Bulletin of the American Physical Society
77th Annual Meeting of the Division of Fluid Dynamics
Sunday–Tuesday, November 24–26, 2024; Salt Lake City, Utah
Session L13: CFD: General II |
Hide Abstracts |
Chair: Jared Whitehead, Brigham Young University Room: 155 C |
Monday, November 25, 2024 8:00AM - 8:13AM |
L13.00001: A modern concept of Lagrangian hydrodynamics Jesse Canfield, Len Margolin We give a modern overview of Lagrangian hydrodynamics as it is implemented in Lagrangian codes for modeling compressible fluid dynamics. The main result is to show that artificial viscosity, often treated as a numerical knob to control unphysical oscillations near shocks, actually represents a physical process and it is required to produce accurate simulation results for any compressible flow. In the first part, we review the origins of two numerical tools, artificial viscosity and finite volume methods. Then, we mathematically derive a set of partial differential equations (PDE) with a high-order finite volume approximation that contains a new length scale, the observer, as part of the equations that arises representing the discretization. Associated with that length scale, there are new inviscid fluxes which are the artificial viscosity that was first formulated by Richtmyer and von Neumann. Additionally an artificial heat flux postulated by Noh appears that is not typically included in Lagrangian codes. The results are discussed in the context of bi–velocity hydrodynamics. Finally, the conclusion introduces speculations about the direction of future developments in multidimensional Lagrangian codes as computers get faster and have larger memories. |
Monday, November 25, 2024 8:13AM - 8:26AM |
L13.00002: Numerical Modeling and Convergence Analysis of the Coupled Cahn-Hilliard-Navier-Stokes System Using a Combined Scalar Auxiliary Variable and HDG Scheme Sanchita Chakraborty, Yukun Yue, Adam Distler, Alexis Liu In this presentation, we introduce an overview and a novel numerical approach for solving the Cahn-Hilliard-Navier-Stokes system. Our method integrates the Scalar Auxiliary Variable (SAV) technique with a Hybridized-Discontinuous Galerkin (HDG) scheme. Initially, we focus on the development and analysis of a first-order numerical scheme and establish its convergence properties. Through extensive numerical simulations, we validate the effectiveness of the proposed scheme in accurately simulating the Cahn-Hilliard-Navier-Stokes system. Additionally, we provide an analytical convergence proof to support our findings. This work enhances our understanding of the system dynamics and offers a robust numerical framework for future simulations. |
Monday, November 25, 2024 8:26AM - 8:39AM |
L13.00003: Simulation of a bubble curtain in coastal engineering applications for the retention of suspended particles, erosion and sedimentation. Ernesto Ramon Cornelio, Alicia Aguilar Corona, Bernardo Figueroa Espinoza A numerical study has been carried out using the OpenFoam open source code to evaluate the effectiveness of a bubble curtain immersed in a liquid with suspended particles, with the purpose of investigating its feasibility in coastal engineering applications for the retention of these particles. 2D numerical simulations were carried out to determine the size of the recirculations generated by the curtain at different depths and their influence on the suspended particles, analyzing the percentage retention of these under various scenarios. In addition, the behavior and configuration of the curtain was investigated against the action of waves, currents and their combination. The results obtained were contrasted with data available in the specialized literature. |
Monday, November 25, 2024 8:39AM - 8:52AM |
L13.00004: Novel efficient, high-resolution and structure-preserving convection schemes for computational fluid dynamics Xi Deng, Zhenhua Jiang, Omar K Matar In this work, we present our recent efforts to develop high-resolution, scalar-structure-preserving schemes within a three-cell-based compact stencil. Our approach begins with unifying existing three-cell-based non-linear reconstruction schemes into a single framework through the definition of a Unified Normalised-Variable Diagram (UND). Using UND, the reconstruction operator can be designed directly in the normalised-variable space to satisfy desirable properties such as Total Variation Diminishing (TVD) and Essentially Non-Oscillatory (ENO). The resolution of the reconstruction operator is enhanced by introducing anti-diffusion errors to UND, while preserving the scalar structure through fine-tuning these errors. The resultant novel scheme is termed ROUND (Reconstruction Operator on Unified Normalised-Variable Diagram). The proposed ROUND schemes are evaluated using benchmark tests, demonstrating their superior performance in terms of accuracy and resolution. Notably, in some cases, the low-dissipative ROUND schemes achieve comparable or even better resolution than the classic fifth-order Weighted ENO (WENO). We further extend and implement the ROUND scheme on unstructured grids within the OpenFOAM framework. Compared to conventional second-order schemes in OpenFOAM, ROUND significantly reduces numerical errors at a similar computational cost. Additionally, ROUND offers improved structure-preserving properties over conventional schemes. The performance of ROUND schemes is further evaluated by simulating high-speed compressible single-phase and multiphase flows. We also demonstrate the efficacy of ROUND schemes by extending them to the Finite Difference Method Immersed Boundary Method (FDM-IBM) and Discontinuous Galerkin methods, thereby enhancing robustness and numerical resolution. |
Monday, November 25, 2024 8:52AM - 9:05AM |
L13.00005: Numerical stability analysis of a coupled fluid-elastic structure system using the Method of Regularized Stokeslets. Dana Ferranti, Sarah D Olson The method of regularized Stokeslets (MRS) is a widely used numerical method for simulating fluid-structure interaction in the Stokes flow regime (Reynolds number Re=0). The method works by desingularizing forces distributed over a curve or surface discretization of the structure, resulting in a smooth velocity field that can be directly evaluated at any point in the domain. While several works have been devoted to improving the error that arises from the spatial discretization or regularization, relatively fewer works have worked on issues of numerical stability. This is particularly important in the modeling of deformable bodies (e.g. flagella or cilia), where a common approach is to model elasticity by connecting points on the body with virtual Hookean springs. This approach results in a stiff system that requires extremely small time steps to maintain stability. Here, we present work from our investigation into the nature of these stability issues. We start with a linear stability analysis of an elastic membrane immersed in a Stokes flow subject to small perturbations from its configuration at equilibrium. For a particular choice of regularized delta function and time stepping scheme, this allows us to estimate a critical time step at which the system becomes unstable. The analysis is accompanied by numerical simulations where we initialize the elastic membrane with small deformations for validation, and with larger deformations to understand how our insights can be used in practice. |
Monday, November 25, 2024 9:05AM - 9:18AM |
L13.00006: Effect of Numerical Dissipation on Long Time Energy Dissipation Rate Kiera E Harmatz-Kean When modeling fluids, accurately capturing the energy dissipation rate is vital to ensure the correct long time behavior of the simulation. Many popular turbulence models, e.g., eddy viscosity models, may overdissipate in the long time average, pushing flow to lower Reynolds number states. Other sources of dissipation, e.g., numerical dissipation can also contribute to incorrect behavior in simulations. Here, we examine the effect of numerical dissiaption on the long time energy dissipationr ate. A-stable timestepping methods, which are necessary for high Reynolds number simulations are known to be dissipative. We explore the effect of the time discretization on the energy dissipation rate for a simple numerical experiment for three dimensional flow. |
Monday, November 25, 2024 9:18AM - 9:31AM |
L13.00007: Analysis of Particle Contamination in Plasma Enhanced Chemical Vapor Deposition Chamber Using Computational Fluid Dynamics Han Sol Lee, Han Seo Ko, Dong Kee Sohn Defects arising from particle contamination in Plasma Enhanced Chemical Vapor Deposition (PECVD) chambers pose critical quality issues in semiconductor devices, significantly impacting their performance and reliability. In showerhead-type PECVD systems, the showerhead serves as both the gas distributor and electrode. Particles tend to form at the showerhead edges due to interactions between plasma, and particles can also be generated on the chamber wall because of the by-products. Periodic purging is essential to reduce particle contamination within the chamber, as particle behavior is influenced by temperature and pressure variations. Computational Fluid Dynamics (CFD) was employed to analyze the gas flow within the chamber under various pressure and temperature conditions. The simulation data guided our experiments, revealing that a pressure of 100 Torr and a temperature of 723 K are optimal for minimizing particle contamination. Our findings provide crucial insights for improving PECVD chamber designs and operational protocols, thereby enhancing the overall quality and reliability of semiconductor devices. By optimizing the operating conditions, manufacturers can achieve higher yields and better device performance, ultimately contributing to advancements in semiconductor technology. |
Monday, November 25, 2024 9:31AM - 9:44AM |
L13.00008: A fully-implicit method to simulate convection in spherical shells Bhargav Mantravadi, Stefano Zampini, Pankaj Jagad, Peter J Schmid We present a fully-implicit method to simulate Boussinesq convection in spherical shells. This method combines the discrete exterior calculus and second-order central finite difference methods. The governing equations are formulated as a non-linear function and are solved fully implicitly using the backward Euler time integration scheme, where the Jacobian is hand-coded and utilized in the integration. We use a triangulated spherical surface mesh that is distributed across the MPI ranks, and a Chebyshev grid along the radial direction for higher resolution of the boundary layers. We apply the method of manufactured solutions (MMS) and observe second-order spatial convergence. The method is verified through various test cases, including the determination of critical Rayleigh numbers for spherical shells of varying aspect ratios, simulation of spherical spiral rolls in moderately thin shells, and Nusselt-Rayleigh number scaling. We also present strong and weak scaling tests conducted on the Shaheen-III supercomputer. |
Monday, November 25, 2024 9:44AM - 9:57AM |
L13.00009: Inexact Newton-Krylov for Fractional Step Method in Incompressible Flow Solver based on the Adaptive Mesh Refinement for Exascale (AMReX) Framework Thien-Tam Thien Nguyen, Andy Nonaka, Trung Bao Le We present a novel approach for simulating incompressible flows with moving boundaries. Our method uniquely combines staggered/non-staggered grid layouts with a projection method, storing fluxes at volume faces and pressure fields at volume centers. This hybrid approach allows flexible boundary condition prescription on moving bodies while maintaining exact incompressibility. A key innovation is our momentum equation solver, which employs an Inexact Newton method, offering improved convergence and efficiency over traditional iterative schemes such as the Runge-Kutta method. The Adaptive Mesh Refinement for Exascale (AMReX) framework continues to be integrated for leveraging fast linear solvers, extensibility to AMR, and parallelization in multi-GPU HPC clusters. The accuracy of our schemes has been validated through two standard 2D problems: (i) lid-driven cavity flow; and (ii) Taylor-Green vortex. The large-scale performance of our code is important in many problems in fluid-structure interaction of biological flows. Finally, our scaling tests are performed to benchmark performance improvements across various grid sizes and heterogeneous computing infrastructures. This work is supported by the NSF grant number 1946202 ND-ACES and a start-up package of Trung Le from North Dakota State University. The authors acknowledge the use of computational resources at the Center for Computationally Assisted Science and Technology CCAST-NDSU, which is supported by the NSF MRI 2019077. |
Monday, November 25, 2024 9:57AM - 10:10AM |
L13.00010: Analysis of the Principle of Minimum Pressure Gradient Method on Airfoil Behavior through Stall Sweety Sarker, Michael Kinzel Understanding overall airfoil aerodynamic performance is critical to design wings and lifting bodies as it is a first-order driver of vehicle efficiency. Traditionally, aerodynamics is assessed by examining the lift-to-drag ratio, integral forces, moment measurements, and/or pressure distribution patterns. In this research, we aim to explore the Principle of Minimum Pressure Gradient (PMPG) [1] and if it can provide new aerodynamic insight. PMPG is a newly introduced concept that proposes Navier Stokes solutions are minimizations of the pressure gradient norm in various fluid flows. PMPG offers a potentially transformative approach to fluid mechanics as it proposes a new minimization scalar that is not directly the traditional Navier-Stokes equations. In this work, we investigate PMPG on three airfoils (NACA 63-412, NACA 63-412 with flap 15 degree, and Eppler 25) using computational fluid dynamics (CFD) over a wide range of angles of attack (α) to characterize the flow separation, lift, drag, pitch moment, and lift-to-drag ratio from attached flow through stall. The innovative aspect of this effort aims to exploit data-access provided by a CFD solver to directly assess the flow-field domain and develop an understanding of the PMPG as a new metric to characterize aerodynamic efficiency. Some preliminary results of the analysis suggest that the PMPG develops connections between , Cl, Cd, flow-separation point, Cl/Cd, and the PMPG integral. The PMPG integral indicates a nonlinear behavior and a loss of the PMPG integral as stall is approached. This character will be evaluated quantitatively and qualitatively to better understand the metric as an aerodynamic tool to drive aerodynamic design. |
Monday, November 25, 2024 10:10AM - 10:23AM |
L13.00011: Extending the Pele Suite through Python and pyAMReX Nicholas T Wimer, Riley Fisher, Marc Day The Pele Suite is a collection of adaptive mesh refinement CFD codes designed for handling real-world, complex turbulent reacting flows while effectively scaling to exascale computing resources. These codes are highly performant and scalable across a variety of computing architectures thanks to the AMReX framewok on which they are built. In this talk we will discuss our recent work in extending the capabilities of these codes using pyAMReX, exposing our underlying CFD algorithms to Python, opening up a range of new functionalities. We present the overall framework of the extension process and showcase some of the new capabilities enabled by expanding into the Python ecosystem. |
Monday, November 25, 2024 10:23AM - 10:36AM |
L13.00012: Robust Data-Driven Turbulence Modeling for RANS Closures Using a SciML Approach for Validation Uma Balakrishnan, William J Rider, Matthew Barone, Eric Parish Scientific Machine Learning is revolutionizing scientific domains crucial for national security by enhancing analytical capabilities. The Reynolds-averaged Navier–Stokes (RANS) equations, essential for simulating compressible fluid flows, suffer from model-form errors that limit their applicability. Addressing this, Parish et al. [AIAA 2023-2126] introduced a data-driven turbulence modeling strategy to improve RANS models. Utilizing multi-step training on eight diverse datasets, (channel flows at different Reynolds numbers, duct flow, periodic hill, and hypersonic boundary layers) the study demonstrated the model's efficacy. The research focuses on predicting Reynolds stress term discrepancies, emphasizing hyperparameter sensitivity and out-of-distribution dataset performance across various combinations of training datasets. Extensive validation efforts ensure the reliability of the machine learning models in capturing elusive model-form errors. The findings show robust improvements in wall-bounded flows, jet flows, and hypersonic boundary layers, advancing turbulence modeling comprehension. Future phases will further test dataset and hyperparameter sensitivity to ensure credibility and applicability. This study significantly enhances RANS simulations' accuracy and reliability, impacting broader fluid dynamics and machine learning fields. This work is supported by the DOE-NNSA ASC program. |
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. |
© 2025 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