Stochastic Vortex Structure Model for Adhesive Particles in Turbulent Flows

点击翻译按钮获取中文摘要

Marshall, Jeffrey 1332472Particles present in turbulent flows are transported because of a large-scale diffusive process caused by exchange of fluid between turbulent eddies and by a small-scale anti-diffusive process. The latter (which we will refer to as particle dispersion) is associated with the tendency of heavy particles to be thrown out of turbulent eddies, forming high-concentration particle sheets in the interstitial region between the turbulent eddies. This small-scale heterogeneity is known to cause fluctuations in the particle concentration of several orders of magnitude, which in turn have a dramatic effect on processes that involve particle collision and agglomeration. While stochastic Lagrangian models (SLMs) have been shown to accurately simulate large-scale particle diffusion, these methods lack the spatial correlation necessary to account for small-scale concentration heterogeneity. The work proposed here will replace random forcing used in stochastic Lagrangian models with a spatially-correlated stochastic forcing based on approximating the turbulent eddies by a set of random vortex structures.Intellectual Merit :The proposed research will develop a stochastic vortex structure (SVS) model for simulating collision and agglomeration of adhesive particles in turbulent flows. The SVS model will generate a turbulent velocity field with correct spatial correlation over a length scale interval ranging between the energy-containing turbulence integral scale and a minimum scale that is associated with the length scale at which the eddy Stokes number is equal to unity. The proposed stochastic model will be validated by comparison to direct numerical simulations of particle agglomerate formation in homogeneous turbulence and turbulent shear flows. Simulations will be performed using a discrete element method for adhesive particles, a fast multipole method for solution of the vorticity-induced velocity field, and a pseudospectral method for direct numerical simulation (DNS) of turbulence. Broader Impacts :Adhesive particles dominate critical problems ranging from bioengineering systems (blood flow, GI flows, microorganism suspensions), manufacturing processes (electrospray, electrocoating, dust mitigation), energy generation (pulverized coal combustion, biomass combustion, ash mitigation), environmental processes (sediment transport, volcanic processes, filtration processes), and nano- and microscale technology (nanoparticle dispersion, nanotube alignment, self-assembly). The proposed research will develop and validate a novel structurally-based stochastic model that will enable an entirely new simulation approach for a wide range of problems involving adhesive particles.The research will be shared with a minority-serving institution via an existing arrangement between University of Vermont and City University of New York.