NPSHr 3% in pump simulation

pump_passion · December 29, 2011, 05:59

Hi guys,

I have made several test for estimation of pump head-capacity and NPSHr 3% curves running both steady state and transient simulations.

The results for the head-capacity curve are really satisfactory (with a maximum error of 4% between the tested and simulated results) for all the five cases analized (involving different pump types: 2 end-suction pumps and 3 between bearing pump with suction volute and prerotation).

The real problem is the estimation of NPSHr 3%, since the simulated results are too much optimistic if compared to the tested ones. In general, the simulated NPSHr 3% is in a range of 30% to 70% below the tested one (both running transient and steady state simulations).

I run all the simulations with the Rayleigh Plesset cavitation model setting a mean diameter of 2E-6 [m] and a saturation pressure of 3574 [Pa], typical value of water.

Does anyone have the same issues calculating NPSHr with a 3% head drop?

Has someone found/built a cavitation model that gives more reliable results?

Many thanks for your help.

Regards

Graham81 · January 3, 2012, 03:27

Dear pump_passion (great name by the way),

Instead of answering your question, Id like to ask you one...how great is that?
We use a different software package (Simerics Pumplinx) for all cavitation simulations (including 3% head drop) because I always seem to run into stability issues when running CFX cavitating simulations (especially the ones with a lot of it). Even if I induce cavitation very slowly by sequentially lowering the inlet pressure and run transient it is quite unstable. Before Glenn refers me to the stability FAQ (

), could you give me one of your definition files so I can see where I go wrong maybe?

thanks in advance

ghorrocks · January 8, 2012, 20:29

Here is the stability FAQ at Graham's request

http://www.cfd-online.com/Wiki/Ansys...gence_criteria

The cavitation numbers (diameters, vapour pressure) can be heavily affected by any contaminants in the water. They are based on pure water. Are your experiments on pure water?

Creador28 · January 12, 2012, 13:12

Las simulaciones con CFX dan muy buenos resultados digamos un 4% de error cerca al BEP cuanto mas a la izquierda o derecha te encuentres de ese punto tus resultados pierden presicion.

aerospace84 · August 7, 2014, 05:37

hello to all!! for my job i'm running simulations on cavitation..usually i get the right results, other time, simulations go wrong not going to convergence.
i use fluent to do them...i'm quite curios about pumplinx...does it works well?? also in the determination of Q-H?

thank you!

ghorrocks · August 7, 2014, 07:29

This is the CFX forum. I don't think you will find much about Fluent or pumplinx here.

aerospace84 · August 7, 2014, 09:33

probably yes!
howver yes i'm not in the wrong place working daily with fluent and cfx both for my work. and finding advise and good suggestion here!
i've just seen your quote with a hint about pumplinx...so i've thought it was worth to ask an opinion about it!

December 29, 2011, 05:59	NPSHr 3% in pump simulation	#1
pump_passion New Member NA Join Date: Jul 2009 Posts: 11 Rep Power: 17	Hi guys, I have made several test for estimation of pump head-capacity and NPSHr 3% curves running both steady state and transient simulations. The results for the head-capacity curve are really satisfactory (with a maximum error of 4% between the tested and simulated results) for all the five cases analized (involving different pump types: 2 end-suction pumps and 3 between bearing pump with suction volute and prerotation). The real problem is the estimation of NPSHr 3%, since the simulated results are too much optimistic if compared to the tested ones. In general, the simulated NPSHr 3% is in a range of 30% to 70% below the tested one (both running transient and steady state simulations). I run all the simulations with the Rayleigh Plesset cavitation model setting a mean diameter of 2E-6 [m] and a saturation pressure of 3574 [Pa], typical value of water. Does anyone have the same issues calculating NPSHr with a 3% head drop? Has someone found/built a cavitation model that gives more reliable results? Many thanks for your help. Regards

January 3, 2012, 03:27		#2
Graham81 Member Join Date: Dec 2009 Posts: 57 Rep Power: 16	Dear pump_passion (great name by the way), Instead of answering your question, Id like to ask you one...how great is that? We use a different software package (Simerics Pumplinx) for all cavitation simulations (including 3% head drop) because I always seem to run into stability issues when running CFX cavitating simulations (especially the ones with a lot of it). Even if I induce cavitation very slowly by sequentially lowering the inlet pressure and run transient it is quite unstable. Before Glenn refers me to the stability FAQ (), could you give me one of your definition files so I can see where I go wrong maybe? thanks in advance

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
How many prism layers should be added in a pump simulation?	Jasmine	CFX	5	April 3, 2020 20:58
Centrifugal Pump Simulation Problem!	warex	FloEFD, FloWorks & FloTHERM	29	September 23, 2014 11:27
Positive displacement pump simulation	jsm	STAR-CCM+	0	October 5, 2009 04:14
A pump paradox	bearcat	Main CFD Forum	7	July 14, 2009 21:02
3-D Contaminant Dispersal Simulation	Apple L S Chan	Main CFD Forum	1	December 23, 1998 11:06

January 8, 2012, 20:29		#3
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,854 Rep Power: 144	Here is the stability FAQ at Graham's request http://www.cfd-online.com/Wiki/Ansys...gence_criteria The cavitation numbers (diameters, vapour pressure) can be heavily affected by any contaminants in the water. They are based on pure water. Are your experiments on pure water?

January 12, 2012, 13:12		#4
Creador28 New Member Victor Anicama Join Date: Jul 2009 Location: Lima - Perú Posts: 1 Rep Power: 0	Las simulaciones con CFX dan muy buenos resultados digamos un 4% de error cerca al BEP cuanto mas a la izquierda o derecha te encuentres de ese punto tus resultados pierden presicion.

August 7, 2014, 05:37		#5
aerospace84 New Member Luke Join Date: Nov 2012 Location: italy Posts: 24 Rep Power: 14	hello to all!! for my job i'm running simulations on cavitation..usually i get the right results, other time, simulations go wrong not going to convergence. i use fluent to do them...i'm quite curios about pumplinx...does it works well?? also in the determination of Q-H? thank you!

August 7, 2014, 07:29		#6
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,854 Rep Power: 144	This is the CFX forum. I don't think you will find much about Fluent or pumplinx here.

August 7, 2014, 09:33		#7
aerospace84 New Member Luke Join Date: Nov 2012 Location: italy Posts: 24 Rep Power: 14	probably yes! howver yes i'm not in the wrong place working daily with fluent and cfx both for my work. and finding advise and good suggestion here! i've just seen your quote with a hint about pumplinx...so i've thought it was worth to ask an opinion about it!