Imaging Pyrometry of Smoldering Wood Embers

The wild fire problem in U.S. and across the world has caused the losses of human lives and property. Firebrands can dramatically increase the hazards of wildland fires. While embers have been extensively studied (e.g. firebrand generation, transport, ignition, size, mass, and moisture contents etc.), little is known about their temperatures. Also, all past works failed to measure firebrand temperature relying on thermocouple and IR camera, which are not accurate and have many drawbacks. Therefore, in this dissertation, to address this an imaging stationary ember and airborne firebrand pyrometer was developed using an inexpensive digital color camera. The camera response was calibrated with a blackbody furnace at 600 – 1200 °C. The embers were 6.4 mm maple rods with lengths of 2 cm. Temperatures were obtained from ratios of green/red pixel values and from grayscale pixel values. Ratio pyrometry is more accurate when ember emissivity times ash transmittance is not unity, but grayscale pyrometry has signal-to-noise ratios 18 times as high. Thus, a hybrid pyrometer was developed that has the advantages of both, providing a spatial resolution of 17 µm, a signal-to-noise ratio of 530, and an estimated uncertainty of 20 C. The measured ember temperatures were between 750 – 1070 °C with a mean of 930 °C. Comparing the ratio and grayscale temperatures indicates the mean visible emissivity times transmittance was 0.73. Temperatures were also measured with fine bare-wire thermocouples, which were found to quench smolder reactions, make imperfect thermal contact, and underpredict the mean ember temperature by more than 200 °C. The pyrometry was also performed on a pendulum firebrand with different velocities imitating airborne firebrands in real fire scenario. The temperature increases as the velocity of pendulum firebrands increases. Ratio pyrometry determined mean temperatures of pendulum firebrand between 878 – 1064 C. Grayscale pyrometry temperatures were lower. The relationship between velocities and temperatures were quantified. The pyrometry was additionally performed on smoldering fuels such as a rolled paper, incense, maple rod, ashless filter paper, and rattan sticks with different air jet velocities to explore the smoldering extinction at high air velocity. To summarize, the main achievement of my Ph. D researches was to develop a new diagnostic of ember and firebrand temperature and emissivity and was successfully completed. The results would be a stepping stone to nearby future exploration of wildfire. The hazards of various firebrand materials and moisture contents could be better assessed in different wind velocities. Computational fire models could be improved. The firebrand size could be simultaneously measured with the same device, both key firebrand attributes could be determined in near real time