|
[Sponsors] |
August 20, 2011, 04:41 |
Transient simulation in CFX
|
#1 |
Member
Join Date: Sep 2010
Posts: 45
Rep Power: 16 |
Hi all, I am doing transient simulation of a compressor using CFX 12.1. I set the time step to be 20 steps in a passage. But the tips show up as "The Maximum mach num exceed 3" after a few steps,then the mach number go larger and larger until it flows up.
I try to set the time step smaller, i.e. 30 steps in a passage, or to use the adaptive timesteps, but problem is still there. I have no idea now, what should i do next? Thanks a lot, BlanceChen |
|
August 21, 2011, 19:11 |
|
#2 |
Member
Mauricio Labarca
Join Date: Aug 2009
Posts: 35
Rep Power: 17 |
a high Mach number means an acceleration of fluid in your domain. Try simplyfying a bit the model and also trying a different turbulence model and options.
|
|
August 26, 2011, 11:58 |
|
#3 |
New Member
Join Date: Jul 2011
Posts: 2
Rep Power: 0 |
Probably you have a sharp corner or some geometry ending in tangent.
As Mauricio told, try to simplify your geometry in these points building right angles. |
|
August 27, 2011, 23:33 |
|
#4 |
Member
Join Date: Sep 2010
Posts: 45
Rep Power: 16 |
In a casual test, I do a simulation both in Win CFX 12.1 and Linux CFX 12.1, the problem still appears in the Linux CFX 12.1, but fine with Win CFX12.1. Is it the problem of the Linux CFX12.1? Both sets of the simulation are exactly the same, except the CPU used to compute.
|
|
August 28, 2011, 07:40 |
|
#5 |
Super Moderator
Glenn Horrocks
Join Date: Mar 2009
Location: Sydney, Australia
Posts: 17,870
Rep Power: 144 |
It is very unusual but not impossible for a simulation to have different results in Windows and Linux. If you can isolate what is causing it you can report it as a bug - but this will take some detective work.
Also be aware that you are not running the current version of the software (V13 SP2). There is a good chance your bug has been fixed in the newer releases so I suggest you upgrade. |
|
August 28, 2011, 22:45 |
|
#6 |
Member
Join Date: Sep 2010
Posts: 45
Rep Power: 16 |
I do another test by an example transient simulation on the Linux CFX12.1, found no problem there. Then it should be some mistake i made in my simulation settings, or maybe the grid is not good enough? Which will be the most possible to cause such problem...It is really unbelievable that the same simulation can be continued on the Win CFX, but stopped on Linux CFX.
|
|
August 29, 2011, 19:36 |
|
#7 |
Super Moderator
Glenn Horrocks
Join Date: Mar 2009
Location: Sydney, Australia
Posts: 17,870
Rep Power: 144 |
As I said, as you appear to have found a bug in CFX I would stop going round in circles exploring this bug and upgrade to the most recent version where it is probably fixed. I don't know about you but I do not have time to waste chasing bugs which have probably already been fixed.
|
|
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
CFX steady simulation | gharek | CFX | 1 | April 7, 2010 19:41 |
chart for transiente simulation - CFX Post | rogbrito | CFX | 0 | June 22, 2009 13:12 |
Problems on H2/air CFX simulation | xulixian | OpenFOAM Running, Solving & CFD | 2 | April 14, 2009 16:00 |
tracer simulation for transient problem. | John | CFX | 0 | August 12, 2005 11:37 |
transient simulation: natural convection problem? | Basics | CFX | 3 | September 25, 2002 10:42 |