element deletion 1

[Jason kamrani]

Hello,

I have a question regarding element deletion and would appreciate any comments.
I wrote a vusdlfd code and defined a damage model in a way that after some level of deformation, stiffness of the material reduces to almost zero. Also, in order to delete the elements once some criteria is met, I used "*depvar, delete = no. of a state variable that is either one or zero". Once I ran the model I saw that right after a few elements are removed simulation diverges giving the "highly distorted elements" error. In order to figure out the problem first I excluded the reduction of the elastic property and just kept the element deletion part. This is what happens, once the criteria for element deletion is met in an element, that element is not removed but is shrieked to become a flat element (elements are initially hexahedral). Sample after this looks like as if it has a hole in it. Has anyone faced such a case or has any suggestions? I would appreciate any help.
The attached picture, is the half of a dog-bone sample and the deleted elements look like a hole while there are just a few elements removed.

Thanks,

 

[Mustaine3]

I'm not familiar with element removal through subroutines, but maybe it is just a visualization issue. A/CAE uses the STATUS variable to know what elements are alive or dead and can hide or show them. Have you requested STATUS as field output variable to get it written to .odb?

 

[Jason kamrani]

Hi. thanks for the comment.
Yes, I have.

 

[FEA way]

There might be a problem with the implementation of element deletion in your model using subroutine. Check this thread, it may help you:

https://www.eng-tips.com/viewthread.cfm?qid=462351

Also take a look at the documentation chapter "User-defined mechanical material behavior".

 

[Jason kamrani]

thanks.
this is actually the thread that I got the idea of element deletion through vusdfld from. It seems that it is working for them but is not for me and I wonder why.