Problem with Solid/Shell Mesh Composite Sandwich

Julian_21834
Julian_21834 Altair Community Member
edited July 2023 in Community Q&A

Hey guys!

I have a problem with my Sandwich model in Hypermesh (Nastran). I am looking for a solution for quite a while but all other related posts did not help me with my issue.

 

My model is a sandwich structure (HC core with CFRP plies on both sides) and I need the shear stress inside of the HC core. Therefore I am modelling my core as a solid and the plies with PCOMPG attached to the solid core. I checked all the elements for connectivity/dublicates as well as free edges and faces but I cannot find the problem in my mesh. Every time I run the model Nastran gives me the error "USER FATAL MESSAGE 9050 - RUN TERMINATED DUE TO EXCESSIVE PIVOT RATIOS IN MATRIX KLL" and I know that there has to be a problem with either my mesh or my constraints. I fixed the outer face of the model in all dof and the force is attached to the center nodes. (I modelled it with an rbe3 element as well but I wanted to make sure that that's not part of the problem). Therefore I don't think that my model is not properly constrained (I ran a pure shell model as well and that one works just fine).

Here is a picture of my HM model: image

 

image

 

I ran a modal analysis but it did not produce any results since the analysis ran out of memory so I could not check for rigid body movements.

And here is what I get when I run the analysis (101) with param, bailout, -1 (solution of the CFRP sheets looks good, so the picture shows the core by itself):

 

image

 

So I know that it has something to do with the solid core nodes (two different honeycomb structures (red and pink) as well as a carbon insert (blue) but I cannot figure out what's the issue. Most nodes look like they have very little stiffness but are connected (looks like a cone) but some do really weird stuff.

Do you maybe have an idea what I need to check or how I could identify the problem? Maybe you guys even now right away what's the issue.

 

Thank you so much in advance!

 

 

Edit:

I ran the analysis with the hc core as Mat1 (isotropic material with the same properties as before) and everything worked well. So it looks like it is not a mesh problem but rather some solid element / material property thing. Before that I modeled the two different honeycomb solids with MATORT but that always ends up with the Error I described in the beginning.

Which Material Model would be the best to set up all the properties of an aluminium honeycomb in all three directions? 

Tagged:

Answers

Categories