 2 0 -8 20 -8 1 2 -8 2 0 w15 0.002184
0 0 0 0 0 0 0 1 0 0 2 -8 2 1 -8 20 -8 0 2 -8 2 w16 0.002184
0 0 0 0 0 0 0 0 1 0 0 2 -8 0 1 -8 19 0 0 2 -8 w17 0.002184
0 0 0 2 -8 0 0 0 0 1 0 0 0 -8 2 0 0 20 -8 1 0 w18 0.002184
0 0 0 0 1 0 0 0 0 0 1 0 0 2 -8 2 0 -8 19 -8 1 w19 0.002184
0 0 0 0 0 0 0 0 0 0 0 1 0 0 2 -8 2 1 -8 19 -8 w20 0.002184
0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 2 -8 0 1 -8 18 w21 0.002184

Matrix [B] = .
0.002184 𝑚
.

 
3 
 
 CE 501 Numerical Analysis

On solving by using Matlab

Generate mesh & Solve the model

Values of deflection  at all nodes 

4 
 CE 501 Numerical Analysis

Output results & graphs from Matlab

W No. Value (m)

w1 0.0886
w2 0.0846
w3 0.0728
w4 0.0538
w5 0.0288
w6 0.0808
w7 0.0696
w8 0.0514
w9 0.0275
w10 0.0696
w11 a) Displacement (Mesh-plot)
0.0599
w12 0.0443
w13 0.0238
w14 0.0514
w15 0.0443
w16 0.0329
w17 0.0177
w18 0.0275
w19 0.0238
w20 0.0177
w21 0.0095

b) Displacement (Contour-Lines)

5 
 CE 501 Numerical Analysis

.Classical Solution from Timoshenko and Woinowsky-Krieger, 1987

Maximum deflection at midspan of plate
0.00406 𝑞 𝐿 0.00406 5 10
𝛿 0.08867 𝑚
𝐷 2289.377

Computer Software Solution FEA results from (Autodesk Robot Structural Analysis)
with mesh size division 10×10

Maximum deflection is 0.089 m

6 
 CE 501 Numerical Analysis

● Results Summary:
The maximum deflection obtained from the previous three methods

Maximum deflection 𝛿 (m)

F.D.M Classical solution Computer software (FEA)
0.0886 0.08867 0.089

7