Scholarlink

<p>Recently, utilization of Wood Plastic Composite (WPC) panels has grown in both indoor and outdoor settings. Drilling these panels are essential for installation, assembly, and securing components productivity. This study focuses on developing predictive modeling and optimization for drilling WPC composite panels by considering the machining parameters, mean roughness depth (R<sub>z</sub>), and average surface roughness (R<sub>a</sub>). An experimental design based on an L<sub>27</sub> orthogonal array is employed, along with regression and Artificial Neuro-Fuzzy Inference System (ANFIS) approaches to enhance prediction accuracy for R<sub>a</sub> and R<sub>z</sub>. Also, the key input parameters have included diameter of drill bit (<i>d</i>), feed rate (<i>f</i>), and spindle speed (<i>N</i>). The results have indicated that R<sub>a</sub> and R<sub>z</sub> are mainly influenced by <i>f</i> and <i>d</i>, which increase the roughness owing to enhanced cutting forces and poor chip evacuation. Conversely, higher <i>N</i> and lower <i>f</i> reduce the R<sub>a</sub> and R<sub>z</sub>. Optimized parameters identified through desirability-based methods are <i>N</i> = 3000 rpm, <i>f</i> = 75&#xa0;mm/min, and <i>d</i> = 6&#xa0;mm. Results showed that ANFIS prediction models outperform response surface methodology (RSM) models with higher R<sup>2</sup> values of 0.9525 and 0.9581 for R<sub>a</sub> and R<sub>z</sub> with root mean squared error value of 0.87 and 2.52 (R<sub>a</sub> and R<sub>z</sub>) that are lower than RSM models. This study introduces the ANFIS model approach integrated with particle swarm optimization (PSO) and hippopotamus optimization algorithm (HOA), to enhance the performance of the drilling of WPC. This enables more accurate and reliable optimization of R<sub>a</sub> and R<sub>z</sub> compared to previous single or non-hybrid algorithm approaches.</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

Optimization of surface roughness in drilling wood plastic composite (WPC) using hybrid Taguchi-response surface methodology and metaheuristic algorithms

  • 期刊论文

  • 发表日期：

    2026年1月8日

查看全文
  • Sundarapandiyan Murugesan,
  • Palanikumar Kayaroganam,
  • Senthilkumar Natarajan,
  • Latha Balasubramanian

摘要

Recently, utilization of Wood Plastic Composite (WPC) panels has grown in both indoor and outdoor settings. Drilling these panels are essential for installation, assembly, and securing components productivity. This study focuses on developing predictive modeling and optimization for drilling WPC composite panels by considering the machining parameters, mean roughness depth (Rz), and average surface roughness (Ra). An experimental design based on an L27 orthogonal array is employed, along with regression and Artificial Neuro-Fuzzy Inference System (ANFIS) approaches to enhance prediction accuracy for Ra and Rz. Also, the key input parameters have included diameter of drill bit (d), feed rate (f), and spindle speed (N). The results have indicated that Ra and Rz are mainly influenced by f and d, which increase the roughness owing to enhanced cutting forces and poor chip evacuation. Conversely, higher N and lower f reduce the Ra and Rz. Optimized parameters identified through desirability-based methods are N = 3000 rpm, f = 75 mm/min, and d = 6 mm. Results showed that ANFIS prediction models outperform response surface methodology (RSM) models with higher R2 values of 0.9525 and 0.9581 for Ra and Rz with root mean squared error value of 0.87 and 2.52 (Ra and Rz) that are lower than RSM models. This study introduces the ANFIS model approach integrated with particle swarm optimization (PSO) and hippopotamus optimization algorithm (HOA), to enhance the performance of the drilling of WPC. This enables more accurate and reliable optimization of Ra and Rz compared to previous single or non-hybrid algorithm approaches.