Part 5

Chapter 17

Numerical Integration
Formulas

Prof. Hae
Hae--Jin Choi

[email protected]

School of Mechanical Engineering

Chung-Ang University
Numerical Methods 2010-2 1
 Chapter Objectives
l Recognizing that Newton-Cotes integration formulas are based on the
strategy of replacing a complicated function or tabulated data with a
polynomial that is easy to integrate.
l Knowing how to implement the following single application Newton-
Cotes formulas:
§ Trapezoidal rule
§ Simpson’s 1/3 rule
§ Simpson’s 3/8 rule
l Knowing how to implement the following composite Newton-Cotes
formulas:
§ Trapezoidal rule
§ Simpson’s 3/8 rule
l Recognizing that even-segment-odd-point formulas like Simpson’s 1/3
rule achieve higher than expected accuracy.
l Knowing how to use the trapezoidal rule to integrate unequally spaced
data.
l Understanding the difference between open and closed integration
formulas. School of Mechanical Engineering
Chung-Ang University
Numerical Methods 2010-2 2
 Integration
l Integration: b
I= ò f (x) dx
a
is the total value, or summation, of f(x) dx over the range from a to b:
l I represents the area under the curve f(x) between x= a and b.

School of Mechanical Engineering

Chung-Ang University
Numerical Methods 2010-2 3
 Integration in Engineering and Science
l This integral can be evaluated over a line, an area, or a volume.
l For example the total mass of gas contained in a volume is given as
the product of the density and the volume. However, suppose that
the density varies from location to location within a volume, it is
necessary to sum the product
n
mass = å ri DVi
i =1

l For a continuous case, the integration is expressed by

mass = òòò r ( x, y, z )dxdydz mass = òòò r (V )dV

V

l There is strong analogy between summation and integration

à Basis of numerical integration

School of Mechanical Engineering

Chung-Ang University
Numerical Methods 2010-2 4
 Newton--Cotes Formulas
Newton
l The Newton-Cotes formulas are the most common numerical
integration schemes.
l Generally, they are based on replacing a complicated function
or tabulated data with a polynomial that is easy to integrate:
b b
I= ò f (x) dx @ ò f (x) dx
a a n

where fn(x) is an nth order interpolating polynomial.

f n ( x) = a0 + a1 x + L + a n -1 x n -1 + a n x n
�
First-order Parabola
polynomial
(line)

School of Mechanical Engineering

Chung-Ang University
Numerical Methods 2010-2 5
 The Trapezoidal Rule

l The integral can be

approximated using a series of
polynomials applied piecewise
to the function or data over
segments of constant length.
l For example, three straight line
segments are used to
approximate the integral.
Higher-order polynomial can be
used for the same purpose.

School of Mechanical Engineering

Chung-Ang University
Numerical Methods 2010-2 6
 The Trapezoidal Rule
l The trapezoidal rule is the first
of the Newton-Cotes closed
integration formulas; it uses a
straight-line approximation for
the function:
b
I = ò f n ( x ) dx
a

é
b f (b) - f ( a ) ù
I = ò ê f (a) + ( x - a )ú dx
a
ë b-a û
f ( a ) + f (b)
I = (b - a ) ® I = (b - a )(average height)
2

School of Mechanical Engineering

Chung-Ang University
Numerical Methods 2010-2 7
 Error of the Trapezoidal Rule
l An estimate for the local truncation
error of a single application of the
trapezoidal rule is:
1 3
Et = - f ¢¢(x )(b - a)
12
where x is somewhere between a
and b.

� l This formula indicates that the error

is dependent upon the curvature of
the actual function as well as the
distance between the points.
l Error can thus be reduced by
breaking the curve into parts.

School of Mechanical Engineering

Chung-Ang University
Numerical Methods 2010-2 8
 Example 17.1 (1/2)
l Q. Use the trapezoidal rule to numerically integrate the following
equation from a = 0 to b = 0.8. The true solution is 1.640533.

f ( x) = 0.2 + 2.5 x - 200 x 2 + 675 x 3 - 900 x 4 + 400 x 5

0.2 + 0.232
Sol.) I = (0.8 - 0) = 0.1728
2
® Et = 1.640533 – 0.1728 = 1.467733
® et = 89.5%

School of Mechanical Engineering

Chung-Ang University
Numerical Methods 2010-2 9
 Example 17.1 (2/2)
Approximate error:
f ¢¢( x) = -400 + 4,050 x - 10,800 x 2 + 8,000 x 3
0.8
ò0 (-400 + 4,050 x - 10,800 x 2 + 8,000 x 3 )dx
f ¢¢( x) = = -60
0.8 - 0
1
E a = - (-60)(0.8) 3 = 2.56
12

note : this value is of the same order of magnitude and sign as

the true error. Average second derivative is not an
accurate approximation of f’’(ξ), so a discrepancy exists
and Ea rather Et.

School of Mechanical Engineering

Chung-Ang University
Numerical Methods 2010-2 10
 Composite Trapezoidal Rule
l One way to improve the accuracy is to divide
the integration interval from a to b into a
number of segments and apply the method to
each segment.
l Assuming n+1 data points are evenly spaced,
there will be n intervals over which to
integrate. n

(b - a ) 3 n
å f ¢¢(ξ ) i
l The total integral can be calculated by Et = - å f ¢¢(ξ )
i f ¢¢ @ i =1
12n 3 i =1 n
integrating each subinterval and then adding
them together: n
(b - a )3
å f ¢¢(ξ i ) @ nf ¢¢ Ea = -
12n 2
f ¢¢
xn x1 x2 xn i =1
I = ò f n ( x ) dx = ò f n ( x ) dx + ò f n ( x ) dx + L + ò f n ( x ) dx
x0 x0 x1 xn-1

f ( x0 ) + f ( x1 ) f ( x1 ) + f ( x2 ) f ( xn -1 ) + f ( xn ) If the number of
I = ( x1 - x0 ) + ( x2 - x1 ) + L + ( xn - xn -1 )
2 2 2 segments is doubled,
n -1
f ( x0 ) + 2å f ( xi ) + f ( xn ) the error will be
n -1
hé ù quartered
I = ê f ( x0 ) + 2å f ( xi ) + f ( xn ) ú ® I = (b - a) i =1
2ë 1 2 3 2n
û Width 14444244443
i =1
Average height

School of Mechanical Engineering

Chung-Ang University
Numerical Methods 2010-2 11
 MATLAB Program

School of Mechanical Engineering

Chung-Ang University
Numerical Methods 2010-2 12
 Example 17.2 (1/2)
l Q. Use the two-segment and composite trapezoidal rule to estimate
the integral of the function from a = 0 to b = 0.8. The exact value is
1.640533.
f ( x) = 0.2 + 2.5 x - 200 x 2 + 675 x 3 - 900 x 4 + 400 x 5

Sol.
For n = 2 (h = 0.4)
f (0) = 0.2 f (0.4) = 2.456 f (0.8) = 0.232
0.2 + 2(2.456) + 0.232
I = 0.8 = 1.0688
4
E t = 1.640533 - 1.0688 = 0.57173 ε t = 34.9%
0.8 2
Ea = - 2
(-60) = 0.64
12(2)
School of Mechanical Engineering
Chung-Ang University
Numerical Methods 2010-2 13
 Example 17.2 (2/2)
<Results for the composite trapezoidal rule to estimate the
integral of f(x) = 0.2 + 25x – 200x2 + 675x3 – 900x4
from x = 0 to 0.8. The exact value is 1.640533>
à As the number of segments increases, the error decreases.

n h I et (%)
2 0.4 1.0688 34.9
3 0.2667 1.3695 16.5
4 0.2 1.4848 9.5
5 0.16 1.5399 6.1
6 0.1333 1.5703 4.3
7 0.1143 1.5887 3.2
8 0.1 1.6008 2.4
9 0.0889 1.6091 1.9
10 0.08 1.6150 1.6
School of Mechanical Engineering
Chung-Ang University
Numerical Methods 2010-2 14
 Simpson’s Rules
l One drawback of the trapezoidal rule is that the error is related to the
second derivative of the function.
l More complicated approximation formulas can improve the accuracy for
curves - these include using (a) 2nd and (b) 3rd order polynomials.
l The formulas that result from taking the integrals under these
polynomials are called Simpson’s rules.

School of Mechanical Engineering

Chung-Ang University
Numerical Methods 2010-2 15
 Simpson’s 1/3 Rule
l Simpson’s 1/3 rule corresponds to using second-order
polynomials. Using the Lagrange form for a quadratic fit of three
points:

fn (x) =
(x - x1 ) (x - x2 ) f x + (x - x0 ) (x - x2 ) f x + (x - x0 ) (x - x1 ) f x
( ) ( ) ( )
(x0 - x1 ) (x0 - x2 ) 0 (x1 - x0 ) (x1 - x2 ) 1 (x2 - x0 ) (x2 - x1 ) 2
l Integration over the three points simplifies to:
x2
� I = ò f n ( x ) dx
x0

h f ( x0 ) + 4 f ( x1 ) + f ( x2 )
I= éë f ( x0 ) + 4 f ( x1 ) + f ( x2 ) ùû = I = (b - a )
3 6

Where, h = (b – a)/2, a = x0, b = x2, and x1 = (a + b)/2

School of Mechanical Engineering

Chung-Ang University
Numerical Methods 2010-2 16
 Error of Simpson’s 1/3 Rule
l An estimate for the local truncation error of a single application of
Simpson’s 1/3 rule is:
1 5
Et = - f (4 ) (x )(b - a)
2880
where again x is somewhere between a and b.
l This formula indicates that the error is dependent upon the fourth-
� of the actual function as well as the distance between
derivative
the points.
l Note that the error is dependent on the fifth power of the step size
(rather than the third for the trapezoidal rule).
l Error can thus be reduced by breaking the curve into parts.

School of Mechanical Engineering

Chung-Ang University
Numerical Methods 2010-2 17
 Example 17.3
l Q. Use Simpson 1/3 rule to integrate the following equation
from a = 0 to b = 0.8. Exact solution is 1.640533.
f ( x) = 0.2 + 2.5 x - 200 x 2 + 675 x 3 - 900 x 4 + 400 x 5
sol) for n = 2 (h = 0.4)에 대해서
f (0) = 0.2 f (0.4) = 2.456 f (0.8) = 0.232
0.2 + 4(2.456) + 0.232
I = 0.8 = 1.367467
6

® This shows improved results compared to the trapezoidal rule.

Et = 1.640533 - 1.367467 = 0.2730667 εt = 16.6%

0.85
Ea = - (-2400) = 0.2730667
2880
School of Mechanical Engineering
Chung-Ang University
Numerical Methods 2010-2 18
 Composite Simpson’s 1/3 Rule
l For improved results, Simpson’s
1/3 rule can be used on a set of
subintervals in much the same
way the trapezoidal rule was,
except there must be an odd
number of points..
l Because of the heavy weighting of
the internal points, the formula is a
little more complicated than for the
trapezoidal rule:
xn x2 x4 xn
I= ò x0
fn (x) dx = ò x0
fn (x ) dx + ò x2
fn (x ) dx +L+ ò x n-2
fn (x) dx
h h h
I=
3
[ f (x0 ) + 4 f (x1 ) + f (x2 )]+ [ f (x2 ) + 4 f (x3 ) + f (x4 )]+L+ [ f (xn-2 ) + 4 f (xn-1 ) + f (xn )]
3 3
é ù (b - a )5 ( 4)
n-1 n-2
hê
I = ê f (x0 ) + 4 å f (xi ) + 2 å f (xi ) + f (xn )ú
ú Ea = - 4
f
3 180n
i=1 j=2
êë i, odd j, even úû School of Mechanical Engineering
Chung-Ang University
Numerical Methods 2010-2 19
 Example 17.4
l Q. Use composite Simpson 1/3 rule to integrate the following
equation from a = 0 to b = 0.8. Exact solution is 1.640533.
f ( x) = 0.2 + 2.5 x - 200 x 2 + 675 x 3 - 900 x 4 + 400 x 5
Sol.) for n = 4 (h = 0.2)
f (0) = 0.2 f (0.2) = 1.288
f (0.4) = 2.456 f (0.6) = 3.464
f (0.8) = 0.232
0.2 + 4(1.288 + 3.464) + 2(2.456) + 0.232
I = 0.8 = 1.623467
12
Et = 1.640533 - 1.623467 = 0.017067 εt = 1.04%

0.85
Estimated error: Ea = - 4
(-2400) = 0.017067
180(4)
School of Mechanical Engineering
Chung-Ang University
Numerical Methods 2010-2 20
 Example 17.4
l The composite version of Simpson 1/3 rule is superior to
the trapezoidal rule for most applications.
l It is limited to cases where the values are equispaced.,
even number of segments, and odd number of points.
l Odd segment and even point formula is known as
Simpson 3/8 formula.

School of Mechanical Engineering

Chung-Ang University
Numerical Methods 2010-2 21
 Simpson’s 3/8 Rule
l Simpson’s 3/8 rule corresponds to
using third-order polynomials to fit
four points. Integration over the
four points simplifies to:
x3
I= ò x0
fn (x) dx
3h
I=
8
[ f (x0 ) + 3 f (x1 ) + 3 f (x2 ) + f (x3 )]

l Simpson’s 3/8 rule is generally

� used in concert with Simpson’s 1/3
rule when the number of segments
is odd.

School of Mechanical Engineering

Chung-Ang University
Numerical Methods 2010-2 22
 Higher--Order Formulas
Higher
l Higher-order Newton-Cotes formulas may also be used
- in general, the higher the order of the polynomial used, the
higher the derivative of the function in the error estimate and the
higher the power of the step size.

l As in Simpson’s 1/3 and 3/8 rule, the even-segment-odd-point

formulas have truncation errors that are the same order as
formulas adding one more point. For this reason, the even-
segment-odd-point formulas are usually the methods of
preference.

School of Mechanical Engineering

Chung-Ang University
Numerical Methods 2010-2 23
 Example 17.5 (1/2)
l Q. (a) Use Simpson 3/8 to integrate from a = 0 to b = 0.8.
(b) Use it in conjunction with Simpson 1/3 for five segment
integration.
f ( x) = 0.2 + 2.5 x - 200 x 2 + 675 x 3 - 900 x 4 + 400 x 5
Sol.) (a) For n = 3 (h = 0.2667)
f (0) = 0.2 f (0.2667) = 1.432724
f (0.5333) = 3.487177 f (0.8) = 0.232
0.2 + 3(1.432724 + 3.487177) + 0.232
I = 0.8 = 1.51970
8
(b) For n = 5 (h = 0.16)
f (0) = 0.2 f (0.16) = 1.296919
f (0.32) = 1.743393 f (0.48) = 3.186015
f (0.64) = 3.181929 f (0.80) = 0.232
School of Mechanical Engineering
Chung-Ang University
Numerical Methods 2010-2 24
 Example 17.5 (2/2)
The integral for the first two segments using Simpson 1/3

0.2 + 4(1.296919) + 1.743393

I = 0.32 = 0.3803237
6
For the last three segments, the Simpson 3/8

1.743393 + 3(3.186015 + 3.181929) + 0.232

I = 0.48 = 1.264754
8

The total integral is by summing the two results.

I = 0.3803237 + 1.264754 = 1.645077

School of Mechanical Engineering

Chung-Ang University
Numerical Methods 2010-2 25
 Summary
<Newton-Cotes integration: h = (b – a)/n>

poi
n name 공 식 절단오차
nts
Trapezoidal f ( x 0 ) + f ( x1 )
1 2 (b - a) - (1 / 12 )h 3 f ¢¢(x)
rull 2

f ( x0 ) + 4 f ( x1 ) + f ( x 2 ) - (1 / 90 )h 5 f ( 4)
( x)
2 3 Simpson 1/3 (b - a)
6

f ( x0 ) + 3 f ( x1 ) + 3 f ( x 2 ) + f ( x3 ) - (3 / 80 )h 5 f ( 4)
(x)
3 4 Simpson 3/8 (b - a)
8
7 f ( x0 ) + 32 f ( x1 ) + 12 f ( x2 ) + 32 f ( x3 ) + 7 f ( x4 )
4 5 Boole’s rull (b - a) - (8 / 945)h 7 f (6)
(x)
90
19 f ( x0 ) + 75 f ( x1 ) + 50 f ( x2 ) + 50 f ( x3 ) + 75 f ( x4 ) + 19 f ( x5 )
5 6 (b - a) - (275 / 12,096 )h 7 f ( 6 ) (x)
288

School of Mechanical Engineering

Chung-Ang University
Numerical Methods 2010-2 26
 Integration with Unequal Segments
l Previous formulas were simplified based on
equispaced data points
- though this is not always the case.
l The trapezoidal rule may be used with data containing
unequal segments:

xn x1 x2 xn
I= ò x0
fn (x) dx = ò x0
fn (x ) dx + ò x1
fn (x ) dx +L+ ò x n-1
fn (x) dx
f (x0 ) + f (x1 ) f (x1 ) + f (x2 ) f (xn-1 ) + f (xn )
I = (x1 - x0 ) + (x2 - x1 ) +L+ (xn - xn-1 )
2 2 2

� School of Mechanical Engineering

Chung-Ang University
Numerical Methods 2010-2 27
 Integration Code for Unequal
Segments

School of Mechanical Engineering

Chung-Ang University
Numerical Methods 2010-2 28
 MATLAB Functions (option)
l MATLAB has built-in functions to evaluate integrals
based on the trapezoidal rule
l z = trapz(y)
z = trapz(x, y)
produces the integral of y with respect to x. If x is
omitted, the program assumes h=1.

>> x = [0 .12 .22 .32 .36 .4 .44 .54 .64 .7 .8];

>> y = 0.2 + 25*x - 200*x.^2 + 675*x.^3 - 900*x.^4 + 400*x.^5;
>> trapz(x,y)
ans =
1.5948

School of Mechanical Engineering

Chung-Ang University
Numerical Methods 2010-2 29
 Multiple Integrals
• Multiple integrals can be
determined numerically by first
integrating in one dimension,
then a second, and so on for
all dimensions of the problem.

School of Mechanical Engineering

Chung-Ang University
Numerical Methods 2010-2 30