Scholarlink

<p>This article proposes a modification to the Transient Plane Source (TPS) method for simultaneous estimation of thermophysical properties in low thermal conductivity and moderate-to-high volumetric heat capacity materials like powders, by solving the inverse parameter estimation problem employing Levenberg–Marquardt (LM) algorithm using an integrated COMSOL and MATLAB LiveLink approach. The symmetric placement of thermocouples on either side of the Kapton-insulated TPS sensor further enhances sensitivity and reduces parameter correlation. Sensitivity and correlation analysis was conducted to identify a suitable location for the additional thermocouples, aimed at improving the accuracy of the estimation. The forward model was developed in COMSOL and was connected to the inverse problem solving LM algorithm in MATLAB using LiveLink server. Using the combined dataset consisting of experimental TPS data (obtained from Zhang et al. in Prog Comput Fluid Dyn 13:191–201, 2013) and numerically generated thermocouple data, the proposed inverse framework yielded mean deviations in thermal conductivity and specific heat capacity of less than 1&#xa0;% for both materials. The deviation in the estimated thermal conductivity reported by Zhang et al. (Prog Comput Fluid Dyn 13:191–201, 2013) was 1.92&#xa0;% for stainless steel and 2.86&#xa0;% for ceramic.</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

An integrated COMSOL-MATLAB Livelink Framework for Inverse Analysis of the Transient Plane Source Method Applied to Low-Conductivity Materials with Moderate-To-High Volumetric Heat Capacity

  • 期刊论文

  • 发表日期：

    2026年3月24日

查看全文
  • Haris Rhaj Shanmugavel,
  • Somasundharam Sankaran,
  • Rodrigo G. Dourado da Silva,
  • Elisan S. Magalhães

摘要

This article proposes a modification to the Transient Plane Source (TPS) method for simultaneous estimation of thermophysical properties in low thermal conductivity and moderate-to-high volumetric heat capacity materials like powders, by solving the inverse parameter estimation problem employing Levenberg–Marquardt (LM) algorithm using an integrated COMSOL and MATLAB LiveLink approach. The symmetric placement of thermocouples on either side of the Kapton-insulated TPS sensor further enhances sensitivity and reduces parameter correlation. Sensitivity and correlation analysis was conducted to identify a suitable location for the additional thermocouples, aimed at improving the accuracy of the estimation. The forward model was developed in COMSOL and was connected to the inverse problem solving LM algorithm in MATLAB using LiveLink server. Using the combined dataset consisting of experimental TPS data (obtained from Zhang et al. in Prog Comput Fluid Dyn 13:191–201, 2013) and numerically generated thermocouple data, the proposed inverse framework yielded mean deviations in thermal conductivity and specific heat capacity of less than 1 % for both materials. The deviation in the estimated thermal conductivity reported by Zhang et al. (Prog Comput Fluid Dyn 13:191–201, 2013) was 1.92 % for stainless steel and 2.86 % for ceramic.