Visualization Error with Wall Vents

[tGagnonCorvid]

Hi all, sorry if I'm overlooking something simple here but I think the following behavior is not intended. I was noticing some simulations in which a defined wall vent was always being visualized on one of the compartments regardless of which compartments are defined as the first and second compartment. It is my understanding that this does not affect any of the calculations (the visualization only serves as visualizations), but if that is not the case, I would like to get some eyes on this. The following input file consist of some compartment geometry and two wall vents, each specified between "room3air" and "hallwayair". Upon running and visualizing this simulation both wall vents are placed in "room3air".

vent test.in.txt

Update: I have found an error message which I originally missed which explains this behavior.
"Error: Wall vent 2. Compartment order is incorrect. Wall vents must always be connected with lower numbered vent first."

Is there any way to bypass this? I am generating these *.in files automatically and it is not feasible to parse the compartment order during creation. As a visual artifact this is fine, but will the vent definition between these incorrectly ordered compartments affect the simulation results?

[rmcdermo]

@preneke Can you take a look? Thanks

[drjfloyd]

Why is not feasible to parse the compartment order? In the file you attached you are writing COMP before you write VENT which means you have the COMP order before the VENT lines are written. Couldn't you build an array of COMP names as you write them and then before writing a VENT line test if the first COMP in the VENT has a lower position in the COMP name array. If true you write the VENT with the current order and if false you write the VENT with the opposite order.

[preneke]

The problem is in CEdit and not CFAST proper. I think this can be fixed but it isn't going to be fast. It is only going to impact the visualization and not CFAST. If you can't get the order correct in your program to generate the cases then if you do it by hand you can run a few cases to see the visualization. It may take some time to fix this. I am sorry about the problem.

[tGagnonCorvid]

Thank you! Given that this does not affect the simulation results; I am not too concerned about the visualization!
However, I did want to clarify that the crux of this issue is that regardless of the compartment order specified by the vent definition, both vents are always visualized in whichever compartment is listed first in the "!! Compartments ". Listing / parsing the compartments in a different order does not resolve this issue, as demonstrated by the following examples.

for the following file, both vents are visualized in "COMP_ROOM3AIR"
vent_test.in.txt

for the following file, both vents are visualized in "COMP_HALLWAYAIR"
vent_test_invert.in.txt

Anyways, thank you all for your help!

[gforney]

sorry for the confusion. this was a problem with cfast not cedit. if a vent connects an inside room to the outside then yes the first compartment index needs to be smaller that is the inside room but if both compartments are "inside" then the order doesn't matter. cfast was always putting the compartment with the smaller index first even if neither compartment was the outside. If you can build cfast then update your cfast repo and rebuild. Otherwise a nightly bundle with the fix should be available later today or tomorrow.

let me know if the following image looks correct to you.

[gforney]

try cfast posted here
https://github.com/firemodels/test_bundles/releases/tag/CFAST_TEST

[gforney]

did this fix the problem you were seeing?

[tGagnonCorvid]

Hey, sorry for the late reply. This did appear to fix the visualization issue for the most part. However, and I didn't have any real time to investigate this but wanted to inform you, I began running into issues with cedit as a whole. This is likely due to path variable issue on my part or something similar, but, after installing that sprint build, CEDIT began taking 10's of minutes to open. I also experienced some issues with loading zones.csv in smokeview in some cases. I have since reverted the changes and returned to cfast 774, but if I find some time in the near future I would like to troubleshoot this further. Thanks for the help!