Power law flow

joel

i am running a simulation using power law fluids instead of newtonian viscosity, my variable n is negative except star doesnt like this. My simulation goes unstable can star not deal with n values less than zero? or do i need to fiddle with something else such as relaxation. Cheers