I am trying to solve a problem of thermo-elasto-plasticity. For this aspect I have used the following code by tianjuxie:
His problem also involves the moving of the heat source as well as phase change of material from powder to liquid to solid since he models the laser bed powder fusion process.
I have modified his script by making sure the phase change becomes irrelevant (always solid material properties are used and max temp does not go beyond melt temp) and the heat source does not move. This way my problem is only a problem of thermo-elasto-plasticity.
My code can be found here:
Now, naturally I have a different set of boundary conditions and material properties because I have a test case from Abaqus that I can compare with. My main issue is with convergence, I cannot use a larger time step if my Young’s modulus is too large because this makes the matrix highly ill conditioned. This affects the nonlinear iterations in the mechanical part of the problem because if you use too large a time step, the right hand side can significantly influence the residuum and thus the new right hand side. Thus, the residuum never converges to zero.
Convergence is more likely if the Young Modulus/stiffness is reduced. Nevertheless, tianjuxue’s code runs smoothly and I was wondering if anybody could have a look at my code and see what I could check. So far, I have checked the degree of functions, boundary conditions, plasticity algorithm but nothing has worked. I have also tried different preconditioners and both direct and iterative problem. The thermal part of the problem runs fine.