Bulletin of the American Physical Society
68th Annual Meeting of the APS Division of Fluid Dynamics
Volume 60, Number 21
Sunday–Tuesday, November 22–24, 2015; Boston, Massachusetts
Session G7: CFD: Computational Methods and Modeling of Multiphase Flows III |
Hide Abstracts |
Chair: Pablo Salinas, Imperial College London Room: 107 |
Monday, November 23, 2015 8:00AM - 8:13AM |
G7.00001: A new Control Volume Finite Element Method with Discontinuous Pressure Representation for Multi-phase Flow with Implicit Adaptive time Integration and Dynamic Unstructured mesh Optimization Pablo Salinas, Dimitrios Pavlidis, James Percival, Alexander Adam, Zhihua Xie, Christopher Pain, Matthew Jackson We present a new, high-order, control-volume-finite-element (CVFE) method with discontinuous representation for pressure and velocity to simulate multiphase flow in heterogeneous porous media. Time is discretized using an adaptive, fully implicit method. Heterogeneous geologic features are represented as volumes bounded by surfaces. Our approach conserves mass and does not require the use of CVs that span domain boundaries. Computational efficiency is increased by use of dynamic mesh optimization. We demonstrate that the approach, amongst other features, accurately preserves sharp saturation changes associated with high aspect ratio geologic domains, allowing efficient simulation of flow in highly heterogeneous models. Moreover, accurate solutions are obtained at lower cost than an equivalent fine, fixed mesh and conventional CVFE methods. The use of implicit time integration allows the method to efficiently converge using highly anisotropic meshes without having to reduce the time-step. The work is significant for two key reasons. First, it resolves a long-standing problem associated with the use of classical CVFE methods. Second, it reduces computational cost/increases solution accuracy through the use of dynamic mesh optimization and time-stepping with large Courant number. [Preview Abstract] |
Monday, November 23, 2015 8:13AM - 8:26AM |
G7.00002: Collocated approximations on unstructured grids: a comparison between General Finite Differences (GFD), Moving Least Squares (MLS), and Smoothed Particle Hydrodynamics (SPH) Yaroslav Vasyliv, Alexander Alexeev In the meshfree family of methods, partial differential equations are solved on unstructured grids where a search radius establishes an implicit nodal connectivity used to determine whether to include or exclude neighboring nodes in the constructed approximation. Smoothed Particle Hydrodynamics (SPH) is widely attributed to be the eldest of the meshfree methods dating back to an astrophysics paper published in 1977 by Gingold and Monaghan. However, beating them by five years was Jensen when he published Finite Differences for Arbitrary Grids (FIDAG) in 1972. Ultimately this work and others were generalized by Liszka and Orkisz in 1979 as a weighted least squares formulation solving for the Taylor coefficients and is now commonly known as General Finite Differences (GFD). Shortly after in 1981, Lancaster and Salkauskas introduced the Moving Least Squares (MLS) approximation for surface reconstruction using a weighted least squares formulation where the unknown coefficients are treated as functions varying from node to node in the support domain. Here we examine important differences, similarities and limitations of each method by solving the 2D Poisson equation on unstructured grids. [Preview Abstract] |
Monday, November 23, 2015 8:26AM - 8:39AM |
G7.00003: Multi-resolution flow simulations by smoothed particle hydrodynamics via domain decomposition Xin Bian, Zhen Li, Yu-Hang Tang, George Karniadakis We present a methodology to concurrently couple particle-based methods via a domain decomposition (DD) technique for simulating viscous flows. In particular, we select two resolutions of the smoothed particle hydrodynamics (SPH) method as demonstration. Within the DD framework, a simulation domain is decomposed into two (or more) overlapping sub-domains, each of which has an individual {\it particle scale} determined by the local flow physics. Consistency of the two sub-domains is achieved in the overlap region by matching the two independent simulations based on Lagrangian interpolation of {\it state variables} and {\it fluxes}. The domain decomposition based SPH method (DD-SPH) employs different spatial and temporal resolutions, and hence, each sub-domain has its own smoothing length and time step. As a consequence, particle refinement and de-refinement are performed {\it asynchronously} according to individual time advancement of each sub-domain. The proposed strategy avoids SPH force interactions between different resolutions on purpose, so that coupling, in principle, can go beyond SPH - SPH, and may allow SPH to be coupled with other mesoscopic or microscopic particle methods. The DD-SPH method is validated first for a transient Couette flow, where simulation results base [Preview Abstract] |
Monday, November 23, 2015 8:39AM - 8:52AM |
G7.00004: The effect of density estimation on the conservativeness in Smoothed Particle Hydrodynamics Pranav Suresh, S.S.Prasanna Kumar, B.S.V. Patnaik Smoothed Particle Hydrodynamics (SPH) is a popular mesh-free method for solving a wide range of problems that involve interfaces. In SPH, the Lagrangian nature of the method enables mass conservation to be naturally satisfied. However, satisfying the conservation of momentum and energy are indeed formulation dependent. One major aspect of ensuring conservativeness comes from the density estimation. There are two distinct types of density estimation approaches, namely continuity density approach and summation density approach. Both approaches are indeed popular with single and multi-phase flow communities. In the present study, we assess the role of density evaluation on the conservativeness, using several representative numerical examples. In particular, we have simulated the Rayleighâ€“Taylor instability problem, Non-Boussinesq lock exchange problem, bubble rise in water column etc. Although for shorter time scales of simulation, both methods have similar conservative properties, we observe that for longer time scales, summation-density approach is better. For free surface detection and normal vector computations, efficient computational procedures have been devised. [Preview Abstract] |
Monday, November 23, 2015 8:52AM - 9:05AM |
G7.00005: Advection Scheme for Phase-changing Porous Media Flow of Fluids with Large Density Ratio Duan Zhang, Juan Padrino Many flows in a porous media involve phase changes between fluids with a large density ratio. For instance, in the water-steam phase change the density ratio is about 1000. These phase changes can be results of physical changes, or chemical reactions, such as fuel combustion in a porous media. Based on the mass conservation, the velocity ratio between the fluids is of the same order of the density ratio. As the result the controlling Courant number for the time step in a numerical simulation is determined by the high velocity and low density phase, leading to small time steps. In this work we introduce a numerical approximation to increase the time step by taking advantage of the large density ratio. We provide analytical error estimation for this approximate numerical scheme. Numerical examples show that using this approximation about 40-fold speedup can be achieved at the cost of a few percent error. [Preview Abstract] |
Monday, November 23, 2015 9:05AM - 9:18AM |
G7.00006: Simulations of Coalescence and Breakup of Interfaces Using a 3D Front-tracking Method Jiacai Lu, Gretar Tryggvason Direct Numerical Simulations (DNS) of complex multiphase flows with coalescing and breaking-up of interfaces are conducted using a 3D front-tracking method. Front-tracking method has been successfully used in DNS of turbulent channel bubbly flows and many other multiphase flows, but as the void fraction increases changes in the interface topology, though coalescence and breakup, become more common and have to be accounted for. Topology changes have often been identified as a challenge for front tracking, where the interface is represented using a triangular mesh, but here we present an efficient algorithm to change the topology of triangular elements of interfaces. In the current implementation we have not included any small-scale attractive forces so thin films coalesce either at prescribed times or when their thickness reaches a given value. Simulations of the collisions of two drops and comparisons with experimental results have been used to validate the algorithm but the main applications have been to flow regime transitions in gas-liquid flows in pressure driven channel flows. The evolution of flow, including flow rate, wall shear, projected interface areas, pseudo-turbulence, and the average size of the various flow structures, is examined as the topology of the interface changes through coalescence and breakup. [Preview Abstract] |
Monday, November 23, 2015 9:18AM - 9:31AM |
G7.00007: Addressing the Numerical Challenges Associated With Laser-Induced Melt Convection Brian Weston, Robert Nourgaliev, Jean Pierre Delplanque, Andy Anderson We present a new robust and efficient numerical framework for simulating multi-material flows with phase change. The work is motivated by laser-induced phase change applications, particularly the selective laser melting (SLM) process in additive manufacturing. Physics-based simulations of the laser melt dynamics requires a fully compressible framework, since incompressible flow solvers are inefficient for stiff systems, arising from laser-induced rapid phase change. In this study, the liquid and solid phases are both modeled with the compressible Navier-Stokes equations. The solid phase has an additional combined variable viscosity and drag force model to suppress the velocity in the solid. Our all-speed Navier-Stokes solver is based on a fully-implicit, high-order reconstructed Discontinuous Galerkin method. A Newton-Krylov based framework is used to solve the resulting set of non-linear equations, enabling robust simulations of the highly stiff compressible Navier-Stokes equations. We demonstrate the method's capabilities for phase change on several different melting and freezing configurations, including a three-dimensional laser-induced melt convection problem. Future model enhancements will incorporate material evaporation and rapid solidification. [Preview Abstract] |
Monday, November 23, 2015 9:31AM - 9:44AM |
G7.00008: An inviscid regularization technique for two-phase flows with shocks and turbulence Kamran Mohseni, Teng Li An inviscid regularization technique for the simulation of multiphase flows with sharp interfaces is introduced. This methodology is based on a similar approach successfully used by our group in the past for regularizing single-phase problems with shocks and/or turbulence. The observable divergence theorem is employed to obtain the governing equations, namely the observable Euler and Navier-Stokes equations, from the conservation laws. Results of several inviscid simulations of incompressible and compressible two-phase flows with sharp interfaces are reported and compared with other available techniques. Specifically, simulation results of the Rayleigh-Taylor instability and rising bubble problems in viscous or inviscid fluids are reported. [Preview Abstract] |
Monday, November 23, 2015 9:44AM - 9:57AM |
G7.00009: Modeling of Two-Phase Immiscible Flow with Moving Contact Lines Moataz Abu AlSaud, Cyprien Soulaine, Amir Riaz, Hamdi Tchelepi A new numerical method based on the implicit interface approach on Cartesian grids is proposed for modeling two-phase immiscible flow with moving contact lines. The reinitialization of level-set function by computing the minimum distance to linearly reconstructed interface to obtain signed distance function is extended to include the contact angle boundary condition. The physics of contact line dynamics is implemented using the Cox-Voinov hydrodynamic theory that efficiently captures the effect of the microscopic contact line region. The numerical method is validated through various examples. Parasitic currents are studied in the case of static and constantly advected parabolic interface intersecting the domain boundary with an imposed contact angle. Moving contact line in the viscous dominated regime is studied and verified through comparison with experiments. [Preview Abstract] |
Monday, November 23, 2015 9:57AM - 10:10AM |
G7.00010: A low Mach number preconditioned scheme for a two-phase liquid-gas compressible flow model Marica Pelanti The simulation of liquid-gas flows such as cavitating flows demands numerical methods efficient for a wide range of Mach number regimes, due to the large and rapid variation of the speed of sound in these two-phase flows. When classical upwind finite volume discretizations for compressible flow models are employed, suitable strategies are needed to overcome the well known difficulty of loss of accuracy encountered at low Mach number by these methods. In this work we present a novel finite volume wave propagation scheme with low Mach number preconditioning for the numerical approximation of a six-equation two-phase liquid-gas compressible flow model with stiff mechanical relaxation. A Turkel-type preconditioner is designed to correct the acoustic fields at low Mach number, by altering the numerical dissipation tensor of the scheme. We present numerical results for two-dimensional liquid-gas nozzle flow tests both for low Mach number regimes and for transonic regimes with shock formation, which show the effectiveness and accuracy of the proposed preconditioned method. In particular, in the low Mach number limit the order of pressure perturbations at the discrete level agrees with the theoretical results for the continuous two-phase flow model. [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. |
© 2019 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
1 Research Road, Ridge, NY 11961-2701
(631) 591-4000
Office of Public Affairs
529 14th St NW, Suite 1050, Washington, D.C. 20045-2001
(202) 662-8700