CFD Online Discussion Forums

CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   CFX (https://www.cfd-online.com/Forums/cfx/)
-   -   How to model flow through "solid" material (https://www.cfd-online.com/Forums/cfx/27115-how-model-flow-through-solid-material.html)

Michael March 9, 2009 15:12

How to model flow through "solid" material
 
Hi all

I am working on modelling 150 persons and would therefore like to make some simplifications. In regards to this i have made a solid volume with the same surface area as 150 persons. Of course the flow field is incorrect and I would therefore like to investigate how porous the solid volume needs to be in order to give reliable results. Is there someone you can guide me how to this in a fluid domain.

kind regards Michael


Glenn Horrocks March 9, 2009 17:02

Re: How to model flow through "solid" material
 
Hi,

What is the application? Are you trying to model the heat load of 150 people in a HVAC simulation? In that case a simple heat source on the floor would probably do it.

Glenn Horrocks

Michael Tøgersen March 10, 2009 02:35

Re: How to model flow through "solid" material
 
Hi Glenn

Yes it is a kind of HVAC simulation where I am studing five different details of 150 persons geometry, solid surface/volumen and three details of a person. Therefore i need to model them as a solid volumen as one of the geometry with different porosities and not only as an heat source.

Glenn Horrocks March 10, 2009 17:02

Re: How to model flow through "solid" material
 
Hi,

You have not explained why putting a heat source on the floor is not suitable. If you want a better model then model the people in the room - a simple model would be to make the people rectangular blocks, a more complex model would be to make realistic models of people. This sounds like a better approach than a porous region.

Glenn Horrocks

Michael Tøgersen March 11, 2009 02:36

Re: How to model flow through "solid" material
 
Hi

I am sorry that I did not explain properly why I think that a heat source on the floor is not the ''right'' solution for me. The reason why is that I what to investigate the behaviour of the flow field in a large enclosure subject for crowding with different geometries of the persons.

Regarding the people I am working with three different levels of detail with use of rectangular blocks and to others by a solid volumen and a surface.

All in all the problem is to define to the solid volumen or surface with different permeabilities to allow a flow through the solid domain represented by the solid volumen or surface.

kind regards Michael

Glenn Horrocks March 11, 2009 21:33

Re: How to model flow through "solid" material
 
Hi,

It still sounds like modelling the people as blocks or other another simple shape would be the best approach.

However, if you still want to use the porous region approach then define the region as a subdomain (you will need the region defined in the mesh) and set it as a porous region.

Glenn Horrocks

Michael March 12, 2009 03:11

Re: How to model flow through "solid" material
 
Hi

I still have some questions about setting up the solid region as a subdomain for my simulation. I have made the enclosure in the workbench and afterwards "cut" out the solid region corresponding to the total area of people. After this I give the surface names and generate the mesh in CFS-Mesh. When I go to PRE in Advanced CFD I can only make one domain covering all surfaces and the solid volume and not define the solid volume as a subdomain.

Therefore I made the enclosure by it self and generate a mesh for it. The same is done for the solid volume. Then I have two meshes which I import in PRE in Advanced CFD. This gives me the opportunity to define two domains one for the fluid field - the enclosure, and the solid volume. Could this be the "right" way to do it?

But then the two meshes seems to merge where the solid volumen is located in the enclosure. Also it seems like it is not possible to define inflation layers around the solid volumen.

The problems seems to carry on for me :/

Kind regards Michael



All times are GMT -4. The time now is 18:40.