AMS DIvergence with only RADIOSS v14.0?
Altair EmployeeHi,
I have started applying AMS in RADIOSS v14.0 .Initially the simulation is running and after some time (around 2 hours). Later an error is sprouting .Pl see the below error message.
** ERROR : AMS IS LIKELY DIVERGING:
... RUNNING DIAGNOSIS
** INFO : COULD NOT IDENTIFY THE ISSUE.
Can you suggest a possible soultion to curb this? The values recorded in .OUT file are normal. in the sense there is no abnormalities in %energy error or mass error%. However it suprises me that after almost half of the simulation is done with AMS,divergence happens.
I have converted the same input files(Starter and Engine files,) into v13.0 and run in RADIOSS v13.0. The simulation is running with no error. I am using a time step of 5microseconds for the simulation. I have laso tried with smaller and larger time steps just in case to verify if it is because of DTMIN value. But the scenario is same even with that case.
I have also gone through the release notes for v14.0 . It has included some relevant points while using INTER7 and INTER2 which I have taken care.
Thanks for your support in advance.
Answers
-
Maximum allowed iterations before sending a divergence message is 1000 by default. Anything above this maximum value, RADIOSS stops with this error message. 30 iterations or less is considered as a good convergence.
Monitoring the number of iterations taken in your analysis before showing this error.
Choose a minimum time step, Tmin, from 10 to 20 times than the one used in /DT/NODA/CST. Tsca sued is 0.67 for /AMS.
Please recheck the Tmin value and try rerunning the analysis.
0 -
Hi,
the target time step for AMS is x12 times the reference time step used for /DT/NODA/CST. Hence the Tmin shouldn't be a prooblem. Also I have run the same model with RADIOSS v13.0 and it is working good with that. However if I solve same .rad files with RADIOSS v14.0. I am getting this error. How could the version change cause divergence in AMS. If there is a problem with either starter or engine files. It could cause problem in both versions of RADIOSS. Is there anyother way to look at the issue?
0 -
Can you share the model files (starter and engine files).?
0 -
Hi,
I couldn't share the files (I'm not authorized to share). However I could look into them if you could specify me the key areas to be checked.. I have checked for interface definitions and also the model has no initial penetrations or intersections.Pl let me know the possible areas that could cause divergence...
Thanks in advance :-) .
0 -
Also Im using the Tsca as 0.67 to my simulation.
0 -
Hi Raghav,
The model file will be safe if you can share via the dropbox link provided. So that we can try the same model in the latest solver patch.
Meanwhile please check the HW Solvers patch you are using now.
0 -
Hi,
Do you mean the version of HW on which I'm working when you refer to solvers patch? or if it is something else.Pl let me know how to verify that.
0 -
In Help menu of HyperMesh, click on Updates and System Information. Please copy it.
Please share the out files (both 13.0 and 14.0) to my dropbox.
0 -
------------------------------------------------------------
hgtrans: 14.0.0.88 (Feb 5 2016)
hmopengl: 14.0.0.88 (Feb 5 2016)
hstdss: 14.0.0.88.872435 (Feb 5 2016)
hvp: 14.0.0.88 (Feb 5 2016)
hw: 14.0.0.88 (Feb 5 2016)
hwx: 14.0.0.88 (Feb 5 2016)
0 -
I hope this helps .. I have tried /DT/NODA/CST with version 14.0 it is working good and the results are on par with the simulation run with v13.0. However there is an increase in the elapsed time. In version 13.0 the simulation run with /DT/NODA/CST has taken 8hr27min. However for the same time step & the same model with v14.0 it has taken 8hr55min. Could you tell me why there is an increase on simulation time just with the version change?
0 -
Hi Raghav,
We will update you on this soon.
0 -
Hi ,
Any solution for the issue?
0 -
Hi Raghav,
We will fix this issue in the next Student Edition..
0