CFD Online Discussion Forums - Problem with centrifugal pump calculation

The idea is to create a predicted pump curve which gives information about the maximum pressure at a fixed flow and a fixed inlet pressure. Based on this information I like to calculate the pump power (mechanic and hydraulic), and the efficiency.

I use Floworks to perform the analysis. The medium is water (pre defined in floworks). First I like to set up a basic calculation method, and then I can analyze some pumps and compare results tot actual test data.

For calculating the maximum outlet pressure at fixed flow and inlet pressure I set the following boundary conditions:
-At inlet a total pressure of 1 bar
-At oulet a outlet volume flow of 2 m3/sec
-The impeller walls which are in contact with the fluid are defined as ideal wall
-The pump casing and the inlet pipe are defined as real wall

I have set the following goals:
-Bulk av static pressure at inlet and outlet
-Equation goal to calculate the pressure drop (used for convergence)
-Inlet volume flow (used for convergence)
-Outlet volume flow (used for convergence)

When I run the calculation I can’t get a good result. The most strange is that the inlet flow is not the same as the outlet flow. During the calculation the inlet flow varies from -8 m3/sec up to 12 m3/sec. After 400+ iterations the inlet flow keeps bouncing up and down. This indicates there is something wrong. The inlet and outlet flow should be exactly the same.

When I use the same model to solve the volume flow at fixed inlet and outlet pressure, there is no problem. Also when I solve the inlet pressure at fixed outlet pressure and fixed inlet volume flow, there is no problem.

Does anyone has an idea what is going wrong over here?