A: the Reynolds Stress Model and considering the


A: Introduction to Episode

1.1: This career episode pertains to the Final Year Project I completed
as part of my Bachelors degree in Mechanical Engineering. The project lasted
one year, from August 2011 to May 2012. The title of the project was “Life
Prediction of Tarbela Dam Tunnel 4 by measuring the erosion rate using
numerical models”. It was performed at Ghulam Ishaq Khan Institute of
Engineering Science and Technology (GIKI). The results were shared with Water
and Power Development Authority Pakistan. The project team consisted of 4
people, including me, sharing responsibilities as a team, and we were working
under our designated Advisor.

We Will Write a Custom Essay Specifically
For You For Only $13.90/page!

order now


B: Background of Episode

1.2: At  GIKI, we were required to complete a
year-long engineering project during our senior year as part of our degree. It
was part of the requirements that the project be done as a team, and be of
industrial and/or research relevance. We were meant to form a team, select a
suitable topic and request a faculty member to accept the project as our
advisor based on its merits.


1.3: Tarbela dam is one of
the largest earth filled dam in the world. The inflow of sediments in the
Tarbela reservoir has resulted in reduction in water storage capacity. During
recent years, a reasonable increase of sediment particles in the tunnel has
been observed. This is damaging to the tunnels, power generating units and is a
severe threat to the plant equipment. No comprehensive studies have been
performed on the turbulent flow through the Tarbela dam tunnel 4 with the added
penstock. In this project, turbulent flow in Tunnel 4 of the Tarbela Dam with
and without the added penstock was analyzed using the Reynolds Stress Model and
considering the effect of sediments particle. Results were presented for three
different water heads in the reservoir i.e. considering summer, winter and
average seasons and for one-way and two-way/full coupling for sediments
particle tracking/deposition. The effect of cavitation erosion and damage to
the tunnel due to erosion was investigated and results were compared with the
WAPDA reports. Moreover, pressure, velocity and erosion rate results were
observed to obtain the complete behavior of the turbulent flow of water in the


1.4: My responsibilities
in this project mainly included work on the core engineering aspects of the
problem as well as preparing relevant presentations and ensuring timely completion
of milestones. My engineering work included analyzing different mesh types
using tetrahedral and prism elements for use in ANSYS numerical analysis,
working on comparing the results of 3 phase and 2 phase analyses of fluid flows
through the tunnel using ANSYS CFX®. I also ran analyses for the
effects of cavitation related failures on the tunnel. Lastly, I was responsible
for validating the results of the simulations.


C: Personal Engineering Activity

1.5: With the
end goal of estimating erosion in mind, there were several approaches we could
have taken. I understood that the most direct way of approaching the problem
was to model the physics of the flow field. In order to do so, I identified,
according to my knowledge of Finite Element Analysis, that I had to perform
mesh sensitivity analysis. The purpose of mesh sensitivity analysis is to
determine the elements required to generate a solution that is independent of
the number of elements used. Starting with a coarse mes, I increased the number
of elements in the region of interest until the solution from each grid was
unchanged for successive grid refinements.


1.6: For
turbulence modeling, I considered two turbulence models, the Reynolds
Stress Model and the K- ? model. I decided to use the Reynolds Stress Model
over the other one as it can predict flow separation better and reproduce
anisotropy of the flow in the turbulent boundary layer, and due to the presence
of sharp bends, high velocity and highly swirling flows of the tunnel under
consideration. In order to model erosion, I studied erosion models, within the
context of my knowledge of Fluid Dynamics, and their implementations in ANSYS
CFX®. The
erosion model I decided to use in ANSYS CFX® was the Finnie model since in my
research I found it to be one of the most basic and easy to use erosion models.
It requires lesser number of coefficients to calculate the erosion rate in
comparison to other erosion models like Tabakoff. In ANSYS CFX® I used the
Finnie model in conjunction with Lagrangian particle tracking and
Eulerian-Eulerian multiphase approaches.


1.7: The next step was to conduct the
multiphase analyses. The tunnel mainly carried water, along with sediments, and
certain amounts of water vapor. I figured that from our options it would be
best if we could use the 2 phase model as it is simpler and requires less time
for computation. However, to do so, I had to show that the model gave us
results accurate enough, since the 2 phase model does not account for water
vapor. Therefore, I ran 2 phase and 3 phase analyses for different pressure
heads to get water and sediment velocities. I found the results to be close
enough for us to go ahead with choosing the 2 phase model.


1.8: After we ran the simulations, the
final issue was about validating our results. To that end, I met with WAPDA
engineers to get their measurements for tunnel 4. The measurements I obtained
were for the existing tunnel. In order to validate our models, we first had to
make sure that the results of our simulations matched those of the engineers at
WAPDA. However, our simulations had been run using a modified model of the
tunnel with an added penstock. The addition of penstock was a future project in
order to install turbines and generate electricity from tunnel 4, while
currently the tunnel was being used for irrigation purposes only, and the
measurements corresponded to this current model. So I modified the model to
match the current physical one and re-ran the simulations for the measurements
that I obtained from WAPDA. The simulated results for pressures and velocities
matched the measurements from WAPDA within the required margin for error. This
validated our model.





D: Summary

This project was my first experience of working on a large engineering
project. It helped me understand better everything that I had previously
learned about engineering, teamwork and project management, as I encountered
real world examples and was able to refine my concepts in relation to how they
applied to the real world as opposed to remaining abstract constructions. I was
able to perform all my work and deliver well within deadlines, and the methods
I used and solutions I arrived at were well regarded. We were able to use our
results to publish a scientific paper.