CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   Main CFD Forum (http://www.cfd-online.com/Forums/main/)
-   -   Is it possible in a week? (http://www.cfd-online.com/Forums/main/87343-possible-week.html)

mechfries April 18, 2011 07:56

Is it possible in a week?
 
So randomly my fluid mechanics professor came to class and said ' if you know abaqus, i have something which would pay you good decent money'

i was like.. umm... sounds cool.

Problem, I have no experience in CFD software nor any of my classmates.

My question is it can i learn ansys or abaqus in a week?? Which one should I basically is better? I know each software is best suited for different problem but as a beginner - HARDWORKING beginner- which one should i go for? I have ansys installed and I believe ansys is the most popular CFD out there and I assume loads of tutorial out there. Can I learn ansys in a week? I only need to learn the fluid mehanic part.

Alanseu April 18, 2011 10:42

ABAQUS is very great for the nonlinear analysis.
In a week, er...i don't know, maybe you can, try your best
After all, there are many common respects for the FEM softwares.;)
Quote:

Originally Posted by mechfries (Post 304040)
So randomly my fluid mechanics professor came to class and said ' if you know abaqus, i have something which would pay you good decent money'

i was like.. umm... sounds cool.

Problem, I have no experience in CFD software nor any of my classmates.

My question is it can i learn ansys or abaqus in a week?? Which one should I basically is better? I know each software is best suited for different problem but as a beginner - HARDWORKING beginner- which one should i go for? I have ansys installed and I believe ansys is the most popular CFD out there and I assume loads of tutorial out there. Can I learn ansys in a week? I only need to learn the fluid mehanic part.


mechfries April 18, 2011 13:25

what do u mean very good for nonlinear? Will it get work done faster or is it less equation and more click click click click?

I know everyone in CFD uses Ansys which could mean the program is big -pretty big installation- and verryryryr complex. I have COSMOL install also and after watching some tutorial on their site, it looks a good program.

Do you know about COSMOL? I think I'm going to try ansys tonight and see how far I get. If it seems a bit too complicated, going to try Abaqus or COSMOL. I just like the way COSMOL gui looks. :)

Alanseu April 18, 2011 20:48

The Abaqus is good at analyzing the nonlinearity problems such as the material nonlinearity, structure nonlinearity and the geometry nonlinearity.
Sorry, I don't know the COSMOL, many people in my lab use the abaqus, ansys, patran/nastran. :)
Quote:

Originally Posted by mechfries (Post 304098)
what do u mean very good for nonlinear? Will it get work done faster or is it less equation and more click click click click?

I know everyone in CFD uses Ansys which could mean the program is big -pretty big installation- and verryryryr complex. I have COSMOL install also and after watching some tutorial on their site, it looks a good program.

Do you know about COSMOL? I think I'm going to try ansys tonight and see how far I get. If it seems a bit too complicated, going to try Abaqus or COSMOL. I just like the way COSMOL gui looks. :)


jbritton April 19, 2011 06:41

Im not sure Abaqus actually has a CFD side to it, as far as I know it is a structural software for mechanical analysis, usually if you were to use abaqus you would have a separate CFD solver such as CFX Fluent star CCM, so think you should clarify if you Fluid mechanics professor is actually referring to a mechanical problem he has.

Might be, he already has some CFD results and wants to couple them with mechanical results.

But yeah as far as I know abaqus is a high end non linear mechanical analysis tool. You could not learn how to use it in a week. To be honest same with ANSYS mechanical, its much easier to use but dependent on the complexity of the problem you would probably be unable to solve a complex non linear problem with no help after a week.

Comsol I dont use but as far as I know isnt really in the same league as the other codes i have mentioned, it also involves a lot more coding from what ive heard (Though that may well be wrong) an experienced comsol user is prob best to ask.

I would use ANSYS to do this if I had the choice, but TBH dont think you'll be able to learn either their structural package or their CFD packages to a particularly high level in a week. You could prob learn to do a static structural and thermal and modal analysis, and maybe a basic flow simulation in a week. But if you havnt had any previous experiences of correctly meshing CAE models your results will be questionable. Non linear structural analysis would take at least 2 weeks to learn as so many problems can occur to disrupt the convergence and stop you from getting a solution, and it can take years to gain the experience to be able to quickly identify these and fix them.

mechfries April 19, 2011 13:32

Thanks for your reply. After just looking at Ansys, I had a big 'WTF' face. The program is enormous. I was thinking 2 weeks is nothing, probably will take a month to learn it. I'm going to start with the official Ansys tutorial or documentation. I'm going to sticking to the mechanical and fluid part only. I'm not in the electromagnetic field, well not to in-dept in it.

But before I start doing anything, I have project of my own and why I requrie is to have nice animation showing how the 3d magnetic field is affected by the heat propagation into it, is it possible?

Martin Hegedus April 19, 2011 16:09

I think the question which should also be asked is, "Can a good answer be obtained in a few weeks?" The answer is probably no. Good answers usually require a lot of grid work, a lot of grid points or cells, and a few trail and error runs. One must know what they are looking for in a solution. This takes experience. A few solutions should also be converged to machine zero (i.e. 1.0e-12). Don't expect to get a good 3D answer on a laptop. A top of the line workstation (12 cores, 48GB memory) might be able to obtain one solution point (where only loads have converged) on an adequately meshed grid in a day. Of course, this depends on your problem. This also depends on if you are looking at overall values or local values such as pressure or separation. If you are trying to capture local quantities, then that can be a lot of work and time.

This doesn't mean you should not take the job. Just manage your, and the client's, expectations.

mechfries April 20, 2011 12:21

I would need a supercomputer if working commercially but I'm just making all this just to show that I know how to use these software with nice graphics and animation during presentation. I don't actually require any specific results only need to know what happens when the system turns own and must be able to explain the specific. So returing to my question, is there a way to simulate electromagnetic field and heat thermal in ansys or will I need a different program?


All times are GMT -4. The time now is 03:56.