UNS: Collaborative Research: Experiments and Theory of Nonequilibrium Processes in Turbulent Combustion

点击翻译按钮获取中文摘要

1511834 (Raman, University of Michigan-Ann Arbor)/1511025 (Clemens, University of Texas-Austin)Fluid flows are complicated under conditions found in engines and power plants, due to fuel combustion. The research team proposes to investigate the fluid flow phenomenon by using the state-of-art computer simulation techniques with the help of advanced laser-based measurements. The results will be used for constructing models for the flame and flow predictions under such complex conditions. Both graduate and undergraduate students will be involved in the research. The PIs also propose K-12 activities.A novel multi-scale modeling approach is proposed to elucidate energy transfer mechanisms between scales in low-Mach number and highly turbulent reacting flows with premixed and non-premixed injections. For the experimental effort, advanced, laser-based diagnostic tools are proposed: PIV (particle imaging velocimetry) and Rayleigh and Raman scattering for velocity field and scalar/species concentration measurements, respectively. The experimental data will provide large ensembles of statistical data that have direct bearing on the presence and impact of non-equilibrium in the velocity and scalar fields. They will also be used for comparison with result from CFD (Computational Fluid Dynamics) simulation, based on DNS (Direct Numerical Simulation), to understand the interaction between combustion and turbulence and between phenomena occurring at various length scales in turbulent combustion. Furthermore, by using time and length scales characterizing non-equilibrium generating impulses, the structure and evolution of turbulent flames will be studied. A variational multiscale approach will be used as a framework for modeling non-equilibrium turbulent combustion, where the energy transfer across impulse-generating scales will be modeled explicitly. A small-parameter expansion based unsteady dissipation rate model will be developed. The experiments will be critical to the modeling effort since a far larger range of conditions can be investigated than is possible with DNS.