Outflow drag

Luk_Fiz · September 16, 2009, 08:23

Hi all,
I have a simple question that is maybe more related to physics not CFD itself. The simulated arrangement is very simple: Outflow of the water to air through the pipe (it is a kind of valve), the outlet of the pipe is approximately 5 mm. The velocity of the water is approximately 100m/s. Physical testing shows, that the outflow with this velocity produces spray or mist rather than water flow.
My question is: what is a most appropiate outlet boundary condition (simlating only a pipe, not a pipe outlet and some air volume)?. Setting "Opening" with relative pressure 0 overpredicts mass flow. I guess that there is some kind of aerodynamic drag of such a spray but looking through literature I cannot find anything seriously usefull. Can anyone tell anything about the "drag" of the waterflow into a gas?

Thanks a lot,
Luk

ghorrocks · September 16, 2009, 18:59

It might also be because the 0 pressure at the outlet assumption is not correct. Try moving the domain boundary beyond the outlet a bit and try again - it means you may have to model it as two phase flow also.

Luk_Fiz · September 17, 2009, 04:04

Hi,
Yes - it is my opinion that 0 pressure is wrong. However I wanted to omit simulation of 2 phases because spray started by the outflow takes a few cubic meters of the air volume. Because of that I am studing papers of "spray formation", "nozzle exit loss" etc. but so far I didnt find anything usable.

Luk

ghorrocks · September 17, 2009, 07:33

I suspect the pressure profile in the outlet pipe is not simple so the only way of being accurate is to include it in the model. You probably won't need to model the entire spray formation, just get far enough away from the outlet so the pressure profile in the outlet is correct. You may well be able to have most of the flow still in a jet as it leaves the domain.

September 16, 2009, 08:23	Outflow drag	#1
Luk_Fiz Member Lukasz Join Date: Mar 2009 Posts: 67 Rep Power: 17	Hi all, I have a simple question that is maybe more related to physics not CFD itself. The simulated arrangement is very simple: Outflow of the water to air through the pipe (it is a kind of valve), the outlet of the pipe is approximately 5 mm. The velocity of the water is approximately 100m/s. Physical testing shows, that the outflow with this velocity produces spray or mist rather than water flow. My question is: what is a most appropiate outlet boundary condition (simlating only a pipe, not a pipe outlet and some air volume)?. Setting "Opening" with relative pressure 0 overpredicts mass flow. I guess that there is some kind of aerodynamic drag of such a spray but looking through literature I cannot find anything seriously usefull. Can anyone tell anything about the "drag" of the waterflow into a gas? Thanks a lot, Luk

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Pressure drag, friction drag and total drag?	Cheng	CFX	9	January 26, 2024 13:46
Drag coefficient for parcels in dieselFoam	sebastian_vogl	OpenFOAM Running, Solving & CFD	5	December 31, 2008 12:19
Drag formulation	REMY	FLUENT	0	March 10, 2002 14:05
Drag formulation	remy	Main CFD Forum	0	March 10, 2002 13:05
Inviscid Drag at subsonic, subcritical Mach #	Axel Rohde	Main CFD Forum	1	November 19, 2001 12:19

September 16, 2009, 18:59		#2
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,700 Rep Power: 143	It might also be because the 0 pressure at the outlet assumption is not correct. Try moving the domain boundary beyond the outlet a bit and try again - it means you may have to model it as two phase flow also.

September 17, 2009, 04:04		#3
Luk_Fiz Member Lukasz Join Date: Mar 2009 Posts: 67 Rep Power: 17	Hi, Yes - it is my opinion that 0 pressure is wrong. However I wanted to omit simulation of 2 phases because spray started by the outflow takes a few cubic meters of the air volume. Because of that I am studing papers of "spray formation", "nozzle exit loss" etc. but so far I didnt find anything usable. Luk

September 17, 2009, 07:33		#4
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,700 Rep Power: 143	I suspect the pressure profile in the outlet pipe is not simple so the only way of being accurate is to include it in the model. You probably won't need to model the entire spray formation, just get far enough away from the outlet so the pressure profile in the outlet is correct. You may well be able to have most of the flow still in a jet as it leaves the domain.