I have a fast surface reaction, and the reaction rate should be determined by mass diffusion from gas phase. Currently I'm using the method of setting preexponent factor to a large value. I can set it until to 10E26, but if large than this, Fluent will apear the error of "Gauss matrix singular". Even at this large preexponent value, the surface concentration of reactant is not close to zero. Is there any good way to prevent this situation? Thanks

I am not an expert in the subjet, but if you reaction is infinatly fast, then will the concentration of reactant not be 0 at the reaction surface. If so why are you not simply specifying this as a Boundary condition.

You do not necessarily have to specifiy an extremely high value like 10e26 - you just want to specify a value sufficient that its mmuch higher than the diffusion rate.

I find a value around 1e10 is sufficient for my problems. I would imagine in your case a value less than 1e15 would be OK, but you should check. If its too high you affect the solver and it won't solve (as you found out).

The concentration at the surface will never be zero, but I guess it should decrease as the pre-exponent factor increases. What values are you getting? Remeber the surface rate should in fact balance the diffusion rate - so the surface concentration will depend on your arrhenius factors and the boudnary layer regime - turbulent, laminar etc.

Greg