The shockless explosion combustion (SEC) is a novel approach to constant volume combustion in gas turbines. It promises an efficiency gain comparable to that of pulse detonation combustion (PDC), but without the drawbacks associated with detonations. It utilizes homogeneous combustion of a volume of fuel/air to avoid strong shock waves, similar to the RCCI process in internal combustion engines. Recharging is handled analogous to a pulse jet through the pressure waves in the combustion chamber. To achieve homogeneous auto ignition, the process involves setting up a stratified layer of fuel/air such that it homogeneously auto ignites in an approximately constant volume combustion (CVC). This becomes feasible by using fuels with small dependence of their auto ignition time on temperature, e.g. blends involving fuels with negative temperature coefficient (NTC) behaviour. The ignition process of such fuels is complex and often involves multi-stage ignition on time scales comparable to the acoustic time scale. It is hence expected that even though a SEC effectively is CVC, the ignition can not be modeled in 0D, but that it instead involves complex interaction between gas dynamics and chemical kinetics. The stratification process therefore has to be numerically optimized in CFD calculations. Optimization, especially if whole cycles are to be simulated, requires small kinetics models, even if restricted to one dimension, to be computationally feasible. On the other hand, the interaction of kinetics and gas dynamics at ignition rules out an easy to evaluate optimization goal for reduction of the chemical kinetics using offline methods like directed relation graphs and the techniques based on sensitivity analysis introduced by Williams/Peters. Even if sufficient computation power was invested, the accuracy constraints on the auto ignition times severely limit the usability of a conventional reduced mechanism. Online tools like CSP or ILDM would be an option for practical purposes, but do not provide insight into the ignition process and are therefore of little help for fundamental research. For a CFD simulation of a SEC, a new ansatz has therefore been developed. We exploit the constraint of a small temperature dependence of the auto ignition time on temperature to introduce a model specialized for SEC simulation that is sufficiently small to allow optimization of a fuel/air stratification, yet features correct auto ignition delay times for each ignition stage to the accuracy of experimental measurements. We then proceed to present simulation results which a posteriori justify our approach and demonstrate that shockless explosion combustion is feasible.

This content is only available via PDF.
You do not currently have access to this content.