著者
斉藤 雄一 堀内 三郎
出版者
一般社団法人 日本建築学会
雑誌
日本建築学会論文報告集 (ISSN:03871185)
巻号頁・発行日
vol.283, pp.150-159, 1979

We made a research on the fire spreading speed of wooden houses in the downwind eddy behind a fire-resistant building. This report runs as follows : -(1) Eddy areas were observed carefully, and a reattachment point were obtained in connection with the aspect ratio of the building. The reattachment point is the downwind limit of the eddy. (2) Distributions of the wind velocity in the eddy were searched. Its maximum value appears in the middle point between the building front and the reattachment point. (3) The relations between V_∞ and V_i were investigated in the wind tunnel. V_∞ means the velocity of wind which blows in the sky higher than the building and is free from the influence of it. V_i is that of the inverse flow in the downwind eddy behind the building. A formula combining V_∞ and V_i was obtained. (4) The buring tests were made in following two ways A & B. A : fire spreading speed from the ignited crib to the adjacent downwind crib. B : fire spreading speed from the adjacent crib to the further adjacent crib. Each case of A & B was done with 3 scaled cribs (i.e. 1/200, 1/100, 1/80 of full scale). (5) The relationships of experimental fire spreading speed in regard to above mentioned 3 scales were obtained by the method of least square. (6) The relationships of (5) were extended to scale 1/1 (full size) by exrtrapolation. Comparing the results gotten by extrapolation wite those by an actual fire spreading speed formula (for instance, Hamada's formula) using V_i instead of V_∞, the former proved to be about 50% of the latter. (7) The fire spreading formula obtained from the above results was applied to actual fires, and the obtained values were compared with the actual values of the fire data. They proved to be approximately equal. Conclusion : The fire spreading speed of wooden houses in the downwind eddy behind a fire-resistant building are scarcely about 50% of ones by another fire spreading formula, using inverse wind velocity V_i instead of fair wind V_∞.