Modelling of the Growth Phase of Dalmarnock Fire Test One

Abstract

The challenge of modelling a well characterized full-scale fire test using computational fluid dynamics is illustrated in this work comparing a priori and a posteriori simulations. In 2006, The Dalmarnock Fire Tests were conducted in two identical 3.5 m 4.75 m 2.5 m concrete enclosures with a real residential fuel load. This data set provides measured data at the highest spatial resolution available from a fire experiment to date. Prior to the tests, an international study of fire modelling was conducted in order to assess the state-of-the-art of fire simulations using a round-robin approach. Each of the seven round-robin teams independently simulated the test scenario a priori using a common detailed description of the compartment geometry, fuel packages, ignition source and ventilation conditions. Most teams decide to use the numerical code Fire Dynamics Simulator (FDSv4). Comparison to the experimental measurements showed a large scatter and considerable disparity (much larger than the error and variability associated to the experiments). The study showed that the accuracy predicting fire growth is poor. A posteriori simulations of the growth phase were conducted afterwards while having full access to all the measurements. No previous fire simulation had this large amount of data available for comparison. Simulations were compared against average and local measurements. The heat release rate is reconstructed from additional laboratory tests and upper and lower bounds for the fire growth are found. Within these bounds and after adjusting uncertain parameters, the level of agreement reached with the measurements was of 10 to 50% for the evolution of the average hot layer temperatures and between 20% and 200% for local temperatures.