Why do I see a difference in total pressure on either side of a frozen rotor domain interface in CFX 10.0?
I've been doing CFX calculations for an axial pump impeller with a diffuser, connected by frozen rotor domain interfaces. I got a well converged solution (rms-residuals < 10^-4). The results look good however when I calculate the massflow average of the total pressure (in stationary frame) at the frozen rotor domain interfaces, there is a significant difference in the calculated value for side 1 and side 2 of the interface. The difference is more than 3%; Int Imp Diff. The interfaces are coinciding and and there is a 100% overlap. What is the explanation for this difference?
|
There could be several reasons for the difference, with the most likely being the mesh on either side of the interface. You should check that on either side of the interface the meshes are simliar in density both circumferentially and hub to shroud. Also the distribution of nodes should be similar in these directions. |
||
|