|
[Sponsors] |
March 7, 2006, 09:12 |
Making program run faster
|
#1 |
Guest
Posts: n/a
|
Are there any easy alterations that u can do on a model so that the iterations rus faster? Because its moving really slow... Im using star-cd 3.26
|
|
March 7, 2006, 10:09 |
Re: Making program run faster
|
#2 |
Guest
Posts: n/a
|
Decrease the number of cells, use simpler or less models, buy more and faster computers....
What are you trying to do? |
|
March 7, 2006, 12:52 |
Re: Making program run faster
|
#3 |
Guest
Posts: n/a
|
Modelling the flow around a formula car... (Not a formula 1 obviously, its a formula student car). So i cant use a different model. Computers are quite fast because they r university computers! How do i decrease the number os cells?
|
|
March 7, 2006, 23:29 |
Re: Making program run faster
|
#4 |
Guest
Posts: n/a
|
You need to re-mesh the car model again, to decrease the number of cells.
|
|
March 8, 2006, 06:09 |
Re: Making program run faster
|
#5 |
Guest
Posts: n/a
|
how many cells you are using. To simulate a 3D car body, it would be advisable to run in parallel mode using as many processors as you can grab. It is usually true if you use 2 processors, time reduced by half, and so on.
|
|
March 8, 2006, 06:35 |
Re: Making program run faster
|
#6 |
Guest
Posts: n/a
|
Thanks!
|
|
March 8, 2006, 06:39 |
Re: Making program run faster
|
#7 |
Guest
Posts: n/a
|
I dont know how many cells the problem is... Anyway I will try running it on two processors then, thanks!
|
|
March 8, 2006, 06:53 |
Re: Making program run faster
|
#8 |
Guest
Posts: n/a
|
Konstantinos, obviously if you want some reasonable results, you NEED and MUST read something about CFD (you'll find some references in the CFD-onine book section situated on the top of this page - right hand side). Additionally you MUST check the manuals. I know its a lot of reading, but STAR-CD is not MS Excel. Wait for your advisor and make him to get the training for you.
good luck matej ps: RTFM stands for 'read the f*** manual' ) which is usually good advice in 99% of cases |
|
March 8, 2006, 10:53 |
Re: Making program run faster
|
#9 |
Guest
Posts: n/a
|
Are you using tets or hex for the case?
If you're using tet mesh (as I always do) I think it would help to use the command TETALIGN ALL which will reorder the vertex numbering and increase uniformity for tet meshes. This in turn would help to reduce memory requirements and computer time. |
|
March 8, 2006, 12:58 |
Re: Making program run faster
|
#10 |
Guest
Posts: n/a
|
I have to agree with matej unfortunately it seems that you dont know a great deal about the subject, "how can I make my simulation run faster" is in the same bracket of questions of "how many cells do I need", "how big should my cells be", "I am studying X and my results are wrong, why" and so on and so forth.......!
|
|
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
near wall faster | azmir | Siemens | 0 | December 26, 2007 20:36 |
Making SIMPLE converge faster | Fabio | Main CFD Forum | 4 | April 9, 2006 12:35 |
Fortran 90 faster than C/C++ | B. R. Guirguis | Main CFD Forum | 48 | March 6, 2006 14:49 |
MicroTunnel V1.2 - Now 6x faster! | Axel Rohde | Main CFD Forum | 0 | September 19, 2001 00:34 |
more RAM or faster CPU?? | Fabrizio Grieco | Siemens | 11 | January 23, 2001 08:35 |