Projet Feux | UMR SPE 6134
Events  | Seminars

Team seminar of 29 May 2017

During the course of this team seminar, three presentations were given, the titles and overviews of which appear below:

Presentation by Lara Leonelli"Characterisation of smoke released by controlled wildland fires" »

Overview: Within the framework of my thesis work, the composition of smoke released by wildland fires was studied, with their toxicity potential measured: In the laboratory, the compounds present in the smoke released through the combustion of two plant species, namely the Montpellier cistus and the Eagle fern (known for its toxic nature) were identified and measured. Emission factors were calculated for all the compounds identified. The aim is to provide input data for fire simulation models to the scientific community and in so doing predict the quantities of each and their impact on health. These emission factors were calculated according to the combustion phases: the preheating phase, the combustion phase with flames and the combustion phase with sooty residue, without flames. In the field, samples were taken during prescribed burning undertaken by firefighters with the aim of characterising the smoke that they inhale over the course of a typical burning day. The data collected was compared to the data obtained through laboratory experiments. In order to determine the firefighters’ level of exposure, the carbon monoxide exhaled by volunteer firefighters was measured, at the start and the end of the day. The objective is to define potential toxicity through exposure level to smoke and thus measure the impact this smoke has on health in the short and medium term. At present, no epidemiological study exists on the long-term impact of smoke inhaled as a result of wildland fires.


Presentation by François Joseph Chatelon"A convective model of surface fire spread" »

In this research carried out by F.J. Chatelon, J.H. Balbi, D. Morvan, J.L. Rossi, T. Marcelli, we aim to present a simplified physical model of the spread of surface fires based on convection. The objective is to test the model's consistency over specific plant strata, comprised of plant elements that can serve as matches, pine needles and even cardboard, regularly stored and placed vertically. In these examples of vertical strata, convection has been known to be the predominant mode of heat transfer. The propagation of this type of stratum has been studied at length in the literature in an effort to understand the mechanisms, particularly regarding the transfer of heat. The specific geometry of the plant bed allows for radiation from the flame to be disregarded, and convection becomes the driving force behind fire spread. Our modelling considers convection as a displacement of hot gases from the inside of the base of the flame to unburned vegetation impelled by wind speed and rate of rise of the gases. These gases therefore produce a contact flame causing the unburned vegetation to ignite. Radiation from the flame is disregarded unlike the radiation from the base of the flame which, although weak, will be retained in the modelling. This model must be simple and solid enough to allow for operational use. Thus, the calculation time must be lower than real time, which requires a simplified model, in other words, one based on strong hypotheses and therefore without partial differential equations. Moreover, the model must be predictive, in other words devoid of model parameters or with universal model parameters (whose value is fixed regardless of the experiment concerned). The ultimate objective remains the coupling of this convective model with the purely radiative model created and improved at the University of Corsica in order to have a complete model taking into account radiation and convection and that can be applied in all circumstances. This convective model will be put to the test in numerous laboratory experiments (172 fires) with a wide variety of characteristics pertaining to the combustive material. The model is also compared to two simple empirical models in the literature. Standard statistical tools (standardised mean square error, correlation coefficient, fractional bias) are used to measure the relation between predicted values ??and experimental values.


Presentation conducted by Xavier SilvaniMesures thermiques sur réseau capteur sans fil pour la surveillance des sites industriels de stockage de fuel »

Overview: An improvement in the safety of industrial facilities initiated, in 2010, the development of wireless sensor networks for the very rapid detection of a breach in oil parks (or other industrial sites, pipeline crossings, confined premises, etc.). The rapidity of the detection response time would represent real progress made (significant reduction in potential losses, reduction of damage to heritage, reduction in insurance contributions). In the context of the study of wildland fires conducted at the Laboratory for Environmental Sciences (UMR CNRS 6134, located in Corte), temperature and heat flow sensors with wireless communication were developed to track the spread of wildland and measure their thermal impacts on targets (dwellings, humans). An initial wireless system based on the Zigbee protocol makes it possible to take immersive thermal measurements of the fire but the measurements are not time-stamped and the consumption of the nodes is high. EDF R&D’s delivery of the OCARI battery to the University of Corsica helped to overcome the obstacles encountered with the Zigbee and to offer higher quality solutions. Both systems were tested by the Forest Fire team on field-level vegetation fire experiments.

Page mise à jour le 04/12/2017 par VAREILLE MATTHIEU