Bulletin of the American Physical Society
72nd Annual Meeting of the APS Division of Fluid Dynamics
Volume 64, Number 13
Saturday–Tuesday, November 23–26, 2019; Seattle, Washington
Session G24: Multiphase Flows: Computational Methods III |
Hide Abstracts |
Chair: Fabian Denner, Magdeburg Room: 606 |
Sunday, November 24, 2019 3:48PM - 4:01PM |
G24.00001: Analysis and performance of high-order finite difference shock capturing schemes for multi-fluid flow computations Khosro Shahbazi Ahigh-order finite difference scheme for compressible multi-fluid and multi-phase dynamics. The scheme is applicable to a wide range compressible multiphase models including mixture theory models, capable of phase change and spontaneous cavitation modeling. The scheme overcomes the difficulty of applying the common flux-based WENO finite difference scheme to multi-fluid problems by applying the reconstruction on primitive variables and, thus avoiding the spurious oscillations inherent in the standard methods. Schemes of orders up to nine are introduced and analyzed. The proposed finite difference schemes are significantly more efficient than the available high-order finite volume schemes in both storage requirement, operation counts, and inter-processor message passing in parallel computations with efficiency gains being higher at higher orders and higher spatial dimensions. For the same level of accuracy, in three-dimensional calculations, a fourfold speedup or higher at fifth-order accuracy or higher over the finite volume scheme is expected. In computations of a two-dimensional shock bubble interaction, good agreements with experimental data are obtained and competitive performance to the high-order finite volume scheme is shown. [Preview Abstract] |
Sunday, November 24, 2019 4:01PM - 4:14PM |
G24.00002: Low-dissipation strategies for simulating compressible turbulent multiphase flows Michael B. Kuhn, Olivier Desjardins Simulating compressible, multiphase flows requires a robust numerical solver, due to the discontinuities introduced by shocks and phase interfaces. Geometric transport schemes and pressure projection schemes both provide stability and the freedom to take larger timesteps, but these also introduce spurious dissipation of kinetic energy. That dissipation diminishes the resolution of turbulent processes, limiting the efficacy of these computational methods in compressible, multiphase applications that prominently feature turbulence, such as transonic atomization. While taking advantage of the benefits of geometric transport and pressure projection, we address and mitigate sources of numerical dissipation, employing a hybrid transport scheme and a modified pressure implementation. Using simulations of droplets in turbulence, we quantify the influence of these aspects of the flow solver and evaluate the scheme by comparing with results from the literature. Finally, we demonstrate the capability of our scheme in simulating transonic atomization. [Preview Abstract] |
Sunday, November 24, 2019 4:14PM - 4:27PM |
G24.00003: Pressure-based algorithm and thermodynamic closure for compressible gas-liquid flows Fabian Denner, Fabien Evrard, Berend van Wachem Simulating compressible gas-liquid flows, e.g. air-water flows, presents considerable numerical challenges due to the stiff pressure-density-temperature relationship of the liquid and the sharp difference in compressibility at the fluid interface. We present a fully-coupled pressure-based algorithm for the simulation of interfacial flows in all Mach number regimes, based on a conservative finite-volume discretisation and a VOF-PLIC method to represent the interface, which treats the continuity equation as an equation for pressure and solves the discrete governing equations in a single linear system of equations. In this contribution, we focus especially on the implementation of the discretised governing equations and on different thermodynamic closures based on the Noble-Abel-stiffened-gas model in fully-compressible and polytropic form. Results of representative test-cases, e.g. pressure-driven bubble collapse or the interaction of shocks with bubbles and drops, are presented to scrutinise the presented algorithm and to highlight the differences caused by the considered thermodynamic closure models. [Preview Abstract] |
Sunday, November 24, 2019 4:27PM - 4:40PM |
G24.00004: Simulating Multi-Species Compressible Reactive Flow at Low-Mach Number with a High-Order Fully-Implicit All-Speed Flow Solver Brian Weston, Robert Nourgaliev, Matt McClelland We present a high-order, fully-implicit all-speed fluid dynamics solver for simulating multi-species compressible reactive flow at very low-Mach numbers. The work is motivated by the development of high-explosive cookoff simulations, which requires modeling multi-species/multi-phase reactive melt convection physics over long time-scales. The governing equations are discretized in space up to 5$^{\mathrm{th}}$-order accuracy with a reconstructed Discontinuous Galerkin method and integrated in time with $L$-stable fully implicit time discretization schemes. The resulting set of non-linear equations is converged using a robust physics-block based preconditioned Newton-Krylov solver. We demonstrate that our fully-implicit flow solver is able to robustly converge multi-species compressible flow calculations with Mach numbers less than 10$^{\mathrm{-5}}$. Furthermore, our fully-implicit framework allows for large time steps relative to fast chemical kinetic timescales, which result in highly stiff linear systems. [Preview Abstract] |
Sunday, November 24, 2019 4:40PM - 4:53PM |
G24.00005: A Dual Scale Model for Reconstructing Sub-Filter Shear Driven Instabilities Austin Goodrich, Marcus Herrmann A method to compute sub-filter shear-induced velocity on a liquid-gas interface for use in a dual-scale LES-DNS model is presented. The method computes velocity perturbation growth rates by constructing a linear eigenvalue problem based on the well known Orr-Sommerfeld equation using a velocity profile approximated with an error function scaled by the far-field velocities and a prescribed boundary layer thickness. The Orr-Sommerfeld equation, along with appropriate boundary and interface conditions, is then solved numerically with a Chebyshev collocation method as outlined by Schmid and Henningson (2001). The eigenfunctions of the Orr-Sommerfeld equations are expanded into Chebyshev polynomials and evaluated at their Gauss-Lobatto points for spectral accuracy, resulting in an algebraic eigenvalue that can be solved using a standard linear algebra package. With the unstable growth rates computed, the streamfunction definitions are used to compute the normal velocities at the liquid-gas interface. The Chebyshev method is tested under a variety of conditions, and results are presented and compared against prior literature. [Preview Abstract] |
Sunday, November 24, 2019 4:53PM - 5:06PM |
G24.00006: A stabilized coupled level set and volume of fluid method for incompressible two-phase flow at high Reynolds number and high density ratio Han Liu, Qiang Gao, Lian Shen When coupled level set and volume of fluid (CLSVOF) method is used coupling with non-conservative schemes, it can suffer from instability issue when the Reynolds number or the density ratio is high. We present an improved CLSVOF method that is able to simulate two-phase flows at very high Reynolds numbers and large density ratios on Cartesian grid while having high accuracy for turbulent flow resolution. To reduce the error near the two-phase interface, a consistent treatment of mass and momentum transport in the conservative form of discrete equations is employed to solve the nonlinear inertial terms of the Navier-Stokes equations. To resolve the discontinuous momentum across the two-phase interface without oscillation while keeping the accuracy of the numerical solution, a WENO scheme is used for the reconstruction of both velocity and density. The accuracy and robustness of this method are validated by benchmark tests. Quantitative comparisons have been made to show the capability of the method handling realistic two-phase flow problems at high Reynolds numbers and high density ratios. [Preview Abstract] |
Sunday, November 24, 2019 5:06PM - 5:19PM |
G24.00007: On the use of traction outlet boundary conditions for turbulent multiphase flows Cyril Bozonnet, Olivier Desjardins, Guillaume Balarac Due to the finite nature of numerical simulations, computational domains need to be truncated and artificial boundary conditions need to be introduced to close the system of equations being solved. In the context of turbulent multiphase flows, this artificial boundary may lead to the development of unstable backflow patterns in outflow regions and can be the source of wave reflection. Moreover, if the flow is incompressible, the position of the artificial boundary can durably impact the upstream flow. The “stabilized traction-free boundary condition” has already been introduced in order to mitigate backflow instabilities. In this talk, we will present the improvements that can be obtained by using a non-zero traction boundary condition. Specifically, error level, effect of domain truncation, and surface wave reflection are analyzed. This novel outlet boundary treatment is presented in the context of pressure projection algorithms and interface capturing methods. [Preview Abstract] |
Sunday, November 24, 2019 5:19PM - 5:32PM |
G24.00008: A numerical model for liquid-vapor flows with arbitrary heat and mass transfer relaxation times and general equation of state Marica Pelanti, Marco De Lorenzo, Philippe Lafon We describe liquid-vapor flows by a single-velocity six-equation two-phase compressible flow model with relaxation source terms accounting for volume, heat and mass transfer. The system of equations is numerically solved by a classical fractional step algorithm, where we alternate between the solution of the homogeneous hyperbolic portion of the model system via a HLLC-type finite volume scheme, and the solution of a sequence of systems of ordinary differential equations for the relaxation source terms driving the flow toward mechanical, thermal and chemical equilibrium. For an accurate description of the thermodynamical processes involved in transient liquid-vapor flow problems it is often important to be able to simulate both instantaneous and finite-rate relaxation processes. In this work we present new numerical relaxation procedures to integrate interphase transfer terms with two significant properties: the capability to describe heat and mass transfer processes of arbitrary relaxation time, and the applicability to a general equation of state. We show the effectiveness of the proposed computational model by presenting several numerical tests in one and two dimensions, including simulations of depressurization and blowdown experiments. [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. |
© 2020 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