Numerical Integration

Lesson 5
Rationale
• Two main reasons for why there is a need to do numerical integration:
1. Analytical integration may be impossible/infeasible
2. It may be necessary to integrate tabulated data rather than known functions
• Applications for integration:
• Ex. Maxwell’s equations – numerical solutions of Maxwell’s equations can be directly used for
many engineering applications
• Integration is involved in practically every physical theory:
• Vibration, distortion under weight, or one of many types of fluid flow − e.g. heat flow, air flow
(over a wing), or water flow (over a ship’s hull, through a pipe, groundwater flow regarding a
contaminant), etc.;
• All these things can be either directly solved by integration (for simple systems), or some type of
numerical integration (for complex systems).
• Numerical integration is essential for evaluation of integrals of functions available only at
discrete points;
• Such functions often arise in the numerical solution of differential equations or from experimental
data taken at discrete intervals
• Engineers therefore often require numerical integration.
Lesson Objectives
• appreciate the need for numerical integration
• evaluate integrals using the trapezoidal rule
• evaluate integrals using the mid-ordinate rule
• evaluate integrals using Simpson’s rule
• apply numerical integration to practical situations
Contents
1. Introduction
2. The Trapezoidal Rule
3. The Mid-ordinate Rule
4. Simpson’s Rule
1. Introduction
• Even with advanced methods of integration there are many mathematical
functions which cannot be integrated by analytical methods and thus
approximate methods have then to be used.
• Approximate methods of definite integrals may be determined by what is
termed numerical integration.
• It may be shown that determining the value of a definite integral is, in fact,
finding the area between a curve, the horizontal axis and the specified
ordinates.
• Three methods of finding approximate areas under curves:
• The trapezoidal rule
• The mid-ordinate rule
• Simpson’s rule
• These rules are used as a basis for numerical integration.
Contents
1. Introduction
2. The Trapezoidal Rule
3. The Mid-ordinate Rule
4. Simpson’s Rule
2. The Trapezoidal Rule
• Let a required definite integral be
𝑏
denoted by 𝑎 𝑦𝑑𝑥 and be
represented by the area under the
graph of 𝑦 = 𝑓(𝑥) between the limits
𝑥 = 𝑎 and 𝑥 = 𝑏 (see Fig. 48.1)
• Let the range of integration be
divided into 𝑛 equal intervals, each of
width 𝑑, such that 𝑛𝑑 = 𝑏 − 𝑎 or
𝑏−𝑎
𝑑=
𝑛
• The ordinates are labelled 𝑦1 , 𝑦2 ,
𝑦3 ,… 𝑦𝑛+1
2. The Trapezoidal Rule
• An approximation to the area under
the curve may be determined by
joining the tops of the ordinates by
straight lines.
• Each interval is thus a trapezium, the
area of which is given by:
1 sum of parallel sides ×
𝐴=
2 perpendicular distance between them
𝑏
𝑦𝑑𝑥
𝑎
1 1
≈ 𝑦1 + 𝑦2 𝑑 + 𝑦2 + 𝑦3 𝑑
2 2
1 1
+ 𝑦3 + 𝑦4 𝑑 + ⋯ + 𝑦𝑛 + 𝑦𝑛+1 𝑑
2 2
1
≈ 𝑑 𝑦1 + 𝑦2 + 𝑦3 + 𝑦4 + ⋯ + 𝑦𝑛 +
2
The Trapezoidal Rule:

𝑏
width of 1 first + last remaining
𝑦𝑑𝑥 = +
𝑎 interval 2 ordinate ordinates
 3 2
Problem 1. (a) Use integration to evaluate 1 𝑥
𝑑𝑥,
correct to 3 decimal places.
• Solution:
Problem 1. (b) Use the Trapezoidal Rule with four
intervals to evaluate the integral in 1(a), correct to 3
decimal places.
• The range of integration is the difference
between the upper and lower limits, 3 − 1 = 2.
• Using the trapezoidal rule with four intervals
3−1
gives an interval width 𝑑 = = 0.5 and
4
ordinates situated at 1.0, 1.5, 2.0, 2.5, and 3.0.
2
• Corresponding values of are shown in the
𝑥
table, each correct to 4 decimal places (one more
decimal place than required in the problem).
Problem 1. (b) Use the Trapezoidal Rule with four
intervals to evaluate the integral in 1(a), correct to 3
decimal places.
• Solution:
Problem 2. Use the trapezoidal rule with eight intervals
3 2
to evaluate 1 𝑑𝑥 correct to 3 decimal places.
𝑥
3−1
• With eight intervals, width of each is 𝑑 = = 0.25
8
• Ordinates: 1.00, 1.25, 1.50, 1.75, 2.00, 2.25, 2.50, 2.75
and 3.00.
2
• Corresponding values of are shown in the table:
𝑥
Problem 2. Use the trapezoidal rule with eight intervals
3 2
to evaluate 1 𝑑𝑥 correct to 3 decimal places.
𝑥
Problem 2. Use the trapezoidal rule with eight intervals
3 2
to evaluate 1 𝑑𝑥 correct to 3 decimal places.
𝑥
• This problem demonstrates that the greater the number of intervals
chosen (i.e. the smaller the interval width) the more accurate will be
the value of the definite integral.
• The exact value is found when the number of intervals is infinite,
which is the basis for the process of integration.
 𝜋/2 1
Problem 3. Use the trapezoidal rule to evaluate 0
𝑑𝑥
1+sin 𝑥
using six intervals. Give the answer correct to 4 significant
figures.
𝜋
−0 𝜋
• With six intervals, width of each is 𝑑 = 2
=
6 12
rad (15°)
1
• Ordinates occur at corresponding values of
1+sin 𝑥
(see table)
 𝜋/2 1
Problem 3. Use the trapezoidal rule to evaluate 0
𝑑𝑥
1+sin 𝑥
using six intervals. Give the answer correct to 4 significant
figures.
Contents
1. Introduction
2. The Trapezoidal Rule
3. The Mid-ordinate Rule
4. Simpson’s Rule
3. The Mid-ordinate Rule
• Let a required definite integral be
𝑏
denoted by 𝑎 𝑦 𝑑𝑥 and represented
by the area under the graph of 𝑦 =
𝑓(𝑥) between the limits 𝑥 = 𝑎 and
𝑥 = 𝑏 (see Fig. 48.2)
• Each interval of width 𝑑 is assumed to
be replaced by a rectangle of height
equal to the ordinate at the middle
point of each interval, shown as 𝑦1 ,
𝑦2 , 𝑦3 ,..., 𝑦𝑛 (see Fig. 48.2)
3. The Mid-ordinate Rule
• Thus
𝑏
𝑦 𝑑𝑥
𝑎
≈ 𝑑𝑦1 + 𝑑𝑦2 + 𝑑𝑦3 + ⋯ + 𝑑𝑦𝑛
≈ 𝑑(𝑦1 + 𝑦2 + 𝑦3 + ⋯ + 𝑦𝑛 )

• The Mid-ordinate Rule states:

𝑏
𝑦 𝑑𝑥 ≈ (width of interval)(sum of mid − ordinates)
𝑎
Problem 4. Use the mid-ordinate rule with (a) four intervals, (b) eight
3 2
intervals, to evaluate 1 𝑑𝑥, correct to 3 decimal places.
𝑥

Solution (a):
3−1
• With four intervals, each will have a width of 𝑑 = = 0.5
4
• The ordinates will occur at 1.0, 1.5, 2.0, 2.5 and 3.0
• The mid-ordinates 𝑦1 , 𝑦2 , 𝑦3 , and 𝑦4 occur at 1.25, 1.75, 2.25
and 2.75
2
• Corresponding values of are shown in the table
𝑥

• From equation (2):

3
2
𝑑𝑥 ≈ 0.5 1.7889 + 1.5119 + 1.3333 + 1.2060 = 𝟐. 𝟗𝟐𝟎
1 𝑥
Correct to 3
decimal places
Problem 4. Use the mid-ordinate rule with (a) four intervals, (b) eight
3 2
intervals, to evaluate 1 𝑑𝑥, correct to 3 decimal places.
𝑥

Solution (b):
3−1
• With eight intervals, each will have a width of 𝑑 = = 0.25
8
• The ordinates will occur at 1.00, 1.25, 1.50, 1.75,…
• The mid-ordinates 𝑦1 , 𝑦2 , 𝑦3 , 𝑦4 ,… occur at 1.125, 1.375,
1.625 and 1.875…
2
• Corresponding values of are shown in the table
𝑥
• From equation (2):
3
2
𝑑𝑥
1 𝑥
≈ 0.25 1.8856 + 1.7056 + 1.5689 + 1.4606 + 1.3720 + 1.2978 + 1.2344
Correct to 3 The greater the number of intervals the
decimal places nearer the result is to the true value.
 −𝑥2
2.4
Problem 5. Evaluate 0
𝑒
𝑑𝑥, correct to 4 significant
3

figures, using the mid-ordinate rule with six intervals.

2.4−0
• With six intervals, each will have a width of 𝑑 = = 0.40
6
• The ordinates will occur at 0, 0.40, 0.80, 1.20, 1.60, 2.00, and
2.40
• The mid-ordinates are at 0.20, 0.60, 1.00, 1.40, 1.80, and 2.20
−𝑥2
• Corresponding values of 𝑒 3 are shown in the table

• From equation (2):

2.4 −𝑥 2
𝑒 3 𝑑𝑥 ≈ 0.40 0.98676 + 0.88692 + 0.71653 + 0.52031 + 0.33960 + 0.19922
0
= 𝟏. 𝟒𝟔𝟎 Correct to 3
decimal places
Contents
1. Introduction
2. The Trapezoidal Rule
3. The Mid-ordinate Rule
4. Simpson’s Rule
4. Simpson’s Rule
• The approximation made with the trapezoidal rule is to join the top of
two successive ordinates by a straight line, i.e. by using a linear
approximation of the form 𝑎 + 𝑏𝑥.

• With Simpson’s Rule, the approximation made is to join the tops of

three successive ordinates by a parabola, i.e. by using a quadratic
approximation of the form 𝑎 + 𝑏𝑥 + 𝑐𝑥 2 .
4. Simpson’s Rule
• Fig. 48.3 shows a parabola 𝑦 = 𝑎 + 𝑏𝑥 + 𝑐𝑥 2 with ordinates 𝑦1 , 𝑦2 and 𝑦3
at 𝑥 = −𝑑, 𝑥 = 0, and 𝑥 = 𝑑, respectively.
• Thus the width of each of the two intervals is 𝑑.
• The area enclosed by the parabola, the 𝑥-axis and ordinates 𝑥 = −𝑑 and
𝑥 = 𝑑 is given by:
𝑑 2 3 𝑑
2
𝑏𝑥 𝑐𝑥
𝑎 + 𝑏𝑥 + 𝑐𝑥 𝑑𝑥 = 𝑎𝑥 + +
−𝑑 2 3 −𝑑
𝑏𝑑 2 𝑐𝑑 3 𝑏𝑑 2 𝑐𝑑 3
= 𝑎𝑑 + + − −𝑎𝑑 + −
2 2 2 2
2 3 1
= 2𝑎𝑑 + 𝑐𝑑 or 𝑑(6𝑎 + 2𝑐𝑑 2 )
3 3
4. Simpson’s Rule
• Since 𝑦 = 𝑎 + 𝑏𝑥 + 𝑐𝑥 2
• at 𝑥 = −𝑑, 𝑦1 = 𝑎 − 𝑏𝑑 + 𝑐𝑑2
• at 𝑥 = 0, 𝑦2 = 𝑎
• at 𝑥 = 𝑑, 𝑦3 = 𝑎 + 𝑏𝑑 + 𝑐𝑑2
• Hence: 𝑦1 + 𝑦3 = 2𝑎 + 2𝑐𝑑2
• and: 𝑦1 + 4𝑦2 + 𝑦3 = 6𝑎 + 2𝑐𝑑2
• Thus the area under the parabola between 𝑥 = −𝑑 and 𝑥 = 𝑑 in Fig.
1
48.3 may be expressed as 𝑑(𝑦1 + 4𝑦2 + 𝑦3 ), from equations (3)
3
and (4)
• The result is seen to be independent of the position of the origin.
 • Let
𝑏
a definite integral be denoted by
𝑦 𝑑𝑥 and represented by the area
4. Simpson’s Rule 𝑎
under the graph of 𝑦 = 𝑓(𝑥) between the
limits 𝑥 = 𝑎 and 𝑥 = 𝑏 (Fig. 48.4)
• The range of integration, 𝑏 − 𝑎, is divided
into an even number of intervals, say 2𝑛,
each of width 𝑑.
• Since an even number of intervals is
specified, an odd number of ordinates,
2𝑛 + 1, exists.
• Let an approximation to the curve over
the first two intervals be2
a parabola of the
form 𝑦 = 𝑎 + 𝑏𝑥 + 𝑐𝑥 , which passes
through the tops of the three ordinates
𝑦1 , 𝑦2 and 𝑦3 .
• Similarly, let an approximation to the
curve over the next two intervals be the
parabola which passes through the tops of
the ordinates 𝑦3 , 𝑦4 and 𝑦5 , and so on.
4. Simpson’s Rule
• Then…
𝑏
𝑦 𝑑𝑥
𝑎
1 1
≈ 𝑑 𝑦1 + 4𝑦2 + 𝑦3 + 𝑑 𝑦3 + 4𝑦4 + 𝑦5
3 3
1
+ 𝑑 𝑦2𝑛−1 + 4𝑦2𝑛 + 𝑦2𝑛+1
3
1
≈ 𝑑 𝑦1 + 𝑦2𝑛+1 + 4 𝑦2 + 4𝑦4 + ⋯ + 𝑦2𝑚
3
+ 2 𝑦3 + 𝑦5 + ⋯ + 𝑦2𝑛−1
4. Simpson’s Rule
• Simpson’s Rule states:
𝑏
𝑦 𝑑𝑥
𝑎
1 width of first + last sum of even sum of odd
≈ +4 +2
3 interval ordinates ordinates ordinates

• Note: Simpson’s rule can only be applied when an even number of

intervals is chosen, i.e. an odd number of ordinates.
Problem 6. Use Simpson’s rule with (a) four intervals, (b) eight
3 2
intervals, to evaluate 1 𝑑𝑥, correct to 3 decimal places.
𝑥

Solution (a):
3−1
• With four intervals, each will have a width of 𝑑 = = 0.5 first
4
even
• The ordinates will occur at 1.0, 1.5, 2.0, 2.5, and 3.0
odd
2
• Corresponding values of are shown in the table even
𝑥
• From equation (5): last
3
2 1
𝑑𝑥 ≈ 0.5 2.0000 + 1.1547 + 4 1.6330 + 1.2649 + 2 1.4142
1 𝑥 3
1 Correct to 3
= 0.5 3.1547 + 11.5916 + 2.8284 = 𝟐. 𝟗𝟐𝟗 decimal places
3
Problem 6. Use Simpson’s rule with (a) four intervals, (b) eight
3 2
intervals, to evaluate 1 𝑑𝑥, correct to 3 decimal places.
𝑥

Solution (b):
3−1
• With eight intervals, each will have a width of 𝑑 = = 0.25
8
• The ordinates will occur at 1.00, 1.25, 1.50, 1.75,…, and 3.0
2
• Corresponding values of are shown in the table
𝑥
• From equation (5):
3
2
𝑑𝑥
1 𝑥
1
≈ 0.25 2.0000 + 1.1547
3
+ 4 1.7889 + 1.5119 + 1.3333 + 1.2060

Correct to 3 decimal places

Problem 6 Notes:
• The latter answer is exactly the same as that obtained by integration.

• In general, Simpson’s rule is regarded as the most accurate of the

three approximate methods used in numerical integration.
 𝜋/3 1
Problem 7. Evaluate 0
1− sin2 𝜃 𝑑𝜃, correct to 3
3
decimal places, using Simpson’s rule with six intervals.
𝜋
−0 𝜋
• With six intervals, each will have a width of 𝑑 = 3
= = 10°
6 18
𝜋 𝜋 𝜋 2𝜋 5𝜋 𝜋
• The ordinates will occur at 0, , , , , , and
18 9 6 9 18 3
1
• Corresponding values of 1 − sin2 𝜃 are shown in the table
3
 𝜋/3 1
Problem 7. Evaluate 0
1− sin2 𝜃 𝑑𝜃, correct to 3
3
decimal places, using Simpson’s rule with six intervals.
• From equation (5):
𝜋/3
1
1 − sin2 𝜃 𝑑𝜃
0 3
1 𝜋
≈ 1.0000 + 0.8660 + 4 0.9950 + 0.9574 + 0.8969 + 2 0.9803 + 0.9286
3 18
1 𝜋
= 1.8660 + 11.3972 + 3.8178 = 𝟎. 𝟗𝟗𝟒 Correct to 3 decimal places
3 18

first even odd even odd even last

Problem 8. An alternating current 𝑖 has the
first
following values at equal intervals of 2.0 even
milliseconds (see table at the right). Charge 𝑞 in odd
12.0
millicoulombs is given by 𝑞 = 0 𝑖 𝑑𝑡. Use even
odd
Simpson’s rule to determine the approximate even
charge in the 12 millisecond period. last
• From equation (5):
12.0
1
𝑞= 𝑖 𝑑𝑡 ≈ 2.0 0 + 0 + 4 3.5 + 10 + 2.0 + 2 8.2 + 7.3
0 3
𝑞 = 𝟔𝟐 𝒎𝑪
end