Fire Safety Science Digital Archive

In designing underground railway systems, fire engineers need to estimate likely fire sizes in the event of a fire involving a train. Traditionally they have done so by taking the complete fire load and dividing by assumed times, and incorporating an assumed efficiency factor. However, fire loads tell nothing about the burning behaviour of materials. Therefore, this approach penalises materials of good fire performance. With new fire test methods, it is now possible to measure the rate of heat release of all materials used in the saloons of trains. But this still does not enable us to predict what size fire will occur in a train, as the link between flammability of individual materials and likely fire size has not been established. At present, simple summation methods are used in some rail specifications. This approach is at odds with what is known about fire growth in enclosures, and may result in grossly incorrect estimates of fire size. CSIRO is conducting a research program that is attempting to establish links between flammability of materials and fire size in train saloons. The goals of the research are to develop techniques for estimating fire sizes that can be used as design fires in tunnel design. So far, the research program has focused on measuring the contribution to fire growth of assemblies of seats and linings. Experiments have been carried out in enclosures designed to have cross-sectional areas the same as train carriages, but of shorter length, and in one actual train carriage. Fire spread and window breakage have been studied. The ISO 9705 room corner test has also been used to study the behaviour of linings. This paper describes experiments that have been performed, and the data obtained. It discusses current methodologies for estimating fire size and compares predictions with experimental results.