Scribd
Upload
54 views

STR 613: Advanced STR 613: Advanced Numerical Analysis Numerical Analysis

This document discusses numerical methods for solving nonlinear equations. It begins by noting that nonlinear equations have many applications but can be much more difficult and computationally expensive to solve than linear equations. The document then provides an example of a simple linkage problem that results in a nonlinear equation. It explains that solving nonlinear equations typically involves two phases: bounding the solution range and then refining the solution. Four common refinement methods are described - interval halving, fixed-point iteration, Newton's method, and the secant method. The bisection method, a type of interval halving, is explained in detail through an example problem. Convergence properties of the bisection method are also summarized.

Uploaded by

Hamza El-masry
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
54 views

STR 613: Advanced STR 613: Advanced Numerical Analysis Numerical Analysis

This document discusses numerical methods for solving nonlinear equations. It begins by noting that nonlinear equations have many applications but can be much more difficult and computationally expensive to solve than linear equations. The document then provides an example of a simple linkage problem that results in a nonlinear equation. It explains that solving nonlinear equations typically involves two phases: bounding the solution range and then refining the solution. Four common refinement methods are described - interval halving, fixed-point iteration, Newton's method, and the secant method. The bisection method, a type of interval halving, is explained in detail through an example problem. Convergence properties of the bisection method are also summarized.

Uploaded by

Hamza El-masry
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/ 4

STR 613: Advanced Website

Numerical Analysis  Assignment 3 is now online


 Due Dec. 16
Instructor
Dr. Ahmed Amir Khalil

Nonlinear equations Difficulty!


 Endless applications
 Structural analysis Nonlinear problems can cost as
1. Material nonlinearity much as 10-100 times as much
a. Plasticity to solve as corresponding linear
b. Nonlinear elasticity
problems!
2. Geometric nonlinearity
We often try to approximate
a. Large deflections
b. Large rotations
nonlinear solutions by linear
solutions

Typical Nonlinear Problem


1 D-
D-O-F Problem Statement

k u

P
(k 0 + k N )u = P
k = k0 + kN
k N = f (u)
k0 constant Given P find u.
kN function of u Assume f(u) is a known function.

1
Nonlinear Equations Nonlinear Equations
 Simple linkage problem
 Simple linkage problem

Nonlinear Equations Nonlinear Equations


 Freudenstein Equation  Numerical solution involves two phases
R1 cos R2 cos + R3 cos( ) = 0 A. Bounding the Solution
B. Refining the Solution
5 5 11
cos cos + cos( ) = 0
3 2 6

 Given

 Find phi

Nonlinear Equations Nonlinear Equations

A. Bounding the solution B. Refining the Solution


 Systematic search Four methods shall be presented:
 f(x) for 1<x<
<x<11 intervals of x  Interval halving.
 f(1
f(1/x) for 1<1/x<
/x<11 intervals of 1/x  Fixed--point iteration.
Fixed
 Newtons method.
 Past experience
 Secant method.
 Solution of a simplified approximate model
 Previous solution in an iterative procedure

2
Bisection Method
Bisection Method

Example: i a f(a) b f(b) c


1 30 -0.0397719 40 0.19496296 35.0
 Solve the four-
four-bar linkage problem for an 2 30 -0.0397719 35 0.06599926 32.5
input of = 40 by interval halving 3 30 -0.0397719 32.5 0.01015060 31.25
4 31.25 -0.01556712 32.5 0.01015060 31.875
5 5 11
f ( ) = cos 40 cos + cos(40 ) = 0 5 31.875 -0.00289347 32.5 0.01015060 32.1875
3 2 6
6 31.875 -0.00289347 32.1875 0.00358236 32.03125
7 31.875 -0.00289347 32.03125 0.00033288 31.953125
.
2 32.015176 -0.00000009 32.01518 0.000000 32.015178
2 1
2 32.015178 -0.00000004 32.01518 0.000000 32.015179
3 1
2 32.015179 -0.00000002 32.01518 0.000000 32.01518
4 1

Bisection Algorithms
Convergence of Bisection
 while (b-a)>tol  fa = f(a);
 m = (a+b)/2;  while (b-a)>
Method
 if f(a)*f(m)<=0 tol + eps*max(|a|,|b|) ba
 b = m;  m = (a + b)/2;
ek = x k r Zero
 else
 a = m;
 fm = f(m);
 if fa*fm<=0
2 k +1
 end;  b = m; ek+1 < c ek
 end;  else
 a = m; fa = fm; ek : error in kth iteration Linear convergence
 r = (a + b)/2; One bit per iteration
 end; r: root Slow but sure
While loop?  end; C: constant
No. of function evaluations?
 r = (a + b)/2;

3
Pitfalls

You might also like