Hello there,
I am looking for a way to deactivate elements when a limit criterion is reached. I still need to find out the limit criterion, but currently I think that it will be either a certain stress or strain.
It would be elegant to exclude the element directly from the stiffness matrix, but if this is not possible, an alternative would be to set the stiffness to zero (or to something very small, so that the solver can still solve the matrix).
Thank you!
Such functionality is not implemented in CalculiX but you could use the UMAT subroutine to zero the stiffness of failed elements.
Here’s a previous discussion about this: Element Removal in CalculiX for Fracture
Hi,
Have youy explore the card *MODEL CHANGE,TYPE=ELEMENT, REMOVE ?
Manual v2.19 part 7.85 *MODEL CHANGE
I’m afraid that it won’t work in this case because it only removes elements that are initially defined by a user. Here the goal is to deactivate/delete elements dynamically, based on a failure criterion.
Hello and thanks for the suggestions.
*MODEL CHANGE doesn’t seem to be meant for my case, because it activates contact elements. But I want to disable solid elements.
I have had a look at the User Material help (http://web.mit.edu/calculix_v2.7/CalculiX/ccx_2.7/doc/ccx/node255.html as well as the chapter 8.5 linked there), but can’t really figure out how to define one. Could someone give me a recommendation on how best to get started with this?
Manual v2.19 part 7.85 *MODEL CHANG descrives also elements removal.??¿
*MODEL CHANGE TYPE=ELEMENT, REMOVE (or ADD)
3,4,5 (List of elements and/or sets of elements to be activated or deactivated)
Remove the elements 3,4,5 at the beggining of the STEP
That is interesting. I still have CCX2.16. It says that *MODEL CHANGE only works for contact elements. In the help of 2.19 it actually says that all elements can be deactivated. Thanks for the correction.
That would really simplify the whole thing. However, there is still the challenge to combine all elements for which the error criterion (e.g. a certain elongation) to one set.
I don’t know of any command that can query the results of the last calculation step while solving.
Could you write a UMAT that calls the *MODEL CHANGE or something like that?