centrifugal pump CFD - torque

phsieh2005 · December 30, 2015, 08:40

Dear CFX experts,

I simulate a centrifugal pump case using MRF - a transient case. The impeller rotates at 4000 rpm. The calculation completed successfully. I output the torque on the impeller at each time step (torque_z()@impeller). During one revolution, the reported torque_z goes from 0.01 N-m to about -0.01 N-m and increases back to 0.01 N-m. I am wondering why this is the case. Shouldn't torque in Z direction be constant?

Any suggestion?

Thanks!

Pei-Ying

ghorrocks · January 2, 2016, 00:01

Why should the torque be constant?

You are going to have to look at what it is modelling and try to work out why the torque is changing. Then you can think about whether it is realistic or not.

phsieh2005 · January 6, 2016, 07:47

HI, Glenn,

It looked like the torque is almost constant because rpm is constant. the origin and rotational axis is offset from the global coordinate, hence, it is wrong to use torque_z() because it references the global coordinate.

I created a local coordinate, called myCoord that fixes on the MRF axis and rotates with the MRF.

Then, in expression, I define torque_z_myCoord()@impeller. Now, I am getting the torque value at each time step almost constant (very small oscillation).

Hope this will help other people running similar cases.

Pei-Ying

ghorrocks · January 7, 2016, 18:32

Obviously you have to align the torque axis with the axis of rotation. You are going to get useless results if you don't do this.

But it is normal for a rotating machine to have torque variations against angle as blades pass stators, wakes move across blades and other similar things. How large these torque variations are vary depending on the type of machine.

December 30, 2015, 08:40	centrifugal pump CFD - torque	#1
phsieh2005 Senior Member Pei-Ying Hsieh Join Date: Mar 2009 Posts: 334 Rep Power: 18	Dear CFX experts, I simulate a centrifugal pump case using MRF - a transient case. The impeller rotates at 4000 rpm. The calculation completed successfully. I output the torque on the impeller at each time step (torque_z()@impeller). During one revolution, the reported torque_z goes from 0.01 N-m to about -0.01 N-m and increases back to 0.01 N-m. I am wondering why this is the case. Shouldn't torque in Z direction be constant? Any suggestion? Thanks! Pei-Ying

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Analysis of hemolysis in a centrifugal blood pump.	ameen@cfd	OpenFOAM Running, Solving & CFD	1	August 16, 2016 02:57
centrifugal Pump Efficiency	A.farid	Main CFD Forum	0	March 31, 2012 07:39
Centrifugal Pump Cavitation problem or not.	ismael.s	CFX	13	February 27, 2012 08:00
CFD analysis of Centrifugal pump at shut off condition	Nitin Dewangan	CFX	2	February 10, 2012 18:48
Which is better to develop in-house CFD code or to buy a available CFD package.	Tareq Al-shaalan	Main CFD Forum	10	June 12, 1999 23:27

January 2, 2016, 00:01		#2
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,703 Rep Power: 143	Why should the torque be constant? You are going to have to look at what it is modelling and try to work out why the torque is changing. Then you can think about whether it is realistic or not.

January 6, 2016, 07:47		#3
phsieh2005 Senior Member Pei-Ying Hsieh Join Date: Mar 2009 Posts: 334 Rep Power: 18	HI, Glenn, It looked like the torque is almost constant because rpm is constant. the origin and rotational axis is offset from the global coordinate, hence, it is wrong to use torque_z() because it references the global coordinate. I created a local coordinate, called myCoord that fixes on the MRF axis and rotates with the MRF. Then, in expression, I define torque_z_myCoord()@impeller. Now, I am getting the torque value at each time step almost constant (very small oscillation). Hope this will help other people running similar cases. Pei-Ying

January 7, 2016, 18:32		#4
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,703 Rep Power: 143	Obviously you have to align the torque axis with the axis of rotation. You are going to get useless results if you don't do this. But it is normal for a rotating machine to have torque variations against angle as blades pass stators, wakes move across blades and other similar things. How large these torque variations are vary depending on the type of machine.