Abu-Bakar, Ariza Sharikin (2015) Characterization of Fire Properties for Coupled Pyrolysis and Combustion Simulation and Their Optimised Use. PhD thesis, Victoria University.

Abstract

Performance-based fire safety system design enables fire safety engineers to assess the performance of buildings and components by taking into account various scenarios of real fires and their severity. This approach offers more flexibility to the fire safety engineer to adopt new design concepts such as aesthetic values, material and energy efficiencies, while complying with regulatory building codes. For such an approach, the testing of fire safety systems either experimentally or numerically is essential and since experimental studies are vastly expensive, fire models are usually used. State-of-the-art Computational Fluid Dynamics (CFD) based fire models typically include pyrolysis and combustion sub-models for predicting fire growth and spread in addition to background mass, momentum and energy balance sub-models. Separately, the phenomena of pyrolysis and combustion of materials are quite complex and during a fire both phenomena occur simultaneously thus compounding the complexity. Simulating these phenomena relies primarily on the prescribed fire properties of combustible materials and any error in providing the input fire properties data may affect the prediction of the fire behaviour. This study has been conducted to characterize the fire properties for coupled pyrolysis and combustion simulation. The overarching objective is to find, given an unknown or novel material, how would a user go about quantifying the representative fire properties and use them optimally? A range of experimental techniques and where necessary, data post-processing methods have been established, developed, selected and implemented to determine critical fire properties. Two bench-scale instruments, the cone calorimeter and hot disk analyser, as well as two miligram-scale instruments, a thermogravimetric analyser and differential scanning calorimeter, have been used for this purpose.

Search Google Scholar

Repository staff login