Artificial Intelligence-Based Inverse Analysis of Flexible Pavement Deteriorations under Moving Loads
Abstract
Fatigue damage and aging are two typical deterioration modes in flexible pavements. Fatigue damage is mainly caused by load repetitions. It results in the degradation of pavement materials and structures which significantly influences the riding quality and safety, and reduces the service life of pavements. As the loading application continues, microcracks in the asphalt layer initiate, develop and connect to form visible macrocracks. Macrocracks further deteriorate to potholes, which finally result in the moisture infiltration, the loss of strength and durability of the pavement. To avoid such problems, fatigue damage should be controlled before macrocracks form. An indicator to reflect the current deterioration condition of the pavement and provide an advance warning of the pavement failure is necessary.
Aging of flexible pavements is caused by physical and chemical processes in asphalt layers such as the oil volatilization, oxidation and steric hardening etc. Asphalt mixtures closer to the pavement surface are more susceptible to aging and have higher modulus. It is one essential reason to explain the modulus gradient in asphalt layers. Flexible pavements suffering from aging tend to deteriorate in terms of cracking such as the fatigue cracking and thermal cracking. Therefore, an effective method to evaluate the aging degree of a flexible pavement is also necessary.
By reviewing current technologies in determining degrees of fatigue damage and aging in flexible pavement materials and structures, motivations and objectives of this research are identified. The dissertation aims to propose methodologies in evaluating these two deteriorations in flexible pavements from pavement responses under moving loads or equivalent dynamic loads, which benefit from fast developing nondestructive testing devices. To achieve this goal, the finite element model updating incorporating surrogate models and artificial intelligence algorithms were applied in the equivalency of moving vehicular loads and stationary dynamic loads, calibration of layer moduli, etc. Deteriorations in flexible pavements can be directly determined from field testing and information recorded in the database as well.
Subject
flexible pavementsfatigue damage
aging
finite element simulation
Artificial Intelligence
moving loads
Citation
Deng, Yong (2020). Artificial Intelligence-Based Inverse Analysis of Flexible Pavement Deteriorations under Moving Loads. Doctoral dissertation, Texas A&M University. Available electronically from https : / /hdl .handle .net /1969 .1 /192250.