Moving PIPE

ghorrocks · November 20, 2014, 06:16

1) Yes, the sliding interface is transient rotor stator.
2, 3) 1 and 2 are moving mesh domains (but stationary frame of reference). 3 is a stationary frame of reference with no moving mesh.
4) No, do not select any solid motion options. Use moving mesh on a solid domain. Do not use a rigid body model.
5) 1 and 2 have mesh deformation on.
6) interface 1 to 2 is stationary I think, if that does not work then TRS. 2 to 3 is TRS.

Martin_Sz · November 20, 2014, 06:36

I have another question
Where do I finf frame of reference option ?
Best regards

ghorrocks · November 20, 2014, 06:51

Frame of reference is where you choose either stationary or rotating. The default is stationary so leave it at the default for all domains.

Martin_Sz · November 24, 2014, 02:22

Hello Glenn,
So I make simulation with idea which You wrote on the lasts posts and I have another problem.
On the solver only six steps go on and on the seventh error jump in with information
cNWDIST
At least one highly skewed element has been detected on a wall boundary, leading to unreliable near-wall distance calculation for the turbulent wall functions. The solver will continue to execute, but convergence and/or accuracy may be affected. Please consider improving the mesh quality. the coordinates of the element face are...
So what can I do at that moment to fix this error ??
Best regards

ghorrocks · November 24, 2014, 05:22

This tells me you have got the mesh motion wrong.

Re-run the simulation saving a results file on each time step which includes the mesh. Then have a look at the time step before it crashes at the location it specifies. If you cannot work out what the problem is post an image of it on the forum.

Martin_Sz · November 24, 2014, 05:36

I increase mesh quality and simulation goes on to the 14th step.
and crashes with element volume error.
The strangest thing is that. On the 14th step and other steps before I haven't got heat transfer on the tube (on moving area of the tube).
When I define moving interface on domain 2, i need to define wall velocity ??
Best regards

ghorrocks · November 24, 2014, 05:46

The error has nothing to do with mesh quality. Have you done what I recommend in my last post?

The error is most probably due to the mesh motion settings you are using. You need to define a displacement function (versus time) and apply that to all the boundaries of the moving domain. If you miss a boundary you can get the error you got - and that includes interface boundaries so I bet you missed one of them.

Martin_Sz · November 24, 2014, 06:03

So on the attachments I'm sending You what i define on interfaces of domains
Best regards

ghorrocks · November 24, 2014, 06:12

I cannot tell much about what is happening from the screen dumps. Please your CCL.

But even more important is the post I put up #25 about how to debug this. It should allow you to identify exactly what the problem is.

Martin_Sz · November 24, 2014, 06:24

can You verify my defined interfaces which I send on the last post (attachments )
Best regards

ghorrocks · November 24, 2014, 06:51

I said:

Quote:

I cannot tell much about what is happening from the screen dumps.

The next sentence meant to say: Please attach your CCL.

But even better: if you do the images I said in post #25 I think we will be able to identify the problem straight away.

Martin_Sz · November 24, 2014, 08:06

I'm sending a ccl of the simulation
Best regards

Martin_Sz · November 24, 2014, 08:11

I change mesh deformation to initial boudaries on subdomain option
Then 13 steps go on and on the 14th pop up an error (attachment)
Best regards

ghorrocks · November 24, 2014, 18:25

As I thought - you have not defined the mesh motion parameters on many boundary surfaces on mesh movement domains.

Boundaries which need it are:
kostka subdom Side 2
subdomena Default

Also the Walek domain (which appears to be your solid) looks like it has no mesh deformation. You need to use mesh deformation on this domain as well.

Martin_Sz · November 25, 2014, 02:34

Hello Glenn,
Thanks for reply
Where can I change on solid domain mesh deformation ??
I dont see this option on solid domain (attachment)
Best regards

ghorrocks · November 25, 2014, 05:07

What version of CFX are you using?

Martin_Sz · November 25, 2014, 05:17

v15
And I have another question.
On Your schematic which domain is moving 1 or 2 ??

ghorrocks · November 25, 2014, 06:52

1 and 2 are stationary frame of reference, moving mesh domains.

OK, good. V15 is the only version which supports motion in solid domains. I am pretty sure you can do moving mesh on solid domains. You might need to edit the CCL to do this, rather than doing it in CFX-Pre. Copy the format used in your moving mesh fluid domain.

Martin_Sz · November 25, 2014, 08:47

So for the first time solution goes on.
And I have interesting results.
Interface between walek and subdomain go on the direction which I defined in mesh motion but solid domain on the end of simulation is on the start position. Only interfaces walek-subdomena with empty area inside goes on.
I dont know what to define mesh deformation on solid. ccl copy paste from fluid domain doesnt work. What can I do ??
Best regards

ghorrocks · November 25, 2014, 18:27

Actually, you might need to use the solid motion options for the solid domain. Have you tried that?

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Help needed in modelling/meshing a body moving in a pipe which is driven by a fluid	hareshram.n	ANSYS	1	July 4, 2024 03:13
simple open ended pipe flow with moving piston	lmarf88	ANSYS	1	November 22, 2013 22:12
moving sphere in a pipe	sajeesh	FLUENT	1	February 25, 2013 09:35
Terrible Mistake In Fluid Dynamics History	Abhi	Main CFD Forum	12	July 8, 2002 10:11
fluid flow fundas	ram	Main CFD Forum	5	June 17, 2000 22:31

November 20, 2014, 06:16		#21
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,870 Rep Power: 144	1) Yes, the sliding interface is transient rotor stator. 2, 3) 1 and 2 are moving mesh domains (but stationary frame of reference). 3 is a stationary frame of reference with no moving mesh. 4) No, do not select any solid motion options. Use moving mesh on a solid domain. Do not use a rigid body model. 5) 1 and 2 have mesh deformation on. 6) interface 1 to 2 is stationary I think, if that does not work then TRS. 2 to 3 is TRS.

November 20, 2014, 06:36		#22
Martin_Sz Senior Member Marcin Join Date: May 2014 Location: Poland, Swiebodzin Posts: 313 Rep Power: 13	I have another question Where do I finf frame of reference option ? Best regards

November 20, 2014, 06:51		#23
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,870 Rep Power: 144	Frame of reference is where you choose either stationary or rotating. The default is stationary so leave it at the default for all domains.

November 24, 2014, 02:22		#24
Martin_Sz Senior Member Marcin Join Date: May 2014 Location: Poland, Swiebodzin Posts: 313 Rep Power: 13	Hello Glenn, So I make simulation with idea which You wrote on the lasts posts and I have another problem. On the solver only six steps go on and on the seventh error jump in with information cNWDIST At least one highly skewed element has been detected on a wall boundary, leading to unreliable near-wall distance calculation for the turbulent wall functions. The solver will continue to execute, but convergence and/or accuracy may be affected. Please consider improving the mesh quality. the coordinates of the element face are... So what can I do at that moment to fix this error ?? Best regards

November 24, 2014, 05:22		#25
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,870 Rep Power: 144	This tells me you have got the mesh motion wrong. Re-run the simulation saving a results file on each time step which includes the mesh. Then have a look at the time step before it crashes at the location it specifies. If you cannot work out what the problem is post an image of it on the forum.

November 24, 2014, 05:36		#26
Martin_Sz Senior Member Marcin Join Date: May 2014 Location: Poland, Swiebodzin Posts: 313 Rep Power: 13	I increase mesh quality and simulation goes on to the 14th step. and crashes with element volume error. The strangest thing is that. On the 14th step and other steps before I haven't got heat transfer on the tube (on moving area of the tube). When I define moving interface on domain 2, i need to define wall velocity ?? Best regards

November 24, 2014, 05:46		#27
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,870 Rep Power: 144	The error has nothing to do with mesh quality. Have you done what I recommend in my last post? The error is most probably due to the mesh motion settings you are using. You need to define a displacement function (versus time) and apply that to all the boundaries of the moving domain. If you miss a boundary you can get the error you got - and that includes interface boundaries so I bet you missed one of them.

November 24, 2014, 06:12		#29
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,870 Rep Power: 144	I cannot tell much about what is happening from the screen dumps. Please your CCL. But even more important is the post I put up #25 about how to debug this. It should allow you to identify exactly what the problem is.

November 24, 2014, 06:24		#30
Martin_Sz Senior Member Marcin Join Date: May 2014 Location: Poland, Swiebodzin Posts: 313 Rep Power: 13	can You verify my defined interfaces which I send on the last post (attachments ) Best regards

November 24, 2014, 18:25		#34
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,870 Rep Power: 144	As I thought - you have not defined the mesh motion parameters on many boundary surfaces on mesh movement domains. Boundaries which need it are: kostka subdom Side 2 subdomena Default Also the Walek domain (which appears to be your solid) looks like it has no mesh deformation. You need to use mesh deformation on this domain as well.

November 25, 2014, 05:07		#36
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,870 Rep Power: 144	What version of CFX are you using?

November 25, 2014, 05:17		#37
Martin_Sz Senior Member Marcin Join Date: May 2014 Location: Poland, Swiebodzin Posts: 313 Rep Power: 13	v15 And I have another question. On Your schematic which domain is moving 1 or 2 ??

November 25, 2014, 06:52		#38
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,870 Rep Power: 144	1 and 2 are stationary frame of reference, moving mesh domains. OK, good. V15 is the only version which supports motion in solid domains. I am pretty sure you can do moving mesh on solid domains. You might need to edit the CCL to do this, rather than doing it in CFX-Pre. Copy the format used in your moving mesh fluid domain.

November 25, 2014, 08:47		#39
Martin_Sz Senior Member Marcin Join Date: May 2014 Location: Poland, Swiebodzin Posts: 313 Rep Power: 13	So for the first time solution goes on. And I have interesting results. Interface between walek and subdomain go on the direction which I defined in mesh motion but solid domain on the end of simulation is on the start position. Only interfaces walek-subdomena with empty area inside goes on. I dont know what to define mesh deformation on solid. ccl copy paste from fluid domain doesnt work. What can I do ?? Best regards

November 25, 2014, 18:27		#40
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,870 Rep Power: 144	Actually, you might need to use the solid motion options for the solid domain. Have you tried that?