Posted by Precise Simulation on
URL: http://forum.featool.com/Time-dependent-solution-tp1598p1605.html

Thank you for attaching a model file. With the default settings, the stationary solver does not converge likely due to the nonlinear radiative term ~T^4, so the solution can not be expected to be correct.

One can try to tune the solver setting by decreasing the nonlinear relaxation parameter and increasing the number of iterations. Or slowly increasing the nonlinear term and using the previously computed solution as initial guess. Also please check if all you boundary coefficients are correct, as some have the coefficient 5E-8 and others 1.5E-8.