Scholarlink

In the construction of roads, soil is commonly utilised as a fill material to support soft soil layers. The unconfined compression strength (UCS) and California bearing ratio (CBR) scores are key parameters. The California Division of Highways in 1928 developed the California bearing ratio method, an empirical technique for designing flexible pavements for roads, railways, and airfields. The subgrade for the pavement serves as the base, and to maximise its strength, it must be well-compacted. The soil’s CBR value is linked to the strength of the subgrade. Conducting the CBR test is labour-intensive and time-consuming, requiring the testing of moulded soil samples. Obtaining accurate CBR values for the soil is always a challenge for engineers. The soil type and other soil properties influence the CBR values. This study aims to establish correlations using adaptive neuro-fuzzy inference system (ANFIS) analysis between various soil index properties and CBR values to obtain more accurate CBR values. The results are satisfactory, showing that the soil’s CBR value is influenced by factors such as organic content, sand fraction, plastic limit (PL), clay fraction, plasticity index, liquid limit (LL), etc. It has been observed that the predicted values align with the experimental values, indicating a strong correlation for CBR.

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

Evaluation of ANFIS-Based Models for Predicting Soil CBR from Index Properties

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Himanshu Kumar Jangir,
  • Anirban Mandal,
  • V. Srinivasan

摘要

In the construction of roads, soil is commonly utilised as a fill material to support soft soil layers. The unconfined compression strength (UCS) and California bearing ratio (CBR) scores are key parameters. The California Division of Highways in 1928 developed the California bearing ratio method, an empirical technique for designing flexible pavements for roads, railways, and airfields. The subgrade for the pavement serves as the base, and to maximise its strength, it must be well-compacted. The soil’s CBR value is linked to the strength of the subgrade. Conducting the CBR test is labour-intensive and time-consuming, requiring the testing of moulded soil samples. Obtaining accurate CBR values for the soil is always a challenge for engineers. The soil type and other soil properties influence the CBR values. This study aims to establish correlations using adaptive neuro-fuzzy inference system (ANFIS) analysis between various soil index properties and CBR values to obtain more accurate CBR values. The results are satisfactory, showing that the soil’s CBR value is influenced by factors such as organic content, sand fraction, plastic limit (PL), clay fraction, plasticity index, liquid limit (LL), etc. It has been observed that the predicted values align with the experimental values, indicating a strong correlation for CBR.