Scholarlink

<p>The Titanium metal matrix hybrid composite was developed and its wear behaviour was investigated in this study. Zirconium Oxide (ZnO<sub>2</sub>) and Graphene (Gr) were used as reinforcement particles. The liquid metallurgical (stir casting) method was implemented in developing the composite. Taguchi’s analysis and ANOVA methods were used to design the experiments and to identify the most influencing parameters on wear rate. ANOVA analysis results indicates that the Normal load (43.64%) and sliding distance (36.77%) were most influence on wear rate followed by sliding velocity (2.04%) and interaction effect shows insignificant. The wear patterns of SEM images shows the abrasion wear mechanisms with layers of a material separate were observed during the testing. Further delamination was restricted due to self-lubrication properties of the reinforcement particles. Advanced Machine learning algorithms like KNN, Support vector Machine and XG Boost methods were accurately classified the wear data and KNN Model gave the accuracy of 78.68%, SVM classifier showed 91.80% and XG Boost classifier showed 93.44%. The statistical results and the Machine learning model predictions are within the acceptable limit.</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

Systematic Wear Rate Analysis of Ti–6Al–4 V Based Hybrid Composites Using Taguchi Optimizer and Machine Learning Classifiers

  • 期刊论文

  • 发表日期：

    2025年2月10日

查看全文
  • T. S. Sachit,
  • R. Suresh,
  • T. Nagaraj

摘要

The Titanium metal matrix hybrid composite was developed and its wear behaviour was investigated in this study. Zirconium Oxide (ZnO2) and Graphene (Gr) were used as reinforcement particles. The liquid metallurgical (stir casting) method was implemented in developing the composite. Taguchi’s analysis and ANOVA methods were used to design the experiments and to identify the most influencing parameters on wear rate. ANOVA analysis results indicates that the Normal load (43.64%) and sliding distance (36.77%) were most influence on wear rate followed by sliding velocity (2.04%) and interaction effect shows insignificant. The wear patterns of SEM images shows the abrasion wear mechanisms with layers of a material separate were observed during the testing. Further delamination was restricted due to self-lubrication properties of the reinforcement particles. Advanced Machine learning algorithms like KNN, Support vector Machine and XG Boost methods were accurately classified the wear data and KNN Model gave the accuracy of 78.68%, SVM classifier showed 91.80% and XG Boost classifier showed 93.44%. The statistical results and the Machine learning model predictions are within the acceptable limit.