Using Inverse Fire Modeling With Multiple Input Signals to Obtain Heat Release Rates in Compartment Fire Scenarios

Abstract

A set of multi-room compartment fire experiments were conducted to obtain measurements of hot gas layer temperature and depth. These measurements were used as an input to an inverse fire model that coupled a genetic algorithm with a zone fire model to calculate a unique solution to the original fire size and door opening used in the experiments. The objective of this research was to calculate simultaneously the real-time fire size and fire door opening of the experiment using a combination of hot gas layer temperature and hot gas layer height measurements from a multi-room compartment in concert with an inverse fire model. This research focused on increasing the robustness of an inverse fire model (IFM) with respect to physical accuracy and multi-variable calculations. The IFM successfully identified a unique solution and calculated fire size within 10-40% of experimental values.