[Alczem]

Hi all! Thanks to all the people working on OpenFoam and helping the community. I am new on this forum but I have been browsing the topics for quite some time now, and all the answers I find are always appreciated, so thanks for your time.


I have an issue that I can't seem to solve.
For a project, I am trying to use a P1 radiation model with buoyantSimpleFoam, but with no participating media (air inside a closed cylinder). I am aware that viewFactors or fvDom might be more suited and more accurate, but I want to first try this model.


The geometry is a simple cylinder, and one of the extremities generates heat via an externalWallHeatFluxTemperature BC.
The cylinder and the other extremities are also externalWallHeatFluxTemperature BCs, with ambient temperatures.


My issue comes from setting the emissivity and absorptivity in radiationProperties to 0 (or using "none" as the model for AbsorptionEmission"). The heat fluxes (monitored with wallheatflux) no longer match, although with emissivity/absorptivity = 1, they eventually do.

I have reproduced the examples, so all the files (G, boudaryRadiationProperties...) are similar to what we can find in the examples.


Any clue on what is responsible ? Let me know if more details are needed, and thanks for your help!

[dlahaye]

What do you mean by "no longer match"? What is your reference?

[Alczem]

With absorptivity=1 the sum of the heat fluxes on all the patches equals 0, but with absorptivity=0 there is a significant difference between the heat flux entering the domain and the ones leaving (sum not 0).

I use these values to try and see if my settings are relevant, maybe I am mistaken.

[derekm]

P1 is well known to be inaccurate with zero participating media [1]

I have validated the Openfoam P1 model and it appears to be for my geometry to be 1.9 times overestimating radiation with zero participating media. I found for my work it was possible to adjust the emissivity to correct for the error with a "real" 0.95 emissivity being replaced with a 0.6 emissivity. This might not suffice for your application.











Sazhin, S. S., Sazhina, E. M., Faltsi-Saravelou, O., & Wild, P. (1996). The p-1 model for thermal radiation transfer: Advantages and limitations. Fuel, 75(3), 289–294. https://doi.org/10.1016/0016-2361(95)00269-3

[Alczem]

Thanks for sharing your results, it clearly confirms what I found when I investigated the issue. I ended up using the fvDOM model and the results are quite close to the experimental data.

My only remaining issue is that I can't seem to make cyclic or symmetry boundary conditions work with the fvDOM model, whereas the P1 model seemed easier to set up with such conditions. I am still looking but I jut don't have the time right now

Anyways, I appreciate the feedback, cheers!

[derekm]

Originally Posted by Alczem

...I can't seem to make cyclic or symmetry boundary conditions work with the fvDOM model,...

They dont work at all with cyclic or symmetry boundary conditions and FVDOM is expensive in computation

[Alczem]

Originally Posted by derekm

They dont work at all with cyclic or symmetry boundary conditions and FVDOM is expensive in computation

Oh well, I guess I won't have to think about it again, thanks for clarifying once and for all. Let's hope they manage to implement these features in the future updates, the DO model in Fluent works with symmetry and cyclic boundaries.

[altinel]

Originally Posted by derekm

P1 is well known to be inaccurate with zero participating media [1]

I have validated the Openfoam P1 model and it appears to be for my geometry to be 1.9 times overestimating radiation with zero participating media. I found for my work it was possible to adjust the emissivity to correct for the error with a "real" 0.95 emissivity being replaced with a 0.6 emissivity. This might not suffice for your application.











Sazhin, S. S., Sazhina, E. M., Faltsi-Saravelou, O., & Wild, P. (1996). The p-1 model for thermal radiation transfer: Advantages and limitations. Fuel, 75(3), 289–294. https://doi.org/10.1016/0016-2361(95)00269-3

Hello Derek,

could you please tell me where you actually set the "emissivity" in OF-case? In the boundary conditions as a surface emissivity (in 0-Folder) or in the constant/radiationProperties?

The reason I ask is that there is also an "emissivity" in constant/radiationProperties, but it has the unit 1/m, which is a bit confusing.

Thank you very much!

[derekm]

Originally Posted by altinel

Hello Derek,

could you please tell me where you actually set the "emissivity" in OF-case? In the boundary conditions as a surface emissivity (in 0-Folder) or in the constant/radiationProperties?

The reason I ask is that there is also an "emissivity" in constant/radiationProperties, but it has the unit 1/m, which is a bit confusing.

Thank you very much!

constant/radiationProperties


it has those dimensions so it can work for Fluids as well. For solids very confusing

[altinel]

Originally Posted by derekm

constant/radiationProperties


it has those dimensions so it can work for Fluids as well. For solids very confusing

Hello Derek,

thank you very much for your quick reply!

So, for a case with multiple regions (one solid-region and one fluid-region), can I directly consider the emissivity of my solid in constant/solid/radiationProperties to be dimensionless? I just need to get the emissivity of my solid (e.g. aluminum) and then set this dimensionless value in constant/solid/radiationProperties as emissivity?

So does the unit 1/m refer only to the length dependence of the emissivity of fluids?

[derekm]

Originally Posted by altinel

Hello Derek,

thank you very much for your quick reply!

So, for a case with multiple regions (one solid-region and one fluid-region), can I directly consider the emissivity of my solid in constant/solid/radiationProperties to be dimensionless? I just need to get the emissivity of my solid (e.g. aluminum) and then set this dimensionless value in constant/solid/radiationProperties as emissivity?

So does the unit 1/m refer only to the length dependence of the emissivity of fluids?

its what has worked for me.

[derekm]

Originally Posted by altinel

Hello Derek,

thank you very much for your quick reply!

So, for a case with multiple regions (one solid-region and one fluid-region), can I directly consider the emissivity of my solid in constant/solid/radiationProperties to be dimensionless? I just need to get the emissivity of my solid (e.g. aluminum) and then set this dimensionless value in constant/solid/radiationProperties as emissivity?

So does the unit 1/m refer only to the length dependence of the emissivity of fluids?

That is correct when you specify the solid radiation model as opaque

[altinel]

Originally Posted by derekm

That is correct when you specify the solid radiation model as opaque

Hi Derek,

okay, thank you very much for the clarifications!