# Continuum between volumes (default wall condition)

 Register Blogs Members List Search Today's Posts Mark Forums Read

 August 11, 2010, 22:58 Continuum between volumes (default wall condition) #1 Senior Member   Mohsin Mukhtar Join Date: Mar 2010 Location: South Korea Posts: 249 Rep Power: 8 I have 2 cylinders. 1 cylinder is small and is at some distance inside of the bigger cylinder. The flow comes from the smaller cylnder and goes thorugh the bigger cylinder. But When I export the mesh a Default face wall boundary is created at the end face of the small cylinder. GAMBIT is not taking it as a continuum by default. Can anybody please tell me how i can force continuum between them so that the face is not considered as wall by default. Thanks Mohsin

 August 12, 2010, 01:43 #2 Super Moderator     Maxime Perelli Join Date: Mar 2009 Location: Switzerland Posts: 2,973 Rep Power: 30 post a picture and explain how you build your geometry. Did you split /substract volumes? __________________ In memory of my friend Hervé: CFD engineer & freerider

August 12, 2010, 02:51
#3
Senior Member

Mohsin Mukhtar
Join Date: Mar 2010
Location: South Korea
Posts: 249
Rep Power: 8
Max Thank you for ur reply.

Attached is the geometry. I was talking about the upper 2 cylinders. The middle (smaller cylinder) is through in the bigger cylinder. when i export it to fluent it takes the lower end face of the middle cylinder as wall (Air is flowing through the lower end so I need Gambit to consider it as a continuum not a wall). How should i specify that
Attached Images
 cyclone.jpg (45.6 KB, 31 views)

 August 12, 2010, 03:05 #4 Super Moderator     Maxime Perelli Join Date: Mar 2009 Location: Switzerland Posts: 2,973 Rep Power: 30 try to move the small cylinder. If you can move it, then your geometry isn't ok. In this cas you will need to split the biggest cylinder with the smallest. (but the question is what are you supposed to modelise) __________________ In memory of my friend Hervé: CFD engineer & freerider

 August 12, 2010, 03:45 #5 Senior Member   Mohsin Mukhtar Join Date: Mar 2010 Location: South Korea Posts: 249 Rep Power: 8 Max thanks for ur reply. The middle cylinder is moving. Actually I m simulating cyclone. The same cyclone which in the following tutorial he is making us learn to simulate. http://www.itc.polsl.pl/gwecel/Mns/Fluent_tutorial.pdf on page 14, he says about connect disconnet faces in order to make the middle cylinder as a continuum. "After operation two volumes are linked by one face forcing later the same mesh to be generated for both volumes at that face. Not connected faces will be by default treated as wall" could u please explain that.

 August 12, 2010, 04:04 #6 Super Moderator     Maxime Perelli Join Date: Mar 2009 Location: Switzerland Posts: 2,973 Rep Power: 30 If the faces aren't connected, then you have 2 superposed meshes, treated as wall by Fluent. In your case I would split the small cylinder with the big one (enable Retain option). And in the BC you need to enforce gambit setting the cover surface as wall. With split, gambit set automatically all internal surfaces (as Internal) __________________ In memory of my friend Hervé: CFD engineer & freerider

 August 12, 2010, 04:20 #7 Senior Member   Mohsin Mukhtar Join Date: Mar 2010 Location: South Korea Posts: 249 Rep Power: 8 Max I m sorry I didnt get your point. I dont want the cover face as wall. If I split the middle cylinder wityh the bigger cylinder then 2 cylinder volumes are created but still Gambit treats the cover face as wall. Could you please elaborate

 August 12, 2010, 04:58 #8 Super Moderator     Maxime Perelli Join Date: Mar 2009 Location: Switzerland Posts: 2,973 Rep Power: 30 Faces resulting from Split are treated as Internal (not wall) If you get walls, then splits aren't correct. But if you don't want the cover as wall, and the top as wall, then go and unite the 2 cylinders __________________ In memory of my friend Hervé: CFD engineer & freerider

August 12, 2010, 05:19
#9
Senior Member

Mohsin Mukhtar
Join Date: Mar 2010
Location: South Korea
Posts: 249
Rep Power: 8

Max yes faces resulting from split operation are treated as internal.But the face (which i have marked in picture) cannot to be splitted. I united the 2 cylinders but the lower face (as shown in picture) is always treated as wall.

Could you tell me how can i split it (the lower face shown in figure) so that it is not treated as wall.

I have tried number of ways to split it but in vain.
Attached Images
 cyclone.png (10.1 KB, 11 views)

 August 12, 2010, 05:47 #10 Super Moderator     Maxime Perelli Join Date: Mar 2009 Location: Switzerland Posts: 2,973 Rep Power: 30 check if don't have superposed surface at this place. If yes than connect them (surface/connect) __________________ In memory of my friend Hervé: CFD engineer & freerider

 August 12, 2010, 08:15 #11 Senior Member   Mohsin Mukhtar Join Date: Mar 2010 Location: South Korea Posts: 249 Rep Power: 8 There is only 1 face at that place and no other face is superimposed.

 August 13, 2010, 01:24 #12 Super Moderator     Maxime Perelli Join Date: Mar 2009 Location: Switzerland Posts: 2,973 Rep Power: 30 Ok if you split volume 1 (biggest cylinder) with volume 4, you don't have problem. I exported the mesh and I don't have any unwanted wall. But if you need the cover surface (surface 11 ) inside the biggest volume, then you need to define it as wall in the BC You will also have a problem with skewness because of small angle between edges 18 and 24. You can sweep face 15 in x-direction with amplitude 0.01. And aplly a size function for refining the mesh along the suface 0.01x0.89 __________________ In memory of my friend Hervé: CFD engineer & freerider

August 13, 2010, 01:40
#13
Senior Member

Mohsin Mukhtar
Join Date: Mar 2010
Location: South Korea
Posts: 249
Rep Power: 8
Thank you so much Max. I am really grateful.

Quote:
 Originally Posted by -mAx- But if you need the cover surface (surface 11 ) inside the biggest volume, then you need to define it as wall in the BC
When i define wall inside and export in fluent. Fluent gives me an extra boundary condition "wall-shadow". What influence it may have on the result? and how to remove it?

 August 13, 2010, 02:10 #14 Super Moderator     Maxime Perelli Join Date: Mar 2009 Location: Switzerland Posts: 2,973 Rep Power: 30 Then something isn't correct. But the issue with the top of small cylinder is ok, now? I saw you set the BC before split operations. I would alway set the BC once your mesh is ready. In your case, in the BC Panel select delete All, and redefine the BC. If you still have the problem, send me the journal file describing all the steps till mesh export. __________________ In memory of my friend Hervé: CFD engineer & freerider

 August 13, 2010, 02:28 #15 Senior Member   Mohsin Mukhtar Join Date: Mar 2010 Location: South Korea Posts: 249 Rep Power: 8 The upper cylinder problem is solved. Thankyou. as now its just a fluid zone not occupied with wall faces. For Fluent problem which gives additional shadow wall term I got the following from CFD online discussions: you will only get a shadow wall if the mesh on a common wall is the same. That is to say if their are two volume zones that are connected and separated by one wall then FLuent will split the wall into two, i.e wall and then its shadow. So the two volume zones each have their own identical wall. Reference: Fluent not creating Shadow wall It just seperates the 2 fluid zones. So i may not be worried about it? I tried what you asked me to do. i-e first I created gerometry and did all split operations, then mesh (checked with just very coarse mesh with less interval count) and then applied boundary conditions. But stuill fluent creates a shadow wall.

 August 13, 2010, 03:07 #16 Super Moderator     Maxime Perelli Join Date: Mar 2009 Location: Switzerland Posts: 2,973 Rep Power: 30 Ok, then you can try to iterate. Else you can try spliting the volume at the same z-level than your top surface. (copy and move outer edges from big cylinder, create surface from them and split big volume with this surface). Then the fluid volume will be completly splitted, and you shouldn't get the wall shadows. __________________ In memory of my friend Hervé: CFD engineer & freerider

August 13, 2010, 03:50
#17
Senior Member

Mohsin Mukhtar
Join Date: Mar 2010
Location: South Korea
Posts: 249
Rep Power: 8
Great Max.

I created a circular face and moved it until the lower face of cover and split it with the bigger cylinder. After that i splitted the middle cylinder with the bigger one and got the desired result. I have attached the picture.

Max you told me that

"You will also have a problem with skewness because of small angle between edges 18 and 24. You can sweep face 15 in x-direction with amplitude 0.01. And aplly a size function for refining the mesh along the suface 0.01x0.89 "

I splitted the face in x direction but it will create 2 faces and the problem with the original face (having sharp angles ) will remain there.
Also which size function is preferable in this case "Mesh, curvature or fixed"

Thank you for ur assistance.

P.S. I am unable to increase your reputation more now
Attached Images
 Cyclone.jpg (40.9 KB, 8 views)

August 13, 2010, 04:12
#18
Super Moderator

Maxime Perelli
Join Date: Mar 2009
Location: Switzerland
Posts: 2,973
Rep Power: 30
Quote:
 Originally Posted by Mohsin I splitted the face in x direction but it will create 2 faces and the problem with the original face (having sharp angles ) will remain there. Also which size function is preferable in this case "Mesh, curvature or fixed"
Merge the 2 surfaces on both sides from the edge giving you trouble --> the edge will disappear
Regarding size funtion you can use fixed with start size 0.01 for instance
__________________
In memory of my friend Hervé: CFD engineer & freerider

 August 13, 2010, 04:30 #19 Senior Member   Mohsin Mukhtar Join Date: Mar 2010 Location: South Korea Posts: 249 Rep Power: 8 Error: tried to merge faces that are not connected. Tried to merge faces with diferent upper topoplogy or with nion manifold neighbours(share the common boundary with another face) Which face you are refereing to. The upper face or or side faces.

 August 13, 2010, 04:58 #20 Super Moderator     Maxime Perelli Join Date: Mar 2009 Location: Switzerland Posts: 2,973 Rep Power: 30 upper faces (z=constant) __________________ In memory of my friend Hervé: CFD engineer & freerider

 Thread Tools Display Modes Linear Mode

 Posting Rules You may not post new threads You may not post replies You may not post attachments You may not edit your posts BB code is On Smilies are On [IMG] code is On HTML code is OffTrackbacks are On Pingbacks are On Refbacks are On Forum Rules

 Similar Threads Thread Thread Starter Forum Replies Last Post mahdi haddad FLUENT 1 November 11, 2014 07:13 shekharc Main CFD Forum 3 December 2, 2009 09:20 mskim FLUENT 0 February 14, 2009 17:37 Jay FLUENT 0 April 14, 2007 18:06 Rashad FLUENT 1 October 9, 2006 18:31

All times are GMT -4. The time now is 10:58.