Journal of Computational Physics最新文献

{"title":"A total-shear-stress-conserved wall model for large-eddy simulation of high-Reynolds number wall turbulence","authors":"Huan-Cong Liu ,&nbsp;Chun-Xiao Xu ,&nbsp;Wei-Xi Huang","doi":"10.1016/j.jcp.2025.114029","DOIUrl":"10.1016/j.jcp.2025.114029","url":null,"abstract":"<div><div>Wall-modeled large-eddy simulation (WMLES) is widely recognized as a useful method for simulation of turbulent flows at high Reynolds numbers. Nevertheless, a continual issue in different wall models is the shift of the mean velocity profile from the wall-model/RANS (Reynolds-averaged Navier-Stokes) region to the LES region. This phenomenon, referred to as logarithmic layer mismatch (LLM), occurs in both wall shear stress models and hybrid RANS/LES models. Many efforts have been made to explain and resolve this mismatch, including decreasing the high correlation between the wall shear stress and the velocity at the matching layer, modifying the subgrid-scale (SGS) eddy viscosity, and adding a stochastic forcing. It is widely believed that the inclusion of the resolved Reynolds shear stress (or the convection term) is essential to eliminate the LLM, as it prevents the overestimation of the modeled Reynolds shear stress and promotes the generation of the small-scale flow structures in the near-wall region. In this work, by comparing three different SGS eddy viscosity models, we demonstrate that ensuring the total-shear-stress-conserved (TSSC) constraint is key to resolving the LLM. Under the TSSC framework, the effect of the convection term on LLM can be quantitatively assessed. Furthermore, a modified SGS eddy viscosity modification model that adheres to the TSSC constraint is tested at different Reynolds numbers (<span><math><mi>R</mi><msub><mrow><mi>e</mi></mrow><mrow><mi>τ</mi></mrow></msub><mo>=</mo><mn>1000</mn><mo>,</mo><mn>2000</mn><mo>,</mo><mn>4200</mn></math></span>). Our results demonstrate the robust performance of the present model in predicting skin friction and low-order turbulence statistics, even under a relatively low grid resolution (<span><math><msubsup><mrow><mi>Δ</mi></mrow><mrow><mi>x</mi></mrow><mrow><mo>+</mo></mrow></msubsup><mo>,</mo><msubsup><mrow><mi>Δ</mi></mrow><mrow><mi>z</mi></mrow><mrow><mo>+</mo></mrow></msubsup><mo>≲</mo><mn>500</mn></math></span>, <span><math><mn>2</mn><mo>≤</mo><msub><mrow><mi>Δ</mi></mrow><mrow><mi>x</mi></mrow></msub><mo>/</mo><msub><mrow><mi>Δ</mi></mrow><mrow><mi>y</mi><mo>,</mo><mi>m</mi><mi>a</mi><mi>t</mi></mrow></msub><mo>≤</mo><mn>4</mn></math></span>, where <span><math><msub><mrow><mi>Δ</mi></mrow><mrow><mi>y</mi><mo>,</mo><mi>m</mi><mi>a</mi><mi>t</mi></mrow></msub></math></span> is the wall-normal grid spacing in the wall-model region).</div></div>","PeriodicalId":352,"journal":{"name":"Journal of Computational Physics","volume":"534 ","pages":"Article 114029"},"PeriodicalIF":3.8,"publicationDate":"2025-04-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143887905","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"物理与天体物理","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Learning the solution operator of two-dimensional incompressible Navier-Stokes equations using physics-aware convolutional neural networks","authors":"Viktor Grimm ,&nbsp;Alexander Heinlein ,&nbsp;Axel Klawonn","doi":"10.1016/j.jcp.2025.114027","DOIUrl":"10.1016/j.jcp.2025.114027","url":null,"abstract":"<div><div>In recent years, the concept of introducing physics to machine learning has become widely popular. Most physics-inclusive ML-techniques however are still limited to a single geometry or a set of parametrizable geometries. Thus, there remains the need to train a new model for a new geometry, even if it is only slightly modified. With this work we introduce a technique with which it is possible to learn approximate solutions to the steady-state Navier–Stokes equations in varying geometries without the need of parametrization. This technique is based on a combination of a U-Net-like CNN and well established discretization methods from the field of the finite difference method. The results of our physics-aware CNN are compared to a state-of-the-art data-based approach. Additionally, it is also shown how our approach performs when combined with the data-based approach.</div></div>","PeriodicalId":352,"journal":{"name":"Journal of Computational Physics","volume":"535 ","pages":"Article 114027"},"PeriodicalIF":3.8,"publicationDate":"2025-04-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143913174","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"物理与天体物理","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A FEM towards 3D multi-component elastic interface problems and phononic crystals with nested and intersected scatterer geometries","authors":"Jiaxin Yan ,&nbsp;Liqun Wang ,&nbsp;Yifan Zhang ,&nbsp;Meiling Zhao ,&nbsp;Liwei Shi","doi":"10.1016/j.jcp.2025.114017","DOIUrl":"10.1016/j.jcp.2025.114017","url":null,"abstract":"<div><div>Developing a high efficiency and high precision numerical method for the 3D three-component elasticity interface problems with Bloch-periodic boundary conditions is challenging because of the coupled vector components of elasticity equations, the complex spatial structures and interfacial jump conditions, as well as the periodic boundary conditions. In this paper, we propose a novel Petrov-Galerkin finite element interface method to solve this problem. We choose the standard finite element basis function to be the basis of the test function, which is independent of the interface conditions and satisfies the periodic boundary conditions. Piecewise linear functions independent of the boundary conditions are constructed as the basis of the solution, which satisfy the jump conditions. The proposed method utilizes the non-body-fitted grid and projected grid to simplify the calculation. To our best knowledge, this is the first time that 3D three-component elasticity interface problems with triple junction points are solved by using non-body-fitted grids. Numerical experiments show that the proposed method can achieve near second-order accuracy in the <span><math><msup><mrow><mi>L</mi></mrow><mrow><mo>∞</mo></mrow></msup></math></span> error norm and first-order accuracy in the <span><math><msup><mrow><mi>H</mi></mrow><mrow><mn>1</mn></mrow></msup></math></span> norm for interface problems with matrix coefficients and arbitrarily complex interfaces. With these properties, the method can be applied to the time-harmonic elastic wave equations for the band structure computation of 3D three-component phononic crystals with multiple scatterers of nested and intersected geometries. By the calculation and analysis of band structures, the influences of material properties and structural parameters are fully discussed.</div></div>","PeriodicalId":352,"journal":{"name":"Journal of Computational Physics","volume":"534 ","pages":"Article 114017"},"PeriodicalIF":3.8,"publicationDate":"2025-04-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143874250","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"物理与天体物理","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"An adaptive Dirichlet-to-Neumann finite element method for the thermoelastic scattering problem","authors":"Yu Wang ,&nbsp;Peijun Li ,&nbsp;Liwei Xu ,&nbsp;Tao Yin","doi":"10.1016/j.jcp.2025.114016","DOIUrl":"10.1016/j.jcp.2025.114016","url":null,"abstract":"<div><div>This paper presents the analysis and computation of an adaptive Dirichlet-to-Neumann (DtN) finite element method for solving the two-dimensional thermoelastic wave scattering problem. Using the Helmholtz decomposition, the vectorial coupled governing equations of thermoelastic waves are transformed into three Helmholtz equations for scalar potentials with distinct wavenumbers. The DtN map and the corresponding transparent boundary condition are derived through Fourier series expansions of the potentials. Well-posedness results are established for both the variational problem and its truncated formulation, which accounts for the truncation of the DtN map. Both a priori and a posteriori error estimates are established, accounting for the truncation of the DtN operator and the finite element discretization. Numerical experiments are conducted to validate the theoretical findings.</div></div>","PeriodicalId":352,"journal":{"name":"Journal of Computational Physics","volume":"534 ","pages":"Article 114016"},"PeriodicalIF":3.8,"publicationDate":"2025-04-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143863536","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"物理与天体物理","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Neumann series-based neural operator for solving 2D inverse medium problem","authors":"Ziyang Liu ,&nbsp;Fukai Chen ,&nbsp;Junqing Chen ,&nbsp;Lingyun Qiu ,&nbsp;Zuoqiang Shi","doi":"10.1016/j.jcp.2025.114025","DOIUrl":"10.1016/j.jcp.2025.114025","url":null,"abstract":"<div><div>The inverse medium problem, inherently ill-posed and nonlinear, poses significant computational challenges. We adopt a physics-assisted approach, utilizing a neural operator as a surrogate solver for the forward problem to accelerate reconstruction. Existing neural network methods fail to effectively solve the forward problem when simultaneously handling source and scatterer parameters as inputs. To overcome this, we propose integrating a Neumann series structure to efficiently manage such multi-input scenarios. Extensive experimental results demonstrate the framework's superior computational efficiency, robust generalization, and adaptability, offering valuable insights for solving similar inverse problems.</div></div>","PeriodicalId":352,"journal":{"name":"Journal of Computational Physics","volume":"534 ","pages":"Article 114025"},"PeriodicalIF":3.8,"publicationDate":"2025-04-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143868084","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"物理与天体物理","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A double-phase double-layer nonlocal general particle dynamics for modeling submarine landslides","authors":"Jin-Hu Pan ,&nbsp;Xiao-Ping Zhou","doi":"10.1016/j.jcp.2025.114028","DOIUrl":"10.1016/j.jcp.2025.114028","url":null,"abstract":"<div><div>The existing nonlocal methods rely on a single-layer theory, which limits the capability to capture the variation of free surface and to simulate dynamic coupling problems between fluid and soil in submarine landslides. To overcome the above limitations, a double-phase double-layer nonlocal general particle dynamics (NGPD) method is proposed in this paper. Within the developed framework, the entire problem domain is divided into two computational layers that are allowed to overlap, the fluid layer and the soil layer. Each phase satisfies its own laws of motion within its respective computational layer. The numerical stability of the novel NGPD method is enhanced by several key stabilization techniques, such as artificial terms and efficient particle shifting technique (PST). The boundary conditions of the fluid phase and soil skeleton in the proposed method are introduced in detail. Among them, a novel stress boundary for soils is proposed to avoid the particle penetration phenomenon. The Graphics Processing Unit (GPU) acceleration based on the Taichi kernel is leveraged in this work to achieve a parallel solution. To validate the performance of the proposed method in simulating submarine landslide problems, four benchmark examples, including a classical underwater soil column problem with analytical solution and three experimental submerged landslide examples, are studied by the proposed method. The numerical results demonstrate that the proposed method possesses the excellent ability to address water-soil interaction in hydromechanical geotechnical problems. Finally, the proposed method is further applied to model a practical submarine landslide and a submarine retrogressive landslide induced by earthquake.</div></div>","PeriodicalId":352,"journal":{"name":"Journal of Computational Physics","volume":"534 ","pages":"Article 114028"},"PeriodicalIF":3.8,"publicationDate":"2025-04-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143895835","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"物理与天体物理","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A stable and efficient semi-implicit coupling method for fluid-structure interaction problems with immersed boundaries in a hybrid CPU-GPU framework","authors":"Yuhang Zeng ,&nbsp;Yan Wang ,&nbsp;Haizhuan Yuan","doi":"10.1016/j.jcp.2025.114026","DOIUrl":"10.1016/j.jcp.2025.114026","url":null,"abstract":"<div><div>This paper presents a stable and efficient semi-implicit coupling immersed boundary method (IBM) for simulating fluid-structure interaction problems in a hybrid CPU-GPU framework. The method enhances numerical stability by constructing and applying implicit hydrodynamic force schemes and significantly improves computational efficiency by proposing GPU-based parallel strategies. To enhance stability performance, the hydrodynamic forces obtained by the decoupled velocity correction relationships in IBM are treated implicitly as unknowns and formulated as a function of unknown structural velocities and the predicted flow field. Both the hydrodynamic forces and the equations of structural dynamics (SD) are solved simultaneously. The entire solution procedures are realized in a hybrid CPU-GPU heterogeneous parallel framework. To guarantee thread safety and minimum data transfer between CPU and GPU, the unique correspondence between computational tasks and threads is established, optimizing the overall computational performance. The accuracy, stability, and efficiency of the present method are systematically and rigorously examined by numerical simulations of a variety of rigid and deformable FSI problems in both 2D and 3D cases, providing a comprehensive understanding of its performance. It is demonstrated that the present method not only enlarges the time step by more than 50 % compared with the conventional explicit coupling method but is also suitable for FSI problems with arbitrary solid-to-fluid density ratios. Furthermore, the computational efficiency is also enhanced by 35 to 380 times. The present CPU/GPU-based semi-implicit coupling method is particularly promising in simulating both challenging rigid and deformable FSI problems, demonstrating its wide applicability.</div></div>","PeriodicalId":352,"journal":{"name":"Journal of Computational Physics","volume":"534 ","pages":"Article 114026"},"PeriodicalIF":3.8,"publicationDate":"2025-04-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143878668","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"物理与天体物理","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"An imbalanced learning-based sampling method for physics-informed neural networks","authors":"Jiaqi Luo ,&nbsp;Yahong Yang ,&nbsp;Yuan Yuan ,&nbsp;Shixin Xu ,&nbsp;Wenrui Hao","doi":"10.1016/j.jcp.2025.114010","DOIUrl":"10.1016/j.jcp.2025.114010","url":null,"abstract":"<div><div>This paper introduces Residual-based Smote (RSmote), an innovative local adaptive sampling technique tailored to improve the performance of Physics-Informed Neural Networks (PINNs) through imbalanced learning strategies. Traditional residual-based adaptive sampling methods, while effective in enhancing PINN accuracy, often struggle with efficiency and high memory consumption, particularly in high-dimensional problems. RSmote addresses these challenges by targeting regions with high residuals and employing oversampling techniques from imbalanced learning to refine the sampling process. Our approach is underpinned by a rigorous theoretical analysis that supports the effectiveness of RSmote in managing computational resources more efficiently. Through extensive evaluations, we benchmark RSmote against the state-of-the-art Residual-based Adaptive Distribution (RAD) method across a variety of dimensions and differential equations. The results demonstrate that RSmote not only achieves or exceeds the accuracy of RAD but also significantly reduces memory usage, making it particularly advantageous in high-dimensional scenarios. These contributions position RSmote as a robust and resource-efficient solution for solving complex partial differential equations, especially when computational constraints are a critical consideration.</div></div>","PeriodicalId":352,"journal":{"name":"Journal of Computational Physics","volume":"534 ","pages":"Article 114010"},"PeriodicalIF":3.8,"publicationDate":"2025-04-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143843603","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"物理与天体物理","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A mixed-dimensional model for the electrostatic problem on coupled domains","authors":"Beatrice Crippa ,&nbsp;Anna Scotti ,&nbsp;Andrea Villa","doi":"10.1016/j.jcp.2025.114015","DOIUrl":"10.1016/j.jcp.2025.114015","url":null,"abstract":"<div><div>We derive a mixed-dimensional 3D-1D formulation of the electrostatic equation in two domains with different dielectric constants to compute, with an affordable computational cost, the electric field and potential in the relevant case of thin inclusions in a larger 3D domain. The numerical solution is obtained by Mixed Finite Elements for the 3D problem and Finite Elements on the 1D domain. We analyze some test cases with simple geometries to validate the proposed approach against analytical solutions, and perform comparisons with the fully resolved 3D problem. We treat the case where ramifications are present in the one-dimensional domain and show some results on the geometry of an electrical treeing, a ramified structure that propagates in insulators causing their failure.</div></div>","PeriodicalId":352,"journal":{"name":"Journal of Computational Physics","volume":"534 ","pages":"Article 114015"},"PeriodicalIF":3.8,"publicationDate":"2025-04-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143859943","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"物理与天体物理","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Minimum scale and spatial resolution requirement for direct numerical simulations of compressible turbulence","authors":"Chensheng Luo ,&nbsp;Jian Fang ,&nbsp;Le Fang","doi":"10.1016/j.jcp.2025.114014","DOIUrl":"10.1016/j.jcp.2025.114014","url":null,"abstract":"<div><div>In the direct numerical simulation (DNS) of compressible turbulence using Navier-Stokes equations, due to the incomplete resolution of shocklets, the classical grid resolution criterion based on the usual Kolmogorov length scale appears insufficient for high-order statistics. The present study discusses the minimum scale of compressible turbulence under the continuum assumption, and establishes new spatial resolution requirements for DNS. We first define the minimum shock scale for one-dimensional Burgers turbulence, and derive a spatial resolution criterion essential for fully resolving the second- and third-order velocity gradient moments. We demonstrate that this shock scale definition is also applicable to one-dimensional Navier-Stokes turbulence, and validate the spatial resolution requirement through numerical simulations of the Shu-Osher problem. The analysis is then extended to multi-dimensional turbulence. Through theoretical analysis and numerical studies, we conclude that the minimum local Kolmogorov scale, <span><math><msub><mrow><mi>η</mi></mrow><mrow><mi>min</mi></mrow></msub></math></span>, describes the smallest structure in turbulence and is determined by the strongest shocklet. Furthermore, we establish a requirement of <span><math><msub><mrow><mi>η</mi></mrow><mrow><mi>min</mi></mrow></msub><mo>/</mo><mi>Δ</mi><mi>x</mi><mo>≳</mo><mn>1.5</mn></math></span> for compressible turbulence, and validate it through DNSs of two-dimensional compressible turbulence with different grid resolutions. The present study provides a reference on spatial resolution requirement for DNS of compressible turbulence.</div></div>","PeriodicalId":352,"journal":{"name":"Journal of Computational Physics","volume":"534 ","pages":"Article 114014"},"PeriodicalIF":3.8,"publicationDate":"2025-04-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143854658","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"物理与天体物理","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}