Model Penetrates Master Surfaces during simulation

ARUNS_21340

Hi,

 

I am trying to model deformation of foam in Optistruct (Non linear quasi static, NLSTAT subcase) by imposing displacement of top plate through the foam and constraining the bottom plate.

 

Though my model is converging, you can see the foam penetrates the bottom plate (green).

 

Three contacts have been defined in the model, 1) Top plate (Prop) and Foam (Nodes) with PCONT/PCONTX7  2) Side plate (Prop) and Foam (Prop) with Static Fric and 3) Bottom plate (Prop) and Foam (Prop) with Static Fric. (Master and Slave order being followed).

 

Can you suggest a way to overcome this problem?

 

Thanks,

Altair Forum User

Hi Arun,

 

Sorry for a late reply,

 

Could you please share the model file with us?

 

Please use the file transfer link in my signature to share the model file

Altair Forum User

Hi Arun,

 

I got the files. I will check and update you soon. 

ARUNS_21340

Hi Prakash,

 

Thanks

 

Best,

Arun

Altair Forum User

Hi @Arun S

 

I think the problem is coming from the way the contacts are defined.

 

I see that you have selected all foam nodes for all contacts and this is diverging the results.

 

Try this:

 

Only select those nodes which come in contact with the bottom, side and top faces of solid elements and check if you see any improvement in results. 

ARUNS_21340

Sure. I will keep you posted on the results

 

Thanks,

Arun

ARUNS_21340

Hi Prakash,

 

I did include only the nodes which contact the bottom,top and side plates, but ended up with the same results after the simulations.

 

These additional nodes displaced during compression, reaches the bottom plate and penetrates through irrespective of the contact definition. Any thoughts?

 

Regards,

Arun

 

 

Altair Forum User

Hi Arun,

 

Can you share the updated model files?

ARUNS_21340

Just sent the files.

 

Best,

Arun

Altair Forum User

Hi Arun,

 

I tried with freeze and the runs fine with Freeze.

 

I would suggest to try with slide contact without friction and then you add friction based on the response. 

 

Also, it looks like a large strain problem, so the results from NLSTAT may not be right. Use NLSTAT with LGDISP to run large strain non linear problems. 

 

 

ARUNS_21340

Hi Prakash,

 

Thanks for the suggestion. I would definitely run the model with LGDISP once I have the prelim results.

 

I didn't quite get on how the model runs with Freeze definition, because I gave it a shot earlier. Can you please share the model / contact definitions with the sets?

 

Regards,

Arun

Altair Forum User

Hi Arun,

 

I used the same model which you shared but changed the contact type. 

 

Which version of HyperWorks are you using?

ARUNS_21340

Hi Prakash,

 

I am using HyperWorks 14.0.

 

Thanks

Altair Forum User

Hi Arun,

 

I have updated the model and shared the same with you.

 

Can you try that model and check if that is working as intended?

 

ARUNS_21340

Sure. I will keep you posted.

 

Thanks