Received radiation heat flux at target - weird comparison

[gmsafir]

Hi, I have modeled a simple theoretical case of a radiant panel in YZ plane (dimensions a and b), with a DEVC GAS located at a variable distance x with IOR=-1 and which projection on the panel is the center of the panel. This simple case has for empirical solution the solution provided here (multiply by 4 using symmetry) : https://www.thermalradiation.net/sectionb/B-3.html .The comparison with FDS at 10cm or 4cm resolution (like in the Verification/Radiation case target_test.fds) is not very satisfactory, maybe I missed some important RADI parameters. Here is the comparison (Excel file). The difference is increasing with the aspect ratio. The problem I face is that my method usees a threshold, namely the critical exposure received by a buidling, and changing from analytical to FDS makes a large difference in my FSE studies. Could you help me improve this ?
comparison.xlsx
PR.fds.txt
PS: I used FDS version 6.7.8 (did things change largely on radiative heat transfer in the newest version?)

[Er9y714]

IOR is used for solid phase devices. ORIENTATION is used for gas phase devices. The input file has IOR on RADIATIVE HEAT FLUX GAS devices. Can you try with correct parameter?

[gmsafir]

Hi, no change when replacing IOR=-1 by ORIENTATION=-1,0,0

[drjfloyd]

The hand calc is the incident radiation reaching the surface. RADIATIVE HEAT FLUX GAS is the net radiation at the surface and includes emission from the surface and reflection from the surface.

[gmsafir]

hi, I thought was the net radiation (here no convection) arriving to this device considered as a gas phase cell, not as a surface? What would for keywords in order to retrieve the incident heat flux not on a solid surface, but anywhere in the computational domain?

[mcgratta]

You can use a PROP line to set the emissivity to 1 and convective heat transfer coefficient to 0. Then you will have the net radiative heat flux. If you want just the incoming radiation, add $\sigma T_\infty^4$.

Beyond that, be aware that radiation heat flux is grid dependent like everything else. Try increasing the number of angles and the grid resolution. There are studies like this in the FDS Verification Guide.

[gmsafir]

Thanks for your kind answers. We found the problem : we did not take into account properly the transmissivity. Discovering, from the Verfication cases that, to compare with hand calcs, air humidity was put to value 0, we realized that we needed to compare either with transmissivty=100% or to use the same relation to code transmissivity. That solves the discrepancy, and with same transmissivity, FDS outputs are very similar to the hand calc outputs.