This study aimed to investigate variations and relationships between molecular components and rheological properties during the long-term aging of SBS-modified asphalt by the action of moisture. The chemical components and changes in molecular weight during aqueous PAV aging were observed using a four-component test and GPC test, and these morphological changes were quantified. Viscosity tests, time-scan tests and damage-healing tests were carried out to investigate the rheological properties during the aging process. Finally, the mechanism of asphalt aging under the influence of moisture was discussed. The results showed that moisture facilitated the long-term aging of asphalt and complicated the aging behavior under pressure-aging conditions. The dissolution of certain hydrophilic groups may be responsible for the decrease in resin content. The fatigue and self-healing properties of asphalt were weakened by moisture during the aging process. This was mainly attributed to a decrease in the composition of the asphalt colloidal dispersion medium, which resulted in earlier and faster development of microcracks under repeated loading as well as retardation in the rate of asphalt surface approach, wetting and spreading. Under long-term aging in the presence of moisture, the molecular components of the asphalt showed significant correlations with the rheological properties. The results of this study can contribute to further explaining the influence of moisture on the thermal-oxidative aging of asphalt.

A new adjustable and combined reinforcement method consisting of arch rib and steel beam is proposed in this paper. The principle is to redistribute the internal forces of the structure through the support of the arch rib and to form a combined structure by using steel and concrete. Symmetrically loaded bending tests were carried out with five different combined strengthening schemes. The analysis of the test process was conducted adopting a FEM model, and the equations for calculating the RC beam's load capacity under combined strengthening were derived while considering the slip effect. The experimental results indicated that the combined reinforcement method had a synergistic effect when the ultimate load capacity had increased by more than 95% after reinforcement. As the size of the arch ribs and profile beams increased, a better reinforcement effect was seen. However, beyond a certain range, the RC beams had already broken down and had failed before the reinforcement material could yield. The combined reinforcement method changes the force system of the original RC beam, resulting in a smaller angular inclination trend of the cracks in the original purely curved section and a smaller crack extension. The non-linear finite element results are generally consistent with experimental loads, displacements, and ductility, while the average ratio of experimental to calculated ultimate Load capacity is 0.95, demonstrating that non-linear finite elements and theoretical formulas are accurate.

Introduction: The traditional crack detection method usually requires a tedious process of sensor installation and removal, which seriously affects the efficiency of concrete structure management and maintenance.Methods: For this reason, this paper develops a fast concrete crack detection method based on percussion with an improved convolutional neural network (CNN). By utilizing the percussion method, the sensors do not need to be coupled and installed on the concrete structure, which saves a lot of processes. The sound signals generated by percussion are collected by acoustic pressure sensors, while multiple data enhancement techniques are applied to enrich the data volume and diversity of the collected signals. The Mel-frequency cepstral coefficient (MFCC) of the sound signals are then extracted as inputs to the improved CNN model. The CNN used is mainly applied to initialize the weights by applying the transfer learning technique, and the Squeeze-and-Excitation Networks (SENet) attention mechanism is embedded to improve the model’s focus on important features. Finally, comparative experiments with different frame lengths, different models and different signal-to-noise ratios (SNR) are conducted using the improved CNN.Results: The results show that the model validation process has the least loss and highest accuracy when the input frame length is 1024. The improved CNN has good feature learning ability for MFCC of percussion sound signals for effective recognition of concrete cracks. Compared with Resnet18, random forest and long short-term memory networks, the improved CNN has superior recognition accuracy and stability, and shows better noise robustness in high signal-to-noise ratio (SNR: −6 db∼6 db) environments.Discussion: Therefore, the proposed method has a high potential for future crack detection in concrete structures.