Scholarlink

Expansive clays were chemically altered to meet the field foundation and embankment material requirements. Moreover, the blended clays’ gradation significantly affects the filed densification. However, making gradation distribution curves of the fine-grained soils is cumbersome and time-consuming. In this study, an attempt was made to predict the gradation distribution factors such as coefficient of uniformity (Cu) and coefficient of curvature (Cc) using the support vector machine (SVM). An experimental investigation was carried out on the expansive clays blended with varying fly ash and saw dust ash content such as gradation distribution, cohesion, and friction angle. Furthermore, statistical analysis and prediction models were developed. Performance metrics were considered to evaluate the prediction models. Based on the performance metrics it was found that higher coefficient of determination (R2) values was obtained greater than 0.95. Finally, it was found that the Cu value is strongly reliant on saw dust ash, whereas the frictional angle value is influenced by the Cc value. The prediction models could help estimate the dosage of the additive content to fulfill the requirements of the embankment fill material.

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

Support Vector Machine (SVM) Models for Prediction of Coefficient of Uniformity and Coefficient of Curvature of Soils Blended with Fly Ash and Sawdust Ash

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Shaik Subhan Alisha,
  • Bh. Revathi,
  • T. Venkateswararao,
  • Pilla Sita Rama Murty,
  • Pathan Fayaz

摘要

Expansive clays were chemically altered to meet the field foundation and embankment material requirements. Moreover, the blended clays’ gradation significantly affects the filed densification. However, making gradation distribution curves of the fine-grained soils is cumbersome and time-consuming. In this study, an attempt was made to predict the gradation distribution factors such as coefficient of uniformity (Cu) and coefficient of curvature (Cc) using the support vector machine (SVM). An experimental investigation was carried out on the expansive clays blended with varying fly ash and saw dust ash content such as gradation distribution, cohesion, and friction angle. Furthermore, statistical analysis and prediction models were developed. Performance metrics were considered to evaluate the prediction models. Based on the performance metrics it was found that higher coefficient of determination (R2) values was obtained greater than 0.95. Finally, it was found that the Cu value is strongly reliant on saw dust ash, whereas the frictional angle value is influenced by the Cc value. The prediction models could help estimate the dosage of the additive content to fulfill the requirements of the embankment fill material.