Initial Penetration Error

Hallo,

I am trying to run a nlgeom simulation through Optistruct but the warning states that 'Initial Penetrations In Interface' but i checked using Tools -Penetration - Elements (all), there was no penetrations or interference. Could you please suggest what could be the reason of this warning and how can i avoid it in future ? i face similar situation in many simulations.

Needless to say that although in itself, is just a warning but my simulation always fails after running a couple of hours.

<?xml version="1.0" encoding="UTF-8"?>

Thank you in advance!

Best Regards,

Palash

Find more posts tagged with

English

OptiStruct

Accepted answers

All comments

Altair Forum User

Hi Palash,

Check for all intersections and penetrations including self intersection.

Sometimes even after removing all penetrations and intersections a warning message will be written in the out file. If there are no intersections and penetrations you can ignore the warning message.

Can you let me know what is the error message you get when analysis fails?

tinh

Hi,

Try again with option 'include self-interference'

I think you should rely on solver message

try debugging output files to find out node IDs that have penetration

Because sometimes hypermesh do not show it when checking by panel 'penetration'

someone tell me that hypercrash can check penetration better.

Crash models are usually checked penetration by solver or ansa

I could make a hm template to check penetration follow solver method if this function worked correctly:

using tcl command is also possible, but too long time running...

Altair Forum User

Hallo Prakash and tinh,

Thanks for your answer.

I have checked with ''include self interference'' as well but there are no penetrations (atleast hypermesh say so). i again checked with 'crash module' but still the same message. Attached is the picture.

''try debugging output files to find out node IDs that have penetration'' Sorry, tinh i am not sure how to perform this check. Could you please explain it ?

Prakash, I am also attaching the output file with the most encountered error for me.

In some simulations, i have experienced a huge difference in results from nlquasistat and nlgeom cases. What to trust more...any views ?

Thanks.

Palash

<?xml version="1.0" encoding="UTF-8"?>

Unable to find an attachment - read this blog

tinh

I means some solver (like lsdyna) write out node ids that have initial penetration, but i am not sure with radioss/optistruct. let's wait for admin answer

what kind of interface are you using?

Altair Forum User

No, no IDs will be written in the out file unless the penetration depth is too large.

Regarding the Istop= -2 error, can you check the element quality and also the contact interface stiffness?

Altair Forum User

Prakash,

regarding element quality, i have a 3-d mesh with quad elements. and i can see thet some elements are failing in Jacobian, aspect and skew. There are more then 450000 elements. Is there any work around to ignore this check beacuse it is not easy to repair each and every element.

regarding contact interface stifness, i am not sure how should i check that....can you please explain ?

P.S : i have already activated PARAM - CHECKEL - NO

Thanks.

Palash

Altair Forum User

what is the effect of penetration on fea result ?

Altair Forum User

what is the effect of penetration and intersection on results ?

Altair Forum User

Hi @ank

Most of the times if there is penetration the problem will not converge for impact, damage, etc...

If at all converges there could be singularities.

So it is always better to remove penetration before you run the analysis.