Modelling of the behaviour and impacts of fire

The Modelling of the behaviour and impacts of fire group is comprised of Professor Emeritus in Physics, Jacques-Henri Balbi, two senior lecturers in Physics (62nd Section), Jean Louis Rossi and Thierry Marcelli and a senior lecturer in Applied Mathematics (26th Section), François Joseph Chatelon.

The group's work entailed the development of simplified physical models primarily characterised by extremely rapid calculation times (quicker than in real time) while being able to take into account the most possible physical phenomena.

Thus, these models (fire spread, flame, calculation of safety distances, etc.) can easily be integrated into various tools (simulators, tablet and smartphone software) that can be used in operational situations.

This is the group’s historic topic. 2007 saw the first version of the surface fire spread model launched. The model did not take into account radiation as a means of heat transfer. The model was improved in 2010 by improving the means of taking into account cases where the wind direction and ground slope are not parallel to one another. This version is currently that used by the ForeFire simulator.

Further improvements have been made since then. Let us cite for example the disregarding of non-universal parameters of the model, better means of taking into account the speed of the trailing edge and even the definition of a sub-model of flame height.

Consideration of convective effects is a research aspect initiated in 2010. This work led to a new model of purely convective propagation tested in very specific cases of plant strata (well-ordered and vertically-positioned strata) and thus on the integration of this convective model in the radiative model to obtain a comprehensive model that can be used on any vegetative bed and at any level.

The objective is to obtain a fire behaviour model encompassing knowledge about the physical characteristics of the flame front (slope angle, length, height, temperature, etc.) that ventures beyond the rate of fire spread.

The results yielded by the numerical codes realised internally are compared with those provided by other CFD codes similar to the FireStar or WFDS type.

---

This topic concerns knowledge about the rate of fire spread on flat ground without wind, and the effects of certain parameters on failure of a fire to spread. For example, the effects of plant matter’s moisture content on fire spread are only partially understood. Also, one of the objectives would be to release a critical threshold of the moisture content according to leaf surface area which would be an interesting development in the domain of forest fires.

A scientific partnership with the Aix-Marseille University was formed in order to unite researchers working in related fields.

Fires referred to as "eruptive fires" are one of the major causes of human loss in firefighting efforts. They manifest through a very significant acceleration of the flame front in an unexpected manner. Similar to Portugal, Spain, Croatia and even the United States, Corsica experienced a tragic accident in 2000 with the death of two firefighters in the Palasca municipality (in Balagne), when an entire mountain face suddenly set alight.

Our work focuses on two aspects: an explanation of the phenomenon and prediction of when it is likely to happen.

Our explanation of the phenomenon follows the interpretation that was given at the University of Coimbra, namely a phenomenon of feedback due to the convective flow caused by the flame. Although in most cases this feedback balances out, on occasion this may not be the case and we then observe an acceleration of the rate of spread which causes a change in fire behaviour. The issue is then to identify which conditions lead to this imbalance and thus cause the fire eruption.

First of all, we provide a hazard condition which, depending on the characteristics of the vegetation, the slope of the ground and the flame front, makes it possible to know if the danger of eruption exists. Once this condition has been verified, the numerical code allowing for the angle of the slope on which the eruption will take place can be determined.

The purpose of this topic is to accurately evaluate the safety distance between a flame front and a given target. In a first stage, a study was conducted based on the scenario that the target was a fireman equipped with firefighting equipment. This study led to the development of an analytical model of safety distances which, used in conjunction with the fire behaviour model, allowed for the creation of DIMZAL, software designed to calculate the size of firebreaks that can be used on a tablet or smartphone. This software allows for verification of the value of these distances that have been evaluated either empirically or expertly. In terms of town and country planning, the tool would make it possible to identify "at risk" zones, to position and calculate the size of structures to reduce the intensity of a spreading fire, and the withdrawal of firefighters in secure zones.

In a second stage, our work focuses on quantifying the impact of the fire on a target other than a fireman in a firefighting situation. This could be a building for a study on the wildland/dwelling scenario. An understanding of the amount of heat impacting a home would allow one to determine how much vegetation should be cleared around homes as required by the regulations in force. This work is also important from an economic perspective.

---

 

The purpose of this study, carried out in partnership with our chemist colleagues, is determining the kinetic parameters of thermal degradation for plant matter comprising the Mediterranean landscape (Erica arborea, Cistus Monspeliensis, Arbustus unedo, Pinus Pinaster).

To this end, an innovative "hybrid" kinetic method was developed in the laboratory. This method relies on the identification and separation of the exothermic peaks recorded through plant sample analyses using the Differential Scanning Calorimetry (DSC) method through two separations of two distinct processes. The first separation is purely experimental, while the other is numerical in nature. The latter makes it possible to isolate the curves and to provide a mathematical expression for them. These analytical expressions make it possible to calculate the activation energies of related reactions.

In addition, another part of the study concerns providing simplified global kinetics for the degradation of different species typically found as part of Mediterranean vegetation, with the ultimate goal of integrating such into a fire spread We have put forward modelling capable of providing the analytical expressions of the temperatures of solids according to the degree of progress of the related reactions.