Simulation, Geometry & Mesh Creation: Tool Suggestions & System Requirements

[Mann Harjeet]

Hi Guys,

I have an query. I would like to know the best tool for Simulation, Geomtry and Mesh creation. I want to simulate a duct with a turbine in it as a Part of my research project. I am beginner for the simualtion field so please suggest accordingly the appropriate software. Also please suggest the system specification to run the same tool.

 

[Smed]

Is this an undergrad research project? CFD is an extensive field and each of the pieces that you mentioned; geometry creation, mesh generation, and simulation take years of experience before being able to create something so complex as a turbine. As a beginner, I would suggest that you start with something far more simple, like the well studied lid-driven cavity flow.

There are many different software packages available, but you'll need to see what your University offers. Ansys produces fairly intuitive CFD software.

 

[Mann Harjeet]

Thanks for your reply Smed, Its a post Graduation project. Nothing is available in the institute, Only I have to arrange the software at my own. That why I ask for the most suitable software. About Ansys I have heard that interaction of fluid with turbine is quite difficult.

 

[Vadar2012]

Pointwise is the best mesh generator by far. CFD++ seems to give the best results for most things. ANSYS is a decent beginner grade software, but you might need CFD++ for turbine stuff. You'll probably also need to be running it on a super computer. The mesh you'll require for decent results might need in the area of 64 cpus. If you go this route though, you'll have to learn linux as well.

 

[AlephZero]

ANSYS is a decent beginner grade software

Hm... I wouldn't call a software package that is certified compliant with design standards in the nuclear industry "beginner grade", but YMMV.

 

[Vadar2012]

Hm... I wouldn't call a software package that is certified compliant with design standards in the nuclear industry "beginner grade", but YMMV.

I'm not saying it's bad, it is a hell of a lot easier to use than the other programs that seem to give better resuls, and by better results I mean closer to experimental. In my experience, much closer. It's a damm good program, just the best to use for people starting out. You wouldn't want to start out with CFD++ on a supercomputer having to use a command shell. That'd scare you off forever.

I started off with ACE/FASTRAN, which had it's own mesh generation and computational environments. Very good to learn with, but a terrible terrible program in terms of results.

Note, I did scramjet combustion engine research, not a lot of programs are good at simulating turbulent hypersonic conditions. I'm just referring to ease of use.

 

[boneh3ad]

ANSYS Fluent is a great code. It is also a general code, so it is good at a lot of things but not necessarily great any most of them. There are a lot of areas where other solvers are better. I know, for example, that a lot of academe uses GASP for high-speed aerodynamics problems.

Either way, the more pervasive problem is that the OP is wanting to fund this himself, which is going to be exorbitantly expensive.