Discussion Forum

Hello All,

Thank you in advance for reading this. I have an issue with a solution of a 2D axisymmetric model with coupled heat transfer and solid mechanics effects. A relatively thin cylinder (like a pie chart) is heated with a laser up to temperatures on the order of 2,000 K in a thermo-shock heating profile. COMSOL seems to resolve the heat transfer effects correctly, however I am obtaining strange results for the principal stresses. I have tried changing the mesh and the heating rate, however the issue still persists. It seems that strange structures are forming for different thicknesses. In the attached image one can see the results for thicknesses of 2.05 mm and 14 mm. My question is whether these anomalies are caused by an improper resolution of the physics, or the spatial resolution (mesh), or the temporal resolution (using smaller time steps vs. heating rate), or could it be a solver issue? In the 14 mm case, lines which go through the nodal locations of my mesh are formed, which is why i thought that I had to refine it, but after I did it did not fix the issue. All comments and ideas are welcome.