Defining flashover for fire hazard calculations: Part II

Comparison of available correlations and predictive models used to predict the minimum heat release rate (HRR) necessary to cause flashover show consistent trends for a range of empirical data. Nonetheless, available experimental data for HRR at flashover in compartments of similar geometry and venting show substantial scatter. Both the experimental data and theoretical predictions based on computer modeling indicate that a significant portion of the variability can be accounted for by the time period involved in the flashover. Although typically ignored in the available correlations, qualitatively a clear trend emerges-shorter exposure times increase the needed minimum HRR at flashover, due at least in part to the effects of heat transfer to the compartment surfaces. Additional measurement needs are suggested to facilitate better understanding of conditions leading to flashover. (C) 2003 Elsevier Science Ltd. All rights reserved.