Error in topology optimization with modal analysis as input
Altair EmployeeHi,
Can i get to now in which version do you plan to patch the following error :
The SUBCASE ID in ATT1 field of DRESP1 does not refer to a modal FRF subcase.
which occurs while performing modal analysis with the analysis type as 'normal modes' and a optimization resaponse as 'frequency' . I am facing the error in both HM 17 and 17.1!
Answers
-
0
-
Hi Prakash,
Thanks for the clarification. Also can you please let me know if the parameter TOPDISC has been removed from the DOPTPRM control card in HM 17.1 .I was able to use it in 17 but i am unable to find it in 17.1
Thanks & Regards,
Nipun Dahra
0 -
0
-
Hi Prakash,
I did a topology optimization and now i wanted to validate my design using OSSmooth. I was using the FEA reanalysis option in OSSmooth panel.But when i select the 'Keep smooth narrow layer around'option and choose a threshold value of less then ~0.1 (desired threshold is around 0.04),I am facing a segmentation error while running OSSmooth.I am not getting this error when i choose a value greater or equal to around 0.15 (But those values are undesirable as the optimisation results are discontinuous in space for higher threshold values which are expected as i have to use global mises stress as a optimisation constraint). Also i don't get this error when i switch off the 'Keep smooth narrow layer around' option.
I am using Linux and tried deleting some files created in the working directory to avoid this error but with no luck. I am using HM V2017.1.If required,you can find the corresponding .fem and .sh files below. I request you to please guide me through to solve my problem.
Thanks and Regards,
Nipun
0 -
Hi,
There is no workaround for segmentation other than deleting setting files and temp folders. If you access to any other machine please try on that and check if that helps.
0 -
Hi,
I tried it on other machine but was getting the same error.Since I don't have any access to HM 14 or other older versions can you please try running it on HM 13/14?.
Attached below are the required .fem and .sh files and two pictures showing the required setting in ossmooth.Also attached is the terminal output generated while trying to run ossmooth
0 -
Hi @nipundahra
I tried both in 14.0 and HW 2017.1 and it works fine at my end. I am using a Windows machine.
0 -
Hi. Thanks for trying at your end.
As seen in the terminal output ,it shows some 'self intersection' which i think is because the software is unable to generate correct mesh for reanalysis.For me a crash file is generated after the segmentation eror in which a 2D surface mesh is generated.When i open the crash file separately and try to convert that 2d mesh to 3D,i get the same segmentation error at the % progress as observed in the status bar.
Can you suggest some solution based on this observation?
regards,
Nipun
0 -
Hi,
Can you change the parameters in OSSmooth panel?
Remove split quads and check keep Smooth narrow layer around.
0 -
Hi,
I checked with split quads option which didn't give me an error.I get the error only if i check and use the Smooth narrow layer option.I attached 2 pictures for which i wanted you to check.Can you please try for those settings?
I meant please try using the threshold of 0.04 and check keep smooth layer option on.This is the required setting and giving the error!
I don't get any error when -
1) i switch off the smooth layer option and use any threshold value
2) i switch on the smooth layer option and use the threshold value above 0.15
0 -
Altair Forum User said:
I meant please try using the threshold of 0.04 and check keep smooth layer option on.This is the required setting and giving the error!
No segmentation error or crash except a meshing error but over all works fine at my end on HW 2017.1
<?xml version="1.0" encoding="UTF-8"?>
0 -
Ok. Thank you very much for the support
regards,
Nipun
0 -
0
-
yes, only the smoothening option give me the error.Other options don't throw this error
/emoticons/default_sad.png' srcset='/emoticons/sad@2x.png 2x' title=':(' width='20' />
0 -
Well the only better way is to get the geometry and mesh it for your case suits. There are few areas where mesh is not seen which is not good for further analysis.
0 -
Yes,already started with that thing. Can you please share the.hm file you got after the OSSmooth operation?
0 -
You mean reanalysis or geometry?
0 -
Here is the .hm file from OSSmooth reanalysis
0 -
Hi,
can you please try running ossmooth with threshold of 0.04 on linux machine?
Also i don't think the attached model is for the threshold of 0.04.can you please recheck?
0 -
0
-
Thank you very much
0 -
0
