CFD Online Discussion Forums

CFD Online Discussion Forums (
-   Main CFD Forum (
-   -   Centrifugal pump CFD problem...pls help (

Peter June 24, 2008 02:53

Centrifugal pump CFD problem...pls help
Hi, I'm trying to simulate a centrifugal pump with 0.5m3/s, 2600RPM, and 15bar pressure rise.My boundary conditions are: Inlet:mass flow inlet, Outlet: 15bar static pressure. But I'm not getting the right pressure at the inlet which is supposed to be 1bar instead of same pressure as outlet. Am I putting the right boundary condition at the outlet or inlet? The fluid is water.

Thanks for any help

Balduin Bankerotti June 24, 2008 02:57

Re: Centrifugal pump CFD problem...pls help
what software, which models for the rotating part, ... ?

Peter June 24, 2008 03:40

Re: Centrifugal pump CFD problem...pls help
Hi, I'm using Cosmosfloworks. i set the impeller to be rotating with a rotating volume around it. Multiple reference frame. I'm not sure if my input are not realistic


Ananda Himansu June 24, 2008 12:10

Re: Centrifugal pump CFD problem...pls help
The pressure by itself may not be indicative of pump performance. If there is a converging nozzle at the inlet, and if you have placed your outlet boundary before the diffuser and deswirler, then the inlet and outlet pressures may be of the same order.

Check that the steady mass flow across the outlet matches that across the inlet. Check the mass-averaged specific stagnation enthalpy at the inlet and the outlet. Or you can just check the change in the specific "Bernouilli energy" p/rho + 0.5*v^2, which is effectively the change in the stagnation enthalpy for water, since the temperature does not change significantly. Do not forget to count the total velocity including the swirl. If you measure velocities in a frame in which the impeller blades (and thus forces) are moving, then the increase in the stagnation enthalpy between pump inlet and outlet must equal the specific work done on the water by the impeller. You can calculate the latter using the Euler pump equation, since you know the angular velocity of the impeller, and you know the mean radii at which the water enters the impeller and leaves the impeller. Look in some book on elementary fluid mechanics (such as Munson, Young and Okishii) for an example of applying the Euler pump equation (which is the same as the Euler turbine equation in reverse).

If the stagnation enthalpy of the fluid in the absolute frame is not being increased by the right amount between inlet and outlet, then you have a stalled pump or a boundary condition set in the wrong reference frame or an unconverged steady state or some such.

Ananda Himansu June 25, 2008 11:39

Re: Centrifugal pump CFD problem...pls help
My last post was rather wordy, which seems to be a pattern. hmmm.

Anyway, it boils down to checking mass conservation between inlet and outlet, and checking energy conservation. For the latter, as I said, check the increase in stagnation enthalpy between inlet and outlet. As a first check, compare this with what you wish to happen, namely (15 atm - 1 atm)/density in the appropriate units. Later you can check the Euler pump equation. Since you are dealing with a liquid, perhaps the code defaults to accounting for an earth surface gravity vector in the -z direction. So see if you need to include a gz term in your stagnation enthalpy, i.e. if your outlet is at the end of a tall delivery pipe. Check the coordinate system convention used by the flow code: is your impeller rotation specified with the wrong sign?

All times are GMT -4. The time now is 18:21.