Ignition of solids exposed to transient irradiation
Santamaria Garcia, Simon
This work investigates the ignition of solid materials. Ignition is important for fire safety not only as the initiation of a fire event, but because it plays a critical role in all the phenomena of interest, such as fire growth and flame spread. Ignition is a multi-phase, complex phenomenon which requires the description of different, coupled, physical and chemical processes over a range of time and length scales. Consequently, most ignition models need to implement assumptions about the different mechanisms driving ignition to simplify the phenomena considered. Accordingly, the majority of ignition models neglect gas phase considerations and emphasise the heating and pyrolysis processes. There exist a breadth of models, of various levels of complexity, which seek to describe the pyrolysis processes of charring and non-charring materials. Within constrained experimental conditions, most of these models have performed well when compared to relevant experimental data. However, this thesis shows that the experimental data used to validate pyrolysis and ignition models does not reflect the entire range of heating scenarios. It is demonstrated that many scenarios of interest cannot be represented by assuming the irradiation is constant in time and, additionally, that a material exposed to a time dependent Incident Heat Flux will respond in a fundamentally different manner. This means that in previous studies, the effects of irradiation on the response of the solid have not been completely investigated. When a material is exposed to a constant Incident Heat Flux, the net or absorbed energy flux decreases with time. However, this is not the case when the material is exposed to time dependent Incident Heat Fluxes, since the net energy flux can increase with time. Given that the heating and the thermal degradation processes are driven by the energy absorption, these scenarios need to be investigated. It remains necessary to highlight that, even though the evolution of the absorbed energy flux differs, the fundamental process of interest (ignition) is still described by the attainment of a flammable mixture. This is important since it is necessary to predict the onset of ignition for solids that are exposed to thermal radiation during a fire. Data from flame spread and compartment fire experiments was compared and analysed to show that it is not possible to model the ignition of solids under these scenarios assuming a constant irradiation as the boundary condition for the ignition model. Hence, this thesis employs analytical, numerical and experimental investigations to address these limitations. Using analytical solutions, the thermal response of the inert solid was investigated. For constant irradiation, it is common for analytical ignition models (such as the classical ignition theory) to neglect the surface heat losses. It was shown that this is possible for a limited number of scenarios if a constant Incident Heat Flux is used but, it is not accurate for transient irradiation. A numerical model was used to investigate the assumption that surface heat losses are linearly dependent on temperature. This is implemented through the definition of a total heat transfer coefficient, hT. It was shown that this provides acceptable results, particularly for thermoplastic solids where the surface temperature is limited by the pyrolysis process. This is important since the use of hT eliminates the need to accurately describe the radiative properties of both the heat source and the sample. Piloted experiments with thermoplastic polymers and timber samples were completed. A novel approach to calculate the mass loss rate at ignition based on a linear regression for the time interval close to ignition was developed. It was found that for most materials, the mass loss rate at ignition is independent of the heating scenario. Consequently, the true ignition criterion remains unchanged from simple to complex heating. For PA6, large scatter reflected the influence of surface phenomena, particularly bubbling, on the processes driving ignition. A further study using in-depth temperature measurements showed that, even though this effect is large, it mainly affects the transport of gases to the pilot and not the heating of the solid. Gas phase measurements were completed to evaluate gas phase concentrations and investigate the flammability of the mixture. The data served to evaluate the delay between the onset of pyrolysis and the onset of ignition. Traditional ignition theory as well as the ignition temperature criterion are built on assuming this delay time is negligible. It was shown that this is not the case for samples exposed to transient irradiation. Finally, a new experimental methodology to investigate ignition was developed. The goal is to manipulate the material response. This is achieved by implementing an algorithm that monitors the thermal evolution of the solid and defines the irradiation instantaneously, depending on the difference between the desired response and the measured parameters. Conceptually, this represents an advancement in the use of transient irradiation, since the boundary condition becomes the means to manipulate the response of the solid and not the defining parameter of the study. It was shown that it is possible to manipulate the thermal degradation of the solid, but uncertainties associated to the measuring techniques limit the accuracy of the algorithm. This thesis advances the current knowledge on ignition of solids by: (i) investigating the implications of extending the analytical formulations of classical ignition theory for transient irradiation and showing that such an effort does not yield useful results for the advancement of ignition theory, (ii) analysing experimentally measured thermal radiation and showing the need to investigate the ignition of solids exposed to time-varying irradiation curves, (iii) addressing this need by creating a comprehensive data set of piloted ignition data of thermoplastic and charring solids exposed to transient irradiation, (iv) using these data to evaluate the applicability of common assumptions in ignition modelling, including the use of an ignition temperature criterion and (v) developing a novel experimental methodology that allows for the manipulation of the response of the material, which can be utilised to investigate the effect of surface and localised phenomena on the onset of ignition.