|
[Sponsors] |
July 19, 2009, 21:45 |
Server Rack in a room
|
#1 |
New Member
Jayantha Siriwardana
Join Date: Jul 2009
Posts: 3
Rep Power: 17 |
Hi everybody.
I am trying to simulate airflow inside a room with following conditions 1. I have a rack of servers which have air inlets and air outlets. Air inlets draw cold air into the rack and air outlets dissipate hot air to the room. 2. The room has one air conditioner. So far I have done the following in CFX. I Created a room with air conditioner based on the HVAC simulation tutorial. Next I imported the mesh for the rack separately created and tried a few options to define the air flow through the rack. 1. First thing I did was to define the rack in the solid domain as in another thread which tried to simulate humans inside a room. But with that option I cannot define the air flow through the rack because it doesn't allow to define inlets and outlets. Without defining air inlets and outlets, the simulation works fine. 2. Next thing I did was to define the rack in the fluid domain and define wall boundaries for outer surface. This case I can define air inlets and outlets. But the problem arises then I try to run the simulation. It ends with isolated fluid regions error. Does anybody have a suggestion to get through here? Thanks in advance. |
|
July 20, 2009, 03:14 |
|
#2 |
Super Moderator
Glenn Horrocks
Join Date: Mar 2009
Location: Sydney, Australia
Posts: 17,871
Rep Power: 144 |
Hi,
Have you considered modelling the fans and heat source as a source point rather than inlets and outlet? This is the normal way of doing this type of thing. Read the documentation and search the forum for source terms. Glenn Horrocks |
|
July 20, 2009, 03:25 |
|
#3 |
New Member
Jayantha Siriwardana
Join Date: Jul 2009
Posts: 3
Rep Power: 17 |
Hi Glenn
I am new to this CFD modelling. I was thinking of modelling the heat dissipation as a radiation flux emitting from the rack outlet. I will look into your point too. Thanks. |
|
July 20, 2009, 19:43 |
|
#4 |
Super Moderator
Glenn Horrocks
Join Date: Mar 2009
Location: Sydney, Australia
Posts: 17,871
Rep Power: 144 |
Hi,
Obviously you should model the heat fluxes as they physically exist. The heat load coming out of a computer in a rack will be convective, not radiative so if you use a radiative heat source you will get incorrect results. Having said that, radiation may or may not be significant in general for your simulation. For HVAC flows it depends on the exact situation as to whether radiation is significant. You will have to work that out. Glenn Horrocks |
|
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
HVAC Modeling Humans in Room | LilBort | CFX | 27 | July 31, 2009 12:08 |
Cluster Linux+OpenMOSIX with STAR-CD | CFD_User | Siemens | 9 | July 22, 2008 05:52 |
heat transfer: air in a room and water in a coil | Alberto | CFX | 2 | May 9, 2006 10:56 |
Contaminant Distribution in Air-con Room | Jules | Main CFD Forum | 0 | June 13, 2002 11:55 |
CFD Online Server Load | Jonas Larsson | Main CFD Forum | 0 | March 13, 2001 15:29 |