Fire Safety Science Digital Archive

This paper describes results of preliminary measurement of burning rate (mass loss rate) change In a 0.85 m high, 0.82 m wide, 1.05 m long compartment, a roughly one third scale residential room. Interior wall surfaces including the ceiling were partially covered with 12 mm thick wood to simulate a room fire. These woods were the fuel for the fire. The mass loss rate of wood in the compartment was measured by electrobalances. Wood inside the compartment detached from the outer walls of the compartment and stood independently on the special tray. Thus the mass loss rate was precisely measured by electrobalances. Each outer surface of the compartment was covered with two layers of insulation board to obtain a highly insulated condition. The compartment had a small opening in the middle of the front wall to realize a low-ventilation condition. A rapid increase of burning rate was observed. Two distinct behaviors were observed. One was related to so-called pre-backdraft and the other was to backdraft. It was found that both phenomena needed high mass loss rate before they happened. As there were no big differences in the mass loss rate changes for both phenomena, it seems that both phenomena strongly depended on wood pyrolysis characteristics. Another special experiment was carried out by replacing the opaque insulated side wall with a refractory glass window. Fire ball growth or flame propagation during backdraft was observed visually and recorded by a video. It was found that a small flame appeared near the ceiling at first. Then the backdraft flame propagated through the compartment consuming oxygen and the accumulated excess pyrolyzate in the compartment.