Search

This thesis considers the possibility of a novel propagation mechanism of vapour cloud explosions. The hypothesis is based on the potential of forward thermal radiation from hot combustion products to initiate combustion ahead of the main flame front by igniting particles suspended in the unburnt gas mixture. The formation of such local ignition events may cause an augmented explosion. An experimental investigation is described in which a laser source is used to heat various types of particles in flammable butane-air mixtures. It is found that fine powders can ignite an adjacent charge upon irradiation. Ignition times are established for a range of particles with widely different characteristics (size, type, morphology, etc.). In particular, ignition times scales ~100 ms could be obtained at an irradiance of 600 kW/m^2 using glass substrates coated with a carbon black. The relationship between absorption-emission properties of different powders is investigated further by determining particle temperatures under relevant radiation levels using time-resolved emission spectroscopy. Two different ignition regimes were observed, ignition temperatures appeared to be constant for non-reactive silicon carbide, while for reactive powders, temperature appeared to be not the sole ignition criterion. In response to the lack of radiation measurements for large scale premixed systems, a tool is developed for estimating the radiation flux emitted from such systems along with the corresponding level of irradiance posed on particles raised in the unburnt gas mixture. This information permits an assessment of the potential role of forward thermal radiation on the flame propagation. The comparison of such theoretical estimates with the experimentally measured ignition time data is vital for evaluation purposes. The analysis shows that the estimated and required fluxes are comparatively close and, hence, the proposed flame propagation mechanism is credible.