FATAL MESSAGE 4291 (EQD4S) when using MAT8 for composite

[am1234567]

Hi,

I have a problem when running Nastran SOL 103 for normal modes analysis which I believe happened when I used MAT8. Some problem arises saying "*** USER FATAL MESSAGE 4291 (EQD4S) FOR ELEMENT WITH ID =1 THE MATERIAL ROUTINE -MAT- RETURNS A 2X2 J-MATRIX FOR MATERIAL ID =1 WHICH IS SINGULAR."

The properties included are Ex (40776.576MPa), Ey(40776.576MPa),Poisson's ratio(0.436), shear modulus(14197.9696MPa), and density(1.535e-9tonne/mm^3) accordingly. Also, I attached the file for reference.

I know that orthotopic material has 3 shear moduli, but in the 2D MAT8 case based on my understanding it only requires 1 shear moduli such as G12. Also if I put 3 shear modulus it turns out to work just fine. But I don't have the values for other directions. I only have G12.

I read from this page

https://knowledge.autodesk.com/supp...re/caas/CloudHelp/cloudhelp/2019/ENU/NSTRN...

saying if test data is not available to accurately determine G1Z and G2Z, an approximate value is the in-plane shear modulus G12 which is used by default when PARAM, SHELLTVSMATTYPE is set to FLEXIBLE. When PARAM, SHELLTVSMATTYPE is set to RIGID, G1Z, and G2Z will be penalty values which approximate a rigid transverse shear stiffness. I tried but it doesn't work. Could be I use it wrong.

If possible could someone have any idea on how to solve this please let me know.

Im using virtual lab 13.10 to create the input file and MSC Nastran 2018 for the processing.

Thank you.

 

[nlgyro]

I'm confused by your post...

You say you're using MSC.Nastran 2018 as a solver but you quote articles from Autodesk nastran!

These are not the same two products although they carry the nastran name.

Param's that you have quoted are not available in MSC.Nastran.

I'm assuming that you are using MSC.Nastran as your input file is of a MSC.Nastran format.

You're using PSHELL with MAT8.
Well, the nastran QRG for the pshell entry (remark 9) clearly states and I quote:

"....If MID3 references a MAT8 entry, then G1Z and G2Z must
not be zero."

So no wonder the model does not run.

In the absence of G1Z and G2Z, the best bet is the use a MAT8/PCOMP (I'm again assuming you know the lay-up to generate the pcomp) combination with blank fields for G1Z and G2Z in your MAT8 card.

The resulting ABD matrices that nastran internally generates would reflect the true in-plane stiffness for the laminate.

The absence of G1Z and G2Z would ensure that the [G] matrix is not generated making the laminate infinitely stiff in transverse shear. Keep in mind the presence of G1Z and G2Z on the MAT8 card would generate a [G] matrix thereby adding transverse shear flexibility to the laminate.

You'll have the decide if this suits your modeling assumptions. If not you'll have to make an effort to get the out-of-plane shear stiffness values to model your problem.

 

[am1234567]

hi @nlgyro

Thank you very much for your help. I don't know why I thought like that.
You pointed out the most important things for me. Now the solution had worked just fine.

Thank you again. I really appreciate it.