BackUpRoll Contact problem ***WARNING: SOLVER PROBLEM. NUMERICAL SING

[Fabi0]

Hi
i'm trying to simulate the behaviour of a backuproll under radial load, to undestand the forces distribution on the roller (taking into account outer ring deflections and deformation).

I got error like:

***WARNING: SOLVER PROBLEM. NUMERICAL SINGULARITY WHEN PROCESSING NODE RULLO-1-RAD-4.36 D.O.F. 2 RATIO = 5.89045E+13.

They was along 1 2 or 3 dof, so i provided to create bc for every roller, so that the roller can move only in the direction (inner ring point contact) to (outer ring point contact).

After the change, i got always a lot of singularity, but only in the direction 1:

***WARNING: SOLVER PROBLEM. NUMERICAL SINGULARITY WHEN PROCESSING NODE RULLO-1-RAD-4.36 D.O.F. 1 RATIO = 5.89045E+13.

I can't undestand how can i write better boundary condition. It looks like there's some rigid motion, and it could be, because outer ring is deformable.

What could i do to gave abaqus a way to avoid this singularity?

Fabio

 

[Fabi0]

This is a small part of the file

 

[corus]

It's probably due to the contact. You need to define a step in which every part is restrained or has imposed displacements so that contact forces are initialised between the parts. Only then in the 2nd step can you release some of those imposed displacements to allow the true loads.

Tata

 

[Fabi0]

Corus thanks for the reply!

Do u mean that i've to "use more boundary condition" in the new first step? For example, everything except the outer ring is blocked?

Thx again!

 

[corus]

Blocked would be the wrong word. Think of it as pushing everything together so that all the parts make contact with each other. For 'push' use a displacement value. You'll obviously have to impose other 'fixed' zero restraints so that you have something to push against.

Tata