Discussion Forum

Hi,

I am comfused about how the following PDEs can be implemented in COMSOL 3.5 and 4.0:

Modeling domian: 1D

Dependent variables: u1 and u2

PDEs: for u1 -> d(-K*f(u1,u2)*du1/dx)/dx=S_u1 this is a stationary, coupled with u2, and second-order PDE
with its BCs:
at left point: -K*f(u1,u2)*du1/dx|(x=0)=G_in_t G_in is a fixed value
at right point: u1|(x=1)=p_out p_out is also a fixed value

for u2 -> du2/dt+ d(-K*u2*du1/dx)/dx=S_u2 this is a dynamic, coupled with u1, and first-order PDE
with its BCs:
at left point: -K*u2*du1/dx|(x=0) =G_in_v G_in_v is fixed value

Note:1) K, S_u1 and S_u2 are constant.
2) f(u1, u2) represents a nonlinear relationship function.
3) -K*f(u1,u2)*du1/dx means the mass flux physically
4) -K*u2*du1/dx means the mass flux physically

My solution: I used "general PDE" application mode to model these two PDEs in COMSOL 3.5. As for the u1, it needs two BCs to enclose the problem; As for u2, it needs one BC actually, but in COMSOL it still exists a block to be filled at the right point.

How could I get the successful solving results?

Thank you for your attentation !