903 papers
Computational physics bridges the gap between abstract theory and real-world observation by using powerful computers to solve complex physical problems. This field allows scientists to simulate everything from the collision of subatomic particles to the swirling dynamics of galaxies, offering insights that traditional experiments alone cannot provide.
On Gist.Science, we continuously process every new preprint in this category from arXiv to make these breakthroughs accessible to everyone. Each entry is accompanied by both a clear, plain-language explanation and a detailed technical summary, ensuring that researchers and curious readers alike can grasp the significance of the latest findings without getting lost in dense equations.
Below are the latest papers in computational physics, curated to keep you at the forefront of this rapidly evolving discipline.
By applying Lee-Yang zeros to study bilayer water under nanoconfinement, this paper reveals that 2D melting exhibits field-selective criticality where the solid-hexatic and hexatic-liquid transitions respond differently to temperature and lateral pressure, thereby resolving long-standing contradictions between simulations, models, and experiments by identifying which thermodynamic channel each probe observes.
The paper introduces PhyNex, an autonomous agent that leverages large language models and domain-specific tools to systematically explore and optimize scientific solutions in computational physics, successfully matching or exceeding human-designed state-of-the-art methods across diverse tasks such as dielectric spectrum prediction, Max-Cut heuristics, and quantum battery optimization.
This paper presents a new open-source compositional flow model implemented in the PorePy framework that simulates non-isothermal, multiphase flow and halite precipitation in high-enthalpy fractured geothermal reservoirs, utilizing a robust primary variable formulation and discrete fracture-matrix approach to accurately predict permeability damage and operational challenges.
This perspective paper outlines the necessity of integrating classical simulations, experimental measurements, machine learning, and quantum computing within reproducible, standardized workflows to overcome current limitations and accelerate the reliable discovery of advanced materials.
This study numerically demonstrates that the percolation of rod-like particles through a static bed of spheres is governed by a transition between trapping and passing regimes determined by rod length and pore geometry, where shorter rods move nearly twice as fast as longer ones due to reduced susceptibility to geometric trapping.
This paper proposes an equivariant flow matching framework with optimal-transport coupling to effectively model the multimodal probability distributions of symmetry-breaking bifurcations in nonlinear dynamical systems, demonstrating superior performance over deterministic and variational methods in capturing multistability across various physical problems.
This paper introduces BattMo, a flexible MATLAB-based finite volume toolbox for simulating diverse electrochemical battery cells that supports 3D geometries, coupled thermal and degradation models, and FAIR-compliant parameterization while enabling efficient gradient-based optimization through automatic differentiation.
This paper introduces a feature-preserving latent-EnKF that overcomes the Gaussian assumption limitations of traditional ensemble Kalman filters in compressible flows by performing ensemble updates in a learned low-dimensional latent space, thereby accurately recovering shocks and discontinuities without spurious oscillations.
This paper presents an asymptotic analysis and a high-order boundary integral method using Duffy patches to accurately compute the three-dimensional Neumann Green's function for general curved surfaces by decomposing the solution into singular and regular parts, thereby enabling the resolution of open problems in narrow capture theory.
This paper introduces REMAL, a cost-effective active learning framework that employs multitask Gaussian processes to model the joint residual manifold of coupled engineering systems, enabling efficient recovery of equilibrium states for multidisciplinary design analysis without expensive fixed-point iterations.