Dr. Hossam S.S.

AbdelMeguid
[email protected]
Mansoura University
2011-2012
 Calculus is the mathematics of change.
 Engineers must continuously deal with
systems and processes that change,
making calculus an essential tool of our
profession.

All Copy rights reserved

Dr.Hossam AbdelMeguid 18/09/2011 2
 All Copy rights reserved
Dr.Hossam AbdelMeguid 18/09/2011 3
Definition of derivative : y f  xi  x   f  xi 

rate of change of a x x
dependent variable with
respect to an independent dy f  xi  x   f  xi 
variable  lim x 0
dx x
f (x)
f  xi  x

y

f  xi 
x
All Copy rights reserved
Dr.Hossam AbdelMeguid x
18/09/2011 4
 b
Definition:
I   f  x  dx
it is the total value or a
summation of f(x)dx
over a range of x. In f (x)
fact the integration
symbol is actually a
stylized capital S
intended to signify the
connection between
integration and
summation.
x

All Copy rights reserved

Dr.Hossam AbdelMeguid 18/09/2011 5
 Newton-Cotes Integration Formulas
◦ Trapezoidal rule
◦ Simpson’s Rules
◦ Unequal Segments
◦ Open Integration

 Integration of Equations
◦ Romberg Integration
◦ Gauss Quadrature
◦ Improper Integrals

All Copy rights reserved

Dr.Hossam AbdelMeguid 18/09/2011 6
  Common numerical integration scheme
 Based on the strategy of replacing a complicated
function or tabulated data with some
approximating function that is easy to integrate

b b
I   f  x  dx   f n  x  dx
a a

f n  x   a1 x n  a2 x n 1   an x  an 1 fn(x) is an nth order

polynomial

All Copy rights reserved

18/09/2011 Dr.Hossam AbdelMeguid 7
 The approximation of an integral by the area under
- a first order polynomial
- a second order polynomial
x/10 + sin(x) x/10 + sin(x)

1 1

0.5 0.5

f(x)
f(x)

0 0

-0.5 -0.5

-1 -1

-2 -1 0 1 2 3 4 5 -2 -1 0 1 2 3 4 5
x x

We can also approximated the integral by using a

series of polynomials applied piece
All Copy wise.
rights reserved
Dr.Hossam AbdelMeguid 18/09/2011 8
Newton-Cotes Integration
x/10 + sin(x)

0.5
f(x)

-0.5

-1

-2 -1 0 1 2 3 4 5
x

An approximation of an integral by the

area under straight line segments.
All Copy rights reserved
Dr.Hossam AbdelMeguid 18/09/2011 9
  Closed form - data is at the beginning and end
of the limits of integration
 Open form - integration limits extend beyond
the range of data.
x/10 + sin(x) x/10 + sin(x)

1 1

0.5 0.5

f(x)
f(x)

0 0

-0.5 -0.5

-1 -1

-2 -1 0 1 2 3 4 5 -2 -1 0 1 2 3 4 5
x x

All Copy rights reserved

Dr.Hossam AbdelMeguid 18/09/2011 10
 First of the Newton-Cotes closed integration formulas
 Corresponds to the case where the polynomial is a
first order b b
I   f  x  dx   f1  x  dx
a a

f n  x   a0  a1 x

A straight line can be

represented as:
f(b)
f b  f  a 
f1  x   f  a    x  a
ba f(a)

a
All Copy rights reserved
Dr.Hossam AbdelMeguid 18/09/2011
b 11
 Trapezoidal Rule
b b
I   f  x  dx   f1  x  dx
a a f(b)
b
f b  f  a  f(a)
  f a   x  a  dx
a
ba

Integrate this equation. Results a b

in the trapezoidal rule.

f  a   f b 
I  b  a 

height
2
height
width

Area  base x  average of the heights 

All Copy rights reserved
Dr.Hossam AbdelMeguid 18/09/2011 12
 1
Et   f ' '  b  a 
3

12
where a    b

This indicates that is the function being integrated is linear,

the trapezoidal rule will be exact.

Otherwise, for section with second and higher order

derivatives (that is with curvature) error can occur.

A reasonable estimate of x is the average value of b and a

All Copy rights reserved

Dr.Hossam AbdelMeguid 18/09/2011 13
  Improve the accuracy by dividing the integration interval
into n of smaller segments
 Apply the method to each segment
x1 x2 xn

I  f ( x)dx   f ( x)dx 
x0 x1
 
xn1
f ( x)dx

f  x0   f  x1  f  x1   f  x2  f  xn 1   f  xn 
I h h  h
2 2 2
We can group terms to express a general form
n 1
f  x0   2 f  xi   f  xn 
 b  a
3

I  b  a  i 1
, Et  2
f ''
2n 12n
width
average height
weighted average All Copy rights reserved
Dr.Hossam AbdelMeguid 18/09/2011 14
 Evaluate the following integral using the trapezoidal rule and h = 0.1
1.6 exp(x 2)

I  
2
e x dx 40

35
1
30
b  a 1.6  1
h   0.1  n  6

f(x)=exp(x2)
25

n n 20

n 1 15

f  x0   2 f  xi   f  xn  10

I  b  a  i 1 5

2n 0
0 0.5 1 1.5 2
x

x = 1.0000 1.1000 1.2000 1.3000 1.4000 1.5000 1.6000

Exp(x2)= 2.7183 3.3535 4.2207 5.4195 7.0993 9.4877 12.9358

2.72  2  3.35  4.22  5.42  7.10  9.49   12.94

I  1.6  1  3.74
2 6
Exact = 3.7109
All Copy rights reserved
Dr.Hossam AbdelMeguid Error =0.8%
18/09/2011 15
function [ I ] = TrapezoidalRule( funhandle, a, b ,n )
%TrapezoidalRule a function evaluates the closed integral
%of the funhandlefrom a to b with number of segments n
x=a:(b-a)/n:b;
f=funhandle(x);

I=(b-a)*(f(1)+2*sum(f(2:end-1))+f(end))/(2*n);
end
>> f = inline('exp(x.^2)') >> TrapezoidalRule(f,1,1.6,600)
>> TrapezoidalRule(f,1,1.6,6) ans =
ans = 3.7109
3.7408 >> TrapezoidalRule(f,1,1.6,6000)
>> ans =
TrapezoidalRule(f,1,1.6,60) 3.7109
ans = >>
3.7112

All Copy rights reserved

Dr.Hossam AbdelMeguid 18/09/2011 16
trapz
Trapezoidal numerical integration

Syntax
Z = trapz(Y)
Z = trapz(X,Y)

Description
Z = trapz(Y) computes an approximation of the integral of Y via the trapezoidal method
(with unit spacing).

Z = trapz(X,Y) computes the integral of Y with respect to X using trapezoidal integration.

Inputs X and Y can be complex.

All Copy rights reserved

Dr.Hossam AbdelMeguid 18/09/2011 17
>> f = inline('exp(x.^2)')
f=
Inline function:
f(x) = exp(x.^2)

>> xx=1:(1.6-1)/6:1.6
xx =
1.0000 1.1000 1.2000 1.3000 1.4000 1.5000 1.6000

>> yy=f(xx)
yy =
2.7183 3.3535 4.2207 5.4195 7.0993 9.4877 12.9358

>> trapz(xx,yy)
ans =
3.7408

All Copy rights reserved

Dr.Hossam AbdelMeguid 18/09/2011 18
 Corresponds to the case where the function is a
second order polynomial
b b
I   f  x  dx   f 2  x  dx
a a

f n  x   a0  a1 x  a2 x 2

 Designate a and b as x0 and x2, and estimate

f2(x) as a second order Lagrange polynomial
b b
I   f  x  dx   f  x  dx
a a
2

  x  x1  x  x2 
x2

  f  x0   .......dx
x0   0
x  x1  x0  x2  
All Copy rights reserved
Dr.Hossam AbdelMeguid 18/09/2011 19
 After integration and algebraic manipulation, we
get the following equations
h
I   f  x0   4 f  x1   f  x2  
3
f  x0   4 f  x1   f  x2 
 b  a 
6
width
average height

1
Et   f   b  a 
(4) 5

2880
All Copy rights reserved
Dr.Hossam AbdelMeguid 18/09/2011 20
 x1 x2 xn

I  
x0
f ( x) dx  
x1
f ( x )dx   
xn1
f ( x )dx

n 1 n2
f  x0   4  f  xi   2  f x  f x 
j n

I  b  a 
i 1,3,5.. j  2,4,6..

3n
b  a 
5
 4
Ea  f
180n 4

Number of segments n must be even

The odd points represent the middle term for each
application. Hence carry the weight 4.
The even points are common to adjacent applications and
are counted twice.
All Copy rights reserved
Dr.Hossam AbdelMeguid 18/09/2011 21
 Corresponds to the case where the function is a
third order polynomial
b b
I   f  x  dx   f 3  x  dx
a a

f n  x   a0  a1 x  a2 x  a3 x
2 3

3h
I  f  x0   3 f  x1   3 f  x2   f  x3  
8

All Copy rights reserved

Dr.Hossam AbdelMeguid 18/09/2011 22
 Evaluate the following integral using the Simpson’s 1/3 rule and h = 0.1
1.6 exp(x 2)

I  
2
e x dx 40

35
1
30

b  a 1.6  1

f(x)=exp(x2)
25
h   0.1  n  6 20
n n 15
n 1 n2

 f x   2  f x  f x 
10

f  x0   4 i j n 5

I  b  a 
i 1,3,5.. j  2,4,6.. 0
0 0.5 1 1.5 2
3n x

x = 1.0000 1.1000 1.2000 1.3000 1.4000 1.5000 1.6000

Exp(x2)= 2.7183 3.3535 4.2207 5.4195 7.0993 9.4877 12.9358

2.72  4  3.35  5.42  9.49   2  4.22  7.10   12.94

I  1.6  1  3.7112
3 6
Exact = 3.7109
All Copy rights reserved
Dr.Hossam AbdelMeguid
Error =0.008%23
18/09/2011
function [ I ] = Simpsons13Rule( funhandle, a, b ,n )
if(mod(n,2)~=0)
msgbox('Number of segments n must be even','error','error');
return;
end
x=a: (b-a)/n :b;
f=funhandle(x);

I=(b-a)*(f(1) + 4*sum(f(2:2:n)) + 2*sum(f(3:2:n)) + f(end))/(3*n);

end

>> f = inline('exp(x.^2)') >>Simpsons13Rule(f,1,1.6,10)

>> Simpsons13Rule(f,1,1.6,6) ans =
ans = 3.7109
3.7112

All Copy rights reserved

Dr.Hossam AbdelMeguid 18/09/2011 24
 Experimental and field study data is often
unevenly spaced
 In previous equations we grouped the term (i.e. hi)
which represented segment width.
n 1
f  x0   2 f  xi   f  xn 
I  b  a  i 1

2n
f  x0   f  x1  f  x1   f  x2  f  xn  1   f  xn 
I h h  h
2 2 2
All Copy rights reserved
Dr.Hossam AbdelMeguid 18/09/2011 25
 We should also consider alternately using higher
order equations if we can find data in consecutively
even segments

trapezoidal
1/3 rule
rule
trapezoidal
rule
3/8
rule

All Copy rights reserved

Dr.Hossam AbdelMeguid 18/09/2011 26
 Write a Matlab function to evaluate the
integration of unequal segments of the set of
data X and Y by trapezoidal rule. X and Y are
a vectors of n+1 elements. The first line of
the code is
function [I] = TrapzUnSeg(X, Y)

All Copy rights reserved

Dr.Hossam AbdelMeguid 18/09/2011 27
 Integration of analytical as opposed to tabular
functions
 Romberg Integration
◦ Richardson’s Extrapolation
◦ Romberg Integration Algorithm
 Gauss Quadrature
 Improper Integrals

All Copy rights reserved

Dr.Hossam AbdelMeguid 18/09/2011 28
quad
Numerically evaluate integral, adaptive Simpson quadrature
b
Syntax
q = quad (fun, a, b) q  f  x  dx
a

dblquad
Numerically evaluate double integral over rectangle y xmax
max

  f  x, y  dx dy
Syntax
q = dblquad (fun, xmin, xmax, ymin, ymax) q
ymin xmin

triplequad zmax ymax xmax

Numerically evaluate triple integral
Syntax
q    f  x, y, z  dx dy dz
zmin ymin xmin
q = triplequad (fun, xmin, xmax, ymin, ymax, zmin, zmax)
All Copy rights reserved
Dr.Hossam AbdelMeguid 18/09/2011 29
>> fx = inline('exp(x.^2)');
>> quad(fx, 1,1.6)
ans =
3.7109

>> fxy = inline('exp(x.^2.*y.^2)');

>> dblquad(fxy, 1,1.6, 1,1.6)
ans =
14.4167

>> fxyz = inline('exp(x.^2.*y.^2.*z.^2)');

>> triplequad(fxyz, 1,1.6, 1,1.6, 1,1.6)
ans =

All Copy rights reserved

Dr.Hossam AbdelMeguid 18/09/2011 30
>> syms x y a b c
>> F=a*x*y^2-b*x^2*y-c F  axy 2  bx 2 y  c
F=
- b*x^2*y + a*x*y^2 - c

>> SF = int(F,x,y)
SF =

SF  F dx dy
x*(c - x*((a*y^2)/2 - (b*x*y)/3)) - y*(c - y*((a*y^2)/2 -
(b*y^2)/3))

>> dF = diff(F,x,y)
d 2F
dF 
dF = dx dy
2*a*x*y - b*x^2

All Copy rights reserved

Dr.Hossam AbdelMeguid 18/09/2011 31
 Numerical Differentiation
◦ High accuracy formulas
◦ Richardson’s extrapolation
◦ Unequal spaced data
◦ Uncertain data
 Applied problems

All Copy rights reserved

Dr.Hossam AbdelMeguid 18/09/2011 32
 Background
◦ Often need to approximate derivatives to solve ODE
models
◦ Numerical differentiation: approximate derivatives of a
function using only functional values
◦ Can introduce large errors due to data noise and
numerical inaccuracies

 Definition of derivative

df ( x) f ( x  x)  f ( x)
 lim
dx x 0 x

All Copy rights reserved

Dr.Hossam AbdelMeguid 18/09/2011 33
  Based on the Taylor Series
f ''  xi  2
f  xi 1   f  xi   f '  xi  x  x  .........
2!
◦ Forward finite divided difference
◦ Backward finite divided difference
◦ Center finite divided difference
df ( xi ) f ( xi 1 )  f ( xi )
Forward 
dx x
df ( xi ) f ( xi )  f ( xi 1 )
Backward 
dx x
df ( xi ) f ( xi 1 )  f ( xi 1 )
Central  rights reserved
All Copy
18/09/2011 dxDr.Hossam AbdelMeguid2x 34
f (x) f (x)

true derivative Forward finite divided

difference approx.

x x
f (x ) f (x)

Backward finite divided Centered finite divided

difference approx. difference approx.

x x
All Copy rights reserved
Dr.Hossam AbdelMeguid 18/09/2011 35
 You should be familiar with the following
Tables in your text
◦ Table 23.1: Forward Finite Difference Formulas
◦ Table 23.2: Backward finite difference formulas
◦ Table 23.2: Centered finite difference formulas

All Copy rights reserved

Dr.Hossam AbdelMeguid 18/09/2011 36
 All Copy rights reserved
Dr.Hossam AbdelMeguid 18/09/2011 37
 All Copy rights reserved
Dr.Hossam AbdelMeguid 18/09/2011 38
 All Copy rights reserved
Dr.Hossam AbdelMeguid 18/09/2011 39
The rate of cooling of a body can be
expressed as dT
 k T  Ta 
d
where T =temperature of the body (C), Ta =
temperature of the surrounding medium (C) , Time T (C)
and k =a proportionality constant (per min
minute). Thus, this equation specifies that the
rate of cooling is proportional to the 0 80
difference in the temperatures of the body
and of the surrounding medium. If a metal 5 44.5
ball heated to 80C is dropped into water that
is held constant at Ta = 20C , the temperature 10 30.0
of the ball changes, as
15 24.1
Utilize numerical differentiation to determine
dTldt at each value of time . Plot dTldt versus 20 21.7
T - Ta and employ linear regression to
evaluate k. 25 20.7
All Copy rights reserved
Dr.Hossam AbdelMeguid 18/09/2011 40
Time T (C) method dT/dt Matlab
(min)

0 80 2 point F (44.5-80)/5 -7.1 -10.16

5 44.5 2 point F (30.0-44.5)/5 -2.9 -4.55
10 30.0 2 point F (24.1-30.0)/5 -1.18 -1.74
15 24.1 2 point F (21.7-24.1)/5 -0.48 -0.7
20 21.7 2 point F (20.7-21.7)/5 -0.20 -0.40
25 20.7 2 point B (20.7-21.7)/5 -0.20 0.177

All Copy rights reserved

Dr.Hossam AbdelMeguid 18/09/2011 41
 0

-1

-2

-3
dT/dt

-4

-5

-6

-7

-8
0 10 20 30 40 50 60
T-Ta

0
data 1
linear
dT/dt -2

K=0.12 -4

-6
y = - 0.12*x - 0.029
-8
0 10 20 30 40 50 60
All Copy rights reserved T-Ta
Dr.Hossam AbdelMeguid 18/09/2011 42
Polynomial evaluation- polyval
Syntax: y = polyval(p, x)
returns the value of a polynomial of degree n evaluated at x

Polynomial curve fitting - polyfit

Syntax : p = polyfit(x, y, n)
finds the coefficients of a polynomial p(x) of degree n that fits the
data, p(x(i)) to y(i), in a least squares sense.

Polynomial derivative - polyder

Syntax: k = polyder(p)
calculates the derivative of polynomial p.

Integrate polynomial analytically - polyint

Syntax: I = polyint(p)
returns a polynomial representing the integral of polynomial p
All Copy rights reserved
Dr.Hossam AbdelMeguid 18/09/2011 43
 80 80
data
70 70 polyfit 4

60 60
T(C)

50 50

40 40

30
30

20
20 0 5 10 15 20 25
0 5 10 15 20 25
t(min)
2
by hand
0 by polyfit & polyder
linear
-2

-4
dT/dt

K=0.17 -6
y = - 0.17*x - 0.0065
-8

-10

-12
0 10 20 30 40 50 60
All Copy rights reserved
T-Ta
Dr.Hossam AbdelMeguid 18/09/2011 44
 In addition to unequal spacing, the other problem
related to differentiating empirical data is
measurement error
 Differentiation amplifies error
 Integration tends to be more forgiving
 Primary approach for determining derivatives of
imprecise data is to use least squares regression
to fit a smooth, differentiable function to the data
 In absence of other information, a lower order
polynomial regression is a good first choice
All Copy rights reserved
Dr.Hossam AbdelMeguid 18/09/2011 45
 All Copy rights reserved
Dr.Hossam AbdelMeguid 18/09/2011 46
 THE END

[email protected]

All Copy rights reserved

Dr.Hossam AbdelMeguid 18/09/2011 47