Scholarlink

<p>This article presents the analysis of results from Plate Load Tests (PLT) involving a series of load-unload-reload cycles. A total of 128 tests were conducted to investigate the parameters governing the Modulus Improvement Factor (MIF) based on the performance of unreinforced and geocell-reinforced coarse materials under varied conditions. A soft clay and a sandy soil, prepared at three different relative densities, were used as subgrade. The geocell infill material consisted of sandy soil and Granular Subbase (GSB) material, placed at two different densities each. Two types of HDPE geocells with different dimensions were used. Multilayer Perceptron (MLP) neural networks were used to analyze the combined influence of the various parameters on MIF. The results show that MIF has an increasing trend with increasing subgrade modulus, and a decreasing trend with increasing geocell pocket size and infill material modulus. Overall, the MLP was identified as a suitable tool for parametric analyses to assess the benefits of geosynthetics in roadway applications.</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

Integrating Supervised Machine Learning with Laboratory Data To Evaluate the Modulus Improvement Factor in Geocell-Reinforced Soils

  • 期刊论文

  • 发表日期：

    2025年11月25日

查看全文
  • H. D. Trujillo,
  • J. O. Avesani Neto,
  • J. G. Zornberg

摘要

This article presents the analysis of results from Plate Load Tests (PLT) involving a series of load-unload-reload cycles. A total of 128 tests were conducted to investigate the parameters governing the Modulus Improvement Factor (MIF) based on the performance of unreinforced and geocell-reinforced coarse materials under varied conditions. A soft clay and a sandy soil, prepared at three different relative densities, were used as subgrade. The geocell infill material consisted of sandy soil and Granular Subbase (GSB) material, placed at two different densities each. Two types of HDPE geocells with different dimensions were used. Multilayer Perceptron (MLP) neural networks were used to analyze the combined influence of the various parameters on MIF. The results show that MIF has an increasing trend with increasing subgrade modulus, and a decreasing trend with increasing geocell pocket size and infill material modulus. Overall, the MLP was identified as a suitable tool for parametric analyses to assess the benefits of geosynthetics in roadway applications.