Multi-Scale Assessment of Polymer Pyrolysis for Fire Spread Prediction: Experiments and Modelling (MAP-FIRE)

• Applicants: Professor Dr. Lukas Arnold; Professor Dr.-Ing. Jochen Zehfuß
• Subject Area: Structural Engineering, Building Informatics and Construction Operation
• Term: since 2026
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 555700177

One of the current challenges in fire science is the prediction of fire growth and flame spread. Based on fundamental physical and chemical processes it enables novel engineering applications, like the prediction of the fire development of complex materials in build environments, e.g. vehicles, green facades or application of sprinkler systems. Here, the thermal decomposition (pyrolysis) of a solid combustible is one of the key mechanisms and prerequisites for determining ignition, fire spread or even extinction. Thus, the overall system to be modeled is involving multiple scales -- from pyrolysis on the micro-scale to the mass- and heat-transfer in the real-scale. In common scientific and engineering approaches, a heat release rate (HRR) is specified as an input of a computational fluid dynamics (CFD)-based fire prediction model. However, this does not represent the actual pyrolysis process and thus fixes the dynamics of the fire. With this approach, the predictive capabilities of fire models, especially for scientific investigations, are limited. Advanced pyrolysis models use an Arrhenius-based chemical reaction kinetics scheme, coupled with conservation equations for heat and mass transport, to describe the decomposition of the material. Yet, the determination of the pyrolysis and thermo-physical material parameters poses a major challenge, as they cannot in general be directly measured. Recent methods are based on a hierarchical approach, where inverse modelling is used to derive effective parameter sets, while increasing the investigated scale. However, new research indicates that the transition in scales may not preserve the assumed parameter sensitivity, which leads to a loss of information. Up to now, there exists no consistent experimental and theoretical coverage of the interacting scales. The proposed project aims to contribute to the fundamental understanding of this system. The aim of the proposed project is to develop scale-independent parameterisation strategies for pyrolysis modelling that enable the consideration of transient boundary conditions, to investigate suitable modelling approaches and to improve fundamental weaknesses of (sub)models as well as interaction problems. It is planned to develop new parameterisation strategies and modelling approaches in an iterative multi-scale assessment. The decomposition kinetic, thermophysical and transport properties are determined on the basis of established and new designed small-scale experiments, such as simultaneous thermal analysis with emission gas analysis, differential scanning calorimetry and micro-scale combustion calorimetry. For the multi-scale analysis, experiments are carried out from the micro-scale to the real-scale (full room fire). A novel medium-scale fire spread experiment is developed, allowing to control heating and ignition. Finally, the performance of the fire spread prediction models is assessed in a large-scale blind validation setup.

• DFG Programme: Research Grants