FENG 345 Numerical Methods for

Engineers I

Lecture 4

Roots (Open Methods)

Dr. Osman Nuri Şahin
1
Introduction
• For the bisection method in previous lecture, the root is located within
an interval prescribed by a lower and an upper bound.
• These methods always results in closer estimates of the true value of
the root.

2
Introduction
• The open methods require only a single starting value or two starting
values that do not necessarily bracket the root.
• When the open methods converge
they usually do so much more
quickly than the bisection method.

3
Simple Fixed-point Iteration
• Open methods employ a formula to predict the root.
• Such a formula can be developed for simple fixed-point iteration by
rearranging the function 𝑓(𝑥) = 0 so that 𝑥 is on the left-hand side of
the equation:

• Given an initial guess at the root 𝑥𝑖 , this equation can be used to

compute a new estimate 𝑥𝑖+1 as expressed by the iterative formula

• the approximate error for this equation can be determined using the
error estimator:

4
Example 1
• Use simple fixed-point iteration to locate the root of 𝑓 𝑥 = 𝑒 −𝑥 − 𝑥.
Solution: The function can be separated directly

Starting with an initial guess of 𝑥0 = 0, this iterative equation can be

applied to compute:

5
each iteration brings the estimate closer to the true value of the root: 0.56714329.
Example
• The concepts of convergence and divergence can be depicted graphically.
• pair of equations 𝑦1 = 𝑥 and 𝑦2 = 𝑔 𝑥 .
• These two equations can then be plotted separately.
For the function 𝑓 𝑥 = 𝑒 −𝑥 − 𝑥 ,

Then the two equations

and

6
Example
• If |𝑔′ | < 1, the errors decrease with
each iteration. For |𝑔′ | > 1 the errors
grow.
• convergence occurs when |𝑔′ | < 1.

(a) and (b) convergence and (c) and (d) divergence of simple
fixed-point iteration. 7
NEWTON-RAPHSON
• Perhaps the most widely used of all root-locating formulas is the
Newton-Raphson method.
• If the initial guess at the root is 𝑥𝑖 , a tangent
can be extended from the point 𝑥𝑖 , 𝑓 𝑥𝑖 .
• The point where this tangent crosses the x
axis usually represents an improved estimate
of the root.

8
NEWTON-RAPHSON
• the first derivative at x is equivalent to the
slope:

• which can be rearranged to yield

• which is called the Newton-Raphson formula.

9
Example 2
• Use the Newton-Raphson method to estimate the root of 𝑓 𝑥 = 𝑒 −𝑥 − 𝑥
employing an initial guess of 𝑥0 = 0.
Solution:

Then,

Starting with an initial guess of 𝑥0 = 0, this iterative equation can be applied

to compute
**The approach rapidly converges on the true root. Notice that the true percent relative error at each iteration
decreases much faster than it does in simple fixed-point iteration. 10
Example
• Determine the positive root of 𝑓 𝑥 = 𝑥 10 − 1 using the Newton-
Raphson method and an initial guess of 𝑥 = 0.5.
Solution: The Newton-Raphson formula for this case is

which can be used to compute

• After the first poor prediction, the
technique is converging on the true
root of 1, but at a very slow rate.

11
Example
• The first iteration flings the solution far away from the initial guess to a new
value (x = 51.65) where f (x) has an extremely high value.

12
Example
• There is no general convergence criterion for Newton-Raphson. Its
convergence depends on the nature of the function and on the accuracy of
the initial guess.

13
SECANT METHODS
• A potential problem in implementing the Newton-Raphson method is the
evaluation of the derivative.
• Although this is not inconvenient for polynomials and many other
functions, there are certain functions whose derivatives may be difficult
or inconvenient to evaluate.
• For these cases, the derivative can be approximated by a backward finite
divided difference:

14
SECANT METHODS
• Then

This is the formula for the secant method. Notice that the approach
requires two initial estimates of x.

15
 Non-linear algebraic equations. Newton-Raphson method.

f (x)  0 f (a ) f
Taylor series : f ( x )  f (a )  ( x  a )  ... 0  f  f   x n 1  x n  
f
İ
1
Example 4.1 2 sin  u   u a) Find u. b) Write a MatLAB program to find u.

f (u )  u  2 sin  u   0 f ( x )  x  2 sin  x   0
df 1
Solution: a)  1 cos( x )
dx x

0.6829
f (1)  1  2 sin  1  0.6829
1 1   1.4856
Initial guess x1=1 f (1)  1  cos( 1)  0.4597
1 0.4597
x2=1+1.4856=2.4856 f (2.4856)  2.4856  2 sin  2.4856   0.4856
 0.4856
f (2.4856)  1 
1
cos( 2.4856 )  1.0091 2   0.4812
2.4856 1.0091

x3=2.4856-0.4812=2.0044 f (2.0044)  2.0044  2 sin  2.0044   0.0284

1  0.0284
f (2.0044)  1  cos(2.0044)  0.7814 3   0.0363
2.0044 0.7814

x4=2.0044-0.0363=1.9681 … continued  4  0.0049 x5=1.9681+0.0049=1.9730

5  0.0007 x5=1.9730-0.0037=1.9723
 f
f ( x )  x  2 sin  x   0
df 1
b)  x n 1  x n    1 cos( x )
f dx x

% nr1.m
clc,clear,close all
x=1;xe=[0.001];niter=20;kerr=1; ans = 0 4.0000 -0.0001 1.9724
for n=1:1:niter;
f=x-2*sin(sqrt(x)); fd=1-(1/sqrt(x))*cos(sqrt(x));
eps=-f/fd; x=x+eps; if abs(eps)<xe;kerr=0;break;end
end
[kerr,n,eps,x]
 Example 4.2

Write a MatLAB program

Solution:
h 2
x 2 df 2x x 2
(9  h )  30 (9  x )  30  0  (9  x)  (1)
3 3 dx 3 3
% nr1.m
clc,clear,close all ans = 0 3.0000 -0.0001 2.0269
x=3;xe=[0.001];niter=20;kerr=1;
for n=1:1:niter;
f=pi*x^2*(9-x)/3-30; fd=2*pi*x*(9-x)/3-pi*x^2/3; h=2.0269 m
eps=-f/fd; x=x+eps; if abs(eps)<xe;kerr=0;break;end
end
[kerr,n,eps,x]
 Nonlinear set of algebraic equations. Newton-Raphson method.
f (x)  0 f   f x n 1  x n  
(This subject will be in the last weeks of the semester)
Error equations :
f1 ( x1 , x 2 )  0
 f1 f1  x1 n 1  x1 n  1
f 2 ( x1 , x 2 )  0  x x 2   1    f1 
 1  
 f 2 f 2   2   f 2  x 2 n 1  x 2 n   2
 x1 x 2 

Example 4.3 2u  4.1sin(3v)  3 2 v  3u 2  10.5 Write a MatLAB program to find u and v.
% nr.m
f1  2 x1  4.1sin(3x 2 )  3  0 clc, clear,close all
x=[-1;3];xe=[0.001;0.001];niter=20;kerr=1;
f 2  3x12  2 x 2  10.5  0
%------------------
Error equations : for n=1:1:niter;
%------------------
 2  4.1(3) cos(3x 2 )  1    f1  b=[2*x(1)-4.1*sin(3*x(2))+3;3*x(1)^2+2*x(2)-10.5];b=-b;
6 x    
 1 2  
 2   2  f a=[2,-4.1*3*cos(3*x(2));6*x(1),2];
%------------------
eps=inv(a)*b;x=x+eps;if abs(eps)<xe;kerr=0;break;end
end
x1  1.1993, x 2  3.0925
[kerr,n],[eps,x]
 Home Excercise 4.1

Write a MatLAB program.

Home Excercise 4.2

 Home Excercise 4.1

Write a MatLAB program.

Solution: x=0:1:200;u=6.26./sqrt(x);f=6.26*sqrt(x).*tanh(u)-36;plot (x,f)
gm  gc  e x  ex
v( t )  tanh t  tanh( x )  x x
c  m  e e % nr1.m
clc,clear,close all
g=9.81, c=0.25, t=4, v=36 x=100;xe=[0.001];niter=20;kerr=1;
for n=1:1:niter;
 6.26  u=6.26/sqrt(x);f=6.26*sqrt(x)*tanh(u)-36;
36  6.26 m tanh  f ( x )  6.26 x tanh u ( x )   36 fd=3.13/sqrt(x)*tanh(u)+6.26*sqrt(x)*(-3.13)*x^(-1.5)...
 m
*(((exp(u)+exp(-u))^2-((exp(u)-exp(-u))^2)))/((exp(u)+exp(-u))^2);
eps=-f/fd; x=x+eps; if abs(eps)<xe;kerr=0;break;end
6.26 e u  e u
u(x)  tanh(u )  u u end
x e e [kerr,n,eps,x],vpa(x)

u 2 u 2
1.5 (e  e )  (e  e )
u u
df 1
 3.13 tanh u ( x )   6.26 x (3.13) x ans = 144.93259626492792335739068221301
dx x (e u  e  u ) 2
Home Excercise 4.2 9.81
Solution: 1  90(tan 0 )  (8100)  1.8
(2)(900) cos 2 0

44.145
f  90(tan x )   0.8  0
cos 2 x

90 2 sin x cos x
f   44.145
cos 2 x cos 4 x

th0=0:1:80;x=th0*pi/180;f=90*tan(x)-44.145./(cos(x)).^2+0.8;plot (x,f)

nr1.m
clc,clear,close all
x=30*pi/180;xe=[0.001];niter=20;kerr=1;
for n=1:1:niter;
f=90*tan(x)-44.145/(cos(x))^2+0.8;
fd=90/(cos(x))^2-2*44.145*sin(x)*cos(x)/(cos(x))^4;
eps=-f/fd; x=x+eps; if abs(eps)<xe;kerr=0;break;end
end
[kerr,n,eps,x*180/pi]

0  37.95o for initial guess 30o

0  51.53o for initial guess 60o