Riaz K, Ahmad N. Predicting resilient modulus: A data driven approach integrating physical and numerical techniques.
Heliyon 2024;
10:e25339. [PMID:
38327424 PMCID:
PMC10847910 DOI:
10.1016/j.heliyon.2024.e25339]
[Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 11/07/2023] [Revised: 01/09/2024] [Accepted: 01/24/2024] [Indexed: 02/09/2024] Open
Abstract
Resilient modulus (MR) is an important parameter in the design of pavement that helps to characterize the quality of sub-grade materials. Generally, it is not determined experimentally due to time consuming, uneconomical, laborious and lack of advanced equipment in many laboratories. The aim of this research is to determine MR values using experimental (Ultrasonic pulse velocity (UPV) and Cyclic Triaxial) and Artificial neural network (ANN) techniques. For experimental study twenty-four soil samples comprising of coarse and fine-grained soils were collected from different locations. For ANN modelling, Input variables comprised of essential soil Atterberg limits (liquid limit, plastic limit, plasticity index) and compaction properties (maximum dry density, optimum moisture content). The validation of ANN model is done by comparing its results with the experimentally evaluated MR from UPV and Cyclic Triaxial test. Experimental results showed that Cyclic Triaxial test yielded resilient modulus value that was 5 % more than obtained from the UPV test. Moreover, results showed that modulus of resilience (MR) values determined by UPV, and artificial neural network (ANN) modelling have significant closeness with the cyclic triaxial results of resilient modulus; thus, making it a significant development in predicting resilient modulus efficiently.
Collapse