Computational Process for Quantifying the Impact of Flooding on Remaining Life of Flexible Pavement Structures

Abstract

Field observations have indicated significant damage to pavement structures due to sustained saturation of foundation layers after heavy rainfalls. Highway agencies need to have insight into the influence of rainfall events on their road network to properly plan and update their maintenance operations. Attempts have been made to collect experimental data as a basis for network-level decision-making when pavement sections are subjected to surface infiltration. However, generating a comprehensive database of field measurements is extremely laborious and expensive. This paper presents a computational process which can be used to analyze inundated pavement sections considering a wide range of factors. This computational tool was verified with experimental measurements of a flooded pavement. A relatively large database of a sample pavement section was generated, and a neural network was trained accordingly. Acceptable performance of the designed neural network confirmed the suitability of the generated database for development of data-driven models. Such models enable decision makers to rapidly assess the impact of flooding on the life of pavements in the network at reasonable computational cost and accuracy.