Documentation on `RADTMP`

[johodges]

I have been looking into the absorption coefficient calculation in more detail to better my understanding. The RADTMP parameter is only mentioned in the user guide in Section 14.4 related to absorption and scattering of radiation by particles. My understanding from how RADCAL works is this value is also used in the absorption coefficient calculation with gases and soot. It would be good to discuss RADTMP in Section 14.3 on the absorption coefficient calculation if it is being used here as well.

[shostikk]

You are right. I will add it to 14.3. as well.

[johodges]

Thanks Simo. I had a follow-up question from the documentation of RadCal. My understanding from the RTE section is that the source term used in RadCal uses a black body emission spectra at RADTMP. The absorption coefficient is then the quantity predicted to be absorbed by the heterogeneous mixture of gases along each mixture path defined by the user. This approach makes sense to me if the emission source matches the assumption (i.e., a black body). However, in FDS simulations we are typically interested in the fire itself as the emission source. Wouldn't this lead to fairly different absorption coefficients since the emitter and absorber will have more similar spectral properties than is assumed in RadCal?

[shostikk]

I think I cannot fully understand your question.

In Radcal calculation, intensity I(L) consists of two parts:

1. Transmitted part (emitted by a black wall at RADTMP and absorbed by gas mixture at TMP_G along path L)
2. Emitted part from gas mixture at TMP_G along path L.
   Effective absorption coefficient is a coefficient that gives I(L) when RADTMP and TMP_G are specified.

When we specify local absorption coefficient in RTE solver, we use pick a value that corresponds to local gas mixture and TMP_G equal to local gas temperature. To fulfil Kirchoff's law, we use the same kappa for both absorption and emission terms, but we do not have to do so, because Kirchoff's law only applies to spectral radiation. Technically we could use different kappa for absorption and emission. This could lead to unexpected non-equilibrium situations, though. Have not really tried it.

[johodges]

Sorry the question was unclear. My comment is related to the transmitted part. RADCAL is assuming a full black body emission spectra at RADTMP from a wall. However, in most scenarios of interest in FDS the emission is starting from a gas phase cell with combustion. If we consider a control volume around a cell adjacent to this emission, the source term that is being considered as part of the transmitted calculation will interact with a more limited emission spectra based on the gas phase emission rather than a black body emission. I tried to demonstrate the question in the schematic here.

[shostikk]

To test the idea of 'more realistic incoming spectrum', It might be possible to call RadCall in a different way:
NPT = 2
SEGMENT_LENGTH_M(1) = something related to domain size or characteristic flame size?
SEGMENT_LENGTH_M(2) = PATH_LENGTH
TWALL = 0._EB
TMP_GAS(1) = RADTMP
TMP_GAS(2) = RTMPMIN + K*(RTMPMAX-RTMPMIN)/N_KAPPA_T (inside the T_LOOP_Z loop

PARTIAL_PRESSURE(gas_index,1) should be some characteristic concentration in a flame
PARTIAL_PRESSURE(gas_index,2) should be YY/(YY+MW*X_N2)