<p>Focusing of strong shock waves in a gas-filled thin test section with various forms of the reflector boundary is investigated. The test section is mounted at the end of the horizontal co-axial shock tube. Two different methods to produce shock waves of various forms are implemented. In the first method the reflector boundary of the test section is exchangeable and four different reflectors are used: a circle, a smooth pentagon, a heptagon and an octagon. It is shown that the form of the converging shock wave is influenced both by the shape of the reflector boundary and by the nonlinear dynamic interaction between the shape of the shock and the propagation velocity of the shock front. Further, the reflected outgoing shock wave is affected by the shape of the reflector through the flow ahead of the shock front. In the second method cylindrical obstacles are placed in the test section at various positions and in various patterns, to create disturbances in the flow that will shape the shock wave. It is shown that it is possible to shape the shock wave in a desired way by means of obstacles. The influence of the supports of the inner body of the co-axial shock tube on the form of the shock is also investigated. A square shaped shock wave is observed close to the center of convergence for the circular and octagonal reflector boundaries but not in any other setups. This square-like shape is believed to be caused by the supports for the inner body. The production of light, as a result of shock convergence, has been preliminary investigated. Flashes of light have been observed during the focusing and reflection process.</p>

Focusing of strong shock waves in a gas-filled thin test section with various forms of the reflector boundary is investigated. The test section is mounted at the end of the horizontal co-axial shock tube. Two different methods to produce shock waves of various forms are implemented. In the first method the reflector boundary of the test section is exchangeable and four different reflectors are used: a circle, a smooth pentagon, a heptagon and an octagon. It is shown that the form of the converging shock wave is influenced both by the shape of the reflector boundary and by the nonlinear dynamic interaction between the shape of the shock and the propagation velocity of the shock front. Further, the reflected outgoing shock wave is affected by the shape of the reflector through the flow ahead of the shock front. In the second method cylindrical obstacles are placed in the test section at various positions and in various patterns, to create disturbances in the flow that will shape the shock wave. It is shown that it is possible to shape the shock wave in a desired way by means of obstacles. The influence of the supports of the inner body of the co-axial shock tube on the form of the shock is also investigated. A square shaped shock wave is observed close to the center of convergence for the circular and octagonal reflector boundaries but not in any other setups. This square-like shape is believed to be caused by the supports for the inner body. The production of light, as a result of shock convergence, has been preliminary investigated. Flashes of light have been observed during the focusing and reflection process. / QC 20101126

An axi-symmetric shock-tube model has been developed to simulate the shock-wave propagation and reflection in both non-reactive and reactive flows. Simulations were performed for the full shock-tube geometry of the high-pressure shock tube facility at Texas A&M University. Computations were carried out in the CFD solver FLUENT based on the finite volume approach and the AUSM+ flux differencing scheme. Adaptive mesh refinement (AMR) algorithm was applied to the time-dependent flow fields to accurately capture and resolve the shock and contact discontinuities as well as the very fine scales associated with the viscous and reactive effects. A conjugate heat transfer model has been incorporated which enhanced the credibility of the simulations. The multi-dimensional, time-dependent numerical simulations resolved all of the relevant scales, ranging from the size of the system to the reaction zone scale. The robustness of the numerical model and the accuracy of the simulations were assessed through validation with the analytical ideal shock-tube theory and experimental data. The numerical method is first applied to the problem of axi-symmetric inviscid flow then viscous effects are incorporated through viscous modeling. The non-idealities in the shock tube have been investigated and quantified, notably the non-ideal transient behavior in the shock tube nozzle section, heat transfer effects from the hot gas to the shock tube side walls, the reflected shock/boundary layer interactions or what is known as bifurcation, and the contact surface/bifurcation interaction resulting into driver gas contamination. The non-reactive model is shown to be capable of accurately simulating the shock and expansion wave propagations and reflections as well as the flow non-uniformities behind the reflected shock wave. Both the inviscid and the viscous non-reactive models provided a baseline for the combustion model iii which involves elementary chemical reactions and requires the coupling of the chemistry with the flow fields adding to the complexity of the problem and thereby requiring tremendous computational resources. Combustion modeling focuses on the ignition process behind the reflected shock wave in undiluted and diluted Hydrogen test gas mixtures. Accurate representation of the Shock - tube reactive flow fields is more likely to be achieved by the means of the LES model in conjunction with the EDC model. The shock-tube CFD model developed herein provides valuable information to the interpretation of the shock-tube experimental data and to the understanding of the impact the facility-dependent non-idealities can have on the ignition delay time measurements.

Shock-tube experiments are typically performed at high temperatures (≥1200K) due to test-time constraints. These test times are usually ~1 ms in duration and the source of this short, test-time constraint is loss of temperature due to heat transfer. At short test times, there is very little appreciable heat transfer between the hot gas and the cold walls of the shock tube and a high test temperature can be maintained. However, some experiments are using lower temperatures (approx. 800K) to achieve ignition and require much longer test times (up to 15 ms) to fully study the chemical kinetics and combustion chemistry of a reaction in a shock-tube experiment. Using mathematical models, analysis was performed studying the effects of temperature, pressure, shock-tube inner diameter, and test-port location at various test times (from 1 - 20 ms) on temperature maintenance. Three models, each more complex than the previous, were used to simulate test conditions in the endwall region behind the reflected shock wave with Ar and N2 as bath gases. Temperature profile, thermal BL thickness, and other parametric results are presented herein. It was observed that higher temperatures and lower pressures contributed to a thicker thermal boundary layer, as did shrinking inner diameter. Thus it was found that a test case such as 800K and 50 atm in a 16.2-cm-diameter shock tube in Argon maintained thermal integrity much better than other cases - pronounced by a thermal boundary layer ≤ 1 mm thick and an average temperature ≥ 799.9 K from 1-20 ms.

Les travaux présentés dans ce mémoire de thèse s’inscrivent dans le cadre des études liées aux dégâts sur les structures et les blessures subies par les personnes à la suite d’explosions de charges explosives en milieu confiné et semi-confiné. Afin de mener cette étude, des expériences sont réalisées à petite échelle en laboratoire et sont complétées par des simulations numériques. Les ondes de choc sont obtenues grâce à la détonation d’une charge explosive gazeuse composée de propane-oxygène en proportion stoechiométrique. L’étude consiste donc à réaliser des expériences à petite échelle en laboratoire afin d’apprécier les champs de pression obtenus à la suite de la détonation d’une charge explosive au sein de deux configurations différentes. La première représente un atelier pyrotechnique et la seconde met en jeu un entrepôt de stockage de bouteilles de gaz. Les résultats expérimentaux sont ensuite confrontés à des résultats obtenus par simulations numériques réalisées grâce au logiciel AUTODYN. En complément de ces deux configuration principales, une étude est menée sur l’identification des pics de surpressions réfléchis grâce à une approche expérimentale appelée paroi par paroi. Une étude est également menée sur la détermination d’une équivalence massique entre le TNT et le mélange gazeux utilisé pour les expériences. / The goal of this study is to investigate shock waves propagation, in a geometrically complex confined and semi-confined environment, consecutive to the detonation of a spherical explosive charge. In this objective, small scale experiments are conducted in laboratory and are completed with numerical analysis. Shock waves are generated thanks to spherical detonation of a gas mixture composed of propane-oxygen in stoechiometric proportion. Two main configurations are studied: the first represents a pyrotechnic workshop and the second is a warehouse containing gas cylinder. Experimental and numerical results are then compared. Complementary studies are realised to describe blast wave propagation inside a semi-confined volume thanks to a new experimental approach named wall by wall. Finally, in order to simulate TNT charges detonation by computational means, an important study is conducted to determine a mass equivalent between TNT and gas mixture.