Journal of Computational Physics最新文献

{"title":"High order-accurate solution of scattering integral equations with unbounded solutions at corners","authors":"Constantine Sideris ,&nbsp;Davit Aslanyan ,&nbsp;Oscar P. Bruno","doi":"10.1016/j.jcp.2025.114213","DOIUrl":"10.1016/j.jcp.2025.114213","url":null,"abstract":"<div><div>Although high-order Maxwell integral equation solvers provide significant advantages in terms of speed and accuracy over corresponding low-order integral methods, their performance significantly degrades in presence of non-smooth geometries—owing to field enhancement and singularities that arise at sharp edges and corners which, if left untreated, give rise to significant accuracy losses. The problem is particularly challenging in cases where the density function—that is, the solution to the integral equation—exhibits unbounded singular behavior at edges and corners. While such difficulties can be circumvented in two-dimensional configurations, they constitute an intrinsic feature of existing three-dimensional Maxwell integral formulations, in which the tangential component of the surface current density diverges along edges. In order to tackle the problem this paper restricts attention to the simplest context in which the unbounded-density difficulty arises, namely, integral formulations in 2D space whose integral density blows up at corners; the strategies proposed, however, can be generalized to the 3D context. The novel methodologies presented in this paper yield high-order convergence for such challenging equations and achieve highly accurate solutions (even near edges and corners) without requiring a-priori analysis of the geometry or use of singular bases.</div></div>","PeriodicalId":352,"journal":{"name":"Journal of Computational Physics","volume":"539 ","pages":"Article 114213"},"PeriodicalIF":3.8,"publicationDate":"2025-07-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144605922","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":"Multiscale approximation and two-grid preconditioner for extremely anisotropic heat flow","authors":"Maria Vasilyeva ,&nbsp;Golo Wimmer ,&nbsp;Ben S. Southworth","doi":"10.1016/j.jcp.2025.114201","DOIUrl":"10.1016/j.jcp.2025.114201","url":null,"abstract":"<div><div>We consider anis heat flow with extreme anisotropy, as arises in magnetized plasmas for fusion applications. Such problems pose significant challenges in both obtaining an accurate approximation as well in the construction of an efficient solver. In both cases, the underlying difficulty is in forming an accurate approximation of temperature fields that follow the direction of complex, non-grid-aligned magnetic fields. In this work, we construct a highly accurate coarse grid approximation using spectral multiscale basis functions based on local anisotropic normalized Laplacians. We show that the local generalized spectral problems yield local modes that align with magnetic fields, and provide an excellent coarse-grid approximation of the problem. We then utilize this spectral coarse space as an approximation in itself, and as the coarse-grid in a two-level spectral preconditioner. Numerical results are presented for several magnetic field distributions and anisotropy ratios up to <span><math><msup><mn>10</mn><mn>12</mn></msup></math></span>, showing highly accurate results with a large system size reduction, and two-grid preconditioning that converges in <span><math><mrow><mi>O</mi><mo>(</mo><mn>1</mn><mo>)</mo></mrow></math></span> iterations, independent of anisotropy. Aptara</div></div>","PeriodicalId":352,"journal":{"name":"Journal of Computational Physics","volume":"538 ","pages":"Article 114201"},"PeriodicalIF":3.8,"publicationDate":"2025-07-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144535934","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 decoupled, linear, unconditionally stable, and fully discrete finite element scheme for two-phase ferrofluid flows with different densities and viscosities","authors":"Xiaoyong Chen ,&nbsp;Rui Li ,&nbsp;Jian Li ,&nbsp;Xiaoming He","doi":"10.1016/j.jcp.2025.114209","DOIUrl":"10.1016/j.jcp.2025.114209","url":null,"abstract":"<div><div>In this paper, a model describing the behavior of two-phase ferrofluid flows with different densities and viscosities is established by using phase field techniques. This model is a coupled nonlinear multiphysics PDE system consisting of Cahn-Hilliard equations, Navier-Stokes equations, magnetization equation and magnetostatic equation. By reformulating the magnetic potential equation, applying the artificial compressibility method, utilizing the implicit-explicit scheme for treating the nonlinear terms, and adding several stabilization terms, we propose a linear, decoupled and fully discrete finite element method approximation for the established model. And it is strictly proved to be unconditionally stable and uniquely solvable at each time step. Furthermore, the proposed scheme does not impose any artificial boundary conditions on the pressure. In order to accurately capture the diffuse interface in the numerical simulation, we also apply the adaptive mesh strategy to locally refine the mesh around the interfacial region. Several informative numerical experiments, including an accuracy test, deformation of a ferrofluid droplet, one or two air bubbles rising in ferrofluids, a controllable ferrofluid droplet in a Y-shape domain, and the Rosensweig instability under uniformly or nonuniformly applied magnetic field, are performed to illustrate various features of the proposed model and scheme, especially their applicability for the cases of high density ratio and high viscosity ratio.</div></div>","PeriodicalId":352,"journal":{"name":"Journal of Computational Physics","volume":"539 ","pages":"Article 114209"},"PeriodicalIF":3.8,"publicationDate":"2025-07-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144572060","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":"TDDM: A transfer learning framework for physics-guided 3D acoustic scattering inversion","authors":"Yunwen Yin ,&nbsp;Liang Yan","doi":"10.1016/j.jcp.2025.114211","DOIUrl":"10.1016/j.jcp.2025.114211","url":null,"abstract":"<div><div>Recovering obstacle shapes in three-dimensional acoustic inverse scattering is a challenging task due to its inherent ill-posedness and the limited availability of observational data. Traditional deep learning approaches often require large labeled datasets, lack physical interpretability, and struggle to generalize across diverse scenarios. In recent work Yin and Yan (2025), a physics-aware deep learning framework was proposed to address these limitations by incorporating physical constraints into the learning process, enhancing interpretability and robustness. Here we describe two extensions of that work. First, we tackle the complexity of 3D scattering inversion by introducing a physics-informed loss function that incorporates both the far-field operator and the Herglotz operator. This effectively embeds physical constraints into the network, mitigating the ill-posedness of the problem without relying on labeled data. Second, we employ a transfer learning strategy that utilizes a pre-trained network, constructed offline with minimal source-domain data. During the online phase, a loss-based discriminator adaptively determines whether fine-tuning is necessary. When fine-tuning is required, only a small set of network parameters is updated, ensuring computational efficiency. The resulting Transfer learning-based Deep Decomposition Method (TDDM) is a fully unsupervised, physics-aware framework that enhances both efficiency and generalization capabilities. Numerical experiments demonstrate that TDDM achieves accurate and rapid inversions for source tasks while effectively generalizing to diverse target scenarios.</div></div>","PeriodicalId":352,"journal":{"name":"Journal of Computational Physics","volume":"539 ","pages":"Article 114211"},"PeriodicalIF":3.8,"publicationDate":"2025-07-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144570010","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 kinetic energy and entropy preserving scheme for compressible magnetohydrodynamics","authors":"Takashi Shiroto,&nbsp;Tomo-Hiko Watanabe","doi":"10.1016/j.jcp.2025.114212","DOIUrl":"10.1016/j.jcp.2025.114212","url":null,"abstract":"<div><div>This paper extends the kinetic energy and entropy preserving (KEEP) scheme, which was originally designed for compressible fluid dynamics, to include compressible magnetohydrodynamics (MHD). In MHD, the magnetic energy included in the total energy equation is consistent with that derived from the induction equation. Discretizing the spatial derivative in the induction equation using the central finite difference preserves the solenoidal constraint of the magnetic field. Numerical fluxes for the momentum and the total energy related to the magnetic field can be derived for the momentum and total energy from the Lorentz force and the solenoidal constraint, even in discrete form. Numerical results confirm that the KEEP scheme enables non-dissipative and stable MHD simulations, unlike the conventional central difference method.</div></div>","PeriodicalId":352,"journal":{"name":"Journal of Computational Physics","volume":"539 ","pages":"Article 114212"},"PeriodicalIF":3.8,"publicationDate":"2025-07-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144549553","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":"Weighted essentially non-oscillatory generalized finite difference method for hyperbolic conservation laws","authors":"Zhaoji Xia,&nbsp;Yinhua Xia","doi":"10.1016/j.jcp.2025.114207","DOIUrl":"10.1016/j.jcp.2025.114207","url":null,"abstract":"<div><div>This paper introduces a weighted essentially non-oscillatory (WENO) generalized finite difference method (GFDM) for solving hyperbolic conservation laws. The proposed method accommodates both unstructured meshes and scattered point clouds, ensuring the preservation of free-stream solutions. Drawing from an alternative WENO scheme formulation, the flux functions are disassembled into low-order and high-order constituents. Diverging from conventional finite difference methods, the WENO reconstruction is executed directly on the numerical solutions, rather than the flux functions, with Riemann solvers employed exclusively for the low-order components, thereby diminishing computational overhead. This paper outlines the process for stencil selection, WENO reconstruction, and scheme formulation. Numerical examples for one- and two-dimensional conservation laws validate the high-order accuracy and non-oscillatory properties of the method.</div></div>","PeriodicalId":352,"journal":{"name":"Journal of Computational Physics","volume":"539 ","pages":"Article 114207"},"PeriodicalIF":3.8,"publicationDate":"2025-07-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144557465","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":"Neural triangular map for density estimation and sampling with application to Bayesian inference","authors":"Dawen Wu ,&nbsp;Ludovic Chamoin ,&nbsp;Stéphane Bressan","doi":"10.1016/j.jcp.2025.114208","DOIUrl":"10.1016/j.jcp.2025.114208","url":null,"abstract":"<div><div>In this paper, we address two fundamental problems in computational probability: density estimation and sampling. Specifically, we consider the problem setup where an unnormalized target distribution is known, and the goal is to find a deterministic coupling that transports a specified reference distribution to the target distribution. To achieve this, we introduce the Neural Triangular Map (NTM), which utilizes a triangular function structure with neural networks as the underlying basis functions. The NTM is trained by minimizing the Kullback–Leibler divergence between the density induced by the map and the target density, subject to a constraint ensuring the map’s invertibility; both the objective and constraint functions are defined using the unnormalized target probability density. The NTM combines the computational convenience of the triangular function structure for computing the Jacobian determinant with the powerful approximation capabilities of neural networks. In addition, we integrate the proposed method into the Bayesian inference framework to infer the posterior distribution of model parameters based on noisy observations. Our experiments demonstrate that the NTM outperforms the polynomial-based triangular map and the Markov Chain Monte Carlo method in estimating certain density functions.</div></div>","PeriodicalId":352,"journal":{"name":"Journal of Computational Physics","volume":"539 ","pages":"Article 114208"},"PeriodicalIF":3.8,"publicationDate":"2025-07-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144557463","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":"Time-domain illposedness of effective frequency-domain boundary conditions for quiescent viscothermal acoustics","authors":"Linus Hägg,&nbsp;Martin Berggren","doi":"10.1016/j.jcp.2025.114205","DOIUrl":"10.1016/j.jcp.2025.114205","url":null,"abstract":"<div><div>Accurate simulations of sound propagation in narrow geometries need to account for viscous and thermal losses. In this respect, effective boundary conditions that model viscothermal losses in frequency-domain acoustics have recently gained in popularity. Here, we investigate the time-domain analogue of one such boundary condition. We find that the thermal part of the boundary condition is passive in time domain as expected, while the viscous part is not. More precisely, we demonstrate that the viscous part is responsible for exponentially growing normal modes with unbounded temporal growth rates, which indicates ill-posedness of the considered model. A finite-difference-time-domain scheme is developed for simulations of lossy sound propagation in a duct. If viscous losses are neglected the obtained transmission characteristics are found to be in excellent agreement with frequency-domain simulations. In the general case, the simulations experience an instability much in line with the theoretical findings.</div></div>","PeriodicalId":352,"journal":{"name":"Journal of Computational Physics","volume":"539 ","pages":"Article 114205"},"PeriodicalIF":3.8,"publicationDate":"2025-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144549552","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 low-dissipation and accurate Riemann SPH for multiphase flows with large density ratios and strong shocks","authors":"Xiang-Li Fang ,&nbsp;Ping-Ping Wang ,&nbsp;Zi-Fei Meng ,&nbsp;Fu-Ren Ming ,&nbsp;A-Man Zhang","doi":"10.1016/j.jcp.2025.114210","DOIUrl":"10.1016/j.jcp.2025.114210","url":null,"abstract":"<div><div>In this paper, a low-dissipation and accurate Riemann SPH is developed and applied to multiphase flows with large density ratios and strong shocks. To improve the accuracy of the pressure gradient approximation, the more accurate pressure differencing formulation (PDF) is adopted in the Riemann SPH framework, which allows obtaining smoother pressure and velocity fields. Furthermore, to mitigate the excessive numerical dissipation inherent in the Riemann SPH, a numerical limiter is embedded within the Riemann solver. Additionally, a renormalization operator associated with particle distributions is employed to achieve higher-order accuracy in kernel gradient computation, particularly when particles are disorderedly distributed. The accuracy and robustness of the improved SPH method are first validated by the benchmark test of the Sedov point-explosion with large discontinuities, demonstrating the model’s advantage in capturing smooth velocity fields for light fluids. Subsequently, simulations of typical multiphase flow scenarios are conducted, and the SPH results are compared with reference solutions from other methods to showcase the accuracy and superiority of the proposed SPH approach. Finally, the applicability of the present model to three-dimensional cases is further verified through the simulation of 3D bubble oscillation, highlighting the effectiveness and necessity of both the limiter and renormalization operator in enhancing numerical accuracy and stability.</div></div>","PeriodicalId":352,"journal":{"name":"Journal of Computational Physics","volume":"539 ","pages":"Article 114210"},"PeriodicalIF":3.8,"publicationDate":"2025-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144694297","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 highly accurate and robust finite volume-based phase field method for interfacial mass transfer in two-phase flows","authors":"Yujian Wan (万宇健) ,&nbsp;Liangqi Zhang (张良奇) ,&nbsp;Liming Chen (陈黎明) ,&nbsp;Xiaoshuang Wang (王小双) ,&nbsp;Yao Xiao (肖姚) ,&nbsp;Zhong Zeng (曾忠)","doi":"10.1016/j.jcp.2025.114200","DOIUrl":"10.1016/j.jcp.2025.114200","url":null,"abstract":"<div><div>In this work, we propose a phase-field model for interfacial mass transport in two-phase flows and implement it within the finite volume framework. We specifically reformulate the model proposed by Kou et al. by introducing a consistent mass flux and eliminating unphysical coupling terms. This reformulation ensures momentum conservation across the interfacial region and prevents the unphysical influence of solute transport on interface evolution. The interfacial mass transfer is represented by a compact one-scalar model, grounded in thermodynamically consistent phase field theory. Specifically, two Cahn-Hilliard equations are derived to describe the interfacial dynamics and solute transport across the entire domain. The finite volume method is then employed to solve these governing equations in their conservative form, with the convection terms consistently handled by the 5th order WENO-Z scheme. Furthermore, we meticulously design the algorithm to enforce the volume conservation condition, adapt it to a collocated grid, and optimize the solution of the Cahn-Hilliard equation. The theoretical model and algorithm design confer several advantages to our method: (i) it simultaneously satisfies Henry's law and Fick's law in different regions, while avoiding the complexity of a two-scalar formulation; (ii) it effectively prevents unphysical mass transfer across the interface, known as numerical leakage; (iii) it accurately predicts both interface evolution and solute distribution, even in challenging scenarios with high material property contrast and large Henry coefficient. Numerical examples are provided to demonstrate the method's accuracy in capturing interfacial dynamics and solute distribution, its effectiveness in minimizing mass leakage, and its robustness in challenging cases, as compared to reference solutions. The final two tests, involving bubble rise with or without interfacial mass transfer, illustrate the potential applicability of our method in industrial practice.</div></div>","PeriodicalId":352,"journal":{"name":"Journal of Computational Physics","volume":"539 ","pages":"Article 114200"},"PeriodicalIF":3.8,"publicationDate":"2025-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144572061","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}