Aerospace and Mechanical Engineering

2ndExam of Computational Fluid Dynamics

February 2nd, 2024, 8:00 AM
Rooms: F3, F4
Closed Book Exam
Duration: 2 hours
Name: ________________________________________________________________________ Number: ______________

We are all expected to abide by the IST standards for academic excellence and integrity.

Problem 1 (1.5+0.5+1.5+1.5+1.0 = 6.0 Marks)

Consider the homogeneous conditions in a room with Volume 𝑉, ventilated with a Flow rate 𝐹̇ .
Due to the dilution, an initial concentration of some pollutant in the room is, 𝐶 , is changing with
time. The concentration of that pollutant in the incoming flow is zero, 𝐶𝐼𝑁 = 0. For the initial
condition, consider 𝐶 (𝑡 = 0) = 1000.
The figures below illustrate the system and the exact solution.

Under this conditions the ODE governing the process is:

𝑑𝐶 𝐹̇ 𝑑𝐶 𝑑𝐶
=− 𝐶 → = −𝛼𝐶 → = −0.02𝐶
𝑑𝑡 𝑉 𝑑𝑡 𝑑𝑡
a) Discretize the equation using:
i) First order Euler explicit

ii) Crank-Nicholson

iii) Adams-Bashforth second order. Also refer how to discretize the first time step

1
 Formulae

b) Discretizing the time line of 300 s with 2 time steps (Δ𝑡 = 150 𝑠), and using first order
explicit Euler, perform the 2 time steps and verify that the process is unstable.

c) Conduct a linear stability analysis to the ODE and define the range of stability according
to some pertinent parameters of your choice.

2
Now consider a ‘FULLY IMPICITICY’ within the Finite Volume framework, where the time line is
discretized as usual for the space line. The figures below illustrate the relation between the
Control Volumes and the corresponding time.

d) Apply the Finite Volume Method and discretize using second order upwind for the first
derivative. We recall that Second Order Upwind is a simple Linear Extrapolation.

e) Fill the first 2 lines of the system of equations:

⋱ ⋱ =
⋱ ⋱ ⋮ ⋮

3
4
Name: ________________________________________________________________________ Number: ______________

Problem 2 (2.0 Marks)

Consider the incompressible flow around a bluff body as sketched below, in side view and in top
view, respectively.
The body is located in a plenum and the incoming flow is illustrated with the red arrows.
Say, justifying, the type of boundary condition to be imposed in each one of the 6 boundaries of
the domain.
For a typical length 𝑳, estimate, justifying, the value, or the process to find a value, for: 𝑳𝑼 , 𝑳𝑫 , 𝑳𝑻 ,
𝑳𝑩 and 𝑳𝑺 .

5
Name: ________________________________________________________________________ Number: ______________

Problem 3 (1.0+1.0 = 2.0 Marks)

Consider the mesh below:

a) Explain if this mesh is a Delaunay triangulation, or not.

b) With the target of mesh improvement, it was decided to obtain a more uniform mesh. The
Bowyer-Watson algorithm was chosen as the technique to be considered in mesh
improvement. Insert a point inside this mesh to reach the referred target. Clarify all steps
and conclude.

6
Name: ________________________________________________________________________ Number: ______________

Problem 4 (0.5+1.0+1.5+1.5+0.75+0.75 = 6.0 Marks)

Consider the unstructured mesh on the

right where the central control volume is
to be taken as P.
The pertinent points on the surrounding
were properly labeled.
A local Coordinate system, presented in
red color and centered on the control
volume P, may be the support for all the
computations.
The height of the octagons is 1 and all
octagons and quadrilaterals are regular.

As a methodology, the subscript PA, for

instance, refers to the entities at face
separating control volumes P and A, or to
the distance vector of A relative to P.

For the sake of simplicity, some quantities/entities were previously computed and are
presented below. Again, to clarify, if 𝜙𝑃𝐴 is the interpolation of 𝜙 to the left face shared by
volumes P and A, then: 𝜙𝑃𝐴 = 𝑔𝑃𝐴 𝜙𝑃 + (1 − 𝑔𝑃𝐴 )𝜙𝐴

Centroid
𝐶𝑃 = (0; 0)
Normal √2
𝐶𝐴 = (− ; 0)
𝑛⃗𝑃𝐴 = (−1; 0) 2
√2 √2 √2 √2
𝑛⃗𝑃𝐵 = (− ; ) 𝐶𝐵 = (− ; )
2 2 2 2
𝑛⃗𝑃𝐶 = (0; 1) √2
𝐶𝐶 = (0; )
√2 √2 2
𝑛⃗𝑃𝐷 = ( ; )
2 2 √2 √2
𝑛⃗𝑃𝐸 = (1; 0) 𝐶𝐷 = ( ; )
2 2
√2 √2 √2
𝑛⃗𝑃𝐹 = ( ; − ) 𝐶𝐸 = ( ; 0)
2 2 2
𝑛⃗𝑃𝐺 = (0; −1)
√2 √2
√2 √2 𝐶𝐹 = ( ; − )
𝑛⃗𝑃𝐻 = (− ;− ) 2 2
2 2
√2
𝐶𝐺 = (0; − )
2
√2 √2
𝐶𝐻 = (− ;− )
2 2

7
 √2
a) Prove that 𝑔𝑃𝐴 = 1 − and 𝑔𝑃𝐵 = 0.5.
2

b) Determine the x component of the Cell ‘average’ Gradient for cell P and verify that:
̅̅̅̅̅̅̅̅ √2 𝜙𝐵 𝜙𝐷 𝜙𝐹 𝜙𝐻
∇(𝜙)𝑃 |𝑥 = (−𝜙𝐴 − + + 𝜙𝐸 + − )
4 2 2 2 2
Using the generic function 𝜙(𝑥; 𝑦) = 𝑥, conclude about the order of accuracy of this
component of the cell gradient in cell P.

Formulae
1 1
̅̅̅̅̅̅
∇𝜙𝑃 = ∫ ∇𝜙 𝑑𝑉 ̅̅̅̅̅̅
∇𝜙𝑃 = ∫ 𝜙 𝑑𝑺
𝑉𝑃 𝑉𝑃 𝑉𝑃 𝜕𝑉𝑃
1
̅̅̅̅̅̅
∇𝜙𝑃 𝑉𝑃 = ∑ ∫ 𝜙 𝑑𝑺 ̅̅̅̅̅̅
∇𝜙𝑃 = ∑ ̅̅𝜙̅̅𝑓 𝑺𝒇
𝑓𝑎𝑐𝑒 𝑉𝑃
𝜕𝑉𝑃 𝑓=𝑛𝑏(𝑃)

8
 c) Consider the transport equation:
𝜕𝜙
+U⃗ ∙ (∇(𝜙)) = ∇ ∙ (∇(𝜙))
𝜕𝑡
⃗ = (1.0; 1.0) constant in time and uniform in space.
with the velocity field 𝑈
Applying the Finite Volume method to a generic cell P, the process follows as referred in
the next 8 steps.
i) Fill the green column of the FV process with the label in the green table below.

1 𝜕𝜙
⃗ ∙ (∇(𝜙)) + ∇ ∙ (∇𝜙)
= −𝑈
𝜕𝑡
2 𝑡+Δ𝑡
𝜕𝜙 𝑡+Δ𝑡
∫ 𝑑𝑉) 𝑑𝑡 = ∫
(∫ (− ∫ 𝑈 ⃗ ∙ (∇(𝜙)) 𝑑𝑉 + ∫ ∇ ∙ (∇𝜙) 𝑑𝑉) 𝑑𝑡
𝑡 𝑉 𝜕𝑡 𝑡 𝑉 𝑉
3 𝑡+Δ𝑡
𝜕𝜙 𝑡+Δ𝑡
∫ (∫ 𝑑𝑉) 𝑑𝑡 = ∫ (− ∫ 𝜙(𝑈 ⃗ ∙ 𝑛⃗) 𝑑𝑆 + ∫ (∇𝜙) ∙ 𝑛⃗ 𝑑𝑆) 𝑑𝑡
𝑡 𝑉 𝜕𝑡 𝑡 𝑆 𝑆
𝑡+Δ𝑡 𝑡+Δ𝑡
𝜕𝜙
Δ𝑉 ∫ 𝑑𝑡 = ∫ (− ∫ 𝜙(𝑈 ⃗ ∙ 𝑛⃗ ) 𝑑𝑆 + ∫ (∇𝜙) ∙ 𝑛⃗ 𝑑𝑆) 𝑑𝑡
4 𝜕𝑡
𝑡 𝑡 𝑆 𝑆

5 𝑡+Δ𝑡
Δ𝑉(𝜙𝑃𝑛+1 − 𝜙𝑃𝑛 ) = ∫ ⃗ ∙ 𝑛⃗)𝑑𝑆 + ∫ (∇𝜙) ∙ 𝑛⃗ 𝑑𝑆) 𝑑𝑡
(− ∫ 𝜙(𝑈
𝑡 𝑆 𝑆
𝑛
6
⃗ ∙ 𝑛⃗) + (∇𝜙) ∙ 𝑛⃗) 𝑑𝑆)
Δ𝑉(𝜙𝑃𝑛+1 − 𝜙𝑃𝑛 ) = Δ𝑡 (∫ (−𝜙(𝑈
𝑆

7 ⃗ 𝑖 ∙ ⃗⃗⃗
Δ𝑉(𝜙𝑃𝑛+1 − 𝜙𝑃𝑛 ) = Δ𝑡 [ ∑ [−𝜙𝑖𝑛 (𝑈 𝑛𝑖 ) + (∇𝜙)𝑛𝑖 ∙ ⃗⃗⃗
𝑛𝑖 ]𝐴𝑖 ]
𝑖∈𝑓𝑎𝑐𝑒𝑠
8 Δ𝑡
⃗ 𝑖 ∙ ⃗⃗⃗
𝜙𝑃𝑛+1 = 𝜙𝑃𝑛 + ∑ [−𝜙𝑖𝑛 (𝑈 𝑛𝑖 ) + (∇𝜙)𝑛𝑖 ∙ ⃗⃗⃗
𝑛𝑖 ]𝐴𝑖
Δ𝑉
𝑖∈𝑓𝑎𝑐𝑒𝑠

A B C D E F G H

(up to second order)

Original formulation
Integrating in space

Exact integration of

Numerical fluxes
stationary mesh
integration on a
Explicit scheme
Gauss theorem

Second order
Simplifying

and time

LHS

ii) Chose one step: from 4 to 5; from 5 to 6; or from 6 to 7; and fully explain what
does it mean and the related conditions.

9
d) For the steady version of the previous equation, and under the FV framework, discretize
the equation in cell P using first order upwind for the convective fluxes and second order
for the diffusive fluxes.

𝜕𝜙
e) Regarding the red cell, assuming 𝜕𝑛⃗ = 0 as boundary condition, discretize the 3
convective fluxes and explain how to discretize the 3 diffusive fluxes.

10
f) The flux limiters are non-linear schemes often exposed in the literature for 1 direction of
a uniform Cartesian mesh. Under the conditions presented before for the velocity field,
consider the application of MINMOD flux limiter to the green face. Fully explain how to
discretize 𝜙𝑓 in this face.

Formulae

11
12
Name: ________________________________________________________________________ Number: ______________

Problem 5 (10 x 0.4 = 4.0 Marks)

1- What is the connection between finite-volume methods and conservation equations?

__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________

2- Give two of the main advantages of finite-volume methods compared with finite-
difference methods?

__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________

3- Give an explanation how the lowest order finite-volume (called an upwind scheme) to the
advection equation ∂tQ−a∂xQ=0 leads to a finite-difference algorithm with a forward (or
backward) spatial differencing scheme, depending on the advection direction.

__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________

4- What is numerical diffusion and why is it relevant for finite-volume methods?

__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________

5- Determine the stability limit of the first order explicit upwind scheme.

__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________

13
 𝜕2 𝑄 𝜕2 𝑄
6- Show a finite-difference solution to the equation 2 − 𝜕𝑥2 = 0.
𝜕𝑡

__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________

7- Consider the the wave equation utt = uxx in a rectangular domain (x, t) ∈ Ω = [ − 1, 1] × [0,
1], where u(x, t) is the unknown function and U(x, t) is the known solution with the initial
conditions and boundary condition:
u(x, 0) = U(x, 0)
ut(x, 0) = Ut(x, 0)
u( − 1, t) = U( − 1, t)
u(1, t) = U(1, t)
Obtain the finite difference equation for n=1 and n=2 at a generic point i, using the finite
volume.

__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________

8- In an iterative method for the solution of the governing equations explain if the mesh
structure has influence on the convergence rate of the solution.

__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________

9- Explain in a qualitative way what the Riemann problem is and why it is so essential for
finite-volume schemes.

__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________

10- The finite-volume method extrapolates cell averages. What strategies do you see to
extend the method to high-order accuracy?

__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________
__________________________________________________________________________________________________________

14