CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   CFX (http://www.cfd-online.com/Forums/cfx/)
-   -   Ambiguity in understanding interfaces in CFX (http://www.cfd-online.com/Forums/cfx/84502-ambiguity-understanding-interfaces-cfx.html)

saisanthoshm88 February 1, 2011 03:05

Ambiguity in understanding interfaces in CFX
 
I had a difficulty in understanding the concept of interfaces in CFX. My question is that if I had a surface common to two volumes (with in the same fluid domain) but there isn’t any fluid transfer (or) heat transfer occurring across it (I mean if the surface is supposed to act as a wall) and if I’m not even interested to model it as a Thin material ( well I mean the surface is of zero thickness) (or) give a thermal contact resistance to it. The whole domain is tet meshed and as such there is no issue of dissimilar mesh types.

Then in such case do I still need to have the surface as an interface in CFX, of course by setting it as a No slip wall in the additional interface models , just because it gets represented by two different meshes in CFX, one from volume1 and the other from volume2 and because the element size on these two meshes representing the surface is different so there is a node disconnectivity between these meshes as such?

100tinela February 1, 2011 13:10

Quote:

Originally Posted by saisanthoshm88 (Post 293050)
I had a difficulty in understanding the concept of interfaces in CFX. My question is that if I had a surface common to two volumes (with in the same fluid domain) but there isnít any fluid transfer (or) heat transfer occurring across it (I mean if the surface is supposed to act as a wall) and if Iím not even interested to model it as a Thin material (or) give a thermal contact resistance to it. The whole domain is tet meshed and as such there is no issue of dissimilar mesh types.

Then in such case do I still need to have the surface as an interface in CFX, of course by setting it as a No slip wall in the additional interface models , just because it gets represented by two different meshes in CFX, one from volume1 and the other from volume2 and because the element size on these two meshes representing the surface is different so there is a node disconnectivity between these meshes as such?

Hi,

It seems to be a case of multiphase flow, am I right? If it is, perhaps you don't need to generate two different meshes. You can use a step function to define this two distincts regions and if the interface is stable you can use mesh adaptation to refine the mesh on it. You can also set your case as free surface and set homogeneous model (without interaction between the two fluids).

You can find more information about it on CFX help, just type "homogeneous model" or "free surface". There also some tutorials available, like "Free Surface Flow Over a Bump".

Att.
Willian

saisanthoshm88 February 1, 2011 13:49

Nope it's not a problem of multi-phase flow , it's just a simple problem but the geometry was quite large so I thought of cutting it down to simpler volumes and then work upon it in CFX something like a 'divide and conquer' strategy and the flow physics involved is not so complex it's just a sort of conventional problem.

Any way I really appreciate your help in a way that your suggestions are a bit advanced for me to get along as I'm still a newbie to CFD but when I come across to deal with multi-phase flows later on I think your suggestions help me learn better so thanks for the help indeed.

joey2007 February 1, 2011 16:55

If I remind correctly the modelling of guiding baffles with neglected thickness is shown in the the combustor tutorial. For the meshing use internal surface in the geometry.

In conjunction with workbench meshing use 13.0. Bug in 12.1.

saisanthoshm88 February 1, 2011 21:02

Well the surface i'm talking about is of zero thickness and I use ICEM CFD 12.1 as the meshing tool.

ghorrocks February 2, 2011 06:30

It is late at night and I am tired. I do not understand your question, can you rephrase it?

But if one or two meshed solid regions share a surface you can either set the shared surface as a wall (then it acts as a wall) or an interface (then it allows fluid to pass through it). But I am not sure that is what you are asking.

saisanthoshm88 February 2, 2011 06:53

oh sorry for troubling you Glenn, I didn't really have a track of timings at your place. Well, my question is like when do we really need to set an interface as a No slip wall in the additional interface models.

ghorrocks February 2, 2011 06:57

No problem. My comment about it being late was my excuse for not thinking hard enough to understand your question.

Interfaces and walls are different things. You do not normal define no slip conditions on interfaces. If you want the flow to join then leave it as an interface with the default options. If you want a wall then change it to a wall.

saisanthoshm88 February 2, 2011 07:53

ok I was just getting confused when I've gone through a CFX ppt , available on the link: www.cadfamily.com/Download.asp?ID=298701

just have a look into the 12th slide on thin wall modeling in the ppt.

joey2007 February 2, 2011 15:05

You are right it is not intuitive. ANSYS CFX differs between thin wall and thin surface. Thin wall has heat transfer and is setup via interfaces. Thin surface is an inner wall without any transfer. Tutorial 20 in 13 shows how to setup when the mesh has the required geometry.

Do not know how to realize in ICEM. Maybe you require two surface in two parts.


All times are GMT -4. The time now is 05:51.