arXiv subjects

Xiangyu Hu

Publications and source records attributed to Xiangyu Hu.

81 records · Page 5Linked to original sources

arXiv2019

A weakly compressible SPH method for violent multi-phase flows with high density ratio

The weakly compressible SPH (WCSPH) method is known suffering from low computational efficiency, or unnatural voids and unrealistic phase separation when it is applied to simulate highly violent multi-phase flows with high density ratio, such as that between water and air. In this paper, to remedy these issues, we propose a multi-phase WCSPH method based on a low-dissipation Riemann solver and the transport-velocity formulation. The two-phase Riemann problem is first constructed to handle the pairwise interaction between fluid particles, then modified for the fluid-wall interaction to impose the solid wall boundary condition. Since the method uses the same artificial speed of sound for both heavy and light phases, the computational efficiency increases greatly. Furthermore, due to the transport-velocity formulation employed for the light phase and application of the two-phase Riemann problem, the unnatural voids and unrealistic phase separation are effectively eliminated. The method is validated with several 2- and 3D cases involving violent water-air flows. The results have been compared with existing experimental data, previous numerical and analytical solutions, where the proposed method demonstrates good robustness, improved or comparable accuracy, respectively, comparing to previous methods with same choice of sound speed or those with much less computational efficiency.

physics.comp-ph↗

arXiv2018

Free-stream preserving linear-upwind and WENO schemes on curvilinear grids

Applying high-order finite-difference schemes, like the extensively used linear-upwind or WENO schemes, to curvilinear grids can be problematic. The geometrically induced error from grid Jacobian and metrics evaluation can pollute the flow field, and degrade the accuracy or cause the simulation failure even when uniform flow imposed, i.e. free-stream preserving problem. In order to address this issue, a method for general linear-upwind and WENO schemes preserving free-stream on stationary curvilinear grids is proposed. Following Lax-Friedrichs splitting, this method rewrites the numerical flux into a central term, which achieves free-stream preserving by using symmetrical conservative metric method, and a numerical dissipative term with a local difference form of conservative variables for neighboring grid-point pairs. In order to achieve free-stream preservation for the latter term, the local difference are modified to share the same Jacobian and metric terms evaluated by high order schemes. In addition, a simple hybrid scheme switching between linear-upwind and WENO schemes is proposed of improving computational efficiency and reducing numerical dissipation. A number of testing cases including free-stream, isentropic vortex convection, double Mach reflection and flow past a cylinder are computed to verify the effectiveness of this method.

physics.comp-ph↗

arXiv2018

Positivity-preserving method for multi-resolution simulations of compressible flows

We present a positivity-preserving method for multi-resolution simulations of compressible flows involving extreme conditions such as near vacuum and strong discontinuities. The novelty of this work is due to two aspects. First we extend the positivity-preserving flux limiter of (Hu et al., J Comput Phys 242, 2013) to the multi-resolution framework by modifying the prediction operator, based on the same limiting concept. Second, we develop a positivity-preserving local time stepping scheme for adaptive time marching. Instead of using fixed hierarchical time steps, the local time stepping scheme dynamically adjusts the time steps of all multi-resolution levels to maintain positivity. The method is validated and its capabilities are demonstrated by a range of test cases.

physics.comp-ph↗

arXiv2018

Liquid Splash Modeling with Neural Networks

This paper proposes a new data-driven approach to model detailed splashes for liquid simulations with neural networks. Our model learns to generate small-scale splash detail for the fluid-implicit-particle method using training data acquired from physically parametrized, high resolution simulations. We use neural networks to model the regression of splash formation using a classifier together with a velocity modifier. For the velocity modification, we employ a heteroscedastic model. We evaluate our method for different spatial scales, simulation setups, and solvers. Our simulation results demonstrate that our model significantly improves visual fidelity with a large amount of realistic droplet formation and yields splash detail much more efficiently than finer discretizations.

cs.GR↗

arXiv2018

A network partition method for solving large-scale complex nonlinear processes

A numerical framework based on network partition and operator splitting is developed to solve nonlinear differential equations of large-scale dynamic processes encountered in physics, chemistry and biology. Under the assumption that those dynamic processes can be characterized by sparse networks, we minimize the number of splitting for constructing subproblems by network partition. Then the numerical simulation of the original system is simplified by solving a small number of subproblems, with each containing uncorrelated elementary processes. In this way, numerical difficulties of conventional methods encountered in large-scale systems such as numerical instability, negative solutions, and convergence issue are avoided. In addition, parallel simulations for each subproblem can be achieved, which is beneficial for large-scale systems. Examples with complex underlying nonlinear processes, including chemical reactions and reaction-diffusion on networks, demonstrate that this method generates convergent solution in a efficient and robust way.

physics.comp-ph↗

arXiv2018

A variational-level-set based partitioning method for block-structured meshes

We propose a numerical method for solving block-structured mesh partitioning problems based on the variational level-set method of (Zhao et al., J Comput Phys 127, 1996) which has been widely used in many partitioning problems such as image segmentation and shape optimization. Here, the variational model and its level-set formulation have been simplified that only one single level-set function is evolved. Thus, the numerical implementation becomes simple, and the computational and memory overhead are significantly alleviated, making this method suitable for solving realistic block-structured mesh partitioning problems where a large number of regions is required. We start to verify this method by a range of two-dimensional and three-dimensional uniform mesh partitioning cases. The results agree with the theoretical solutions very well and converge rapidly. More complex cases, including block-structured adaptive mesh partitioning for single-phase and multi-phase multi-resolution simulations, confirm the accuracy, robustness and good convergence property. The measured CPU time shows that this method is efficient for both two-dimensional and three-dimensional realistic partitioning problems in parallel computing. The proposed method has the potential to be extended to solve other partitioning problems by replacing the energy functional.

physics.comp-ph↗

arXiv2017

A conservative sharp-interface method for compressible multi-material flows

In this paper we develop a conservative sharp-interface method dedicated to simulating multiple compressible fluids. Numerical treatments for a cut cell shared by more than two materials are proposed. First, we simplify the interface interaction inside such a cell with a reduced model to avoid explicit interface reconstruction and complex flux calculation. Second, conservation is strictly preserved by an efficient conservation correction procedure for the cut cell. To improve the robustness, a multi-material scale separation model is developed to consistently remove non-resolved interface scales. In addition, the multi-resolution method and local time-stepping scheme are incorporated into the proposed multi-material method to speed up the high-resolution simulations. Various numerical test cases, including the multi-material shock tube problem, inertial confinement fusion implosion, triple-point shock interaction and shock interaction with multi-material bubbles, show that the method is suitable for a wide range of complex compressible multi-material flows.

physics.comp-ph↗

arXiv2017

A consistent analytical formulation for volume-estimation of geometries enclosed by implicitly defined surfaces

We have derived an analytical formulation for estimating the volume of geometries enclosed by implicitly defined surfaces. The novelty of this work is due to two aspects. First we provide a general analytical formulation for all two-dimensional cases, and for elementary three three-dimensional cases by which the volume of general three-dimensional cases can be computed. Second, our method addresses the inconsistency issue due to mesh refinement. It is demonstrated by several two-dimensional and three-dimensional cases that this analytical formulation exhibits 2nd-order accuracy.

math.NA↗

arXiv2017

High-resolution transport of regional level sets for evolving complex interface networks

In this paper we describe a high-resolution transport formulation of the regional level-set approach for an improved prediction of the evolution of complex interface networks. The novelty of this method is twofold: (i) construction of local level sets and reconstruction of a global regional level sets, (ii) locally transporting the interface network by employing high-order spatial discretization schemes for improved representation of complex topologies. Various numerical test cases of multi-region flow problems, including triple-point advection, single vortex flow, mean curvature flow, normal driven flow and dry foam dynamics, show that the method is accurate and suitable for a wide range of complex interface-network evolutions. Its overall computational cost is comparable to the Semi-Lagrangian regional level-set method while the prediction accuracy is significantly improved. The approach thus offers a \textbf{viable} alternative to previous interface-network level-set method.

math.NA↗