JOURNAL OF COMPUTATIONAL PHYSICS计算物理杂志
JOURNAL OF COMPUTATIONAL PHYSICS(英文缩写 J COMPUT PHYS),ISSN 0021-9991,eISSN 1090-2716,中文译名:计算物理杂志 是一本学术期刊。本页汇总该期刊的最新影响因子、分区信息以及最新收录于 PubMed 的文献,帮助您快速了解期刊全貌。
发文量统计区间：2025-11-01 至 2026-10-31，按本站收录文献的发表日期统计。
期刊介绍
历年影响因子趋势
| JCR 数据年份 | 影响因子 | JCR 分区 |
|---|---|---|
| 2021 | 4.645 | Q1 |
| 2022 | 4.100 | Q1 |
| 2023 | 3.800 | Q1 |
| 2024 | 3.800 | Q1 |
| 2025 | 3.900 | Q1 |
JOURNAL OF COMPUTATIONAL PHYSICS 最新收录文献
-
1. D-Flow SGLD: Source-Space Posterior Sampling for Scientific Inverse Problems with Flow Matching.
1. D-Flow SGLD：基于流匹配的源空间后验采样用于科学反问题PMID:日期:2026-09-04Data assimilation and scientific inverse problems require reconstructing high-dimensional physical states from sparse and noisy observations, ideally with uncertainty-aware posterior samples that remain faithful to learned priors and governing physics. While training-free conditional generation is well developed for diffusion models, corresponding conditioning and posterior sampling strategies for Flow Matching (FM) priors remain comparatively under-explored, especially on scientific benchmarks where fidelity must be assessed beyond measurement misfit. In this work, we study training-free conditional generation for scientific inverse problems under FM priors and organize existing inference-time strategies by where measurement information is injected: (i) guided transport dynamics that perturb sampling trajectories using likelihood information, and (ii) source-distribution inference that performs posterior inference over the source variable while keeping the learned transport fixed. Building on the latter, we propose D-Flow SGLD, a source-space posterior sampling method that augments differentiable source inference with preconditioned stochastic gradient Langevin dynamics, enabling scalable exploration of the source posterior induced by new measurement operators without retraining the prior or modifying the learned FM dynamics. We benchmark representative methods from both families on a hierarchy of problems: 2D toy posteriors, chaotic Kuramoto-Sivashinsky trajectories, and wall-bounded turbulence reconstruction. Across these settings, we quantify trade-offs among measurement assimilation, posterior diversity, and physics/statistics fidelity, and establish D-Flow SGLD as a practical FM-compatible posterior sampler for scientific inverse problems.
-
2. An Immersed Interface Method for Incompressible Flows and Near Contact.
PMID:日期:2026-08-29We present an enhanced immersed interface method for simulating incompressible fluid flows in thin gaps between closely spaced immersed boundaries. This regime, common in engineered structures such as tribological interfaces and bearing assemblies, poses significant computational challenges because of limitations in grid resolution and the prohibitive cost of mesh refinement near contact. The immersed interface method imposes jump conditions that capture stress discontinuities generated by forces that are concentrated along immersed boundaries. Our approach introduces a bilinear velocity interpolation operator in two spatial dimensions that incorporates jump conditions from multiple nearby interfaces if they occupy the same interpolation stencil. Furthermore, we show that the methodology developed for resolving distinct interfaces in near contact also improves accuracy for single-interface geometries with intrinsically sharp features, in which multiple interface segments intersect a single grid cell. Numerical results demonstrate substantial improvements in both interface and Eulerian velocity accuracy compared with a lubrication-corrected immersed boundary method, even for interfaces whose separation is one-fiftieth of a computational grid cell. The results also show that this formulation effectively resolves flows near sharp geometric features. The proposed method improves upon previous interpolation schemes and eliminates the need for prior knowledge of interface orientation or geometry. This makes it broadly applicable to a wide range of fluid-structure interaction problems involving near-contact dynamics.
-
3. An effective and accurate semi-implicit time integration scheme for dynamics in nearly- and fully-incompressible hyperelastic solids.
PMID:日期:2026-06-22The choice of numerical integrator in approximating solutions to dynamic partial differential equations depends on the smallest time scale of the problem at hand. Large-scale deformations in elastic solids contain both shear waves and bulk waves, the latter of which can travel infinitely fast in incompressible materials. Explicit schemes, which are favored for their efficiency in resolving low-speed dynamics, are bound by time step size restrictions that inversely scale with the fastest wave speed. Implicit schemes can enable larger time step sizes regardless of the wave speeds present, though they are much more computationally expensive. Semi-implicit methods, which are more stable than explicit methods and more efficient than implicit methods, are emerging in the literature, though their applicability to nonlinear elasticity is not extensively studied. In this research, we develop and investigate the functionality of two time integration schemes for the resolution of large-scale dynamics in nearly- and fully-incompressible hyperelastic solids: a Modified Semi-implicit Backward Differentiation Formula integrator (MSBDF2) and a forward Euler / Semi-implicit Backward Differentiation Formula Runge-Kutta integrator (FEBDF2). We prove and empirically verify second-order accuracy for both schemes. The stability properties of both methods are derived and numerically verified. We find that MSBDF2 is not conditionally stable, while FEBDF2 has a maximum time step size that inversely scales with the shear wave speed and is unaffected by the bulk wave speed-the desired stability property of a semi-implicit scheme.
-
4. HeartSimSage: Attention-Enhanced Graph Neural Networks for Accelerating Cardiac Mechanics Modeling.
PMID:日期:2026-04-17Finite element analysis (FEA) is a powerful tool that forms the cornerstone of modeling cardiac biomechanics. However, FEA is computationally expensive for creating digital twins, which typically involves performing tens or hundreds of FEA simulations to estimate tissue parameters, limiting its clinical application. We have developed an attention-enhanced graph neural network (GNN)-based FEA emulator, , to rapidly predict passive biventricular myocardial displacements from patient-specific geometries, chamber pressures, and material properties. Designed to overcome the limitations of current emulators, can effectively handle diverse three-dimensional (3D) biventricular geometries, mesh topologies, fiber directions, structurally based constitutive models, and physiological boundary conditions. It also supports flexible mesh structures, allowing variable node count, ordering, and element connectivity. To optimize information propagation, we developed a neighboring connection strategy inspired by Graph Sample and Aggregate (GraphSAGE) that prioritizes local node interactions while maintaining mid-to-long-range dependencies. Additionally, we integrated Laplace-Dirichlet solutions to enhance spatial encoding and employed subset-based training to improve computational efficiency. Incorporating the attention mechanism allows to adaptively weigh neighbor contributions and filter out irrelevant information flow, enhancing prediction accuracy. As a result, errors in the predicted biventricular myocardial displacements by were limited to a median of 0.280 mm with an interquartile range of [0.167, 0.484] mm compared to traditional FEA, while achieving a computational speedup of approximately 13000X on a GPU and 190X on a CPU. We validated our model on a published left ventricle dataset and analyzed the model's sensitivity to hyperparameters, neighboring connection strategies, and the attention mechanism.
-
5. Improving the robustness of the immersed interface method through regularized velocity reconstruction.
PMID:日期:2025-12-03Robust, broadly applicable fluid-structure interaction (FSI) algorithms remain a challenge for computational mechanics. Efforts in this area are driven by the need to enhance predictive accuracy and efficiency in FSI simulations, align with experimental observations, and unravel complex multiscale and multiphysics phenomena, while addressing challenges in developing more robust and efficient methodologies. In previous work, we introduced an immersed interface method (IIM) for discrete surfaces and an extension based on an immersed Lagrangia-Eulerian (ILE) coupling strategy for modeling FSI involving complex geometries. The ability of the method to sharply resolve stress discontinuities induced by singular immersed boundary forces in the presence of low-regularity geometrical representations enables it to model complex three-dimensional geometries in diverse engineering applications. Although the IIM we previously introduced offers many advantages compared to other FSI algorithms, it also imposes a restrictive mesh factor ratio, requiring the surface mesh to be coarser than the background fluid grid to ensure stability. This is because if the mesh factor ratio constraint is not satisfied, parts of the structure motion are not controlled by the discrete FSI coupling operators. This constraint can significantly increase computational costs, particularly in applications involving multiscale geometries with highly localized complexity or fine-scale features. To address this limitation, we devise a stabilization strategy for the velocity interpolation operator inspired by Tikhonov regularization. The effectiveness of the stabilization scheme is evaluated using benchmark problems with stationary interfaces and FSI models involving both rigid-body dynamics and elastodynamic structures models. This study demonstrates that using a stabilized velocity interpolation operator in the IIM enables a broader range of structure-to-fluid grid-size ratios without compromising accuracy or altering the flow dynamics. Our approach significantly broadens the applicability of the method to real-world FSI problems involving complex geometries and dynamic conditions, offering a robust and practical solution.
-
6. Composite B-spline regularized delta functions for the immersed boundary method: Divergence-free interpolation and gradient-preserving force spreading.
PMID:日期:2025-12-02This paper presents an approach to enhance volume conservation in the immersed boundary (IB) method by employing regularized delta functions derived from composite B-splines. These delta functions are constructed using tensor product kernels, similar to the conventional IB method. However, the kernels are B-splines whose polynomial degree varies according to the normal and tangential directions of each velocity component. The conventional IB method, while effective for fluid-structure interaction applications, has long been challenged by poor volume conservation which is particularly evident in simulations of pressurized, closed membranes. We demonstrate that composite B-spline regularized delta functions significantly enhance volume conservation through two complementary properties: they provide continuously divergence-free velocity interpolants and maintain the gradient character of forces corresponding to mean pressure jumps across interfaces. By correctly representing these forces as discrete gradients, they eliminate a key source of spurious flows that typically plague immersed boundary computations. Our approach maintains the local nature of the classical IB method, avoiding the computational overhead associated with the non-local Divergence-Free Immersed Boundary (DFIB) method's construction of an explicit velocity potential, which requires additional Poisson solves for interpolation and force spreading operations. Through a series of numerical experiments, we show that sufficiently regular composite B-spline kernels can maintain initial volumes to within machine precision. We provide a detailed analysis of the relationship between kernel regularity and the accuracy of force spreading and velocity interpolation operations. Our findings indicate that composite B-splines of at least regularity produce results comparable to the DFIB method in dynamic simulations, with errors in volume conservation primarily dominated by truncation error of the employed time-stepping scheme. This work offers a computationally efficient alternative for improving volume conservation in IB methods which is particularly beneficial for large-scale, three-dimensional simulations. The proposed approach requires only changing the functional form of the regularized delta function in an existing IB code, making it an accessible improvement for a wide range of applications in computational fluid dynamics and fluid-structure interaction.
-
7. On generalizing the induced surface charge method to heterogeneous Poisson-Boltzmann models for electrostatic free energy calculation.
PMID:日期:2025-10-23The induced surface charges (ISC) method, which computes the induced charges on the molecular surface of macromolecules and uses them via Coulomb's law to calculate the polar solvation energy, was shown to be a robust and almost grid independent approach for electrostatic analysis based on the sharp-interface Poisson-Boltzmann (PB) model. Besides being physically intuitive, the ISC method avoids using the potential near the point charges, which is singular at each atom center. However, the ISC method cannot be physically generalized to heterogeneous dielectric PB models, due to the non-existence of a dielectric boundary. In this work, a novel far-field (FF) method is proposed to calculate the polar solvation free energy, which is derived through reformulating the energy functionals of nonlinear PB potential in solvent and vacuum states. Built upon a rigorous mathematical analysis, the FF method reconstructs the free energies by using far-field solutions outside the solute so that the self-energy terms generated by the singular charges are avoided, just as in the ISC method. Being valid for both sharp-interface and heterogeneous PB models, the performance of the proposed FF method has been validated by considering diffuse interface, Gaussian and super-Gaussian PB models for Kirkwood spheres and various protein systems. Comparison with grid-energy cancellation and regularization methods is also considered. The robustness of the FF method in treating a non-rigid biomolecule with different molecular structures in solvent and vacuum states has been explored, taking advantage of the fact that the far-field potential is insensitive to perturbations of singular charge locations.
-
8. Laplacian Eigenfunction-Based Neural Operator for Learning Nonlinear Reaction-Diffusion Dynamics.
PMID:日期:2025-10-07Learning reaction-diffusion equations has become increasingly important across scientific and engineering disciplines, including fluid dynamics, materials science, and biological systems. In this work, we propose the (LE-NO), a novel framework designed to efficiently learn nonlinear reaction terms in reaction-diffusion equations. LE-NO models the nonlinear operator on the right-hand side using a data-driven approach, with Laplacian eigenfunctions serving as the basis. This spectral representation enables efficient approximation of the nonlinear terms, reduces computational complexity through direct inversion of the Laplacian matrix, and alleviates challenges associated with limited data and large neural network architectures-issues commonly encountered in operator learning. We demonstrate that LE-NO generalizes well across varying boundary conditions and provides interpretable representations of learned dynamics. Numerical experiments in mathematical physics showcase the effectiveness of LE-NO in capturing complex nonlinear behavior, offering a powerful and robust tool for the discovery and prediction of reaction-diffusion dynamics.
-
9. An efficient adaptive algorithm for photon-electron coupled Boltzmann equation in radiation therapy.
9. 一种用于放射治疗中光子-电子耦合玻尔兹曼方程的高效自适应算法PMID:日期:2025-06-06In this paper, we propose an efficient adaptive algorithm for the photon-electron coupled Boltzmann equations in radiation therapy. The algorithm employs adaptive mesh technologies for both angular and spatial discretization and effectively solves the equation across various domains with a reduced number of degrees of freedom and computational time while maintaining accuracy. For spatial discretization, we employ an adaptive discontinuous Galerkin scheme based on material redefinition to automatically conduct adaptive refinement during the computation process. For angular discretization, we utilize high-order discretization near the incident direction, while low-order discretization is applied to other directions. We compare its performance with the classical Monte Carlo simulations for both mono-energetic and multi-energetic photon beams in various media. Numerical results demonstrate the efficiency and accuracy of the algorithm.
-
10. Stability and Robustness of Time-discretization Schemes for the Allen-Cahn Equation via Bifurcation and Perturbation Analysis.
10. 通过分岔和扰动分析探讨Allen-Cahn方程时间离散格式的稳定性和鲁棒性PMID:日期:2024-11-08The Allen-Cahn equation is a fundamental model for phase transitions, offering critical insights into the dynamics of interface evolution in various physical systems. This paper investigates the stability and robustness of frequently utilized time-discretization numerical schemes for solving the Allen-Cahn equation, with focuses on the Backward Euler, Crank-Nicolson (CN), convex splitting of modified CN, and Diagonally Implicit Runge-Kutta (DIRK) methods. Our stability analysis reveals that the Convex Splitting of the Modified CN scheme exhibits unconditional stability, allowing greater flexibility in time step size selection, while the other schemes are conditionally stable. Additionally, our robustness analysis highlights that the Backward Euler method converges to correct physical solutions regardless of initial conditions. In contrast, all other methods studied in this work show sensitivity to initial conditions and may converge to incorrect physical solutions if the initial conditions are not carefully chosen. This study introduces a comprehensive approach to assessing stability and robustness in numerical methods for solving the Allen-Cahn equation, providing a new perspective for evaluating numerical techniques for general nonlinear differential equations.