Boundaries conditions to pump characterization

stephane.jorand · September 15, 2014, 06:37

Hello,

I have to characterize a centrifugal pump by CFD but I don't have any boundaries conditions.
I just know the pump rotation is between 500 and 2000 rpm.
Is it right to set total pressure = 0 at inlet and static pressure = 0 at outlet and then the solver give me the massflow and the pressure developed by the pump?

ghorrocks · September 15, 2014, 07:43

No, this does not sound right.

Pumps operate on a pump curve which relates flow rate to pressure difference. Where on the curve you sit will depend on the pump, the motor power driving it and the system it is connected to.

So doing a single simulation at essentially no pressure difference sounds pretty meaningless. You need to load the pump up with either a flow rate or pressure difference representative of the system it will be running in, and that will give you a single point on the pressure versus flow rate curve for the pump.

September 15, 2014, 06:37	Boundaries conditions to pump characterization	#1
stephane.jorand New Member SteJor Join Date: Jun 2013 Posts: 1 Rep Power: 0	Hello, I have to characterize a centrifugal pump by CFD but I don't have any boundaries conditions. I just know the pump rotation is between 500 and 2000 rpm. Is it right to set total pressure = 0 at inlet and static pressure = 0 at outlet and then the solver give me the massflow and the pressure developed by the pump?

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Linux commands to define boundary conditions at velocity-inlet & wall boundaries	saayo3	Fluent UDF and Scheme Programming	1	June 29, 2013 12:33
Water subcooled boiling	Attesz	CFX	7	January 5, 2013 03:32
boundaries conditions for an oil -gas separator	juliom	CFX	4	May 15, 2011 07:11
Wall Heat Flux Boundaries conditions	Abigail	FLUENT	2	July 24, 2007 09:34
periodic boundaries - flow through a net	PK	FLUENT	0	July 12, 2007 11:58

September 15, 2014, 07:43		#2
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,703 Rep Power: 143	No, this does not sound right. Pumps operate on a pump curve which relates flow rate to pressure difference. Where on the curve you sit will depend on the pump, the motor power driving it and the system it is connected to. So doing a single simulation at essentially no pressure difference sounds pretty meaningless. You need to load the pump up with either a flow rate or pressure difference representative of the system it will be running in, and that will give you a single point on the pressure versus flow rate curve for the pump.