The combustion of solid polymers, is a complicated process involving physical and chemical phenomena that are only partially understood. The burning behaviour of polymeric materials is understood in terms of their ability to generate flammable volatile liquids and gases under the action of heat and their subsequent ignition. Kashiwagi (1994) has highlighted the depth and complexity of the physical phenomena that are responsible for the production of volatiles from a solid under the action of heat. The importance of subsurface pyrolysis in practical fire-test conditions has been demonstrated experimentally for the case of polymethylmethacrylate (PMMA) (Kashiwagi and Ohlemiller 1982; Vovelle et al 1987). Mathematical models that tackle the difficulties associated with solid- and gas-phase chemical and physical processes have been reviewed by Di Blasi (1993).
Although the phrase `polymer flammability' is widely used it has no intrinsic scientific meaning; the flammability of a given polymer depends upon both the physical state of the product in which it appears and the particular fire scenario that is considered. Thus the number of fire tests in use is at least in the hundreds (Troitzsch 1990).
In general, the traditional tests express their results in terms of certain observations or measurements. These are then used to derive a relative ranking scale or index on which to evaluate materials. Unfortunately, the bases of these ranking scales are arbitrary, and therefore, results from one test do not necessarily agree with another (Emmons, 1974), nor do they reflect how a material might behave in a real fire. A major step forward in the systematic quantification of the flammability of polymeric materials has been the increasing adoption of the cone calorimeter, developed in the 1980s at the National Bureau of Standards (Babrauskas 1984; Babrauskas and Parker 1986). Part of the design strategy in its development was to ensure that it was more amenable to mathematical modelling then the, predominantly Bunsen-burner based, tests of previous decades. In particular, the radiative heat source has been designed to ensure that the heat flux across the surface of the sample is uniform, that specimens experience primarily one-dimensional heat transfer, and that edge effects are minimised (Babrauskas and Parker 1987).
As mentioned above there are many ways in which the flammability of a product can be assessed. Within the cone calorimeter these include, but are not limited to, the ease of ignition (the critical heat flux), the rate of flame spread, the time-to-ignition, the heat release rate, the toxicity of the combustion gases and the amount of smoke generated. Heat release rate has been identified as being the single most important variable in characterising the `flammability' of products and their consequent fire hazard (Babrauskas and Peacock, 1992). It is usually assumed in polymer combustion models, and in the analysis of experimental data, that gaseous kinetics occur on a faster timescale than the degradation kinetics. The rate of heat release in the flame is therefore controlled by the rate of flow of volatiles into the flame and not by the gaseous kinetics.