Text "reversed flow" in TUI

SJSW · August 28, 2014, 04:36

Hi~
I found there is some text "reversed flow in 1111 face on pressure-outlet 2222" in the Text-User-Interface (TUI) window.

The solution is converged.
Is this normal when the viscous model, such as k-epsilon model, is activated?
On the other hand, is this bad for laminar flow?

Thanks for your kind suggestions.

marauder · August 28, 2014, 06:40

Reversed flow will depend on the flow at the outlet, most probable case would be that you did not allow the flow to fully develop at the outlet i.e., just after a bend or other geometric condition which can be corrected by extending the geometry at outlet & allow the flow to develop. Normally this would not affect your solution or convergence as this is the actual case that is happening at the outlet.

A pressure contour or image of your geometry will help in understanding your case.

larsschwarzer · August 28, 2014, 09:25

Normally, you should try to avoid reversed flow on the outlet - what the message means is that part of what you think is an outlet is actually an inlet for your calculation.

An example of this would be a vortex cut in two by the outlet boundary. The problem that arises is that the CFD-solver does not know what is happening outside of your boundaries (i.e. how the flow physically continues). Specifically, it does not have enough information to accurately compute the inflow on your (defined) outlet. This can of course lead to wrong flow field solutions. The error may or may not be negligible -- the point is you can't really tell.

The practical (engineering) thing to do is to go by marauder's solution and extend your flow domain. For information regarding the physics behind the problem, check textbooks on (numerical) fluid mechanics. If I remember correctly, there is also some brief information on the subject in Fluent's manual.

SJSW · October 1, 2014, 03:17

Thank you all!
Now I know how to avoid the reversed flow.

hwet · October 2, 2014, 10:23

just adding, if you see vector contours you will be able to see a few coming inside from the outlet, imagine because the flow is turbulent the pressures and velocities are at points both higher and lower inside your body than wot u set up at the outlet, hence the reversed flow, which is OK to have in yur simulation

August 28, 2014, 04:36	Text "reversed flow" in TUI	#1
SJSW Senior Member Join Date: Jun 2014 Location: Taiwan Posts: 100 Rep Power: 11	Hi~ I found there is some text "reversed flow in 1111 face on pressure-outlet 2222" in the Text-User-Interface (TUI) window. The solution is converged. Is this normal when the viscous model, such as k-epsilon model, is activated? On the other hand, is this bad for laminar flow? Thanks for your kind suggestions.

August 28, 2014, 06:40		#2
marauder Member Anonymous Join Date: Mar 2014 Posts: 84 Rep Power: 12	Reversed flow will depend on the flow at the outlet, most probable case would be that you did not allow the flow to fully develop at the outlet i.e., just after a bend or other geometric condition which can be corrected by extending the geometry at outlet & allow the flow to develop. Normally this would not affect your solution or convergence as this is the actual case that is happening at the outlet. A pressure contour or image of your geometry will help in understanding your case. SJSW likes this.

August 28, 2014, 09:25		#3
larsschwarzer New Member Lars Schwarzer Join Date: Apr 2013 Posts: 8 Rep Power: 13	Normally, you should try to avoid reversed flow on the outlet - what the message means is that part of what you think is an outlet is actually an inlet for your calculation. An example of this would be a vortex cut in two by the outlet boundary. The problem that arises is that the CFD-solver does not know what is happening outside of your boundaries (i.e. how the flow physically continues). Specifically, it does not have enough information to accurately compute the inflow on your (defined) outlet. This can of course lead to wrong flow field solutions. The error may or may not be negligible -- the point is you can't really tell. The practical (engineering) thing to do is to go by marauder's solution and extend your flow domain. For information regarding the physics behind the problem, check textbooks on (numerical) fluid mechanics. If I remember correctly, there is also some brief information on the subject in Fluent's manual. marauder and SJSW like this.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
How to deselect a pre-selected boundary in Text User Interface (TUI)	asal	FLUENT	1	October 8, 2017 08:50
Rounding Values in Post	stifelistefan	CFX	4	March 5, 2014 08:10
TUI Commands from GUI?	Carlos	FLUENT	6	May 22, 2013 18:05
Patching with TUI	Lovermore	FLUENT	1	July 22, 2010 22:28
Define animate using TUI	Zurlugh	FLUENT	1	July 24, 2008 08:55

October 1, 2014, 03:17		#4
SJSW Senior Member Join Date: Jun 2014 Location: Taiwan Posts: 100 Rep Power: 11	Thank you all! Now I know how to avoid the reversed flow.

October 2, 2014, 10:23		#5
hwet Senior Member Join Date: Mar 2014 Posts: 375 Rep Power: 13	just adding, if you see vector contours you will be able to see a few coming inside from the outlet, imagine because the flow is turbulent the pressures and velocities are at points both higher and lower inside your body than wot u set up at the outlet, hence the reversed flow, which is OK to have in yur simulation