A battery fire on a train is not “just another small fire.” When a lithium-ion battery in an e-scooter or e-bike fails, the rail car can behave like a long pipe that moves smoke fast, limits escape options, and compresses decision-making into minutes.

We sit down with Adam Barowy from UL Research Institutes FSRI to unpack new full-scale passenger rail car burn tests using real micro-mobility devices and realistic storage locations. We talk through what thermal runaway looks like before flames, why that venting phase is a crucial warning sign, and what changes once flaming ignition starts. Adam shares the data that surprised even seasoned fire researchers: smoke can spread from one end of the car to the other in about 30 seconds after flaming ignition, floor-level visibility can collapse in roughly two to three minutes, and toxic exposure can become a serious egress limiter on the same timescale as train stopping and evacuation.

We also zoom out to the operational and societal reality. Rail operators want to support first mile last mile travel and riders who depend on e-mobility for work, yet they need policies that actually reduce risk. We cover practical options like limiting device size, avoiding carriage in the first or last car when exits are constrained, improving passenger education, requiring battery safety certifications, and exploring segregation strategies that keep devices away from passengers without pushing the problem underground.

I promised you links, so here they are:

• Summary of the research on trains and batteries
• The full report
• Li-Ion battery safety guide

Cover image created from pictures from their report linked above! Following Adam's recommendation and taking your resources for a creative spin :)

----
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.

In this episode of fire fundamentals we sit down with Professor Simo Hostikka from Aalto University to cover radiation in fires, both from the angle of physical phenomena and ways to model it. In this episode we cover following topics:

• feel less mysterious, from blackbody basics and role of radiation actually does inside the CFD N-S equation.
• Spectrum and emissivity to real engineering outcomes like heat flux, tenability
• Radiation’s two roles in fire CFD: target heat flux and the gas energy source term
• Emission versus absorption and why Kirchhoff’s law is spectral, not just a single number
• Spectrum intuition using Planck, Wien’s law, and why T to the fourth explodes heat flux
• View factors as a hazard mental model for layers, panels, and distance effects
• Why gases are strongly non-gray while soot often looks smooth and easier to approximate
• How FDS uses the finite angle method, why 104 directions exists, and how updates are staged in time, how to manage spatial and temporal resolution of the radiation
• Ray effect and numerical diffusion, when you can see the error and when you cannot
• Other radiation models such as Monte Carlo, and when they are worth it.
  
  

----
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.

While we can get pretty far with a very simple approximation of what a fire is in our fire cfd, at some point our simplications are not enough. And there is a plenty of features and phenomena, for which we simply need a better tool to handle - carbon monoxide, soot, extinction, flashover behavior, and what happens when ventilation disappears. At the IAFSS symposium, we sit down with Professor Bart Merci (Ghent University), fresh off delivering the Howard Emmons Invited Plenary Lecture, to talk about what it really takes to model turbulent combustion in real fires without asking practitioners to become full-time combustion scientists.

We start with the engineering reality check: you do not get unlimited mesh resolution, unlimited runtime, or the luxury of endless sensitivity studies. As Bart says - "you need to pick your battles". That practical constraint shapes everything, from whether LES is a smart choice to how you treat the “unseen” physics inside a CFD cell. Bart breaks down turbulence in plain terms, explains why the largest eddies dominate entrainment and smoke movement, and shows how mesh decisions can quietly decide whether LES outperforms unsteady RANS in practical smoke control and compartment fire problems.

Then we go deep on sub-grid combustion models. We unpack why infinitely fast chemistry can be acceptable in well-ventilated flames yet collapses in under-ventilated conditions, where toxicity, soot, and extinction dominate the risk picture. Bart explains a finite-rate, autoignition-informed approach that uses detailed chemistry offline to tune simplified reactions, then applies flamelet concepts and turbulence measures to predict reaction rates and species production inside each cell, including ignition and extinction behavior without relying on a guessed “critical flame temperature.”

We close with what’s next: validation in compartments, microgravity as a brutal test of “universality,” and why advanced non-intrusive diagnostics could finally improve near-wall heat transfer and flame-surface interaction. If you care about CFD, FDS modeling limits, fire dynamics, and the future of practical fire safety engineering, you’ll want this one.

If you would like to read more on the topic, here is Bart's paper that accompanied his brilliant lecture. Figure 3 is what we discuss at the end of the episode.

----
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.