Modeling thermal decomposition of materials in the event of fire

To be able to predict the fire behavior of natural or synthetic combustible materials, it is necessary to have a detailed understanding of the production of combustible gases by a condensed phase during thermal decomposition. Indeed, the flame is fed by various gaseous species escaping from the solid during pyrolysis. The kinetics and nature of these gases depend on a number of parameters, first and foremost the temperature and atmosphere in which the phenomenon takes place.

Establishing a relationship between the local temperature (resulting from the local energy exchange balance), the local atmosphere and the gases produced that are likely to supply the flame with combustible vapours is therefore an essential step in fire modelling. This involves first identifying the physico-chemical phenomena involved. Historically, several techniques have been established, grouped into two main families: the isoconversion approach and the modelling approach. The difficulty of modeling lies in the complexity of the phenomena involved and the simplification assumptions made. The isoconversion approach is based on a graphical analysis of thermogravimetric analysis results, while the modelling approach calls on an assumed reaction scheme and the resolution of a pre-established kinetic model. Current numerical tools, based on these different methods, have enabled the development of techniques for determining the kinetic parameters of these complex systems.

Algorithms linking temperature to pyrolysis rate are beginning to be implemented in numerical fire modeling tools. Ultimately, an approach based on the study of the main materials involved in a fire and the determination of the associated kinetic parameters provides a satisfactory representation of fire behavior at the material level. Additional analyses are then required, such as the introduction of different physical phenomena, in order to represent behavior on the scale of the system, i.e. the real fire.