Heft,
One trick that I have tried is to set the coeffiecint of thermal expansion to 0 for the part being compressed and then apply a global temperature load to the model. In this situation, only the copper insert will be heated and will expand. The glass (cte = 0) essentially does not see an...