Reducing jet aircraft noise by harnessing the hetergeneous XK nodes on Blue Waters
Environmental pollution is a pacing item worldwide and, for certain communities, the noise pollution caused by commercial or military aircraft use leads to adverse psycho-physical reactions, lawsuits, and the retrofitting of homes with noise-reducing treatments. On active Naval carrier decks, aircraft noise-caused hearing loss occurs within 2 minutes without hearing protection and in several weeks with it; hearing loss now accounts for nearly $1 billion annually in military healthcare. The main source of aircraft generated noise is the jet engine exhaust, where the unsteady motion of the engine's exhaust gases generate sound. The conceptual challenge of reducing turbulent jet noise is immense: the fundamental flow is turbulent and we, as a scientific community, do not understand how a turbulent flow generates sound.
Without a guiding theory, reducing aircraft noise has been left to trial-and-error experiments and, more recently, simulations. The turbulence-induced sound is generated over a very large region leading to a multi-scale nonlinear fluid dynamics problem where the relevant energy is contained over six to ten decades of spatial and temporal scales. Simulations that capture the full range of spatial and temporal scales are beyond current computing capacity but remain our best hope for quieter aircraft. A paradigm shift in computational science, enabled by the XK nodes on Blue Waters, is desperately needed to advance the fields of compressible turbulence and aeroacoustics to reduce the noise from jet aircraft.
This project will impact three critical areas. By targeting the heterogeneous XK nodes on Blue Waters we will lead the development of power efficient, high-performance scientific codes that can run across thousands of heterogeneous nodes and thus direct emerging programming models and numerical algorithms compatible with forthcoming hardware complexity at the exascale. Second, the scale of Blue Waters will enable the largest simulations of compressible turbulent jet noise ever that will advance the science of flow-generated sound through carefully conducted simulations and guided post-processing of the 100s to 1,000s of terabytes of data generated. Third, new engine nozzle designs will be develped that will reduce turbulent jet noise and improve the quality of life of airport communities and military personnel.