Doctor of Philosophy (PhD)
Civil and Environmental Engineering
The objective of this study was to formulate a Convolutional Neural Networks (CNN) model and to develop a decision-making tool using Artificial Neural Networks (ANN) to identify top-down, bottom-up, and cement treated (CT) reflective cracking in in-service flexible pavements. The CNN’s architecture consisted of five convolutional layers with three max-pooling layers and three fully connected layers. Input variables for the ANN model were pavement age, asphalt concrete (AC) thickness, annual average daily traffic (AADT), type of base, crack orientation, and crack location. The ANN network architecture consisted of an input layer of six neurons, a hidden layer of ten neurons, and a target layer of three neurons. The developed CNN model was found to achieve an accuracy of 93.8% and 91.0% in the testing and validation phases, respectively. The ANN based decision-making tool achieved an overall accuracy of 92% indicating its effectiveness in crack identification and classification.
In the second phase of the study, the flexible pavement responses under a dual tire assembly were analyzed to identify the critical stress mechanisms for bottom-up and top-down cracking. Higher tensile strains were observed to occur underneath the tire ribs than away from them supporting the argument that both surface initiated and bottom-up fatigue cracking develop in or near the wheel paths. The incorporation of surface transverse tangential stresses increased the surface tensile strains near the tire ribs by approximately 68%, 63%, and 53% respectively for low, medium, and high volume flexible pavements indicating an increased potential for the initiation and development of top-down cracking when tangential stresses are considered. In contrast, this effect was observed to be minimal for the tensile strains at the bottom of the asphalt layer, which are the main pavement responses used in the prediction of fatigue cracking.
Shrinkage cracking in cement treated base (CTB) was also modeled in finite element using displacement boundary conditions. The tensile stresses due to shrinkage strains in the cement treated base were observed to be comparable to the tensile strength of CTB at 7 days and higher at 56 days indicating the potential development of shrinkage cracks.
Dhakal, Nirmal, "Identification of Top-down, Bottom-up, and Cement-Treated Reflective Cracks Using Convolutional Neural Network and Artificial Neural Network" (2020). LSU Doctoral Dissertations. 5270.
Available for download on Saturday, August 07, 2021