Scholarlink

Abstract <p>In this work, the Green–Naghdi type III (GN-III) and Three-Phase-Lag (TPL) thermoelastic theories are used to investigate Rayleigh-type surface wave propagation in a transversely isotropic piezo-thermoelastic half-space. Phase velocity, attenuation, and specific loss may be thoroughly evaluated thanks to the derivation of secular equations for electrically open/shorted and thermally insulated/isothermal boundary conditions. The findings indicate that attenuation and loss show a substantial dependency on boundary restrictions and the chosen thermoelastic model, but phase velocity increases with inclination angle and stabilises for high wave numbers. The TPL framework predicts somewhat greater velocities and damping because of thermal relaxation effects, while electrically shorted isothermal surfaces produce the lowest dissipation. By explicitly incorporating governing equations into its loss function, a Physics-Informed Neural Network (PINN) is utilised to overcome the computational burden of solving difficult transcendental equations. The PINN provides an effective stand-in for optimisation and real-time diagnostics in SAW sensors, ultrasonic devices, and smart piezoelectric materials by precisely reconstructing dispersion trends from sparse analytical data.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Rayleigh-Type Surface Waves in Piezo-Thermoelastic Materials: A Comparative Study Using Green–Naghdi III and Three-Phase-Lag Models with Machine-Learning Surrogates

  • 期刊论文

  • 发表日期：

    2025年12月1日

查看全文
  • Seema,
  • Abhinav Singhal

摘要

Abstract

In this work, the Green–Naghdi type III (GN-III) and Three-Phase-Lag (TPL) thermoelastic theories are used to investigate Rayleigh-type surface wave propagation in a transversely isotropic piezo-thermoelastic half-space. Phase velocity, attenuation, and specific loss may be thoroughly evaluated thanks to the derivation of secular equations for electrically open/shorted and thermally insulated/isothermal boundary conditions. The findings indicate that attenuation and loss show a substantial dependency on boundary restrictions and the chosen thermoelastic model, but phase velocity increases with inclination angle and stabilises for high wave numbers. The TPL framework predicts somewhat greater velocities and damping because of thermal relaxation effects, while electrically shorted isothermal surfaces produce the lowest dissipation. By explicitly incorporating governing equations into its loss function, a Physics-Informed Neural Network (PINN) is utilised to overcome the computational burden of solving difficult transcendental equations. The PINN provides an effective stand-in for optimisation and real-time diagnostics in SAW sensors, ultrasonic devices, and smart piezoelectric materials by precisely reconstructing dispersion trends from sparse analytical data.