Thumbnail Image
Publication or External Link
Baldwin, James H
Sunderland, Peter B
A major mechanism for wildland fire spread are spot fires, where small combusted organic particulate (firebrands) are lofted and transported to a remote location where they can then ignite new fires. The modeling of these spot fire ignitions is limited by the unknown surface temperature and emissivity of firebrands, which is challenging to measure due to the small size of firebrands (precluding the use of intrusive temperature methods such as thermocouples) as well as the dependency of conventional non-intrusive temperature measurements (e.g. Infrared Imagers) on emissivity. A solution to this is presented in Color Pyrometry, which uses color pixel intensities to determine an object's temperature based on a calibration against an object of known temperature/emissivity. The presented method is a Ratio Pyrometry approach between green and red pixel intensities normalized to camera settings, which demonstrates the benefit of being independent of object emissivity as validated by Planck's Law, and is based on a Blackbody Furnace calibration. To determine the method's applicability to realistic firebrand imaging conditions, which would provide the most comprehensive understanding of firebrand ignition, the individual impact of firebrand movement speed on the pyrometry's surface temperature predictions is considered. An apparatus is developed that decouples firebrand movement speed from the surface wind speed (which is known to impact firebrand surface temperature) as well as allows for modulation of the firebrand's simulated movement speed, and involves rotating the imaging device about a fixed axis relative to a stationary firebrand. Five trials at a set orientation were conducted to verify the apparatus' repeatability, and subsequent trials of varying rotation speed, distance, applied wind speed, and mounting orientation were conducted. Both qualitatively and through a statistical analysis consisting of ANOVA and non-parametric distribution testing, firebrand movement speed and orientation are shown to have no individual impact on surface temperature. Average ember surface temperatures were found to be 922.1 ± 20.4 °C with a 1 m/s applied wind speed and 955.0 ± 20.2 °C with a 2 m/s applied wind speed, which is in agreement with previous studies. It is proven that the presented Pyrometry method's results are independent of a major complicating factor associated with realistic firebrands, which thereby further supports future efforts into wildland fire spread modeling.