reversed flow on faces limited in 643 cells in region out

emanuele · September 19, 2009, 06:10

Hello everybody,
i am a newby and i dont know if introducing myself somewhere is required.
Anyway i am 27years old marine engineer from italy approaching for the first time on cfd planet.
I am asking for your help because any boat model i am trying to validate with cfd,return a message reporting what stated in the title of this thread "reversed flow on faces limited in 643 cells in region out"...
I tryed to change volume control dimensions up to 2,5 time boat length,width and depth...but the message won't disappear....although performing grid parameters of my multiphase.

So now i am trying with a second order solver...and the number of faces affected seem to get smaller iteration by iteration.
Can anyone explain me what's the meaning of the above mentioned message and what's the smartest solution to fix similiar problems without affecting solution and its transient?

i used velocity inlet and pressure outlet

Many thanks in advance

Maddin · September 21, 2009, 11:22

Check your conditions on the outlet boundary.
Sometimes it's normal because the flow reserves before it gets a straight flow.
When I use a Pressure Outlet as a Inlet I get this on every iteration

To check the direction I use vectors on the boundarys.

emanuele · September 21, 2009, 11:29

Quote:

Originally Posted by Maddin

Check your conditions on the outlet boundary.
Sometimes it's normal because the flow reserves before it gets a straight flow.
When I use a Pressure Outlet as a Inlet I get this on every iteration

To check the direction I use vectors on the boundarys.

Thank you for your support....but i imagine reversed flow could affect convergence!isn't it?
I would like to try a polyedral mesh in order to check if multifaces volumes can manage turbulent flow in a most accurate way...actually i m going with trimmer and its refinement on free surface..

Maddin · September 21, 2009, 12:26

Is your problem 2D or 3D?
When this is a 2D problem then your region is wrong.
Show more pictures!

emanuele · September 21, 2009, 13:58

my problem is a 3d problem and it don't wanto to converge!!I checked several times what should be wrong but i discovered nothing....
i performed mesh diagnostics and everything seems to be ok,i enlarged my continuum but nothing....solution keep going in a wrong way.I'll show more pic as soon as i can!
i performed some trials without degrees of freedom in order to be sure it is not a prblem related to initial condition an moments of inertia...but it's the same..

Maddin · September 21, 2009, 15:20

What settings to you use? Wall treatment? y+?
Bring more facts plz

emanuele · September 22, 2009, 10:19

ok,at the link below you can find my summary report

...
thank you for your big interest!!

http://www.megafileupload.com/en/fil...umem-html.html

Maddin · September 22, 2009, 10:50

Look at your courant number, maybe you have so choice a smaller time step.

emanuele · September 22, 2009, 11:31

Quote:

Originally Posted by Maddin

Look at your courant number, maybe you have so choice a smaller time step.

thank you!!i will manage to optimize it!

Maddin · September 22, 2009, 11:39

Tell me when it works, picture looks very nice. Could be a really nice animation

emanuele · September 25, 2009, 05:27

ok...error found...i was using velocity modulus..At inlet flow direction is "inlet to outlet"...and you do not need to specify it using "-" or "+"...just use modulus of velocity
In continum you should specify everytime the direction of the flow....i was using positive velocity at inlet and positive velocity in continuum bu the direction of my cordinate system was with x axis directed in the same direction of boat advancing motion,so,to define velocity in continum i must use a negative value because water come forth my boat...differently i was experiencing vertices waves...the hell..

Maddin · September 25, 2009, 06:17

When I understand you right you didn't had seen/known that the velocity in the initial conditions is according the laboratory coordiante system?

Ok when I look at your picture know I see it.
X-axle shows to the left so you must work with negativ flow when the flow direction is from left to right

emanuele · September 25, 2009, 06:22

Quote:

Originally Posted by Maddin

When I understand you right you didn't had seen/known that the velocity in the initial conditions is according the laboratory coordiante system?

Ok when I look at your picture know I see it.
X-axle shows to the left so you must work with negativ flow when the flow direction is from left to right

yes!but at inlet,flow direction is utomatically defined inlet to outlet..isn't it?

Maddin · September 25, 2009, 06:43

Under Physics Conditions of the Boundary you can set the flow direction

You can also give a vector for the flow or "normal to boundary" or a angle.

emanuele · September 25, 2009, 06:47

ok...the solution is:
be sure x axys si oriented as flow motion direction

...it is easyer...

mazhar16823 · April 21, 2020, 13:44

Quote:

Originally Posted by emanuele

ok...the solution is:
be sure x axys si oriented as flow motion direction

...it is easyer...

so how did you introduce modulus of velocity? In my case, I have the blade which rotates in x-y plane and inlet is in z direction but in the coordinate system ''z'' points opposite to the inflow direction so I used -ve velocity value and it shows me the same message of reversed flow at outlet.

September 19, 2009, 06:10	reversed flow on faces limited in 643 cells in region out	#1
emanuele New Member Join Date: Sep 2009 Posts: 19 Rep Power: 16	Hello everybody, i am a newby and i dont know if introducing myself somewhere is required. Anyway i am 27years old marine engineer from italy approaching for the first time on cfd planet. I am asking for your help because any boat model i am trying to validate with cfd,return a message reporting what stated in the title of this thread "reversed flow on faces limited in 643 cells in region out"... I tryed to change volume control dimensions up to 2,5 time boat length,width and depth...but the message won't disappear....although performing grid parameters of my multiphase. So now i am trying with a second order solver...and the number of faces affected seem to get smaller iteration by iteration. Can anyone explain me what's the meaning of the above mentioned message and what's the smartest solution to fix similiar problems without affecting solution and its transient? i used velocity inlet and pressure outlet Many thanks in advance Last edited by emanuele; September 19, 2009 at 06:29.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Problem in IMPORT of ICEM input file in FLUENT	csvirume	FLUENT	2	September 9, 2009 01:08
Error to re-open fluent case file	J.Gimbun	FLUENT	0	April 27, 2006 08:42
[Commercial meshers] Trimmed cell and embedded refinement mesh conversion issues	michele	OpenFOAM Meshing & Mesh Conversion	2	July 15, 2005 04:15
fluid flow fundas	ram	Main CFD Forum	5	June 17, 2000 21:31
Flow visualization vs. Calculated flow patterns	Francisco Saldarriaga	Main CFD Forum	1	August 2, 1999 23:18

September 21, 2009, 11:22		#2
Maddin Senior Member Join Date: Mar 2009 Posts: 260 Rep Power: 18	Check your conditions on the outlet boundary. Sometimes it's normal because the flow reserves before it gets a straight flow. When I use a Pressure Outlet as a Inlet I get this on every iteration To check the direction I use vectors on the boundarys.

September 21, 2009, 12:26		#4
Maddin Senior Member Join Date: Mar 2009 Posts: 260 Rep Power: 18	Is your problem 2D or 3D? When this is a 2D problem then your region is wrong. Show more pictures!

September 21, 2009, 13:58		#5
emanuele New Member Join Date: Sep 2009 Posts: 19 Rep Power: 16	my problem is a 3d problem and it don't wanto to converge!!I checked several times what should be wrong but i discovered nothing.... i performed mesh diagnostics and everything seems to be ok,i enlarged my continuum but nothing....solution keep going in a wrong way.I'll show more pic as soon as i can! i performed some trials without degrees of freedom in order to be sure it is not a prblem related to initial condition an moments of inertia...but it's the same..

September 21, 2009, 15:20		#6
Maddin Senior Member Join Date: Mar 2009 Posts: 260 Rep Power: 18	What settings to you use? Wall treatment? y+? Bring more facts plz

September 22, 2009, 10:19		#7
emanuele New Member Join Date: Sep 2009 Posts: 19 Rep Power: 16	ok,at the link below you can find my summary report ... thank you for your big interest!! http://www.megafileupload.com/en/fil...umem-html.html

September 22, 2009, 10:50		#8
Maddin Senior Member Join Date: Mar 2009 Posts: 260 Rep Power: 18	Look at your courant number, maybe you have so choice a smaller time step.

September 22, 2009, 11:39		#10
Maddin Senior Member Join Date: Mar 2009 Posts: 260 Rep Power: 18	Tell me when it works, picture looks very nice. Could be a really nice animation

September 25, 2009, 05:27		#11
emanuele New Member Join Date: Sep 2009 Posts: 19 Rep Power: 16	ok...error found...i was using velocity modulus..At inlet flow direction is "inlet to outlet"...and you do not need to specify it using "-" or "+"...just use modulus of velocity In continum you should specify everytime the direction of the flow....i was using positive velocity at inlet and positive velocity in continuum bu the direction of my cordinate system was with x axis directed in the same direction of boat advancing motion,so,to define velocity in continum i must use a negative value because water come forth my boat...differently i was experiencing vertices waves...the hell..

September 25, 2009, 06:17		#12
Maddin Senior Member Join Date: Mar 2009 Posts: 260 Rep Power: 18	When I understand you right you didn't had seen/known that the velocity in the initial conditions is according the laboratory coordiante system? Ok when I look at your picture know I see it. X-axle shows to the left so you must work with negativ flow when the flow direction is from left to right

September 25, 2009, 06:43		#14
Maddin Senior Member Join Date: Mar 2009 Posts: 260 Rep Power: 18	Under Physics Conditions of the Boundary you can set the flow direction You can also give a vector for the flow or "normal to boundary" or a angle.

September 25, 2009, 06:47		#15
emanuele New Member Join Date: Sep 2009 Posts: 19 Rep Power: 16	ok...the solution is: be sure x axys si oriented as flow motion direction ...it is easyer...