Atrium model, false Fan position

[JB]

Hello

I am trying to model a simple geometry of an atrium. I just want to know, which smoke temperatures I could expect and I think that CFAST is the proper tool for this goal. 

It is a long time since I last used CFAST, and I would like to ask for your help and suggestions

• I had to define 9 compartment to properly model the Atrium. One central compartment, Atrium compartment, represents the central "hole" and connects all the other compartments. I have alse created 4 compartments (nord, south, east and west) pro floor in order to represent the geometry. The preliminar results seem quite "realistic" so I think I have chosen the right way. Would you have model the geometry different? Is there other posiblities?
• I have a problem with the ventilation. The air comes in the domain through windows in the lower floor and it is exhausted in the ceiling through fans. I have defined one fan pro compartment in the top floor, Atrium compartment plus nord south east und west, total 5 fans. In Smokeview looks like the fans aren't properly placed (see Picture). I have just defined one fan pro compartment at the desired high (9 m in my case) defined. In this tab there is no option to define an offset. Does anyone an idea, why is this happening? I would expect to see the fans in the middel of the compartments, but this is not the case and I only see 4. Could be that the model is proper defined but Smokeview is not representing the geometry properly?

• I am also becoming the following fail:   

Fatal: Visualization 1 width position is less than 0 m or greater than compartment width.
Fatal: Visualization 0 width position is less than 0 m or greater than compartment depth.

What does it mean? The model seems to run well. Could be the fail related with the previous point, position of the fans?

Thank you very much for your help and suggestions

Juanjo

[Richard Peacock]

I do think it's just a visualization issue.  There is an identified bug in the visualization of mechanical vents which only effects the visualization (we mixed up width and depth position).  It's been fixed but the fixed version hasn't been released yet.  If you can upload your .in file, we can verify that it's just a visualization problem.

Rick Peacock

[JB]

Hello Rick

Thanks for your answer. Attached you find the .in file.

What's about the other issues?

Cheers

Juanjo

[Richard Peacock]

It does look like visualization to me.  Here's a screen grab with the fixed version of CFAST/Smokeview.  The ceiling vents in the upper compartments are placed where they should be, in the center of the ceiling.  At the moment, we do pretty crude placement for ceiling/floor vents and mechanical vents.  Hopefully, we'll introduce offsets for these to place them where desired in an upcoming version of the model.

The error messages came from the 2D visualizations added which specified ALL compartments and a position of 13 m.  Some of the compartments are not that large so the model is giving an error.  I changed it to a simple 3D visualization and the error message went away

We'll be issuing a maintenance release that fixes this in a couple days once we resolve a crash in Smokeview for your test case.

Rick Peacock

[JB]

Hello Richard

Thanks for your answer.

May I have your help to answer my main question, is this the right way to simulate a atrium?

Cheers

Juanjo

[Richard Peacock]

While we cannot comment on the appropriateness of a given simulation, your method is typical of those I've seen to simulate atria. There is a bit of additional uncertainty introduced by having very large "wall vents" between the floors and the atrium since the empirical coefficients for vent flow are based on small vents relative to wall surface area.  Keep this in mind when choosing safety factors for a design.

Richard Peacock