Ch-8, Math-5 Lecture Note Summer 20-21
Ch-8, Math-5 Lecture Note Summer 20-21
8.1 Introduction
where x r are called the nodes which are distributed within the limits of integration [a, b],
ω r are called the weighting factors and E is the error.
L7 - 1
Summer 20-21
∫ cos ¿ ¿
0.2
Solution: Here there are two unknowns a and b. So we will use f ( x )=1 and then f ( x )=x .[If there
are three unknowns , we will use f ( x )=1 , f ( x )=x∧f ( x ) =x2 ].
∫ f ( x ) dx=∫ 1 dx=a+b
0 0
h=a+b ………………………(1)
∫ f ( x ) dx= 34h f ( x o+ h3 )+ h4 (x 1)
xo
Now
0.8
∫ cos ¿ ¿
0.2
2
Here f ( x )=cos √ 1+ x ; x o=0.2 , x 1=0.8 , h=0.6
0.8
∫ cos ¿ ¿
0.2
3∗0.6 0.6
¿ f ( 0.4 ) + f ( 0.8 )
4 4
¿ 0.25619187
L7 - 2
Summer 20-21
If the nodes x r ’s are uniformly distributed in [a, b ] with x 0=a , x n=b and the
spacing h=(b−a )/n , the method is known as Newton-Cotes integration method and has
the order n. When both the end points of the interval are included as nodes, the methods are
called closed type methods, otherwise they are called open type methods.
where.
h=x1 −x 0 .
To make the simplification short and simple, the axis is translated to make
x0 as the
origin. Thus the formula we are looking for is of the form
h
∫0 f (x )dx≈af (0 )+bf (h )
For two unknown parameters we may assume that the method is exact for f (x )=1 and
x .
Thus
h
f (x )=1 , ∫0 dx=[ x ]h0=h=a+b (1)
2 h
x h2
f (x )=1 ,
h
[]
∫0 x dx= 2 = 2 =bh
0
h
b=
or 2 (2)
From (1), we have
h h
a=h−b=h− =
2 2
The Trapezoidal rule becomes
h h
∫0 f (x )dx≈ 2 [f (0)+f (h )]
To find the precision and an estimate of error, let us take the rule is of the form
h h
∫0 f (x )dx = 2 [ f (0 )+f (h)]+E
2
Taking f (x )=kx , we have+
3 3
h h kh kh 1
E=∫ 0 kx dx− kh 2 = − =− kh 3
2
2 3 2 6
Here E≠0 . Hence the degree of precision is 2−1=1 .
L7 - 3
Summer 20-21
''
Note that f ( x )=2 k
''
Assuming that 2 k=f ( e ) for 0≤e≤h we can write the error term as
h
E=− h2 f (e)
''
12 .
The Trapezoidal rule for arbitrary points
x0 , x1 with step size h can be obtained by
identifying the points 0 by x 0 , h by x 1=x 0 +h . In this case f (0)=f ( x 0 ) and
f (h)=f ( x 1 ) . Thus
x1h
f ( x )dx≈ [ f ( x 0 )+f ( x 1 ) ]
∫x
0 2
h
¿ [ f 0+ f 1]
2
where the notation f ( x r )=f r is used.
L7 - 4
Summer 20-21
1.0
Solution: Here subinterval sizes are unequal. Using the Trapezoidal rule in each subinterval
separately, we have
1.0
∫ f ( x ) dx= 0.5−0.4
2
( 1.133+1.083 )+
0.7−0.5
2
(1.287+ 1.133 )
0.4
+ 1.0−0.7
( 1.649+1.287 )
2
0.1 0.2 0.3
¿ ( 2.216 ) + ( 2.420 ) + ( 2.936 )=0.7932
2 2 2
Example 8.3: The table below shows the values of f (x ) at different values of x:
Solution : Simpson’s rule is applied to two consecutive subintervals of equal length. Thus
for the given data we may divide the subintervls as follows:
1.0 0.6 1.0
∫ f ( x ) dx=∫ f ( x ) dx+∫ f ( x ) dx
0.4 0.4 0.6
0.1
¿ [ f ( 0.4 )+ 4∗f ( 0.5 ) +f ( 0.6 ) ]
3
+ 0.2
¿
3
0.1
¿ [ 1.083+4 (1.133 )+1.197 ] + 0.2 [1.197+4 ( 1.377 ) +1.649]
3 3
¿ 0.2271+0.5569=0.784
8.5 Composite Quadrature Rules
To avoid the use of higher order methods and still obtain accurate results, we use the
composite integration methods. We divide the interval [a, b] into a number of subintervals
and evaluate the integral in each required number of subintervals by a particular method.
We divide the interval [a, b] into N subintervals [ x r−1 , x r ] , each of length h=(b−a )/N
,
x 0=a , x N =b and x r =x 0 +rh , r=1, 2, 3, ¿ ⋅⋅¿ , N ¿ . We write
b x1 x2 xN
∫a f ( x)dx=∫ x f ( x)dx+∫ x f ( x)dx+¿⋅¿+∫ x
0 1 N−1
f ( x)dx
Evaluating each integral on the right hand side by the trapezoidal rule
b h h h
∫a f ( x )dx≈ 2 ( f 0+f 1 )+ 2 (f 1+f 21 )+¿⋅¿+ 2 ( f N−1+f N )
h
¿ [ f 0 +f N + 2(f 1 +f 2 +f 3 +¿⋅¿+ f N−1 )]
2
The error in the formula is
b−a 2 ''
ET =− h f (e )
12 , a≤e≤b
We divide the interval [a, b] into N, an even number of subintervals [ x r , x r+1 ] , each of
Example 8.4 : The values of f (x ) are given for different values of x below.
x 0 0.2 0.4 0.6 0.8 1.0 1.2
f(x) 1.000 1.780 1.954 2.000 1.976 1.909 1.814
1.2
Evaluate I = 0∫ f (x )dx
using extrapolation and (i) trapezoidal rule (ii) Simpson’s rule
Solution : (i) Consider the integral with trapezoidal rule.
0. 2
I T (0. 2)= [ 1+2(1. 780+1 . 954+2 . 000+1. 976+1 . 909)+1 . 814 ]
2
=0 . 1×22 . 078=2 .208
L7 - 6
Summer 20-21
0. 4
I T (0. 4 )= [ 1+2×1 . 954+2×1. 976+1 . 814 ]
2 = 0 .2×10 .674=2 .135
Using Richardson extrapolation
I T ( 0 . 2)−I T (0 . 4 ) 2 .208−2 .135
I R =I T ( 0 . 2)+ 2
=2 .208+ =2 . 232
2 −1 3
L7 - 7
Summer 20-21
2
1
1
h= =0 .25 Ι ( 0)( 4 , 0 . 25)= [ f (0 )+f (1 )+2(f (0 . 25)+f (0 .5 )+f (0 .75 ))] = 0. 6606
n = 4, 4 2
( 1) (0 ) Ι( 0 )(2 , 0. 5 )−Ι ( 0) (1 , 1)
Ι (2 , 0 .5 )=Ι (2 , 0. 5 )+
First order extrapolated values are 22 −1
0 .6955−0 .6244
=0 . 6955+ =0 . 6525
3
( 1) (0 ) Ι (0 )(4 , 0 .25 )−Ι ( 0) (2 , 0 . 5 )
Ι ( 4 , 0 .25 )=Ι (4 , 0 .25 )+
And 22−1
0 .6606−0 . 6955
=0 . 6606+ =0 . 6490
3
( 1) (1 )
( 2) (1) Ι (4 , 0 . 25)−Ι (2 , 0 .5 )
Ι ( 4 , 0 .25 )=Ι (4 , 0. 25 )+
Second extrapolated value is 24 −1
0 .6490−0 .6525
=0 . 6490+ =0 .6488
15
Results are summarized below in a Table:
N H O(h 2) O(h 4 ) O(h 6)
1 1 0.8244
2 0.5 0.6955 0.6528
4 0.25 0.6606 0.6490 0.6488
8.7 Numerical Evaluation of Double Integrals
Multiple integrals are evaluated by expressing them as iterated integrals. For example a double
integral can be expressed as follows;
❑ b f 2 ( x)
R
(
∬ f ( x , y ) dA=∫ ∫ f ( x , y ) dy
a f 1 ( x)
) dx
or
❑ d g2 ( y )
∬ f ( x , y ) dA= ∫( ∫ )
f ( x , y ) dx dy
R c g1 ( y )
In computing the iterated integral of the first form, we hold x constant while integrating with respect
to y and then integrate with respect to x(vice versa for second form).
Numerical evaluation of integrals we need to to choose interval length and nodal points where the
functional values to be calculated. For fixed limits, selection of interval length and nodal points are
straight forward but for variable limits interval length and nodal points should be calculated for a
fixed value of other variable. For example in firtst form the interval at x=x r is
f 2 ( x r ) −f 1 ( x r )
k r=
n
Example 8.6
L7 - 8
Summer 20-21
1.0 1.6
2 y2
Using Simpson’s rule with 2-subintervals evaluate the double integral ∫ ∫ dydx
0.6 1.0 x+ y
The integral is over the rectangular region bounded by0.6 ≤ x ≤ 1.0∧1.0 ≤ y ≤ 1.6 .
Using 2-subintervals in each direction, we have
1.0−0.6 1.6−1.0
h= =0.2∧k = =0.3
2 2
2 y2
The integrand is f ( x , y )= .
x+ y
Integration using fixed values of x=0.6 , 0.8∧1.0 are as follows:
1.6
0.3
I ( 0.6 )=∫ f ( 0.6 , y ) dy = [ f ( 0.6,1 .0 ) + 4 f ( 0.6,1 .3 ) +f ( 0.6,1 .6 ) ]
1.0 3
0.3
¿ [ 1.25+ 4 ( 1.779 ) +2.327 ]=1.0693
3
1.6
0.3
I ( 0.8 )=∫ f ( 0.8 , y ) dy = [f ( 0.8,1 .0 ) +4 f ( 0.8,1.3 )+ f ( 0.8,1.6 ) ]
1.0 3
0.3
¿ [ 1.111+ 4 ( 1.610 ) +2.133 ] =0.9684
3
1.6
0.3
I ( 1.0 ) =∫ f ( 1.0 , y ) dy= [f ( 1.0,1.0 ) + 4 f ( 1.0,1 .3 ) +f ( 1.0,1 .6 ) ]
1.0 3
0.3
¿ [ 1.0+ 4 ( 1.47 ) +1.969 ]=0.8849
3
Finally combining the integral for xwe have
1.0 1.6
0.2
I( Simp)=∫ ∫ f ( x , y ) dydx ≈ [I ( 0.6 ) + 4 I ( 0.8 ) + I (1.0 ) ]
0.6 1.0 3
0.2
¿ [ 1.0693+4 ( 0.9684 ) +0.8849 ] =0.38852
3
[MATLAB result is I = 0.38846]
L7 - 9
Summer 20-21
Example 8.7 Using Simpson’s rule with 2-subinervals evaluate the double integral
2
1.4 1+ x
∫ ∫ ( 1+ xy ) dydx
1.0 x
Solution: The region of integration is1.0 ≤ x ≤1.4 , x ≤ y ≤ 1+ x 2
The integrand is f ( x , y )=1+ xy
1.4−1.0
With 2 equal subintervals for x, the interval length is h= =0.2. The end points of
2
subintervals are 1 ,1.2∧1.4 .
The interval length for y is variable depending on the values of x.
(1+ 12)−1
For x=1 , k 1= =0.5∧ y r =1, 1.5 , 2.0
2
The integral over y is
0.5
I ( 1 , 0.5 )= [ f ( 1,1 )+ 4 f (1,1.5 )+ f ( 1,2.0 ) ]
3
0.5
¿
3
[ 2+4 ( 2.5 ) +3 ]=2.5
(1+ 1.22)−1.2
For x=1.2 , k 2= =0.62∧ y r=1.2 , 1.82 ,2.44
2
0.62
The integral over y is I ( 1.2 , 0.62 )= [ f ( 1.2,1 .2 )+ 4 f (1.2,1 .82 ) +f ( 1.2 , 2.44 ) ]
3
0.5
¿
3
[ 2.44+ 4 ( 3.184 )+ 3.928 ]=3.9482
(1+1.4 2)−1.4
For x=1.4 , k 3= =0.78∧ y r=1.4 ,2.18 , 2.96
2
0.78
The integral over y is I ( 1.4 ,0.78 )= [ f ( 1.4,1 .4 ) + 4 f ( 1.4 , 2.18 ) + f (1.4 ,2.96 ) ]
3
0.78
¿ [ 2.96+ 4 ( 4.052 ) +5.144 ] =6.3211
3
∫ ∫ ( 1+ xy ) dydx ≈ 0.2
3
[ 2.5+4 (3.9482 ) +6.3211 ]=1.64092
1.0 x
[I ( exact )=1.64061]
7
1.2 2.44 3.928
1.4 1.4 2.96
0.78 1.4 2.18 4.052 6.32112
1.4 2.96 5.144
Note: Define the function using “anonymous handle”. Format is shown below:
fname=@(argument) <space> formula
operation / (division) and ^ (for power) must preceded with . (dot).
sin (xy )
For example f ( x , y )= may be typed as
x2 + y2
ff=@(x,y) sin(x*y)./(x.^2+y.^2))
This will be treated as ff(x,y).
1
x 2
Example 8.6: Evaluate ∫ e sin x dx numerically to 7 digits using MATLAB command.
0
fun = @(x)exp(x).*sin(x.^2)
int7 =0.662701
2 √x
1
Example 8.8: Evaluate ∫ ∫ dy dx numerically using MATLAB command.
0 −x √ x + y2
2
fun2 = @(x,y)1./sqrt(x.^2+y.^2)
L7 - 11
Summer 20-21
ymin = @(x)-x
ymax = @(x)sqrt(x)
int22 =3.811758
dV
Example 8.9: Evaluate ∭ , where R is the region bounded by the coordinate planes
R ( 1+ x+ y+ z )2
and the plane x + y + z=1, numerically using MATLAB command.
>> clear
>> fun3=@(x,y,z) 1./(x+y+z+1).^2 % define the integrand as a @-function
fun3 = @(x,y,z)1./(x+y+z+1).^2
ymax = @(x)1-x
>>zmax=@(x,y) 1-x-y
zmax = @(x,y)1-x-y
int31 =0.05685282
Exercise 8
L7 - 12
Summer 20-21
1. The table shows the power P supplied to the driving wheels of a car as a function of the
speed v. If the mass of the car is m=2000 kg, determine the time ∆ t it takes for the car
to accelerate from 1m/s to 6 m/s. Use the trapezoidal rule for integration.
Hint:
6s
∆ t=m∫ (v / p)dv
1s
which can be derived from Newton’s law F=m(dv /dt ) and the definition of power
P=Fv.
5. The car gives the velocity v of a moving particle at time t seconds. Find the distance
covered by the particle in 8 seconds.
t 0 2 4 6 8 9
v 4 6 16 34 60 75
L7 - 13
Summer 20-21
2h
(b) ∫ f ( x ) dx ≈ af
0
( h3 )+bf (h)
In each case find its degree of precision.
Use the above quadrature rule to estimate the following integrals to 3 decimal places.
0.8 1.6
2
2 2 1−x
(i) ∫ cos ( √ 1+ x )dx, (ii) ∫ x e dx.
0.2 1
9. Using Simpson’s rule with two subintervals evaluate the following double integrals and
also write MATLAB code to evaluate.
1.5 1 2.0 1.5
2
(a) ∫ ∫ ( x + 2 √ y ) dydx , (b) ∫ ∫ ln (2 x + y ) dxdy ,
1 0 1.4 1.0
2 1 0.8 2
(c) ∫ ∫ ( 1+8 x ) dydx , (d) ∫ ∫ √ x+ y dydx .
1 0 0.2 0
0.5 1 1 2
sin(xy )
(e) ∫ ∫ dydx (f) ∫ ∫ 2 exp( y /x )dydx
0 0 1+ xy 0 1
2 3 1
L7 - 14
Summer 20-21
L7 - 15