[ac_ankit]

Hi everyone please advise me a topic that I can finalize for my university minor project on CFD using gambit and fluent.
Thanx in advance.

[wangyifei]

you mean is that someone can teach you about the fluent and Gambit？

[Far]

I think flat plate and circular cylider with DES model would be the good idea

[ac_ankit]

@ wangyifei you mean is that someone can teach you about the fluent and Gambit？

NO I just need an idea and some relevant details and rest I will do myself.

[ac_ankit]

Originally Posted by Far

I think flat plate and circular cylider with DES model would be the good idea

can you please provide some explanation....

[ac_ankit]

Originally Posted by ac_ankit

you mean is that someone can teach you about the fluent and Gambit？

NO I just need an idea and some relevant details and rest I will do myself.

[truffaldino]

Are you looking for an exercice-type project (i.e. already done before, just to pass exams) or developing something more or less new?

[Far]

Originally Posted by ac_ankit

can you please provide some explanation....

Detached eddy simulation is new relatively new development in CFD, lies in between RANS and LES. In other words it is RANS in boundary layer (you can use more stretched mesh) and switches to the LES (automatically) in the far field. So meshing requirements are bit higher than RANS, but very very less than the Large Eddy Simulation (LES). Disadvantages are:

1. Unsteady problem
2. Requires more RAM
3. More computational time



As you know that Flat plate and Circular cylinder very classical fluid mechanics problems. You can find very good details in any fluid mechanics book or use Google.

[ac_ankit]

Originally Posted by truffaldino

Are you looking for an exercice-type project (i.e. already done before, just to pass exams) or developing something more or less new?

yes, developing something new is what I am looking for.

[truffaldino]

Originally Posted by ac_ankit

yes, developing something new is what I am looking for.

I can propose you something which is not new from research side, but can be very useful from ingeneering/user point of view.

Recently, I have encountered problem analyzing unconventional airfoils at low reynolds numbers of order 10^5. At such reynolds numbers one has huge problems with turbulence modelling, as flow is neither fully turbulent nor laminar: there are separation bubbles and transition to turbulence.

There is a good piece of software, called x-foil, which copes much more better with such analysis than any cfd-sofware, but it only works for conventional airfoils.

In my case I was analysing airfoils with backward facing step on the surface, so I had to use LES with gambit/fluent: I had an idea to write an authomatic software, say in mathlab or scilab, that generates journal files for gambit and fluent so that other non-ultra-specialist people could use them, but had no enough time to finish the project.

In more detail: you can write a program in matlab or scilab (better scilab since it is free), that takes airfoil (or multiple airfoil) coordinates, reynolds number and angle of attack/free stream turbulence intensity, and then generates gambit and fluent journal files for LES simulations.

If you write such a program and then post it somewhere, thousands of persons will be grateful to you, and still such a project looks like an excercise.

I could pass you scilab/journal gambit/fluent files I have stopped at.

Truffaldino