Problem in modeling solidification

S. Balasubramanyam

Hi, I am trying to solve a simple 2d solidification problem in a rectangular domain. There is no inlet or outlet, it is a closed domain. Top and bottom walls are adiabatic. Rihght side boundary is cooler than solidus temperature. Left side boundary is hotter than liquidus temperature. I am trying to find out a transient solution. After 3 iterations I am getting a furious solution in which the minimum temperature falls below the cool wall. I have tried reducing the time step and using underrelaxation factor for conservative mass fraction, it didn't work. Can any one kindly help me?? Thanks in advance. Regards, balu.