If someone ever tells you they have modelled fire spread for a commercial project, with 20 cm grids and using generic materials from (old) FDS database, please do me a favour and redirect them to this episode. Because modelling fire is a seriously challenging thing. And by modelling, I really mean it.  Not to apply a surrogate source based on a statistical overview of how fires looked like in the past for fuels of this kind. Not to omit half of the phenomena because they are too hard and in cone they did not matter anyway... To really model the fire.I have invited prof. Lukas Arnold from Bergishe University Wuppertal and Juelich Forszungszentrum to tap into how challenging modelling even simple materials may be. What are the feedback loops and phenomena one must account for, and how do different scales give us different parts of the answer we need? If you would like to know more about Lukas efforts in modelling, please check two recent papers:- Paper on cone calorimetry of different types of PMMA (the material we have discussed in the podcast episode) - Paper on inverse modelling the pyrolysis kinetics - giving insight into how one obtains the material data you may need for modelling.