While much numerical research has been conducted on the effect of suspended point-like particles in a turbulent flow, only a very limited amount of information is available for situations in which the particle size is comparable to, or exceeds, the length scales of the flow, from the Kolmogorov length on up. The obstacle to overcome is the major computational effort necessary to resolve the detailed flow features in the presence of the complex, time-evolving no-slip boundaries represented by the particles. The PI has developed a powerful computational method to address this difficulty, which has recently been made even more efficient by reprogramming it for GPU-based computers. This method will be used to run truly direct numerical simulations of turbulent fluid flows with thousands of suspended finite-size particles and to extract from the numerical results information on various features of the fluid turbulence modification caused by the particles. In addition to these specific results on disperse particulate flows, one of the aims of the work is to demonstrate the possibility of running routine simulation involving many thousands of particles at volume fractions between 0.05 and 0.20 on GPU-based work stations in a matter of hours or a few days at most.
Natural processes offer a wealth of instances in which solid particles are transported by fluid flows: marine sediments, gravity currents, sand dunes and many others. Such flows are also of great technological relevance, primarily in the energy area where fluidized bed combustors, catalytic oil cracking, coal gasifiers ad chemical looping combustors are prime examples. A reliable theoretical understanding of these systems, which the numerical simulations to be undertaken in this project will provide, is essential for a variety of reasons ranging from the modeling of environmental flows to carbon sequestration, to the development of new or scaled-up designs without expensive preliminary experimentation, and many others.
