Discussion Forum

Dear all, I am making a simulation of a superconductor in the magnetic formulation module of COMSOL. I have coupled the problem with the heat transfer in solids. I set the current density according to (0<T<Tc)(alpha(1-(T/Tc)^2)^1.5)+(T>=Tc)*(0). This should set the current density to be 0 when the T exceeds the critical temperature of the superconductor, Tc. I am using an EJ Power relationship for the resistivity in parallel with a normal resistor, so the resistance is defined by it when the critical temperature is exceeded. However, when using certain values of current density sometimes the model starts to have problems to converge due to it being a 3D model and due to memory limitation I can't make the mesh fine enough to avoid this problem. However as the current concentrates in a point while the problem converges, there is too much heat generation and it throws me an error saying that it tried to evaluate the power of negative number to a non integral number in the expression provided. Therefore, I would like to know if I have written the conditions appropiately as it seems to me that the way I have written the condition should avoid evaluating T for values above Tc.