HT3D heating from center of OBST rather than side

[johodges]

Describe the bug
I am seeing odd heating behavior using HT3D in the attached. The heat source is an INTERNAL_HEAT_SOURCE within the 'CELL' HT3D OBSTs as well as an EXTERNAL_FLUX on the bottom of half of them. These are connected to vertical steel sections which then connect to a horizontal steel section above the heat source. The model is configured with SOLID_PHASE_ONLY=T so all the heating is through the IHS and HT3D. Near the end of the simulation you can see the base of the second vertical wall on the other side of the heat source starts to heat before the horizontal section near it.

Screenshots

Desktop (please complete the following information):

• OS: Linux
• Version: latest master

Additional context

rack_simple.txt

[mcgratta]

SOLID_PHASE_ONLY=T does not turn off radiation. I'm not sure why it doesn't, but until I find out, add

&RADI RADIATION=F /

[johodges]

Thanks, I can confirm that was the issue with this input file. However, that was a simplified case to demonstrate a problem with HT3D I was seeing in something more complex. I am still seeing the issue in the more complex case with RADIATION=F. Let me simplify this one a bit more and then I will upload it.

[johodges]

Here is an example showing the bug I was trying to report. You can see the oddity in the temperature distribution on the RHS where the vertical wall meets the horizontal one. This one has SOLID_PHASE_ONLY=T and RADIATION=F

rack_simple.txt

[mcgratta]

Replacing the current OBST for 'TOP' with

&OBST ID='TOP', XB=0.0,0.49,-0.0425,0.5525,0.28,0.315, MATL_ID='STEEL', SURF_ID='STEEL', CELL_SIZE=0.02125 /

solves the problem, but I need to better establish rules for this. This has to do with relatively thin obstructions criss-crossing, rather than a solid obstruction.

[johodges]

Thanks for looking into it. It might be good to have a call where I can give you a bit more detail on the modeling approach we are using for the battery rack. That additional context may help as we figure out how we want the model to handle this type of behavior. My schedule is flexible after 10am if you have some availability.