The State-of-the-Art Review on Prediction of Subgrade CBR: Past and Present Trends

Abstract

The design thicknesses and subgrade strength of the flexible pavement are largely dependent on the California bearing ratio (CBR). The laboratory approach to determining CBR is labor-intensive. Hence, only a limited number of tests may be conducted throughout the length of roadways. The few tests that are required may not be enough to adequately capture the variations in the field most of the time. As a result, there is a pressing demand for CBR prediction models that could forecast the CBR in no time. In view of this, researchers across the world conducted studies to come up with simple, time-saving, and cost-effective testing methods that could replace the existing standard methodology of testing with different types of field methods, mathematical models based on soil index properties, and some soft computing models. This review study evaluates the effectiveness of several field methods that have received a lot of attention from researchers, including the falling weight deflectometer (FWD), dynamic cone penetrometer (DCP), and Clegg hammer tests. Attempts have also been made to review the pre-existing correlation models based on soil index properties. Meanwhile, this study also explores the competency of several widely referenced machine learning-based models for CBR prediction, which comes under the broader domains of artificial intelligence (AI) and machine learning (ML), including gene expression programming (GEP) and artificial neural networks (ANN). Apart from these, approach of various researchers on use of other AI/ML models such as random forest regression (RFR), decision trees (DT), support vector regression (SVR), and Gaussian process regression (GPR) has also been discussed. Correlations and artificially generated models reported in the literature have been studied, and the most suitable prediction methods among them have been suggested. © The Author(s), under exclusive licence to Indian Geotechnical Society 2024.