see, the results indicated in the error for pressure are, well I don't think I know the name for those numbers. But values of XE+118 mean the simulation was diverging. I would recommend you to update BF, if you haven't already, re-run the case and test after.
Also, you can post your mesh settings here. Most other times is mesh quality causing this.
Kind regards,
Theodore.…
سلام آقای جان بیلی کاوه اشکوه هستم اپراتور سی ان سی از ایران وطراح دکوراسیون داخلی اگر ممکن هستش فایل اصلی میزcosinosa baram mail konid adressmail man hast
(kavehoshkooh2290@gmail.com)بسیار متشکرم
Added by kavehoshkooh at 3:25pm on September 11, 2015
nt B2[i] so B1[i]<=0 means no new connections allowed for point i ,so point i is deleted from B1, B2 updated accordingly.
Initialization:
B1: max number of connections x number of points
B2: all the points
B3: nothing (well null or something, need to create the branch)
Algo:
Get first point in B2, get his allowed number of connections N in B1, find N closest points in B2, create lines in B3, update B2 accordingly. Erase points with max connections (including the first point)
Next
Stop when no points available
At end of loop, B3 stores the created lines.
…
he potential of BF to assess such cases. In your search, try and be specific on what you want, cause validation can focus on codes (i.e software environments like OF and Fluent), solvers (e.g. RNG vs kEpsilon vs kOmega, etc.), meshers, and so many more. Additionally, I'm sure there's a lot of CFD studies of Atrium spaces.
Myself, I haven't been involved in any validation studies as I have always used CFD on the practical side of things. Therefore, I always trusted OF since it has been heavily validated over the years.
The beauty of BF, or at least its end goal, is that you can easily test design alternatives directly from a friendlier and possibly better-known environment of Rhino3D.
I would suggest therefore to just try things out. Design your geometry, in this case the atrium, in Rhino. Decide which are the parameters that you wish to investigate and incorporate those to a GH definition that produces different design alternatives for the range of those parameters (i.e. your parametric model). Then run the cases through BF. There's a couple of examples that come with BF and a few others users are providing either here or on github.
I'm afraid trial and error is painful with CFD but it's the best way forward. Also, I suggest you bookmark cfd-online.com and skim through everything in there. Most if not all of what we are discussing has been discussed there.
Good luck!
Kind regards,
Theodore.
…