Millar, Darin John2013-09-172013-09-1719951173-5996http://hdl.handle.net/10092/8282http://dx.doi.org/10.26021/2007The purpose of this project was to investigate the conditions that exist in a typical residential compartment fire when subject to limited ventilation. Emphasis was on the conditions that leads to a backdraft from natural fuel and the circumstances prior to a smoke explosion. Also included within this report is a comparison of the results with those expected from CFAST, a computer zone model. The first phase of the project was to design a full scale apparatus to safely carry out the experiments. The second phase consisted of seven experiments with variations in ventilation size and location, and fuel elevation. The fuel used was timber cribs that provided a fuel load density of 340 MJ/m². Temperatures, mass loss, pressures and O₂ concentration within the compartment were monitored. The areas used to represent leakage ranged from 0.0154m² to 0.0079m². Ventilation was modelled using two horizontal circular vents located at lowest and highest practical points. In each experiment the vent at the bottom was twice the area of that at the top. Under the conditions experimented no backdraft occurred. During one experiment (number 6) a smoke explosion occurred that caused an instantaneous temperature rise of 140 °K and pressure increase of 30 Pa. Further research in this phenomena is required. Comparisons made with CFAST indicate that it is inadequate in coping with limited ventilation fires in a single compartment situation.enCopyright Darin John MillarReports