Discussion Forum

Hey,

I'm trying to solve a system where I have an aqueous phase flowing over a well filled with an oil phase. Flowing aqueous phase will cause convection in the oil phase. I've tried to implement this by having two Laminar Flow physics, one for each phase. If water is defined as phase 1 and oil as phase 2, the velocity vectors are then u1,v1,w1 and u2,v2,w2 for corresponding phases.

Now I've tried to set the boundary condition between the two fluids as the same tangential velocity at the boundary. I tried to use Moving wall boundary condition, by setting the velocity field as x = u2, y = 0 for aqueous and x = u1, y = 0 for oil phase (flow from x-direction only), but the solver did not converge:

Failed to find a solution.
Maximum number of Newton iterations reached.
There was an error message from the linear solver.
The relative error (4e+002) is greater than the relative tolerance.
Returned solution is not converged.

I also got warning during the iteration:

Inconsistent pointwise unidirectional constraints found.
149 constraint force columns are merged/removed.
Iter ErrEst Damping Stepsize #Res #Jac #Sol LinErr LinRes
1 0.039 0.0100000 0.04 2 1 2 3.2 3e-014
Inconsistent pointwise unidirectional constraints found.
149 constraint force columns are merged/removed.
2 0.039 0.0638383 0.041 3 2 4 0.00057 1e-015
More inconsistent pointwise unidirectional constraints found...
3 0.035 0.2262280 0.043 4 3 6 0.00055 1e-015
4 0.022 0.5825226 0.043 5 4 8 0.0031 1.1e-015
etc. etc. until:
25 1.4e-008 0.0000022 2.1e-008 59 25 83 4e+002 1.1e-008
Stationary Solver 1 in Solver 6: Solution time: 30 s
Physical memory: 1.85 GB
Virtual memory: 1.95 GB

However, I could still see the solution, and the velocity field looked similar to the solution obtained with only one Laminar phase -physics for the whole system (with continuity at the boundary between two phases).

I also tried Sliding wall boundary, by setting the velocity of the tangentially moving wall as u2 for aqueous phase and u1 for oil phase. Now the problem was solved but there was basically no convection in the oil phase, even at high flow rates.

Thirdly, I tried to use the tangential velocity, i.e. I set the sliding wall velocity as u2*tx+v2*ty in the aqueous phase and u1*tx+v1*ty in the oil phase, but the effect was the same as with only u2 and u1.

Any help is greatly appreciated.