建築物構造防火性能驗證技術之火災持續時間研究

Validations on the Fire-proof Performance of Architecture— the Study of the Fire Sustaining Time

本研究係使用ISO 9705 實驗屋，以固定火源（25kW-200 秒） 對固定燃載量（5kg/ ㎡） 之木堆加熱，用來模擬房間火災，主要變數為可燃物（木堆）密度及分佈面積，以田口實驗計畫法之L9（34） 直交表配置實驗因子，探討此兩因素對火災持續時間之影響，再配合統計方式探討火災持續時間之分配模式以及其標準差，並利用耐火性能檢證法計算出之火災持續時間作比對驗證。在本研究條件下，利用檢證法計算所得之火災持續時間的信賴區間為+2σ（ 標準差） ，經與實驗數據之統計結果比較，實驗結果仍有將近0.4σ 的機率大於計算所得之火災持續時間值，建議取+3σ 可確保安全範圍。由木堆燃燒實驗結果顯示，於相同火載量下，木堆分散會減緩燃燒速度以及熱釋放率，於固定火載量條件下，以t2 火災成長曲線分析木堆熱釋放率成長曲線，發現木堆密度對於火災成長影響較小，木堆火災成長主要受到木堆分佈影響。

Combustible wood cribs with fixed fire load (5kg/m2 ) are heated by a fix fire source (25kW-200s) in ISO 9705 room is used to investigate effect of the density and the distributions of the combustible on the fire sustaining time. L9(34 ) orthogonal table from Taguchi method is used arranging experiment parameters, i.e. wood crib packing density and the distribution. The distribution mode of fire sustaining time and its standard deviation are discussed statistically and are compared with the calculated one. The reliability region of calculated fire sustaining time is found to be +2σ . Statistics from experimental results shows that the experimental fire sustaining time is still about 0.4σ probability longer than the calculate one. Besides, from the results of wood crib burning, we found that at a fixed fire-load the fire growth rate and maximum heat release rate are inverse proportional to the density of the combustible. By using the t2 fire assumption on the growth rate of the wood crib, we found the fire growth rate is less correlated with the density of wood crib, but closely related to the distribution of wood crib.