Solver problem numerical singularity when tying surfaces together

[tito662]

one face of a part instance meshed with S8R5(8 node doubly curved thin shell, reduced integration, 5 dof per node) is tied(tie surface) to a face of a part meshed with C3D8R elements.
During the calculation, i have this warning :
SOLVER PROBLEM. NUMERICAL SINGULARITY WHEN PROCESSING NODE
RAIDISSEUR-1.24587 D.O.F. 6 RATIO = 3.21587E+12.
(Raidisseur is the part meshed with S8R5)

How should i resolve this problem? should i use S8R6?

in fact, the results i m getting are a little strange ...is this non accuracy of results due to the singularity?
thank u

 

[bassmanjax]

Are you using a mesh tie (*TIE) or a shell to solid coupling (*SHELL TO SOLID COUPLING)?

Martin

Martin Stokes CEng MIMechE

 

[tito662]

I m using tie
When i changed the shell elements to S8R6 (6 degrees of freedom for the shell), i dont have this warning anymore.
Do u advice me to use shell to solid coupling?

 

[bassmanjax]

I've never really used shell to solid coupling, but I would guess that it would be the preferred method for transitioning a mesh from shells to solids.

Give it a try and compare the two for stress/displacement continuity across the interface.

Martin

Martin Stokes CEng MIMechE