Computational Techniques in Civil Engineering

Tutorials (Water Resources part)

FDM
𝑛
1. Values of discharge (Q) in m3/s at three points in the space-time grid are given as 𝑄𝑖−1 = 35,
𝜕𝑄
𝑄𝑖𝑛 = 35.1 and 𝑄𝑖+1
𝑛
= 35.2. Taking ∆𝑥 = 500𝑚 and ∆𝑡 = 5 minute, compute using following
𝜕𝑥
𝜕2𝑄
schemes: (a) backward difference (b) forward difference, (c) central difference. Also compute 𝜕𝑥 2 .

t
n+1

X
n
i-1 i i+1

2. Values of discharge (Q) in m3/s at three points in the space-time grid are given as 𝑄𝑖𝑛−1 = 45,
𝜕𝑄
𝑄𝑖𝑛 = 46 and 𝑄𝑖𝑛+1 = 47. Taking ∆𝑥 =1 km and ∆𝑡 = 30 minute, compute using following schemes:
𝜕𝑡
(a) backward difference (b) forward difference, (c) central difference

n+1

t
n

X
n-1
i-1 i i+1

3. The values of flow rate at four points in the space-time gird are given below:
𝑄𝑖𝑛 =15 m3/s, 𝑄𝑖𝑛+1 =15.8 m3/s,𝑄𝑖+1
𝑛 𝑛+1
=13.9 m3/s and 𝑄𝑖+1 =14.7 m3/. Taking ∆𝑥 = 500m, ∆𝑡 = 1 hr and 𝜃=
𝜕𝑄 𝜕𝑄
0.55, calculate the values of 𝜕𝑥
and 𝜕𝑡
using implicit four-point method.

t
n+1

X
n
i i+1
4. Consider a rectangular channel, 30m wide, with bed slope of 0.015 and Manning’s n = 0.035. The
following flow rates are given: 𝑄𝑖𝑛+1 =30 m3/s, 𝑄𝑖𝑛 =22 m3/s and 𝑄𝑖+1
𝑛
=20 m3/s. Taking ∆𝑥 = 1500m and ∆𝑡
𝑛+1
= 10 min, determine 𝑄𝑖+1 using finite difference scheme for a linear kinematic wave model. Assume
lateral flow to be zero. The equation for linear kinematic wave model with no lateral flow is
𝑄𝑛 𝑛 +1 𝛽 −1
∆𝑡 𝑛 +1
𝑄𝑖 𝑛
+𝛼𝛽 𝑄𝑖+1 𝑖+1 +𝑄 𝑖
∆𝑥 2
𝑛+1
𝑄𝑖+1 = 𝛽 −1 . Take wetted perimeter is approximately equal to width of channel.
∆𝑡 𝑄 𝑛 +𝑄 𝑛 +1
+𝛼𝛽 𝑖+1 𝑖
∆𝑥 2

5. Solve exercise 4 by using non-linear kinematic wave model using the equation
∆𝑡 𝑛+1 𝑛+1 𝛽 ∆𝑡 𝑛+1 𝑛 𝛽
𝑄𝑖+1 + 𝛼 𝑄𝑖+1 = 𝑄 + 𝛼 𝑄𝑖+1
∆𝑥 ∆𝑥 𝑖
𝑛+1
Take value of𝑄𝑖+1 computed in exercise 4 as initial value and solve by Newton-Raphson iteration.

6. Consider a rectangular channel, 90m wide and 5km long with bed slope of 0.015 and Manning’s n =
0.02. The inflow hydrograph for the channel is given below:
Time (min) 0 5 10 15 20
3
Flow (m /s) 14 19 28 32 40
3
The initial condition is a uniform flow of 14 m /s and there is no lateral flow.
Use the linear kinematic wave model to route the inflow hydrograph through the channel taking ∆𝑥 =
1000m and∆𝑡 = 5 min . The equation for routing is
𝑄𝑛 𝑛 +1 𝛽 −1
∆𝑡 𝑛 +1
𝑄𝑖 𝑛
+𝛼𝛽 𝑄𝑖+1 𝑖+1 +𝑄 𝑖
∆𝑥 2
𝑛+1
𝑄𝑖+1 = 𝛽 −1 . Take wetted perimeter is approximately equal to width of channel.
∆𝑡 𝑄 𝑛 +𝑄 𝑛 +1
+𝛼𝛽 𝑖+1 𝑖
∆𝑥 2

7. A flood of 150 m3/s peak discharge passed a gauging station at 12:00 noon on a river. There is a
community adjacent to the river 7.2 km downstream. What will be the value of peak discharge at that
community at 12:00 noon if the velocity of flow is 1.2 m/s and peak discharge at that community at 9:00
A.M. is 100 m3/s. Assume width of river as inside and use first order accurate numerical scheme of
kinematic wave equation. Take ∆𝑥 = 7.2km and ∆𝑡 = 1 hr.

8. A rectangular channel is 60m wide, has bed slope 1% and Manning’s n 0.04. The discharge through a
section of the channel is 130 m3/s.∆𝑥 is taken 1km for kinematic wave routing. ApplyingCourant
condition for stability of kinematic wave solution, recommend the maximum value of time step in this
condition. Assume hydraulic radius is equal to flow depth.
MOC

1. Following data are given at two points A and B along a pipe of diameter 30cm carrying water: VA
=5m/s, VB = 5.1m/s, PA = 100Kpa, PB= 120Kpa, ∆𝑥 = 100m, ∆𝑡 = 10 Sec, f = 0.02, c = 1200m/s, elevation
difference between A and B = 1m. Using the finite difference form of characteristics equations, compute
velocity and pressure at point P.
t

P
𝑡 + ∆𝑡
+
C C
-

t A B

∆𝑥 ∆𝑥

2. Following data are given at two points A and B along a pipe of diameter 20cm carrying water: QA
=0.3m3/s, QB = 0.33m3/s, HA = 20.5m, HB= 21m, ∆𝑥 = 100m, ∆𝑡 = 10 Sec, f = 0.02, c = 1200m/s, elevation
difference between A and B = 1m. Using the finite difference form of characteristics equations,
compute discharge and head at point P.
t

P
𝑡 + ∆𝑡
+
C -
C

t A B

∆𝑥 ∆𝑥

3. Following data are given at two points A and B along a pipe of diameter 40cm carrying water: QA
=0.5m3/s, QB = 0.55m3/s, HA = 25m, HB= 25.5m, ∆𝑥 = 100m, ∆𝑡 = 10 Sec, f = 0.02, c = 1200m/s, elevation
difference between A and B = 1m. Using the finite difference form of characteristics equations,
compute discharge and head at point P.
t

P
𝑡 + ∆𝑡
+
C -
C

t A L R B

∆𝑥 ∆𝑥

X
4. A pipe conveys water from a reservoir as shown in the figure.

C- C+

100m
L= 600m, D = 400mm

B
A

Take f = 0.02, c = 1200m/s. The hydraulic grade line (HGL) at the reservoir is given as 𝐻𝑃𝐴 = 100 +
2sin⁡(𝜋𝑡). The discharge at the downstream end is zero at all times. By using only one reach, compute
discharge from A and elevation of hydraulic grade line at B at 3Sec using discretized equation of the
Method of characteristics in the form of HGL and discharge .

Groundwater simulation

1. The final value of potential head at four gridsare given: 𝜙𝑖+1,𝑗 = 1.1, 𝜙𝑖,𝑗 −1 = 1.09, 𝜙𝑖−1,𝑗 =
1.15, 𝜙𝑖,𝑗 +1 = 1.07. The aquifer is homogeneous and isotropic. Compute the potential head at grid (i,j)
considering steady flow and no withdrawal from the grid.

j-1

j
Y

j+1

i-1 i i+1

2. Figure below shows a part of grids used for simulating seepage under a dam. The inital value of
potential head at four grids are given: 𝜙𝑖+1,𝑗 = 1, 𝜙𝑖,𝑗 −1 = 1.5, 𝜙𝑖−1,𝑗 = 1.6, 𝜙𝑖,𝑗 +1 = 1.7. The aquifer is
homogeneous and isotropic. Compute the potential head at grid (i,j) considering steady flow and no
withdrawal from the grid.
 j-1

j
Y

j+1
i-1 i i+1

3. A schematic for simulating river stage-water table fluctuation is shown in figure below.

River Barrier

Aquifer
hL 1 2 3 4 5 hR

The following data are given for the simulation: homogeneous and isotropic aquifer, river stage (hL) =
203m, Length of aquifer = 500m, ∆𝑥 = 100m, ∆𝑡 = 1 day, transmissivity of aquifer = 0.03m/s, storage
coefficient = 0.01. The initial value of water table at 5 grids given as 99.1, 99.11, 99.2, 99.3, 99.4m
respectively. Compute coefficients of the 1D implicit finite difference model and display the matrix of
known and unknowns.

Theoretical and derivation type questions

Chapter 5
1. Explain the concept of Finite Difference method (FDM) with illustrations. What are the steps involved
in the implementation of FDM?
2. Explain the following terms.
a. Consistency
b. Convergence
c. Stability
d. Order of accuracy
e. Diffusion and dispersion of solution
3. Differentiate between explicit and implicit schemes with examples.
4. Write down the Saint Venant equations with the meaning of notations used. Develop Diffusive and
kinematic equations from full Saint Venant equations.
5. Develop the first order accurate implicit Finite Difference equation for kinematic wave model in the
linear form. Write down the algorithm for channel routing using the linear scheme.
6. Develop the second order accurateexplicit Finite Difference equation for kinematic wave model in the
linear form.
7. Develop the first order accurate implicit Finite Difference equation for kinematic wave model in the
non-linear form. Write down the algorithm for channel routing using the non-linear scheme.
8. Develop the second order accurate explicit Finite Difference equation for kinematic wave model in
the non-linear form.
9. Developthe Finite Difference equations for the Saint Venant equations using explicit central difference
scheme.
10. Develop the Finite Difference equations for the Saint Venant equations using implicit four point
method and explain briefly the solution procedure of the resulting algebraic equations.

Chapter 6
1. Define characteristic curve with figures.
2. Convert the partial differential form of the unsteady pipe flow equations into the characteristicform.
3. Develop the Finite Difference solution of the characteristicform of unsteady pipe flow equations to
obtain solution in terms of velocity and pressure.
4. Convert the Finite Difference solution obtained in question no. 3 in terms of discharge and head.
5. Prepare an algorithm to compute discharge and head based on the following form of Finite difference
equations for unsteady pipe flow problem using rectangular grid.
𝐻𝑃𝑖 = 𝐻𝑖−1 − 𝐵 𝑄𝑃𝑖 − 𝑄𝑖−1 − 𝑅𝑄𝑖−1 𝑄𝑖−1
𝐻𝑃𝑖 = 𝐻𝑖+1 + 𝐵 𝑄𝑃𝑖 − 𝑄𝑖+1 + 𝑅𝑄𝑖+1 𝑄𝑖+1
Where H = head, Q = discharge, HpiandQpi= head and discharge at point of intersection of two
characteristics, B and R = coefficients. The characteristics pass through the diagonal of the grids.
6. How do you obtain the values of Q and H on time line n if the characteristics do not pass through
diagonals? Explain with sketch and equations.
7. Why is boundary condition required in the solution of unsteady pipe flow problem using method of
characteristics (MOC) ? Explain different types of boundary conditions.

Chapter 7
1. Derivethe Finite Difference equation for 2D groundwater flow simulation.
2. The continuity equation in Finite Difference form for 2D groundwater flow simulation is given by
𝜕𝜙
𝐴𝜙𝑖+1,𝑗 + 𝐵𝜙𝑖,𝑗 −1 +𝐶𝜙𝑖−1,𝑗 +𝐷𝜙𝑖,𝑗 +1 − 𝐸𝜙𝑖,𝑗 = 𝑆 𝜕𝑡 + 𝑞𝐹 where A, B, C, D, E = coefficients, 𝜙 = velocity
potential and S = storage coefficient, qF =groundwater withdrawal rate per unit area.
Considering steady state condition with no withdrawal, write an algorithm to compute potential at each
grid and seepage rate under a dam.
3. Derive the Finite Difference equation for simulating river stage-water table interactions considering
one dimensional flow.