laminar viscous flow in a rectangular duct

Bastian

Hi everybody- I am simulating a laminar flow in a rectangular duct under gravity(eulerian model). When I simulate the flow with a low viscous newtonian fluid like water there is no problem at all. In the same simulation with a higher viscous fluid like engine-oil there occurs an air layer under the fluid. In the eulerian model you can't specify friction at a wall, you can specify a value for the shear stress. When I specify the shear stress to 0 there is no air- but is this the right way to calculate the flow??? I would be very happy about every new idea to deal with this problem- thanks Bastian