COSMOSWorks help - singular matrix error 2

[mfe02]

Hi,
I'm trying to run a study in COSMOSWorks on an assembly and I get an error which says "Singlar Matrix Error. Restraints may not be sufficient"

I cannot figure out what is causing the error. Everything is contrained properly and no parts of the assembly can move. I have tried multiple combinations of restraints and I keep getting the error.

Any help is much appreciated. Thanks.

 

[corus]

In some cases a body may be restrained in 3 translational directions but perhaps not sufficiently to prevent the body from rotating. In some cases you have to imagine if you can rotate a part about any of the 3 axis, as well as if any part can be translated in any of the 3 global directions.

Otherwise a singularity in the stiffness matrix means you have a zero in the diagonal. If it's not restraints then it might be zero stiffness from material property definition.

corus

 

[mfe02]

Thanks corus. I understand what you are saying. I think it is definitely an issue with the stiffness matrix. While running the study, it stops everytime at 95% of the "decomposing stiffness matix" stage. I am using Plain Carbon steel from the Solidworks library. I do not know if is not properly defined or if somehow a property was accidentally changed.

Should I try a different material? Or, do you think it could be the case where the restraints are not sufficient and allow rotation? I am pretty sure the restraints I have do not allow rotation, but I could be wrong - wouldn't be the first time.

Thanks again, I appreciate the response.

 

[corus]

If you've picked a standard material then there shouldn't be a problem so it sounds like restraints, unless your dimensions are in m and you've chosen N/mm^2 for E then you'll be a million out. Also, if you're relying on contact to give restraint then that might be a problem if the parts aren't touching.

corus

 

[cbrn]

Hi,
Corus hit a good point when he pointed out the issue of the contacts: be careful that SW's assembly "constraints" are meaningless in FEM: even if your parts are geometrically fully restrained, if you don't check for correct set up of contacts in Cosmos then you probably will get stuck with rigid body motion. As contacts are resolved by the end of the solution preconditioning, this might explain why you get the error at 90% or more of the preconditioning.

Regards

 

[mfe02]

Ok, I finally figured out what what going on. There were some minor interferences between some bearings, the holes they were in, and the shafts going through them (for a scissor lift). I had it as the exact sizes we will be using, so the bearing is 5/1000th larger than the hole and the shaft has a similar tolerance. I set everything to "ideal" sizes (no tolerances) and everything worked great. These interferences were causing my contact set to not work properly and resulted in the error I was getting.

Thanks to both of you for the helpful posts which led me to figure out what was going wrong - much appreciated!!!