076BAS030 Assignment2 CFD
076BAS030 Assignment2 CFD
076BAS030 Assignment2 CFD
INSTITUTE OF ENGINEERING
PULCHOWK CAMPUS
Aerospace Engineering
Computational Fluid Dynamics
Exercise 2 Solution
Q1. Download the second ANSYS Fluent tutorial from MS Team and study it
carefully. If there are any unclear parts, steps that you cannot follow, steps in
which ANSYS does not behave in the way described in the tutorial, or if you
notice any mistakes or typos, take notes about them and submit them as an
answer to this question. Otherwise, answer this question by writing “Everything
was clear and I did not notice any mistakes”.
Solution:
Q2. The Reynolds number values considered in Tutorial 2 are very low. To
appreciate how low they are, consider the following problems.
We have,
ρU ¿ D
ℜ=
μ
U ¿= 0.002 m/s
Again,
For U ¿ =45 m/ s,
We get:
Re = 1.8×106
We have,
ρU ¿ D
ℜ=
μ
U ¿= 0.00439 m/s
Q3. Let’s solve the problem of Tutorial 2 using a series of systematically refined
meshes and perform a mesh independence study based on 𝐶𝐷. In a typical mesh
independence study, the ratio of number of elements in successive meshes is kept
around 2. For the meshing strategy used in the tutorial, this can be achieved by
increasing the number of elements in both the radial and circumferential
directions by a factor of √2.
A) Perform 𝑅𝑒 = 40 solutions (not 𝑅𝑒 = 80) with these meshes and obtain the
missing 𝐶𝐷 values in the table. The mesh used in Case 6 is the one used in
the tutorial.
Solution:
Cas Division in the Division in the No of Cd %
e radial direction circumferential elements Difference
direction
1 21 7 588 1.6548 5.71
2 30 10 1200 1.6131 3.05
3 43 14 2408 1.5905 1.60
4 60 20 4800 1.5768 0.73
5 85 28 9520 1.5707 0.345
6 120 40 19200 1.5670 0.10
7 170 56 38080 1.5653 0
In general, the more the elements there are on a geometry, the better one can
capture the contours of the geometry, and the more precise results may be
produced. Mesh independence study is performed to determine the dependence
of the results on the mesh density. The plot shows that the slope of the curve
almost remains constant once the no. of element exceeds 5000. So, it can be
said that a minimum of 5000 elements is required to get a mesh independent
Cd. This Cd is about 1.57.
C) Calculate the % differences of each 𝐶𝐷 value with respect to the one of
Case 7 (considering that it is most accurate one) and comment on which of
the meshes are acceptable as far as the accuracy of 𝐶𝐷 is concerned.
Solution:
The % difference of each Cd value with respect to the one of case 7 is
presented in the above table. Looking at the value of the % difference, we can
say cases 4 to 7 are acceptable as far as the accuracy of Cd is concerned since
the % difference is less than 1%. This conclusion is also verified by the graph
of Cd vs No of elements as the curve doesn’t show much noticeable change in
slope once the no. of elements exceeds 5000.
Q4. The object considered in Tutorial 2, the cylinder, is a blunt object. It is not
streamlined. As we know, streamlining a blunt body is known to reduce the drag
force. Following figure taken from Munson’s fluid mechanics book gives an
example of this. Three different two-dimensional objects with the same frontal
area, i.e. the same height, can produce vastly different 𝐶𝐷 values. Unfortunately,
we do not know at which Reynolds number this study is done.
Draw the above three geometries as close as you can to what you see above, put
them into proper fluid domains, generate proper meshes and perform 𝑅𝑒 = 100
simulations. 𝑅𝑒 number is based on the height of the objects, which is the same
for all. Select the fluid properties by yourself.
100mm
d) Provide mesh figures and the number of elements you created for each
case. Provide zoomed in figures to show details, if necessary.
Solution:
Fig: Mesh for Body 1
No. of elements: 14,003
e) Provide couple of post-processing figures for each case and discuss them
briefly.
Solution:
For Body 1:
For Body 2:
For Body 3:
The variation of static pressure along the channel is shown by the static
pressure contour. The plot shows that the static pressure is greater at inlet
where flow just hits the channel. Those points are called stagnation point and
the static pressure is higher as velocity is zero at that point. The results reveal
that the static pressure variation along the channel is influenced by frictional
resistance and momentum exchange. The frictional resistance term diminishes
the static pressure, and the momentum exchange term makes the static
pressure tend to increase.
With no surprise, the Cd of Body 1 is the largest followed by body 2 and body
3 has the least Cd among them. This is mainly due to the streamline effect. The
more you streamline an object by elongating its rear surface, the more you
reduce the size of its wake and the resulting pressure drag.
However, it is quite noticeable from above conclusions that the Cd value has
reduced only by few decimals. The streamlined body must have highly reduced
drag coefficient than a bluff one. That’s true but in our case, the length is least
for body 1 and highest for body 3. The higher length also results in increase in
skin friction drag. That’s why, the larger length of body 3 prevented high
reduction in Cd.
g) Evaluate the drag force acting on each body and provide screenshots of
the Console Tab showing viscous drag, pressure drag and total drag
values. Provide a table showing the values for all bodies. Compare and
discuss.
Solution:
For Body 1:
The Cd value for Body 1 is 1.264.
For total drag force:
2
ρU S C D
FD=
2
We get:
F D =31.995 N
For Body 2:
The Cd value for Body 1 is 1.1737.
For total drag force:
ρU 2 S C D
FD=
2
F D =26.889 N
h) Is the 𝐶d ratio between the objects also hold for the total drag force?
Discuss.
Solution:
For Body 1 and 2:
C D 1 1.264
= = 1.0769
C D 2 1.1737
Similarly,
F D 1 31.995
= = 1.0769
F D 2 29.709
Similarly,
F D 2 29.709
= = 1.1048
F D 3 26.889
Similarly,
F D 1 31.995
= = 1.1898
F D 3 26.889
Equal ratio for both Cd and total drag force indicated that the 𝐶d ratio between
the objects also holds true for the total drag force.