Hi.  I'm struggling to get realistic results when using a bolted connection in a static study that incorporates a temperature load.

Essentially I'm attempting to bolt two metal plates to a metal block, which all have a temperature load of 150C applied to them.  I have uploaded a simplified model which demonstrates the problem that I'm having.  I've run the analysis with no temperature load and everything looks fine, however as soon as I introduce temperature the stresses go through the roof!  Any help or advice would be much appreciated.

Rob.