An experimental and computational study of the mechanisms of detonation transmission in layered hydrogen-oxygen mixtures

This thesis presents the results of an experimental, analytical, and numerical investigation of the mechanisms of {dollar}rm Hsb2{dollar}-{dollar}rm Osb2{dollar} detonation transmission. The experimental configuration is designed to study how a detonation originally propagating into one gaseous mixture of hydrogen and oxygen (primary) is transmitted laterally into a secondary mixture. The configuration and the mixtures used for the experimental and numerical investigations are similar so that the experimental and the numerical results can be compared directly.; The experiments are conducted using mixtures with varying equivalence ratios. A stable wave configuration consisting of an oblique shock and an oblique detonation is observed, and three modes of ignition are identified and explained. The analytical results are based on a shock-polar analysis of the steady-state model for the waves in the primary and secondary mixtures, using the two-gamma method of Liou (1986) to derive the jump conditions across the shocks and detonation waves. The analytical method helps interpreting and predicting the experimental results, provided that steady state is achieved. However, numerical simulations are required for the analysis of the early, unsteady part of the transmission.; An improved chemical model for {dollar}rm Hsb2{dollar}-{dollar}rm Osb2{dollar} combustion is developed and used in one and two-dimensional numerical simulations of detonation waves. The one-dimensional calculations are used to simulate the propagation of a planar detonation wave in a shock tube. The tests of the numerical and physical parameters show that the numerical results agree to within a few percent with the theory of one-dimensional CJ detonations. Two-dimensional simulations of the detonation transmission experiments using the improved chemical model are conducted for the case where both the primary and the secondary mixtures are stoichiometric. In the calculations, the secondary mixture does not ignite behind the transmitted blast wave and the primary detonation gets quenched. Both the absence of ignition in the secondary mixture and the quenching of the primary detonation result from the long ignition times characteristic of the background fluid.; In the experiments conducted with both primary and secondary mixtures stoichiometric, the secondary mixture always ignited directly behind the transmitted blast wave. There are at least three possible explanations for the lack of agreement between the numerical and the experimental results: the kinetic model, the physical model (e.g., the transverse waves were not resolved), and discrepancies between the experimental and the simulated problems (e.g., possible presence of three-dimensional effects). These factors suggest that the agreement between the numerical and the experimental results should be improved by increasing the accuracy of the high-pressure kinetic model and the numerical resolution.