Error#15 - CTETRA element ### is distorted (zero or negative Jacobian)

Brian_20847

Hi everyone,

I am doing nonlinear contact with freeze contact but I have errors in the below image (I am using TETRA4) although I used CHECKEL, NO.

I checked the element quality, it doesn't have any errors. 

Thanks.

Ioannis Makris_22013

Hello Brian,

You are in luck, this is a model that I discussed with a colleague of mine a few days ago.

There are multiple things someone should consider before solving this.

1) There are duplicated contacts. I would suggest creating them again and making sure that contacts are correctly defined. One quick way would be to use the Auto Contact option SimLab offers and quick select the bodies that "touch" to create automatically all the necessary contacts. (after deleting all contacts and creating them again I was able to initialize the solution w/o any errors)

2) Continuous Sliding is not the way to go for every contact. This should not provide errors, yet there will be convergence issues later on. Please make sure you know what type of contact you use. Since you have a snap fit simulation, sliding contact is "mandatory" to that region only.

3) You have an enforced constraint and a constraint at the same node. Its like saying simultaneously, "please move downwards" and "stay still". This is also an issue. The correct way would be to have an enforced constraint with a value > 0 in the direction that you wish and equal to 0 for every other direction.

Last but not least, since this query is also in the support line, I would suggest you speaking with the assigned support agent and if you wish you can ask him to include me in a future meeting.

Kindly,

Ioannis Makris

 

Johan_Dahlberg_2011

Hi Biran
Have you selected all bodies when you do your check?
In my image below I have only marked one body when I do the check and thats ok!
if I would have selected all bodies I would have got error on the other body with distorted elements

Kind regards

Johan

Brian_20847

Hi Johan,

I checked for all bodies. It still has these errors.

Thanks

Ioannis Makris_22013

Hello Brian,

In my experience with Non-Linear solutions, this can happen when contacts are incorrectly defined (duplicated surfaces for instance). Could you please share your model for me to look at?

Kindly,

Ioannis Makris

Ioannis Makris_22013

Hello Brian,

You are in luck, this is a model that I discussed with a colleague of mine a few days ago.

There are multiple things someone should consider before solving this.

1) There are duplicated contacts. I would suggest creating them again and making sure that contacts are correctly defined. One quick way would be to use the Auto Contact option SimLab offers and quick select the bodies that "touch" to create automatically all the necessary contacts. (after deleting all contacts and creating them again I was able to initialize the solution w/o any errors)

2) Continuous Sliding is not the way to go for every contact. This should not provide errors, yet there will be convergence issues later on. Please make sure you know what type of contact you use. Since you have a snap fit simulation, sliding contact is "mandatory" to that region only.

3) You have an enforced constraint and a constraint at the same node. Its like saying simultaneously, "please move downwards" and "stay still". This is also an issue. The correct way would be to have an enforced constraint with a value > 0 in the direction that you wish and equal to 0 for every other direction.

Last but not least, since this query is also in the support line, I would suggest you speaking with the assigned support agent and if you wish you can ask him to include me in a future meeting.

Kindly,

Ioannis Makris