Discussion Forum

Hi all.

To communicate my problem I made a simple bimetallic strip (library Materials Values for Copper and Iron) which is fixed to a surface where a temperature T is introduced - no thermal dissipation is assumed here. So as I change the applied T, the bimetallic strip bends one way or another depending on the applied temperature and the CTEs of the metals.

Fine - easy right? My understanding fades, however, when I notice the relative positions of the vector drawing depicting the "original" (T = T(ref)) condition vs the colorized 'Stress (solid)' plot depicting the strained condition of the strip --- these drawings don't seem to progress as a function of the specified temperature, rather they just "jump" to some final value. See images below to understand what I mean: images are for temperatures 500K 1000K 3000K and 294K and T(ref) From Multiphysics>Thermal Expansion is 293.15K . In reality, that strip should be bending continuously depending on what the temperature of the assembly is - but what I observe in the model is a physical bending of the object which "snaps" to one degree of bending or the other, even if deviation from T(ref) is a fraction of a kelvin. This is clearly not correct.

So, am I missing a setting in 'Solid Mechanics' or something so that this bending be described properly? Surely there is a way to accurately report these expansion values. The von Mises Strain gives numbers that are reasonable -- but the bending trend itself seems wrong.

Thank you for the help!