Catálogo de publicaciones - libros

is a new code that utilizes elements of the Immersed Boundary (IB) and Lattice Boltzmann Method (LBM) as well as a Direct Forcing (DF) scheme. As a computational method, it is very flexible and it appears to be ideal in solving fluid-particle interaction problems including problems with deformable boundaries. Proteus uses a regular Eulerian grid for the flow domain and a regular Lagrangian grid to follow particles that are contained in the flow field. The rigid body conditions for the fluid and the particles are enforced by applying the external force acting on the boundary of particles. A penalty method is used, which assumes that the particle boundary is deformable with a high stiffness constant. The velocity fields for the fluid and particles are solved by incorporating a force density term into the lattice Boltzmann equation. This force term is determined by using a technique that is based on the direct forcing scheme. preserves all the advantages of LBM in tracking a group of particles and, at the same time, provides an alternative and better approach to treating the solid-fluid boundary conditions. Because of this it provides for a smooth boundary interface, with only a few nodes assigned for the size of particles. This new method also solves the problems of fluctuation of the forces and velocities on the particles when the “bounceback” boundary conditions are applied. The method has the capability to simulate deformable particles and fluid-structure deformation. The results of the Proteus code have been validated by comparison with results from other computational methods as well as experimental data. Some of the validation results will be given in the presentation of this paper.

Large-Eddy Simulation (LES) and Discrete Particle Simulation (DPS) are used to highlight effects of fluid-particle and particle-particle interactions on dispersed-phase transport in fully-developed turbulent channel flow. A range of particle Stokes numbers in the simulations are considered that lead to strong changes in particle response. In the absence of inter-particle collisions, the calculations illustrate the characteristic build-up of particles in the near-wall region. While mean shear in the carrier and dispersed phase velocities is an important effect in wall-bounded flows, LES/DPS results show that the particle velocity fluctuations in the wall-normal direction are controlled primarily by the drag force and in equilibrium with the corresponding components of the fluid-particle velocity correlation. Inter-particle collisions provide a redistribution mechanism that reduces the strong anisotropy of the particle velocity fluctuations and substantially elevates cross-stream transport. Spatial properties of the particle velocity field are examined using two-point correlations. The correlation functions are discontinuous at the origin and are consistent with a partitioning of the particle velocity by inertia into a spatially-correlated contribution and random component that is not correlated in space. Perspectives and implications of these findings are also discussed.

We present an explicit finite-difference scheme for direct simulation of the motion of solid particles in a fluid. The method is based on a second-order MacCormack finitedifference solver for the flow, and Newton’s equations for the particles. The fluid is modeled with fully compressible mass and momentum balances; the technique is intended to be used at moderate particle Reynolds number. Several examples are shown, including a single stationary circular particle in a uniform flow between two moving walls, a particle dropped in a stationary fluid at particle Reynolds number of 20, the drafting, kissing, and tumbling of two particles, and 100 particles falling in a closed box.

A stochastic representation of fluid turbulence has been developed to study the behavior of very dilute suspensions of solid spheres in a turbulent flow. Particular emphasis is given to the understanding of deposition in an idealized annular pattern. The accuracy of the stochastic method was checked by comparing with calculations done in a DNS at =150. The striking aspect of the study is that calculations of the dimensionless deposition constant in a horizontal channel are presented for =0 to 3.0, =590 and τ=0 to 20,000. For some runs it was necessary to use computation times of =2×10 in order to insure that a fully-developed condition was realized. Such an extensive study would not be possible if the turbulence was represented by a DNS.

Since their initial fabrication two decades ago by McConnell and coworkers, fluid supported phospholipid bilayers (SLBs) have played a key role in the development of nanoscale assemblies of biological materials on artificial supports. The reason for this is quite straightforward. SLBs can serve as biomimetics for chemical and biological processes which occur in cell membranes. A thin aqueous layer (approximately 1 nm thick) is trapped between the bilayer and the underlying support (Figure 6.1). Thiswater layer acts as a lubricant allowing both leaflets of the bilayer to remain fluid. Consequently, planar supported membranes retain many of the physical properties of free vesicles or even native cell surfaces when the appropriate recognition components are present. Specifically, SLBs are capable of undergoing lateral rearrangements to accommodate binding by aqueous proteins, viruses, toxins, and even cells. As substrate supported entities, they are convenient to study by a host of interface-sensitive techniques and are far less fragile than either unsupported membranes or full-blown cellular systems.

Large-Eddy Simulation is used for the investigation of the breaking of steep water waves on a beach of constant bed slope. The method is built within a multi-fluid flow solver, in which the free surface is tracked using a Volume-of-Fluid method featuring piecewise planar interface reconstructions on a twice-as-fine mesh. The Smagorinsky sub-grid scale model is used for explicit under-resolved turbulence closure, coupled with a new scheme for turbulence decay treatment on the air-side of massively deformable free surfaces. The simulations were conducted for shear Reynolds numbers ≈≈400, based on the mean water depth. The Large-Eddy Simulation formulation in the interface tracking, single-fluid formulation is introduced for this purpose. The approach is demonstrated as a powerful tool for exploring large-scale, interfacial turbulent flows. The discussion focuses on coherent structures formation, the free surface flow effects at breaking, and form drag evolution with the surface.

Turbulent dispersed flows over boundary layers are crucial in a number of industrial and environmental applications. In most applications, the key information is the spatial distribution of inertial particles, which is known to be highly non-homogeneous and may exhibit a complex pattern driven by the structures of the turbulent flow field. Theoretical and experimental evidence shows that fluid motions in turbulent boundary layers are intermittent and have a strongly organized and coherent nature represented by the large scale structures. These structures control the transport of the dispersed species in such a way that the overall distribution will resemble not at all those given by methods in which these motions are ignored.

In this paper, we study from a statistical viewpoint turbulence modulation produced by different-size dispersed particles and we examine how near-wall particle concentration is modified due to the action of particles themselves in modulating turbulence. The physical mechanisms and the statistics proposed are based on Direct Numerical Simulation (DNS) of turbulence and Lagrangian particle tracking, considering a two-way coupling between particles and fluid.

Several Continuous Random Walk (CRW) models were constructed to predict turbulent particle diffusion based only on mean Eulerian fluid statistics. The particles were injected near the wall (=4) of a turbulent boundary layer that is strongly anisotropic and inhomogeneous near the wall. To assess the performance of the models for wide range of particle inertias (Stroke numbers), the CRW results were compared to particle diffusion statistics gathered from a Direct Numerical Simulation (DNS). The results showed that accurate simulation required a modified (non-dimensionalized) Markov chain for the large gradients in turbulence based on fluid-tracer simulations. For finite-inertia particles, a modified drift correction for the Markov chain (developed herein to account for Strokes number effects) was critical to avoiding non-physical particle collection in low-turbulence regions. In both cases, inclusion of anisotropy in the turbulent kinetic energy was found to be important, but the influence of off-diagonal terms was found to be weak.

Much research has been done on the motion of heavy particles in simple vortex flows. In most of this work, particle motion is investigated under the influence of fixed vortices. In the context of astrophysics, the motion of heavy particles in rotating two-dimensional flows has been investigated; the rotation follows from the laws of Kepler. In the present paper, the motion of heavy particles in potential vortex flow in a circular domain is investigated. The vortex describes a circular trajectory due to the presence of the boundary, so that a steadily rotating flow is obtained. In order to isolate the effect of particle inertia, only Stokes drag is taken into account in the equation of motion. The numerical simulations are based on a oneway coupling. They show that small heavy particles accumulate in an ellitpic region of the flow, counterrotating with respect to the vortex. When the particle Stokes number exceeds a threshold, depending on the vortex configuration, particles are expelled from the circular domain. A stability criterion for this particle accumulation is derived analytically. These results are qualitatively comparable to those obtained by others in astrophysics.

We present direct simulations with interface resolution of dense, fluidized solid-liquid suspensions. The flow of interstitial fluid is solved by the lattice-Boltzmann method (LBM). The monodisperse, spherical particles move under the influence of gravity, hydrodynamic forces stemming from the LBM, subgrid-scale lubrication forces, and hard-sphere collisions. The cases we study have been derived from the experimental work by Duru et al.