065 - Understanding mesh sensitivity and model uncertainties with Jason Floyd

Fire Science Show

Will a higher resolution mesh make my CFD more accurate? That is a harmless question, and most of us would tend toward 'I guess yeah'. But let us try and unpack this. Into atoms! What does higher resolution mean? How exactly solver deals with increased spatial discretization and what are the exact consequences of that? What is a high resolution for a tiny orifice and what is a high resolution for a road tunnel? But it gets better... What makes CFD more accurate? Is it better alignment with experimental data - if so, do you know the experimental and numerical uncertainties that allow you to actually compare them? If not, how can you tell if the second try in the mesh sensitivity study is a disturbing result or something well within the numerical uncertainty?
Oh boy, such a simple question and so many things to answer.
And you have guessed it - that is what we are trying to do in this podcast episode with dr Jason Floyd of the UL Fire Safety Research Institute. And on top of that, you will learn a ton about mesh sensitivity and model uncertainties. You will also know why some models are more difficult than others - especially when you start to play with fluid-solid interaction and pyrolysis.
This episode was inspired by Bryan Klein - props to Bryan! He was a guest on the podcast and we have covered a very similar topic in it - you may want to listen to it as well! The trigger to make this episode came from the issue tracker, and you may want to check that thread as well
If you have some great podcast episode ideas, let me know! I'll get this done, I'm doing this whole thing for you.

----
The Fire Science Show is produced by the Fire Science Media in collaboration with OFR Consultants. Thank you to the podcast sponsor for their continuous support towards our mission.