A computational study of the nonlinear rocket-engine combustion instability is presented for an engine with 82 coaxial methane–oxygen injector ports, a choked nozzle, and a combustion chamber that models a Rocketdyne experiment. Computations use a three-dimensional unsteady shear-stress transport delayed detached-eddy simulation method. Four different one-step chemical kinetic rates are simulated on the resolved scale to determine the effect of the burning rate. Application of kinetic rates on the resolved scale give faster rates than expected with turbulent combustion where the final mixing and chemical reaction in the cascade occurs on the smallest scales; thus, reduction of the kinetic rate is examined here. When the normalized kinetic rate is 1, implying the nominal Westbrook–Dryer rate, only the longitudinal-mode instability is observed. As the kinetic rate decreases, less of the combustion occurs upstream near the longitudinal-mode pressure antinode, the longitudinal-mode instability becomes weak, and the first-tangential-mode instability emerges, which fits better with experiment. The proximity of the domain of major heat release to the longitudinal mode pressure antinode is found to be critical. Evidence is presented for the need for an improved combustion model that will provide a more accurate and lower burning rate.
Read full abstract