Venting Optimization of a Pulse

A one-dimensional method-of-characteristics (MOC) code was developed to examine the venting of pulse detonation engines. Comparison with experimental results and twodimensional computational fluid dynamics demonstrates that a reasonably accurate level of simulation can be achieved with a single spatial dimension. A semi-empirical, deflagrative, flame-acceleration model was also constructed and compared to experimental results. While the results were promising, they were not sufficiently accurate to allow for modelling of deflagration-to-detonation transition. Several configurations were then examined with the MOC code to determine which parameters optimized the venting of the exhaust gases. The parameters varied were the location of detonation initiation, the filling velocity, and the distribution of reactants (partial fills). Each configuration was also simulated with a practical, fixed-geometry nozzle that was optimized, and a theoretical, variable-geometry nozzle. The results indicate that a variable nozzle increases the impulse by less than 15 % over a configuration with no nozzle. The impulse gain from a fixed nozzle is about half that of a variable nozzle, with a corresponding decrease in average thrust. For quiescent initial conditions, the differences in impulse between detonations initiated at the closed head and the open tail are negligible, although tail-initiated detonations consistently provided faster blow-down times. With increased filling velocity, tail initiated detonations provide several benefits. These include a smaller decrease in specific impulse, a larger increase in average thrust, a longer cycle time, and better performance with a fixed nozzle. Simulations with partial fills showed that they do not replace nozzles in recovering losses. For fixed nozzles, the partial-fill length can be as much as half the total length of the tube before the average thrust begins to decrease significantly.