In this work the previously developed Lattice Boltzmann-Direct Forcing/ Fictitious Domain (LB-DF/FD) method is adopted to simulate the sedimentation of eight circular particles under gravity at an intermediate Reynolds number of about 248. The particle clustering and the resulting Drafting-Kissing-Tumbling (DKT) motion which takes place for the first time are explored. The effects of initial particle-particle gap on the DKT motion are found significant. In addition, the trajectories of particles are presented under different initial particle-particle gaps, which display totally three kinds of falling patterns provided that no DKT motion takes place, i.e. the concave-down shape, the shape of letter “M” and “in-line” shape. Furthermore, the lateral and vertical hydrodynamic forces on the particles are investigated. It has been found that the value of Strouhal number for all particles is the same which is about 0.157 when initial particle-particle gap is relatively large. The wall effects on falling patterns and particle expansions are examined in the final.

We extend our results on fictitious domain methods for Poisson’s problem to the case of incompressible elasticity, or Stokes’ problem. The mesh is not fitted to the domain boundary. Instead boundary conditions are imposed using a stabilized Nitsche type approach. Control of the non-physical degrees of freedom, i.e., those outside the physical domain, is obtained thanks to a ghost penalty term for both velocities and pressures. Both inf-sup stable and stabilized velocity pressure pairs are considered.

We consider an incompressible flow problem in a N-dimensional fracturedporous domain (Darcy’s problem). The fracture is represented by a(N − 1)-dimensional interface, exchanging fluid with the surroundingmedia. In this paper we consider the lowest-order(ℝ T0, ℙ0) Raviart-Thomas mixed finite elementmethod for the approximation of the coupled Darcy’s flows in the porous media and withinthe fracture, with independent meshes for the respective domains. This is achieved thanksto an enrichment with discontinuous basis functions on triangles crossed by the fractureand a weak imposition of interface conditions. First, we study the stability andconvergence properties of the resulting numerical scheme in the uncoupled case, when theknown solution of the fracture problem provides an immersed boundary condition. We detailthe implementation issues and discuss the algebraic properties of the associated linearsystem. Next, we focus on the coupled problem and propose an iterative porousdomain/fracture domain iterative method to solve for fluid flow in both the porous mediaand the fracture and compare the results with those of a traditional monolithic approach.Numerical results are provided confirming convergence rates and algebraic propertiespredicted by the theory. In particular, we discuss preconditioning and equilibrationtechniques to make the condition number of the discrete problem independent of theposition of the immersed interface. Finally, two and three dimensional simulations ofDarcy’s flow in different configurations (highly and poorly permeable fracture) areanalyzed and discussed.

This paper deals with the mathematical and numerical analysis of asimplified two-dimensional model for the interaction between the windand a sail. The wind is modeled as a steady irrotational plane flow pastthe sail, satisfying the Kutta-Joukowski condition. This conditionguarantees that the flow is not singular at the trailing edge of thesail. Although for the present analysis the position of the sail istaken as data, the final aim of this research is to develop tools tocompute the sail shape under the aerodynamic pressure exerted by thewind. This is the reason why we propose a fictitious domain formulationof the problem, involving the wind velocity stream function and aLagrange multiplier; the latter allows computing the force densityexerted by the wind on the sail. The Kutta-Joukowski condition isimposed in integral form as an additional constraint. The resultingproblem is proved to be well posed under mild assumptions. For thenumerical solution, we propose a finite element method based onpiecewise linear continuous elements to approximate the stream functionand piecewise constant ones for the Lagrange multiplier. Error estimatesare proved for both quantities and a couple of numerical testsconfirming the theoretical results are reported. Finally the method isused to determine the sail shape under the action of the wind.

The previously developed LB-DF/FD method derived from the lattice Boltzmann method and Direct Forcing/Fictitious Domain method is extended to deal with 3D particle’s Brownian motion. In the model the thermal fluctuations are introduced as random forces and torques acting on the Brownian particle. The hydrodynamic interaction is introduced by directly resolving the fluid motions. A sphere fluctuating in a cubic box with the periodic boundary is considered to validate the present model. By examining the velocity autocorrelation function (VCF) and rotational velocity autocorrelation function (RVCF), it has been found that in addition to the two relaxation times, the mass density ratio should be taken into consideration to check the accuracy and effectiveness of the present model. Furthermore, the fluctuation-dissipation theorem and equipartition theorem have been investigated for a single spherical particle. Finally, a Brownian particle trapped in a harmonic potential has been simulated to further demonstrate the ability of the LB-DF/FD model.

The accuracy of the domain embedding method from [A. Rieder, Modél. Math. Anal. Numér.32 (1998) 405-431] for the solution of Dirichlet problemssuffers under a coarse boundary approximation. To overcome this drawback the methodis furnished withan a priori (static) strategy for an adaptive approximation space refinement near the boundary. This is done by selecting suitable wavelet subspaces. Error estimates and numerical experiments validate the proposed adaptive scheme.In contrast to similar, but rather theoretical, concepts already described in theliterature our approach combines a high generality with an easy-to-implement algorithm.