abaqus, cel, coupled eulerian lagrangian, fluid flow

I am trying to model a tube which is filled with liquid that is being crushed.  I am using Coupled Eulerian Lagrangian to model the problem in abaqus, and there are a few distortions/errors in the model.  I am using symmetry to model only quarter of the model.  The liquid actually comes out from the sides of the tubes as the two attached pictures show.  

The eulerian mesh is really fine with nearly 1.5 million elements.  I was curious whether i'm missing an important boundary condition for applying in the eulerian region, or some sort of mesh control.  Any help or suggestions would be greatly appreciated.

Sincerely,
Ledjan