[geo]

I am in my final year of mechanical engineering.

Please suggest me with any good topics for my final year project. I am looking for anything in which I can learn more about CFD as I want to build my career in petroleum industry.

Also please suggest whether it can or not, help to find opportunities in petroleum industries if I am good at CFD.

Thanks a lot!

[zmester]

hi

I think the best way to learn cfd is making your own cfd solver. I did this and i felt that everyday i understand cfd better than the previous day. However it is not easy to estimate the time you need to develope your first code.

[CFDtoy]

Since you are looking at petro industry...here is my thought on final year project.

1. work on a 1D or 2D (based on time requirement) Finite volume based multiphase solver.
2. Run cases of multiphase flows in ducts, orifices, venturis (very common).
3. Report on the effect of void fraction of the incoming vapor bubbles embedded within the flow on the overall flow field. For example, deposition of vapor in the recirculating regions near the expansion region of the venturi or a backward facing step flow config.

You can also take opensource codes and run different flow configurations, if you think, coding may take longer.

Also, 1D modeling is possible for multiphase solutions. It is pretty easy to code and you can make it unique and publish it.

Thanks.

Originally Posted by zmester

hi

I think the best way to learn cfd is making your own cfd solver. I did this and i felt that everyday i understand cfd better than the previous day. However it is not easy to estimate the time you need to develope your first code.

[geo]

Its good to have so quick and interesting reponses. Keep more ideas coming. These forums r gr8. Thanks a lot guyz!

[joe_star]

I'm currently working on a FLUENT based final year project to determine the depressurization time and temperature of a natural gas pipeline. You could carry out some similar form of analysis for your research too

[varunjain89]

Originally Posted by zmester

hi

I think the best way to learn cfd is making your own cfd solver. I did this and i felt that everyday i understand cfd better than the previous day. However it is not easy to estimate the time you need to develope your first code.

hi zmester

i'm also trying to learn CFD, and i'm trying to write my own CFD solver code.... and very true every day i get to learn something new. Though those are very simple silly mistakes, but they do matter a lot.

Currently i am trying to write NS code for compressible flow over flat plate. I am referring to ch 10 in cfd by anderson for help. but i'm really stuck @ a point. I need help with that. I know where's the problem in code, but dont know how to rectify it. Have u tried that code, coz it would be really helpful if u can help me.

thanks a lot.
varun

[zmester]

hi

Sorry to say but I dont know any other codes than mines. I developed my code from the basis. Do you mean there's a code in the book of anderson, and you tried to type it? I think one should begin coding with a Poisson solver rather then an NS solver. And its sound to be normal if it doesnt work for the first several tries. Debuging usually takes more time to me than coding.

[ztdep]

there are many free cfd code in the web. try read them and develop you own