RAM and hyperelastic

[346gnu]

Hi,

I have just run an interesting hyperelastic model, it took 3 days and did not complete. (I can't share the model).

Over the 3 days I periodically monitored the study and watched the RAM get gradually used up (including virtual disk)

eventually it ate the complete 48Gb actual and lots of virtual and completed with a fatal error at step 150 of 201 on pass 2.

Fortunately the info I required happens at about step 140.

Any thoughts?

Creo2.0 M100

13585 bricks and wedges

Thanks

This thread is inactive and closed by the PTC Community Management Team. If you would like to provide a reply and re-open this thread, please notify the moderator and reference the thread. You may also use "Start a topic" button to ask a new question. Please be sure to include what version of the PTC product you are using so another community member knowledgeable about your version may be able to assist.

[gkoch]

Hello Charles,

We have recently received a report that a rather simple model failed LDA contact analysis with hyperelastic materials and development is currently looking into it.

However, in that other case, the analysis failed already after half an hour, so this is very likely something else.

If possible, you should open a case with PTC Technical Support.

I would think, 48 GB should be enough to compute 13585 entities, even if the material contributes to some extra steps during computation.

If you want to track the cause yourself, it is normally best to simplify the model and retry.

If analysis is successful (and significantly faster, not just barely making it), add about half of the details again and try again.

If analysis still fails / needs too long, simplify it more.

Within 3 or 4 iterations you should have a good sense of what is keeping the analysis from completion.

Thanks,

Gunter

[346gnu]

Hi Gunter,

if I open the case I will be asked for the model and I cannot release the information; hence I put the question here.. So this is difficult.

The model geometry is quite simple already. No opportunity for simplification.

There are several contact regions some of which are 'self contact' which simulate initially assumes cannot be valid presumably because the surfaces are too far apart. Even so, all contacts work sensibly.

All measures seem well behaved and values of forces have reasonable correlation with experimental results (a pleasant surprise)

Free body forces sum to zero.

So everything otherwise seems good otherwise

Thanks

Charles

[gkoch]

Hi Charles,

if the geometry is really that simple, can you create a new model which is similar, which you can send?

• If you check the new model and it does not show the same behavior, then the differences between both may give you a clue about what causes the problem. You may then use this knowledge to either modify your test model to reproduce the issue or to change your model to solve the issue (or both!)
• And if it reproduces the issue, it should be fine to send this model instead of the original one.

It is very difficult to diagnose without having the model in front and being able to look at the features/geometry and modify it.

Gunter

[346gnu]

Morning Gunter,

I sent you a PM.

Thanks

Charles