不完美的
热的
机械
功能(生物学)
有限差分法
热方程
物理
接口(物质)
基础(线性代数)
有限差分
传热
数学分析
径向基函数
有限差分格式
有限元法
热力学
材料科学
数值分析
工作(物理)
应用数学
偏微分方程
基函数
接口模型
计算机模拟
标识
DOI:10.1016/j.icheatmasstransfer.2026.112101
摘要
The Guyer–Krumhansl heat conduction model has emerged as a thermodynamically consistent linearized continuum framework for describing rapid thermal transport in heterogeneous and layered materials, porous media, and low-temperature regimes where classical Fourier law fails to capture experimentally observed phenomena. Despite its clear physical relevance, the reliable numerical solution of the associated initial–boundary value problems remains challenging. Standard mesh-refinement-based ( h -version) finite element methods may suffer from poor convergence behavior, spurious oscillations, or loss of accuracy, particularly in parameter regimes corresponding to over-diffusive behavior or strong interfacial effects. These limitations make it necessary to construct more robust and higher-order numerical strategies equipped with much more flexible refinement schemes. In this work, a two-field radial basis function finite difference method based on odd order polyharmonic spline augmented with monomials is developed for the space discretization of the Guyer–Krumhansl model in layered media with Kapitza-type imperfect thermal interface conditions, using the implicit scheme for the time discretization. Both uniform and Halton -type quasi-random space point distributions are employed to investigate the h -convergence behavior of temperature response functions at discrete space points. Particular attention is devoted to the influence of the monomial degree p , the non-classical material parameters such as κ 2 and τ , and the thermal resistance R on convergence order and accuracy. The results demonstrate uniformly stable algebraic convergences consistent with theoretical predictions. Increasing the monomial degree p improves the convergence rate, while the thermal resistance does not affect the order of convergence.
科研通智能强力驱动
Strongly Powered by AbleSci AI