Numerical Integration

BY
TAUSIF SHAHANSHAH
BCE – 3rd YEAR

DEPT. OF CIVIL ENGG.

JADAVPUR UNIVERSITY
1
 WHAT IS NUMERICAL INTEGRATION?
• Approximate computation of a definite integral using numerical
techniques
• Tabulated at regularly spaced intervals
• Gives the approximate calculation
• Different from analytical integration

2
 NUMERICAL INTEGRATION

TRAPEZOIDAL SIMPSON’S RULE GAUSSIAN

RULE QUADRATURE

3
 TRAPEZOIDAL RULE
• Fundamental method of Numerical Integration
• Trapezoids are used for finding the area under curve
• Trapezoids better fits the curve, less error

4
 𝑏 𝑏−𝑎
• Evaluating, I = 𝑎
𝑓 𝑥 𝑑𝑥 by Trapezoidal Rule ∆x =
𝑛
• Consider an expanded view of a general region
• Area of each trapezoid be,
𝑥𝑖 𝑓𝑖−1 + 𝑓𝑖
𝑥𝑖−1
𝑓 𝑥 𝑑𝑥 ≈ (∆𝑥)
2

And,
𝑥𝑖+1 𝑓𝑖 + 𝑓𝑖+1
𝑥𝑖
𝑓 𝑥 𝑑𝑥 ≈ (∆𝑥)
2

• The integral over two panel is given by,

𝑥𝑖+1
𝑓𝑖 + 𝑓𝑖+1 𝑓𝑖−1 + 𝑓𝑖 ∆𝑥
𝑓 𝑥 𝑑𝑥 ≈ ∆𝑥 + ∆𝑥 = (𝑓 + 2𝑓𝑖 + 𝑓𝑖+1 )
𝑥𝑖−1 2 2 2 𝑖−1

• The Trapezoidal Rule approximation to a integral over the entire

Interval is ,
𝑏
∆𝑥
𝑓 𝑥 𝑑𝑥 ≈ { 𝑓0 + 𝑓𝑛 ) + 2(𝑓1 + 𝑓2 + ⋯ + 𝑓𝑛−1 }
𝑎 2
Where 𝑓𝑜 = 𝑓 𝑎 𝑎𝑛𝑑 𝑓𝑛 = 𝑓 𝑏

5
 SIMPSON’S 1/3 RULE
• A further improvement over Trapezoidal rule is Simpson’s Rule
• Based on use of parabolic arcs (quadratic function) to approximate the
curve instead of the straight lines employed in the trapezoid rule
• Connecting 3 points, the number of interval has
to be even

6
• Approximating the area of the panel by the area
under the parabola is given by,
𝑥𝑖+1 ∆𝑥
𝑓
𝑥𝑖−1 2
𝑥 𝑑𝑥 = [𝑓 𝑥𝑖−1 + 4𝑓 𝑥𝑖 + 𝑓(𝑥𝑖+1 )]
3

• The Simpson’s rule approximation to the

integral over the entire interval is
𝑏 ∆𝑥
𝑎
𝑓 𝑥 𝑑𝑥 ≈ 𝑓0 + 4𝑓1 + 2𝑓2 + 4𝑓3 + ⋯ + 2𝑓𝑛−2 + 4𝑓𝑛−1 + 𝑓𝑛
3

Where 𝑓𝑜 = 𝑓 𝑎 𝑎𝑛𝑑 𝑓𝑛 = 𝑓 𝑏

7
 SIMPSON’S 3/8 RULE
• Consists of taking the area under a cubic equation connecting four
points.
• Number of interval must be multiple of three
• The formula for Simpson’s 3/8 rule is
𝑏
3∆𝑥
𝑓 𝑥 𝑑𝑥 = [ 𝑓0 + 𝑓𝑛 + 2 𝑓3 + 𝑓6 + ⋯ + 𝑓𝑛−3
𝑎 8
+3 𝑓1 + 𝑓2 + 𝑓4 + 𝑓5 + ⋯ + 𝑓𝑛−1 ]

8
 GAUSSIAN QUADRATURE
• Sampling points may not be equally spaced in practical problems
• Both positions of the sampling points and the weights have been
optimized
• Gaussian Quadrature formula can be expressed as,
1 𝑛
−1
𝑓 𝑥 𝑑𝑥 = 𝑖=1 𝑤𝑖 𝑓(𝑥𝑖 )
• It produces most accurate approximation

9
One point Gaussian Quadrature

• There is only sampling point 𝑛 = 1

1
• −1
𝑓 𝑥 𝑑𝑥 = 𝑤1 𝑓(𝑥1 )
• For one point Gaussian Quadrature the sampling point 𝑥1 = 0 and
weight 𝑤1 = 2
• For two point Gaussian Quadrature the sampling points 𝑥1 = −1/ 3,
𝑥2 = 1/ 3 and weight 𝑤1 = 𝑤2 = 1

10
 PROBLEMS ON NUMERICAL
INTEGRATION
• EXAMPLE 1:
The cross section of stream 9.0 m wide is measured by means of soundings taken 1.5 m apart. The depths
recorded are 0, 0.45 m, 0.6 m, 1.05 m, 0.69 m, 0.3 m, 0 m. The mean velocity is observed to be 1.02 m per
second. Compute the discharge of the stream.
SOLUTION : Assuming the points at the depths may be connected by
straight lines the cross sectional area may be thought of to be divided
into trapezoids. Hence applying ‘Trapezoidal Rule’
𝑑
𝐶𝑟𝑜𝑠𝑠 𝑠𝑒𝑐𝑡𝑖𝑜𝑛𝑎𝑙 𝐴𝑟𝑒𝑎 = { 𝑦0 + 𝑦6 + 2 𝑦1 + 𝑦2 + 𝑦3 + 𝑦4 + 𝑦5 }
2
1.5
Area = { 0 + 0 + 2 0.45 + 0.60 + 1.05 + 0.69 + 0.30 } = 4.635 m2
2
Rate of flow through stream = 1.02 × 4.635
i.e; 4.7277 m3/s

11
• EXAMPLE 2:
A series of offsets were taken at 10 m intervals in the following order from a chain line to a curved boundary: 0, 7.2 m, 5.4 m,
6.0 m, 6.8 m, 7.4 m, 8.2 m and 0 m. Find the area enclosed between the chain line, the curved boundary and the end offsets.

SOLUTION :Assuming the points defined by the offsets over the

intervals 0-20 m, 20-40 m, 40-60 m to be connected by a parabolic curve,
the cross sectional area may be computed from ‘Simpson’s Rule’ as

𝑑 1
= { 𝑦0 + 4𝑦1 + 𝑦2 + 𝑦2 + 4𝑦3 + 𝑦4 + 𝑦4 + 4𝑦5 + 𝑦6 + 𝑦7 + 4𝑦6 + 𝑦7 }
3 2

𝑑
= { 𝑦0 + 𝑦6 + 2 𝑦2 + 𝑦4 + 4 𝑦1 + 𝑦3 + 𝑦5 + 𝑦7 + 2𝑦6 }
3

10
= { 0 + 8.2 + 2 5.4 + 6.8 + 4 7.2 + 6.0 + 7.4 + 0 + 2 ∗ 8.2}
3
= 458 m2

12
 CONCLUSION
• It is an approximation and will not yield an exact answer
• Used for area and volume calculations
• Land filling and excavation works
• Is an iterative method
• Method with small error considered superior

13
 REFERENCES
• Numerical Methods in Science and Engineering –by S. RAJASEKARAN
• Plane Surveying –by Dr. ALAK DE
• http://www.ce.utexas.edu/prof/mckinney/ce311k/handouts/Integration.pdf
• http://www.csun.edu/~ajp42955/382%20paper%20pdf.pdf
• https://www.slideshare.net/ianowen45/numerical-integration?qid=67fb6114-c743-4862-827f-
491935799fe1&v=&b=&from_search=3
• http://www.sam.nitk.ac.in/sites/default/Numerical_Methods/Numerical_Integration/numerical_integration.pdf

14
THANK YOU

15