As the most common road distress, cracks have a substantial influence on the integrity of pavement structures. Accurate identification of crack existence and quantification of crack geometry are thus critical for the decision-making of maintenance measures. This paper proposes a novel neural network for the detection and evaluation of road cracks at pixel level, which combines the advantage of encoding–decoding network and attention mechanism and can thus extract the crack pixels more accurately and efficiently. The proposed network achieves an excellent detection performance with IOU = 92.85%, precision = 96.90%, recall = 95.36%, F1 = 95.53%. Compared with the other advanced networks, the accuracy of the proposed method is substantially enhanced. The quantitative estimation of key geometrical features of cracks including length, width, and area is successfully realized with the development of a prototype of an intelligent mobile system. Compared with the ground truth, the maximum crack width shows the lowest relative error rate, which ranges −31.75%∼28.57%.

This paper presents a theoretical and experimental study of the nonlinear flexural vibration of a cement-based material with distributed microcracks caused by an important deterioration mechanism, alkali-silica reaction (ASR). The general equation of motion is derived for the flexural vibration of a slender beam with the nonlinear hysteretic constitutive relationship for consolidated materials, and then an approximate formula for excitation-dependent resonance frequency is obtained. A downward shift of the resonance frequency is related to the nonlinearity parameters defined in the constitutive relationship. Vibration experiments are conducted on standard mortar bar samples undergoing progressive ASR damage. The absolute nonlinearity parameters are determined from these experimental results using the theoretical solution in order to investigate their dependence on the damage state of the material. With the progress of the ASR damage, the absolute value of the hysteresis nonlinearity parameter increases by as much as six times from the intact (undamaged) state in the sample with highly reactive aggregate; this is in contrast to a change of about 16% in the linear resonance frequency. It is demonstrated that the combined theoretical and experimental approach developed in this research can be used to quantitatively characterize ASR damage in mortar samples and other cement-based materials.

Quality control of embankment compaction is of great significance to the durability and stability of high-embankment airports. The current method of embankment quality control mainly relies on manual inspection for a limited number of designated locations after the completion of compaction, which cannot provide the information feedback simultaneously during the compaction or guarantee the compaction quality of the entire working surface. There is an urgent need to develop a new quality-control method of high-embankment airports for the full course of the compaction process. In this paper, a management platform is developed to monitor the real-time compaction process of a high-embankment airport and evaluate the compaction quality of the working surface. To mitigate the effect of human factors to a minimal extent, an optimal path algorithm to guide the real-time compaction trajectory of the impact compactor is proposed, and the unmanned vehicle control technology is implemented on the impact compactor. Furthermore, a virtual-reality tool is incorporated in the developed management platform to provide a three-dimensional interactive display for the compaction process. The feasibility and robustness of the developed management platform is validated by a case study in a pilot section of Beijing Daxing International Airport. The data collected in the case study show that the proposed optimal path algorithm and unmanned vehicle control technique enable the construction process to be faster and more efficient and improve the compaction quality of embankment.

Timely crack detection of pavement helps inspectors access road conditions and determine the maintenance strategy, which can reduce repair costs and safety risks. Deep learning has greatly advanced the development of automated crack detection, but there are still challenges that hinder the application of crack segmentation networks in engineering practice such as the bloated models, the class imbalance problem, and the high-performance device dependency. For efficient crack segmentation tasks, this paper proposes a novel high-performance lightweight network termed multi-path convolution feature fusion lightweight network (MCFF-L Net) and utilize the concept of knowledge distillation. The MCFF-L Net with only 1.18 M parameters achieves F1 score of 85.70% and intersection over union (IoU) of 78.22%, which surpasses the popular heavyweight networks and lightweight networks. The proposed network is further implemented on an embedded device of Jetson Xavier NX and the detection speed of pavement cracks can reach 9.71 frames per second (FPS). The combination of embedded device and proposed lightweight networks is suitable for application scenarios where the portable and efficient crack segmentation is needed and the reliable data transmission through network is not accessible.

In this article, a practical analytical method for prediction of the postconstruction settlement of high embankments is proposed through the further simplification of the one-dimensional creep deformation algorithm under the theoretical framework of the unified hardening (UH) model. The feasibility and reliability of the proposed practical method were validated by a case study of the Chengde Airport (Chengde city, Hebei province), in which actual settlement measurement data from three monitoring points were employed. The predicted settlement results of the proposed method were further compared with those from two conventional methods, the three-point method and hyperbola method. The results show that the proposed method generally provides a more reliable prediction of settlements over two conventional methods with a limited amount of available measurement data.