Bulletin of the American Physical Society
69th Annual Meeting of the APS Division of Fluid Dynamics
Volume 61, Number 20
Sunday–Tuesday, November 20–22, 2016; Portland, Oregon
Session L29: CFD: Algorithms I |
Hide Abstracts |
Chair: Anne Staple, Virginia Tech Room: F150 |
Monday, November 21, 2016 4:30PM - 4:43PM |
L29.00001: Acceleration of incremental-pressure-correction incompressible flow computations using a coarse-grid projection method Ali Kashefi, Anne Staples Coarse grid projection (CGP) methodology is a novel multigrid method for systems involving decoupled nonlinear evolution equations and linear elliptic equations. The nonlinear equations are solved on a fine grid and the linear equations are solved on a corresponding coarsened grid. Mapping functions transfer data between the two grids. Here we propose a version of CGP for incompressible flow computations using incremental pressure correction methods, called IFEi-CGP (implicit-time-integration, finite-element, incremental coarse grid projection). Incremental pressure correction schemes solve Poisson's equation for an intermediate variable and not the pressure itself. This fact contributes to IFEi-CGP's efficiency in two ways. First, IFEi-CGP preserves the velocity field accuracy even for a high level of pressure field grid coarsening and thus significant speedup is achieved. Second, because incremental schemes reduce the errors that arise from boundaries with artificial homogenous Neumann conditions, CGP generates undamped flows for simulations with velocity Dirichlet boundary conditions. Comparisons of the data accuracy and CPU times for the incremental-CGP versus non-incremental-CGP computations are presented. [Preview Abstract] |
Monday, November 21, 2016 4:43PM - 4:56PM |
L29.00002: ABSTRACT WITHDRAWN |
Monday, November 21, 2016 4:56PM - 5:09PM |
L29.00003: Simulating incompressible flow on moving meshfree grids using General Finite Differences (GFD) Yaroslav Vasyliv, Alexander Alexeev We simulate incompressible flow around an oscillating cylinder at different Reynolds numbers using General Finite Differences (GFD) on a meshfree grid. We evolve the meshfree grid by treating each grid node as a particle. To compute velocities and accelerations, we consider the particles at a particular instance as Eulerian observation points. The incompressible Navier-Stokes equations are directly discretized using GFD with boundary conditions enforced using a sharp interface treatment. Cloud sizes are set such that the local approximations use only 16 neighbors. To enforce incompressibility, we apply a semi-implicit approximate projection method. To prevent overlapping particles and formation of voids in the grid, we propose a particle regularization scheme based on a local minimization principle. We validate the GFD results for an oscillating cylinder against the lattice Boltzmann method and find good agreement. [Preview Abstract] |
Monday, November 21, 2016 5:09PM - 5:22PM |
L29.00004: A characteristic mapping method for two-dimensional incompressible Euler flows Badal Yadav, Olivier Mercier, Jean-Christophe Nave, Kai Schneider We propose an efficient semi-Lagrangian method for solving the two-dimensional incompressible Euler equations with high precision on a coarse grid. The new approach evolves the flow map using the gradient-augmented level set method (GALSM). Since the flow map can be decomposed into submaps (each over a finite time interval), the error can be controlled by choosing the remapping times appropriately. This leads to a numerical scheme that has exponential resolution in linear time. The computational efficiency and the high precision of the method are illustrated for a vortex merger and a four mode flow. Comparisons with a Cauchy-Lagrangian method are also presented. [Preview Abstract] |
Monday, November 21, 2016 5:22PM - 5:35PM |
L29.00005: Reference Map Technique for Incompressible Fluid--Structure Interaction Problems Chris Rycroft, Chen-Hung Wu, Yue Yu, Ken Kamrin We present a fully Eulerian approach to simulate soft structures immersed in an incompressible fluid. The flow is simulated on a fixed grid using a second order projection method to solve the incompressible Navier--Stokes equations, and the fluid--structure interfaces are modeled using the level set method. By introducing a reference map variable to model finite-deformation constitutive relations in the structure on the same grid as the fluid, the interfacial coupling is highly simplified. This fully Eulerian approach provides a computationally efficient alternative to moving mesh approaches. Example simulations featuring many-body contacts and flexible swimmers will be presented. [Preview Abstract] |
Monday, November 21, 2016 5:35PM - 5:48PM |
L29.00006: Application of a derivative-free global optimization algorithm to the derivation of a new time integration scheme for the simulation of incompressible turbulence. Shahrouz Alimohammadi, Daniele Cavaglieri, Pooriya Beyhaghi, Thomas R. Bewley This work applies a recently developed Derivative-free optimization algorithm to derive a new mixed implicit-explicit (IMEX) time integration scheme for Computational Fluid Dynamics (CFD) simulations. This algorithm allows imposing a specified order of accuracy for the time integration and other important stability properties in the form of nonlinear constraints within the optimization problem. In this procedure, the coefficients of the IMEX scheme should satisfy a set of constraints simultaneously. Therefore, the optimization process, at each iteration, estimates the location of the optimal coefficients using a set of global surrogates, for both the objective and constraint functions, as well as a model of the uncertainty function of these surrogates based on the concept of Delaunay triangulation. This procedure has been proven to converge to the global minimum of the constrained optimization problem provided the constraints and objective functions are twice differentiable. As a result, a new third-order, low-storage IMEX Runge-Kutta time integration scheme is obtained with remarkably fast convergence. Numerical tests are then performed leveraging the turbulent channel flow simulations to validate the theoretical order of accuracy and stability properties of the new scheme. [Preview Abstract] |
Monday, November 21, 2016 5:48PM - 6:01PM |
L29.00007: ABSTRACT WITHDRAWN |
Monday, November 21, 2016 6:01PM - 6:14PM |
L29.00008: A Family of Convective-Like Energy-Stable Outflow Boundary Conditions for Incompressible Flow Simulations on Severely-Truncated Unbounded Domains Suchuan Dong A large class of flow problems are spatially developing and involves physically unbounded domains, e.g. wakes, jets, and shear layers. To numerically simulate such problems, it is necessary to truncate the domain to a finite size, and some open boundary condition (a.k.a. outflow boundary condition) will be required at the artificial boundary. Backflow instability is a commonly encountered issue with outflows or open boundaries at moderate and high Reynolds numbers. Simulations have been observed to instantly blow up when strong vortices or backflows occur at the outflow/open boundary. In this talk we present a family of convective-like open boundary conditions that effectively overcomes the backflow instability. A prominent feature of these boundary conditions is that they all ensure the energy stability of the system, even in situations where strong vortices or backflows occur at the outflow/open boundary. The proposed boundary conditions unify and provide an underlying connection between the usual convective boundary condition and the traction-free boundary condition. Several canonical wake and jet problems in open domains will be presented to demonstrate the accuracy and effectiveness of the method for overcoming backflow instability. [Preview Abstract] |
Monday, November 21, 2016 6:14PM - 6:27PM |
L29.00009: A $\sigma$-coordinate model for 3D free-surface flows using an unstructured finite-volume technique Miguel Uh Zapata The aim of this work is to develop a numerical solution of three-dimensional free-surface flows using a $\sigma$-coordinate model, a projection method and an unstructured finite-volume technique. The coordinate transformation is used in order to overcome difficulties arising from free surface elevation and irregular geometry. The projection method consists to combine the momentum and continuity equations in order to establish a Poisson-type equation for the non-hydrostatic pressure. A cell-centered finite volume method with a triangular mesh in the horizontal direction is used to simulate the flows with free-surfaces, in which the average values of conserved variables are stored at the centre of each element. A parallel algorithm is also presented for the finite volume discretization of the 3D Navier-Stokes equations. The proposed parallel method is formulated by using a multi-color SOR method, a block domain decomposition and interprocessor data communication techniques with Message Passing Interface. The model has been validated by several benchmarks which numerical simulations are in good agreement with the corresponding analytical and existing experimental results. [Preview Abstract] |
Monday, November 21, 2016 6:27PM - 6:40PM |
L29.00010: Coupling LAMMPS with Lattice Boltzmann fluid solver: theory, implementation, and applications. Jifu Tan, Talid Sinno, Scott Diamond Studying of fluid flow coupled with solid has many applications in biological and engineering problems, e.g., blood cell transport, particulate flow, drug delivery. We present a partitioned approach to solve the coupled Multiphysics problem. The fluid motion is solved by the Lattice Boltzmann method, while the solid displacement and deformation is simulated by Large-scale Atomic/Molecular Massively Parallel Simulator (LAMMPS). The coupling is achieved through the immersed boundary method so that the expensive remeshing step is eliminated. The code can model both rigid and deformable solids. The code also shows very good scaling results. It was validated with classic problems such as migration of rigid particles, ellipsoid particle's orbit in shear flow. Examples of the applications in blood flow, drug delivery, platelet adhesion and rupture are also given in the paper. [Preview Abstract] |
Follow Us |
Engage
Become an APS Member |
My APS
Renew Membership |
Information for |
About APSThe American Physical Society (APS) is a non-profit membership organization working to advance the knowledge of physics. |
© 2024 American Physical Society
| All rights reserved | Terms of Use
| Contact Us
Headquarters
1 Physics Ellipse, College Park, MD 20740-3844
(301) 209-3200
Editorial Office
100 Motor Pkwy, Suite 110, Hauppauge, NY 11788
(631) 591-4000
Office of Public Affairs
529 14th St NW, Suite 1050, Washington, D.C. 20045-2001
(202) 662-8700