Mini-symposium S5

S5: Numerical Modelling in Fire Safety Engineering

Co-chairs: 

Xu Dai (University of Liverpool)*

Zhuojun Nan (Delft University of Technology)

Keywords: Fire modelling in computational fluid dynamics (CFD); Structural fire analysis in finite element method (FEM)

ABSTRACT

Recent advances in computational mechanics have opened up exciting new possibilities in how we understand, predict, and design structures in fire. All types of structures are exposed to fire risk, including steel, concrete, steel-concrete composite floor system, timber, and glass.

Current numerical modelling methods, including computational fluid dynamics (CFD) for fire modelling as thermal boundaries, finite element method (FEM) for structural analysis, multi physics coupling via heat transfer, in combination with Artificial Intelligence (AI) and machine learning (ML), is increasingly central to performance-based design for fire safety engineering. 

This mini symposium, proposed jointly by Dr Xu Dai (University of Liverpool) and Dr Zhuojun Nan (TU Delft), aims to bring together leading researchers working at the intersection of structural fire engineering and computational mechanics, to share methods, identify challenges, compare approaches, and foster collaboration. 

Objectives 

• Showcase state-of-the-art numerical methods: in fire safety engineering, including but not limited to travelling fire models, fire-structural coupling, hybrid or surrogate modelling (e.g., AI/ML based), reduced-order and multiscale methods, and models for localised failure and collapse in fire scenarios. 

• Address validation, verification, and uncertainty: since computational methods are only as reliable as their validation, we seek presentations comparing models to experiments (small scale/large scale), methods of uncertainty quantification in input (fire boundary conditions, material properties, fire spread), and sensitivity analyses. 

• Foster cross-disciplinary and international collaboration: bring together structural engineers, fire engineers, computational modellers, and AI/data scientists. Encourage participants to identify shared research gaps, possible joint projects (e.g., between UK and EU/global partners), and standardisation of modelling protocols. 

Potential Topics 

• Travelling fires in large open‐plan compartments, and structural response (recent work by both proposers in CFD and FEM).  

• Predictive models for collapse prediction under fire, including using ML/AI to accelerate simulations. 

• Multiphysics coupling (e.g., thermal, mechanical, fluid flow/smoke spread). 

• Model reduction, surrogate modelling, digital twin frameworks. 

• Benchmarking studies and validation using small/large scale experiments. 

Expected Outcomes 

• Enhanced visibility of numerical modelling challenges and opportunities within UKACM and the broader fire safety research community. 

• A set of recommendations for best practice in modelling workflows, validation, and uncertainty. 

• Networking that could lead to collaborative grant proposals, shared datasets, benchmark problems, and possibly software tool development. 

• High quality submissions from the mini symposium for the special issue of Engineering with Computers following UKACM2026.