Determination of Woody Fuel Flame Properties by Means of Emission Spectroscopy Using a Genetic Algorithm
Résumé
Because radiation from flames is often the dominant mechanism for wildfire spread, detailed information on flame properties is required. The proposed procedure combines a spectrally resolved radiation model for simulating the line-of-sight infrared emission intensity and spectroscopy data and uses a genetic algorithm (GA) to determine a set of flame properties, allowing optimal agreement between model and outdoor experiments. GA calibration and sensitivity analysis were conducted using well-defined reference flames. The combined GA/radiation model was used with emission data to estimate the effective properties of flames from the combustion of woody fuel beds from 0.5 to 4 m in thickness. Experimental results show that the contribution of soot particles to flame emission increases with flame thickness. The GA was found to be robust and efficient in providing relevant flame properties from line-of-sight intensities on the infrared spectrum of radiation.