Cyclic Loading of Rubber Part - Standard or Explicit?

[mizzjoey]

Hi everyone.

My group is trying to model cyclic loading of a rubber mount installed between two steel plates. Due to time constraint, we want to extablish a ref model that will run in the shortest time possible (we have several runs to go through this week). If we run in standard, the job does not solve. Running in explicit brings another issue (on top of the longer run time): artificial energy buildup.

We cannot thoroughly get rid of AE at the end of each step, thus AE from previous step will be carried on to subsequent steps. By the end of the run, the AE looks high enough to make us doubt if the results are accurate.

We are running the loading in several steps.
a- clamp the parts to working height
b- preload
c- load from top
d- normalize
e- load from bottom
f- normalize

Is there another way to do this so that we can get the model to solve in standard (preferably) or run in explicit but sidestep the AE problem?

thanks in advance,
jo

 

[bassmanjax]

Should be able to do that in Standard I reckon..

You might need to provide a bit more info though;

- Why doesn't the job run to completion in Standard? What is the error message?
- What material model are you using for the rubber?
- Is the rubber joined to the plates, or is there contact between the plates and the rubber?

Regards

Martin Stokes CEng MIMechE

 

[mizzjoey]

Thanks for your reply! Very timely.

We have two different types of rubber. One is neo-hookean and another one is defined by a stress-train curve. Both rubber parts have metal inserts in them. We use Tie to model the bond interface. Rubber parts and inserts are deformable (have added *plastic to the steel inserts). The top and bottom steel plates are analytical rigids.

The problem is the excessive deformation of the rubber when clamping the parts to working height. Because the lower mount is pushing up against the upper mount, at the Tie interface of the upper mount the rubber tears away from the steel surface. This causes Standard to quit the run even at the 1st step. Explicit allows this though.

Another thing is we have to maintain the distance between the two clamping steel plates at a certain distance (simulate bolt condition). One more reason why we want to use Standard: so we can utilize the *Gap function.

Any suggestions? Thank you very much..

jo