Determination of Material Properties for Fire Modelling

Abstract: Fire and flame spread are very complex phenomenon to study. Several aspects need to be considered and treated. In order to be able to do a correct analysis of the flame spread phenomenon, which is necessary to be able to predict flame spread, knowledge of material properties are needed. Material properties can be obtained in different ways. Traditionally material data have been obtained by testing products in a flame spread situation deemed to be reasonably similar to the critical scenario or scenarios that is desired to be predicted. This thesis presents different ways of improving and developing the measurements of material properties. The methods show that there are a number of ways by which material property collection could be improved, both with regards to the actual test set-up and measurements, and with regards to the interpretation of test results. Thermal degradation is an important parameter that it is essential to know for a correct modelling of the mass loss rate and thus the heat release rate. Thermal degradation is, of course, highly temperature dependent, and for all materials it is probably possible to determine a basic thermal degradation rate as functions of temperature. Knowledge of the surface temperature is a vital parameter in order to be able to model combustion as a combination of solid phase and gas phase reactions. Although measuring temperatures has been a common activity among fire researchers for many years, the surface temperature of a combustible material is very tricky to measure. In this work attempts have been made to use new technology in order to find better solutions for surface temperature measurements up to and after ignition. Two large scale applications are described where detailed measurements methods have been used in order to determine basic properties. These properties have then been used for full scale calculations with CFD and comparisons with test result data. The methods do not constitute a full description of all data that is needed for flame spread calculations, nor do they necessarily describe the best methods to obtain material properties, but they show that methods exist that can improve the material data and that this work should be continued.