International Journal of Science and Research (IJSR)

ISSN: 2319-7064
ResearchGate Impact Factor (2018): 0.28 | SJIF (2018): 7.426

Finite Element Analysis of a 2-D Linear Static

Structure (Flat Plate) Using Matlab
Ranbir Mukhya1, Areeb Siraj2, Gaurav Arya3
1
Guide, Faculties of Mechanical Engineering Department
2, 3
Department of Mechanical Engineering, University Institute of Engineering and Technology, C.S.J.M University, Kanpur – 208024

Abstract: A finite Element method approach has been developed for structural analysis of various industrial products. It helps to
analyze and solve the problem of structural displacement due to various internal and external forces. And it also helps in the calculation
stresses and strains. This method can be applied to typical MEMS structures such as beam, plate and other complex structures. The
theoretical analysis of the finite element method is well established in the case of the triangular or tetrahedral meshes. In this case we
use 4-node rectangular mesh, for the analysis of a flat plate of uniform thickness with the help of Matlab. Our theory supported by the
some numerical experiments, which are taken from various engineering application, ranging from elasticity and in other forces and
stress analysis.

1. Introduction parts, objects under stress etc. FEA simulations provide a

valuable resource as they remove multiple instances of
Finite Element Analysis (FEA) is a numerical method for creation and testing of hard prototypes for various high
solving problems of engineering and mathematical physics. fidelity situations. For instance, in a frontal crash simulation
Typical problem areas of interest include structural analysis, it is possible to increase prediction accuracy in "important"
heat transfer analysis, fluid flow analysis etc. The finite areas like the front of the car and reduce it in its rear thus
element method formulation of the problem results in a reducing cost of the simulation.
system of algebraic equations. The method yields
approximate values of the unknowns at discrete number of The purpose of finite element analysis (FEA) is to reduce
points over the domain. To solve the problem, it subdivides the number of prototypes and experiments that have to be
a large problem into smaller, simpler parts that are called run when designing, optimizing, or controlling a device or
finite elements. The simple equations that model these finite process.
elements are then assembled into a larger system of
equations that models the entire problem. 2. Terminologies
In the first step above, the element equations are simple 1) Discretization - It is a method of subdividing a complex
equations that locally approximate the original complex structure (structural problem) into a convenient no. of
equations to be studied, where the original equations are smaller components.
often partial differential equations (PDE). To explain the 2) Nodes –A node is a specific point in the finite element at
approximation in this process, FEM is commonly which the value of the field variable is to be explicitly
introduced as a special case of Galerkin method. The calculated.
process, in mathematical language, is to construct an 3) Stiffness Matrix – The stiffness matrix contains the
integral of the inner product of the residual and the weight geometric and material behavior information that
functions and set the integral to zero. In simple terms, it is a indicates the resistance of the element to deformation
procedure that minimizes the error of approximation by when subjected to loading. Such deformation may
fitting trial functions into the PDE. The residual is the error include axial, bending, shear, and torsional effects. For
caused by the trial functions, and the weight functions finite elements used in non-structural analyses, such as
are polynomial approximation functions that project the fluid flow and heat transfer, the term stiffness matrix is
residual. also used, since the matrix represents the resistance of
the element to change when subjected to external
FEA as applied in engineering is a computational tool for influences.
performing engineering analysis. It includes the use of mesh 4) Boundary Conditions - Boundary conditions are a set of
generation techniques for dividing a complex problem into loading, constraints and contact conditions that define
small elements, as well as the use of software program (in the status of your simulation in one possible design
our case MATLAB) coded with FEM algorithm. configuration.
5) Node Numbering:
FEA is a good choice for analyzing problems over
complicated domains like cars and oil pipelines, machinery

Volume 8 Issue 7, July 2019

www.ijsr.net
Licensed Under Creative Commons Attribution CC BY
Paper ID: ART20198931 10.21275/ART20198931 345
 International Journal of Science and Research (IJSR)
ISSN: 2319-7064
ResearchGate Impact Factor (2018): 0.28 | SJIF (2018): 7.426

{F} = Nodal Force Vector (Action)

Basic steps for solving structural problems using FEA 4. Assemble element equations to form global system
1. Domain discretization 5. Apply initial and boundary conditions
2. Select element type 6. Solve assembled system of equations for primary
a) 4-Noded triangle unknown variable (in our case displacement)
b) 8-Noded triangle 7. Solve assembled system of equations for secondary
c) 4-Noded quadrilateral unknown variable
d) 8-Noded quadrilateral (in our case)
3. Derive element equations by variational or energy a) Deflection
methods b) Strain
[K]{U} = {F} c) Stress
[K] = Stiffness or Property Matrix (Property)
{U} = Nodal Displacement Vector (Behavior)

Volume 8 Issue 7, July 2019

www.ijsr.net
Licensed Under Creative Commons Attribution CC BY
Paper ID: ART20198931 10.21275/ART20198931 346
 International Journal of Science and Research (IJSR)
ISSN: 2319-7064
ResearchGate Impact Factor (2018): 0.28 | SJIF (2018): 7.426
Problem Based on 4 node quad. Element for flat plate of uniform thickness

Ques. An aluminum sheet of uniform thickness 10mm fixed at one end, whereas a uniform uni-axial load of 100 KN is
applied on the other end. Using FEA do structural analysis and find primary and secondary variables. Take Young’s modulus
= 2.15*109 and Poisson’s ratio = 0.4

Formulation of Element Equations

Volume 8 Issue 7, July 2019

www.ijsr.net
Licensed Under Creative Commons Attribution CC BY
Paper ID: ART20198931 10.21275/ART20198931 347
 International Journal of Science and Research (IJSR)
ISSN: 2319-7064
ResearchGate Impact Factor (2018): 0.28 | SJIF (2018): 7.426

Volume 8 Issue 7, July 2019

www.ijsr.net
Licensed Under Creative Commons Attribution CC BY
Paper ID: ART20198931 10.21275/ART20198931 348
 International Journal of Science and Research (IJSR)
ISSN: 2319-7064
ResearchGate Impact Factor (2018): 0.28 | SJIF (2018): 7.426

MATLAB Code e(eID).D = zeros(3,3);

% Calculate 'D' matrix
e(eID).D(1,1) = e(eID).E/(1-e(eID).nu^2);
n = [2 2]; e(eID).D(1,2) = e(eID).E/(1-e(eID).nu^2)*e(eID).nu;
L = [.1 .1]; e(eID).D(2,1) = e(eID).E/(1-e(eID).nu^2)*e(eID).nu;
nu = .4; e(eID).D(2,2) = e(eID).E/(1-e(eID).nu^2);
E = 2.15e9; %YOUNG'S MODULUS e(eID).D(3,3) = e(eID).E/(1-
nx = n(1)+1; %No. of nodes by X e(eID).nu^2)*0.5*e(eID).nu;
ny = n(2)+1; %No. of nodes by y eID = eID+1;
nNodes = nx*ny; %total no. of nodes end
ne = n(1)*n(2); %total no. of elements end
bnd = []; %boundary condition
dx = L(1)/n(1); %space step by X
dy = L(2)/n(2); %space step by y for eID= 1:ne
Klocal = zeros(8,8); % Local stiffness matrix
%Gaussian points for gp =1:4 % Loop over all gaussian
Zeta = [-(1/3)^0.5; -(1/3)^0.5; (1/3)^0.5; (1/3)^0.5 ]; points
%local zeta zeta = Zeta(gp,1);
Eta = [-(1/3)^0.5; (1/3)^0.5; (1/3)^0.5; -(1/3)^0.5 ]; %local eta = Eta(gp,1);
eta dNdzeta(1) = -0.25*(1-eta);
W = 1; dNdzeta(2) = -0.25*(1+eta);
U = zeros(2*nNodes,1); %displacement matrix dNdzeta(3) = 0.25*(1+eta);
K = zeros (2*nNodes,2*nNodes); %stiffness matrix dNdzeta(4) = 0.25*(1-eta);
res.F = zeros(2*nNodes,1); %force vector dNdeta(1) = 0.25*(1-zeta);
eID = 1; dNdeta(2) = -0.25*(1-zeta);
dNdeta(3) = -0.25*(1+zeta);
%Loop over all elements dNdeta(4) = 0.25*(1+zeta);
for j=1:n(2) dNdlocal = [dNdzeta; dNdeta];
for i=1:n(1) % Jacobian Matrix
% Node IDs J(1,1) = dNdzeta*e(eID).x;
e(eID).n(1) = (j-1)*nx+i; J(1,2) = dNdzeta*e(eID).y;
e(eID).n(2) = j*nx+i; J(2,1) = dNdeta*e(eID).x;
e(eID).n(3) = j*nx+i+1; J(2,2) = dNdeta*e(eID).y;
e(eID).n(4) = (j-1)*nx+i+1; dNdglobal = J\dNdlocal;
% Global node position
e(eID).x(1,1) = (i-1)*dx; B(1,1)= dNdglobal(1,1);
e(eID).y(1,1) = (j-1)*dy; B(1,2)= 0;
e(eID).x(2,1) = (i-1)*dx; B(1,3)= dNdglobal(1,2);
e(eID).y(2,1) = j*dy; B(1,4)= 0;
e(eID).x(3,1) = i*dx; B(1,5)= dNdglobal(1,3);
e(eID).y(3,1) = j*dy; B(1,6)= 0;
e(eID).x(4,1) = i*dx; B(1,7)= dNdglobal(1,4);
e(eID).y(4,1) = (j-1)*dy; B(1,8)= 0;
e(eID).U = zeros(8,1); % Modal displacement B(2,1)= 0;
e(eID).Umagnitude = zeros(4,1); % Magnitude of B(2,2)= dNdglobal(2,1);
displcement B(2,3)= 0;
e(eID).E = E; % Young's modulus B(2,4)= dNdglobal(2,2);
e(eID).nu = nu; % Nu B(2,5)= 0;

Volume 8 Issue 7, July 2019

www.ijsr.net
Licensed Under Creative Commons Attribution CC BY
Paper ID: ART20198931 10.21275/ART20198931 349
 International Journal of Science and Research (IJSR)
ISSN: 2319-7064
ResearchGate Impact Factor (2018): 0.28 | SJIF (2018): 7.426
B(2,6)= dNdglobal(2,3);
B(2,7)= 0;
B(2,8)= dNdglobal(2,4);
B(3,1)= dNdglobal(2,1);
B(3,2)= dNdglobal(1,1);
B(3,3)= dNdglobal(2,2);
B(3,4)= dNdglobal(1,2);
B(3,5)= dNdglobal(2,3);
B(3,6)= dNdglobal(1,3);
B(3,7)= dNdglobal(2,4);
B(3,8)= dNdglobal(1,4);
Klocal = Klocal+W*det(J)*B.'*e(eID).D*B*1e-8; %
Final local stiffness matrix
end
% Global stiffness matrix
for i=1:4
for j=1:4
K(e(eID).n(i)*2-1:e(eID).n(i)*2,e(eID).n(j)*2-
1:e(eID).n(j)*2)=...
K(e(eID).n(i)*2-1:e(eID).n(i)*2,e(eID).n(j)*2-
1:e(eID).n(j)*2)...
+ Klocal(i*2-1:i*2,j*2-1:j*2);
end
end
end
F = zeros(2*nNodes,1);
U = zeros(2*nNodes,1);
Del = zeros(2*nNodes,1);
Strain = zeros(2*nNodes,1);
Stress = zeros(2*nNodes,1);
for i = 1:6:2*nNodes
F(i+4,1) = 100;
F = F+F(i,1);
end
RCOND = 1.908653e-19;
U = (K\F)*RCOND;
%{
Due to the warning given below U is multiplied by RCOND
= 1.908653e-19

Warning: Matrix is close to singular or badly scaled

Results may be inaccurate. RCOND = 1.908653e-19
%}
Del = diff(U);
Strain = Del/L(1);
Sress = Strain*E;

Warning: Matrix is close to singular or badly scaled. Results

may be inaccurate. RCOND = 1.908653e-19.
Published with MATLAB® R2014a

3. Results 4. Conclusion
The conclusion that can be drawn from the above results is;
 Finite element analysis is an essential tool for helping us
in determining the cause of the problems and
recommending the solutions
 FEM analysis of structural failure should be adopted as
standard tool in failure analysis
 With a trained engineer, FEM is quick and easy to deploy
 With the exponential increase in computing power FEM is
economical to carry out.

Volume 8 Issue 7, July 2019

www.ijsr.net
Licensed Under Creative Commons Attribution CC BY
Paper ID: ART20198931 10.21275/ART20198931 350
 International Journal of Science and Research (IJSR)
ISSN: 2319-7064
ResearchGate Impact Factor (2018): 0.28 | SJIF (2018): 7.426

References
Websites
[1] https://nptel.ac.in/reviewed_pdfs/105106051/lec36.pdf
[2] https://www.researchgate.net/.../268188312_A_Four-
node_quadrilateral_element_for_fi...
[3] https://www.emeraldinsight.com/doi/full/10.1108/EC-
04-2015-0102?fullSc=1
[4] web.mit.edu/.../The_finite_element_method_with_ove
rlapping_elements_a_new_para...

Books
[5] Finite element method by J.N Reddy
[6] Computer aided design by R.K Srivastava
[7] Finite element analysis by P. Seshu

Programming books for Matlab

[8] Understanding MATLAB by S.N Alam and S.S Alam
[9] Introduction to MATLAB by Mathworks

Volume 8 Issue 7, July 2019

www.ijsr.net
Licensed Under Creative Commons Attribution CC BY
Paper ID: ART20198931 10.21275/ART20198931 351