Hi everybody,
I would like to have your thoughts on the way we model fire.
So, I have this fire analysis to complete and I am running multiple test on the side to have a better understanding of what the software does with each heat transfer mode, since I cannot stop having unreal smoke temperature results.
My main problem is how to obtain relevant temperature values without changing my volumetric heat load BC randomly. When I say unreal or relevant, I mean regarding fire standards available (NFPA, BR, CIBSE) and fire simulations that have been already done (NIST). And even if I do choose this guessing type of approach, I dont know how I would justify the heat load value.
I read everything I could on the proper way to model a "fire", such as a resistance part with high conductivity surrounded by suppressed walls:
- Resistance part is heat up thanks to the heat generation BC,
- Air is allowed to pass through the resistance part, and air take away some heat with it,
- Air has also its own conductivity and take away some heat by conduction as well,
- Surrounding walls do not participate to the heat transfer because they are unmeshed.
But the problem is no one ever point out a temperature problem neither talk about radiation heat transfer mode in this type of approach, and according to me, not the one you want to forget when modeling a fire.
So on my side, resistance part always remains too hot, and as a result, produce high temperature profile all around the plume. And I am not able to fit any standards' design requirements.
In Autodesk CFD, a resistance material assigned to a part cannot transmit radiation heat loads. I know it can has an emissivity value assigned, thanks to the air, but it does not help any further.
So to model a fire in another way, what do you think of the following approach instead:
- Forget the resistance part and the suppressed walls
- Model a cylindrical solid material instead, with custom emissivity and conduction values
- Assign a heat load BC, scalar, and allow radiation heat loads to be calculated
Thus, we give up the idea of air going through an imaginary part which heats up and slows down the air, but we only autorise the air to flow around a hot burning material representing the fuel itself (as it would be if a solid material surrounded by flames was burning in real life). I think that way would allows radiation to be taken into account.
I'd love to hear and learn from the community who has ever worked on a fire analysis in Autodesk CFD, so please, fell free to comment
