CFD Online Discussion Forums

CFD Online Discussion Forums (
-   CFX (
-   -   Radiation: definitions (

visatron January 5, 2004 14:34

Radiation: definitions
Hi there,

I have a question on the meaning of the different variables available in POST when performing a calculation including radiation:

I want to know the net heat flux due to radiation through a wall. I have 3 radiation variables available:

Incident radiation

Radiation intensity

Radiation emission

To my opinion, the incident radiation and the radiation emission are the radiative flux into the wall and out of the wall, respectively. So this would mean that the difference between those values should give my desired value, am I correct???? Then one question remains: what is the intensity: to my knowledge this should be a function of the viewing angle, but in CFX it is a single value. It's also not the net radiative influx (ie its not similar to the difference of the emission and incident radiation). Can someone shed some light on my problem??



Juan Carlos January 5, 2004 22:39

Re: Radiation: definitions
It seems you are working with CFX-5.5/5.5.1.. That means you are using the P-1 model which assumes a isotropic radiation field; therefore, the radiation intensity is independent of direction.

Incident radiation is by definition the integral over all solid angles of the radiation intensity which is quite different from the irradiation (incident radiative flux) because of a dot product (cosine) in the integrand. Therefore, it is of no use for the wall radiative heat flux for general cases. When using the P-1 model, it could be used but with lot of care.

Radiative Emission is evaluated as n^2*sigma*T^4.. I cannot recall if it is also multiplied by 4*pi

If you are interested in the wall radiative heat flow at boundaries you can look at the flow summary for I-Radiation in the out file. You must multiply the reported value by 4*pi (please see documentation for out file reports) in 5.5/5.5.1, no longer the case since 5.6..

Regards, Juan Carlos

visatron January 6, 2004 04:00

Re: Radiation: definitions
Thanks. So I have no option to evaluate the wall radiative heat flow in POST? I would like to know this value at my boundary as a function of the position, and I-Radiation gives me total quantities..

And how is this different in CFX 5.6? We have CFX 5.6 available, but we still haven't converted our routines from 5.5.1 to 5.6, that's why we sometimes still use 5.5.1



Juan Carlos January 6, 2004 13:34

Re: Radiation: definitions
In CFX-5.6:

1 - the summary for I-Radiation already includes the 4*pi factor.

2 - in Post, you can visualize the Wall Radiative Heat Flux field on walls as the name implies. Using the calculator you can compute an AreaInt of the WRHF on a given boundary and compare to the total value reported in the output file

3 - The Radiation Emission is no longer available. Instead the Volumetric Emitted/Absorbed Radiation fields are written to the results file.

4 - Two additional models are also available: Discrete Transfer and Monte Carlo.

5 - There is full support for non-gray media using either Multiband, or the Weighted Sum of Gray Gases.

6 - If using MC, you can include radiation through solids as well.

7 - if using DT/MC you can include radiation source in subdomains and boundaries.

8 - ....Please read the solver theory/modelling sections.

visatron January 7, 2004 15:02

Re: Radiation: definitions
Thanks Juan,

Just one more small question, couldn't find it in the manual:

Is the following true for CFX 5.5.1?

Wall Radiative Heat Flux=Incident Radiation-Radiative Emission (At the walls of course)




Juan Carlos January 7, 2004 22:11

Re: Radiation: definitions
Sorry, but no..

Look in the manual for the boundary condition used for the P1 model.. From there you should be able to estimate (not exactly as the solver) the Wall Radiative Heat Flux in Post..

Juan Carlos..

visatron January 8, 2004 03:59

Re: Radiation: definitions
Thanks. I looked in the part referenced by you. I found that the incident and outgoing fluxes are given by 2 equations, in which T_r, q_r, q_r^in and q_r^out are involved (page 465). Solving these equations for q_r (the net flux) and eliminating T_r gives me:


So when q_r^out is the emitted radiation and q_r^in is the incident radiation this would mean that the radiative wall heat flux=emitted-incident radiation...

Juan Carlos January 8, 2004 16:25

Re: Radiation: definitions
Nothing wrong with your algebra, but with the interpretation of the Incident Radiation/Radiative Emission fields.

In different terms, you could say that radiative heat flux is = coef * (Emitted radiation - Absorbed radiation), where

Emitted radiation = emissivity * sigma * n^2 * T^4

Absorbed radiation = emissivity * pi * Radiation Intensity

Coef = 1. for models different than P1

Coef = 2/(2-Emissivity) for P1

Neither Radiative Emission (4*sigma*n^2*T^4) or Incident Radiation (4*pi*Radiation Intensity) are the Emitted Radiation or Absorbed Radiation I just described...

Based on Radiative Emission/Incident Radiation you can then write the radiative heat flux as:

q_rad = coef * (Incident Radiation - Radiative Emission)


coef = emissivity / 2*(2-emissivity)

I may have inverted the sign convention, but it is a trivial change from here..

Still, the values you will get from CFX-Post will not exactly match those in the outfile because Incident Radiation/Radiative Emission are vertex (conservative) and are not based on the wall temperature which is slightly different..

For a detailed explanation of the P1 bc model, you could also look at Radiative Heat Transfer (1st/2nd Edition) by M. Modest.

Hope this is clear, and does not make it more confusing..

Juan Carlos

All times are GMT -4. The time now is 21:12.