Title: Prediction of Subgrade California Bearing Ratio Using Falling Weight Deflectometer for Pavement Evaluation in Kenya
Authors: Catherine Mugai, Kepha Abongo and Brian Jacks Odero
Volume: 9
Issue: 8
Pages: 30-38
Publication Date: 2025/08/28
Abstract:
This study aimed to develop a method for determining the California Bearing Ratio of subgrade soils during pavement evaluations without the need for traditional, destructive trenching and laboratory testing. Subgrade California Bearing Ratio is a key parameter for assessing pavement foundation strength, with the Road Design Manual classifying subgrades from S1 to S6 based on California Bearing Ratio values. Eighteen road sections were purposively selected for analysis due to data availability, including Falling Weight Deflectometer readings, trench logs, laboratory soil tests, and construction data. Falling Weight Deflectometer data was collected at 100-meter intervals using a PRIMAX Falling Weight Deflectometer device, while trial pits were excavated every 5 km for sampling and laboratory testing. Soil classification followed the Unified Soil Classification System, with Lean Clay and Elastic Silt being predominant. Plasticity Index, Optimum Moisture Content, and Maximum Dry Density ranged from 7-23%, 12-27%, and 1250-1821 kg/m³ respectively. For modeling purposes, deflections from the 4th geophone (D4), positioned 600 mm from the load center, were selected due to consistent curve behavior in deflection bowl and surface modulus graphs. The laboratory-derived California Bearing Ratio values served as the dependent variable. Data from all road sections were randomly split into training (80%) and testing (20%) sets. Models developed included Linear Regression, Random Forest, and Artificial Neural Networks. These models yielded R² values between 0.55 and 0.61, and Root Mean Square Errors of 7 to 7.2, indicating acceptable predictive performance. Predicted California Bearing Ratio values showed strong correlation with laboratory values, with Pearson correlation coefficients ranging from 0.81 to 0.84. The results demonstrate that Falling Weight Deflectometer data, particularly from the D4 geophone, can effectively be used to estimate subgrade California Bearing Ratio, offering a viable, non-destructive alternative for pavement evaluation.