[Hazem_9]

I have made a model for centrifugal pump of impeller speed 2960 rpm using CFX. Inlet and impeller domains meshes are done using Turbo-Grid, while Volute mesh is done using Ansys mesher. I defined the impeller as a rotating domain of speed 2960 rpm, while Inlet and Volute as stationary domains. Boundary condition: flow rate at inlet, static pressure at discharge and all other faces are no slip walls.
The model residuals converges, results are excellent and head performance curve is very near to the experimental curve.

But the problem is that results of shaft power in simulation is half the value of the experimental data ... consequently, the efficiency calculation was overestimated

Power calculated in CFD-post using the following equation
Power = Torque * Rotational velocity = Number of Blades * Blade torque * Rotational velocity

These are the exact expression written in CFD-Post:

• tBladeRow = torque_z()@Impeller Blade
• tAllBlades = tBladeRow * BladesNumber
• Power = tAllBlades * Absolute Omega / 1 [rad]

__________________________________________________ ______________
Moreover, I tried to solve the same pump model at different speeds. I found that:

• At speed= 1500 rpm --> Shaft power results in simulation half the experimental data value
• At speed= 2960 rpm --> Shaft power results in simulation is half the experimental data value
• At speed= 6000 rpm --> Shaft power results in simulation is half the experimental data value

I want to know how to solve this issue to get correct estimation for shaft power and efficiency

[Evgeniy_IV]

Dear Hazem_9
When you use expression which you wrote (Power = tAllBlades * Absolute Omega / 1 [rad]) you calculate only hydraulic effective power for impeller but not for pump.If you want to calculation power pump you must take into account power losses in your gap between disc of impeller and wall of volute and also loss of friction.