Discussion Groups

Simulation CFD

Reply
Product Support
wildej
Posts: 733
Registered: ‎08-25-2011

Re: Natural ventilation

12-04-2013 11:29 AM in reply to: anasabukhadra

Hi,

 

I have just shared a Dropbox folder with you, please place the CFZ file in there - the invite will come from my personal email just FYI.

 

Kind regards,

Jon



Jon Wilde
Please use plain text.
Contributor
anasabukhadra
Posts: 16
Registered: ‎10-18-2013

Re: Natural ventilation

12-04-2013 11:35 AM in reply to: wildej
Done. Thank you!
Please use plain text.
Product Support
wildej
Posts: 733
Registered: ‎08-25-2011

Re: Natural ventilation

12-05-2013 12:13 AM in reply to: wildej

Morning,

 

Just so you are aware, the Dropbox folder is still empty, did you copy the file over?

 

Kind regards,

Jon



Jon Wilde
Please use plain text.
Contributor
anasabukhadra
Posts: 16
Registered: ‎10-18-2013

Re: Natural ventilation

12-05-2013 12:23 AM in reply to: wildej
I did! I will upload it again now.
Please use plain text.
Product Support
wildej
Posts: 733
Registered: ‎08-25-2011

Re: Natural ventilation

12-05-2013 05:58 AM in reply to: wildej

Thanks for sharing the file, I have a few comments:

 

  1. You cannot have internal surface Boundary Conditions - please remove the internal film coefficients and P=0. CFD will calculate the film coeff's for you and the P=0 is simply not needed, the 2 air parts are already connected
  2. You do not need so much air upstream, nothing is happening there, it would be far more useful downstream. It is best to draw the external volume in CAD - here you have a floating building as the external domain in CFD is not permitted to touch the building
  3. It is liklely you will need to refine the mesh on the external air, see how the results look but I think they may look rather coarse
  4. The real issue is the geometry though, take a look at the mesh preview - there are some very small surfaces here, which is causing an incredibly fine mesh - this really needs to be tidied/simplified (see below)
  5. You cannot run Auto Forced Convection with natural convection as it splits the flow and thermal simulations up. We need them to be linked
  6. I suggest you test this without the lengthy raditaion calcuation first, that is just adding to the difficulty here

tiny surfaces.png

 

Take a close look at the geometry here, it does not look like I would expect, it is not smooth - this is beneath one of the curved 'lids'

 

Strange Geometry.png

 

We also have an infinitely small angle beneath the persons legs where they meet the chair, this ought to be square to prevent the mesher from struggling

 

Infintely small angle.png

 

I hope these points help. Sorry I cannot get it running right away. I tried a manual mesh also which often circumvents some geometry issues but not with this model.

 

Kind regards,

Jon

 

 

 



Jon Wilde
Please use plain text.