Scribd
Upload
44 views

Pre-Analysis: Example: Steady One-Dimensional Heat Conduction in A Bar

1. The document outlines the key steps in a finite element analysis of a 1D heat conduction problem: defining the mathematical model, discretizing the model into elements, deriving equations relating nodal temperatures, and inverting the equations to solve for temperatures. 2. It discusses how increasing the number of elements or order of polynomials used to approximate temperatures can reduce errors in the solution. 3. The finite element solution satisfies an integral form of the governing equations and exactly satisfies essential boundary conditions but only approximately satisfies natural boundary conditions like heat fluxes.

Uploaded by

Anas Elshabli
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
44 views

Pre-Analysis: Example: Steady One-Dimensional Heat Conduction in A Bar

1. The document outlines the key steps in a finite element analysis of a 1D heat conduction problem: defining the mathematical model, discretizing the model into elements, deriving equations relating nodal temperatures, and inverting the equations to solve for temperatures. 2. It discusses how increasing the number of elements or order of polynomials used to approximate temperatures can reduce errors in the solution. 3. The finite element solution satisfies an integral form of the governing equations and exactly satisfies essential boundary conditions but only approximately satisfies natural boundary conditions like heat fluxes.

Uploaded by

Anas Elshabli
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 12

Pre-Analysis

1. Mathematical model
2. Numerical solution procedure
3. Hand-calculations of expected results/trends

Example: Steady One-Dimensional


Heat Conduction in a Bar
y

x
z
L
We are interested in finding the temperature
distribution in the bar due to heat conduction

1
Energy Conservation for an
Infinitesimal Control Volume
Infinitesimal
“Control Volume”

Δ
x
Δ
Δ

Mathematical Model: Governing


Equation and Boundary Conditions
• Governing equation
 + = 0, 0≤ ≤
• Boundary conditions
 0 =
q L = q = −
• Exact solution is straightforward

2
Numerical Solution:
Discretization
• Reduce the problem to determining temperature
values at selected locations (“nodes”)
T

1 2 3 4 x

x
We have assumed a shape for ( ) consisting of piecewise
polynomials

How to Find Nodal


Temperatures ?
System of
Mathematical Model Invert Nodal
algebraic
(Boundary Value temperatures
equations in nodal
Problem)
temperatures
Post
Piecewise
={ } processing
polynomial
approximation for T
( )
Each algebraic
equation will
relate a nodal
1 2 3 4 temperature to
its neighbors

3
How to Derive System of Algebraic
Equations?
Piecewise Weighted Integral Piecewise polynomial
polynomial Form approximation for T
approximation for T

∫ ∫ + =0
(x) is an
+ =0 + =0 arbitrary piecewise
polynomial function
( ) is an
System of algebraic arbitrary function
eqs. in nodal System of algebraic eqs.
temperatures in nodal temperatures

How to Derive System of Algebraic


Equations?
Piecewise polynomial
approximation for T

∫ + dx = 0
(x) is an
arbitrary piecewise
polynomial function

System of algebraic eqs. x


in nodal temperatures

4
Integration by Parts

• ∫ + =0

• w k − ∫ k +
∫ =0

dT dT
w k − k + =0

1 2 3 4

w + + 0.5 Δ − +

+ + + Δ +

+ + + Δ +

w + + 0.5 Δ + =0

5
dT dT
w k − k + =0

1 2 3 4

⋯+w + + 0.5 Δ +

dT dT
w k − k + =0

1 2 3 4

6
dT dT
w k − k + =0

1 2 3 4

w + + 0.5 Δ − +

+ + + Δ +

+ + + Δ +

w + + 0.5 Δ + =0

={ }

dT dT
w k − k + =0

1 2 3 4

={ }

7
dT dT
w k − k + =0

1 2 3 4

w + + 0.5 Δ − +

+ + + Δ +

+ + + Δ +

w + + 0.5 Δ + =0

dT dT
w k − k + =0

1 2 3 4

+ = 0.5 Δ − ={ }

+ + = Δ

+ + = Δ

+ = 0.5 Δ +

8
Essential Boundary Conditions

=T + = 0.5QΔ −

+ + = Δ

+ + = Δ

+ = 0.5 Δ +

Comparison of Finite-Element and Exact


Solutions
• Nodal temperature values are exact
– Unusual property of 1D FE solution
• Temperature boundary condition is
satisfied exactly
• Flux boundary condition is satisfied
approximately

9
Comparison of / between Finite-
Element and Exact Solutions
• Error in / > Error in
• Energy is not conserved for
each element

“Reaction” at Left Boundary

 =− =
− 5.5 W/m
• Energy is
conserved for
the bar

10
How to Improve the Polynomial
Approximation?
• Increase no. of elements Original Mesh
• Increase order of polynomial 4
1 2 3
within each element
– Use more nodes per Refined Mesh
element
1 2 3 4 5 6 7

Second-Order Element

Error Reduction: Results


1 element, second-
3 elements 6 elements
order polynomial

11
Finite-Element Analysis: Summary of
the Big Ideas
• Mathematical model to be solved is usually a boundary value
problem
• Reduce the problem to solving selected variable(s) at selected
locations (nodes)
• Assume a shape for selected variable(s) within each element
• Derive system of algebraic equations relating neighboring nodal
values
• Invert this system to determine selected variable(s) at nodes
• Derive everything else from selected variable(s) at nodes

Finite-Element Analysis: Summary of


the Big Ideas
• Reduce error by using more elements and/or
increasing the order of interpolation
• Finite-element solution doesn’t satisfy the
differential equation(s)
– Satisfies a special weighted integral form
• Essential boundary conditions are satisfied
exactly
• Natural or gradient boundary conditions are
satisfied approximately

12

You might also like