Fiber Bragg grating (FBG) exhibits strong resistance to electromagnetic interference and excellent linear strain response, making it highly promising for structural health monitoring (SHM) in pavement. This research investigates the strain transfer characteristics of embedded FBG in pavement structure and materials by using the relevant theoretical models. Results indicate adhesive layer thickness and sheath modulus are the primary factors influencing the strain transfer coefficient. A thinner adhesive layer and high modulus of sheath enhance the coefficient. Additionally, the strain distribution of sheath significantly affects the transfer efficiency. When the stress level near the grating region is lower than the both ends, the coefficient increases and even exceeds 1, which typically occurs under multi-axle conditions. As for asphalt mixture, high temperature leads to lower efficiency, while accumulated plastic strain improves it. Although the increased load frequency results a higher strain transfer coefficient, the magnitude of this change is negligible. By employing polynomial fitting to the sheath strain distribution, the boundary condition of theoretical equation could be removed. The theoretical and numerical results of strain transfer coefficient for pavement embedded FBG demonstrate good consistency, indicating the polynomial fitting is adoptable for the theoretical calculation with non-uniform strain distribution. This study utilizes the FEM to clarify the evolution of FBG strain transfer in pavement structures and materials, providing a theoretical basis for the design and implementation of embedded FBG in pavement.

In order to study the mechanical properties of graded crushed stone, the discrete element method is used to simulate the CBR test of graded crushed stone. Aiming at the composition structure of graded crushed stone material, the PFC3D simulation software is used to construct the test model, and the process of constructing the virtual specimen model of the graded crushed stone discrete element model is discussed in detail. A servo mechanism is used to control the speed of the wall in the software, so as to control the virtual confining pressure imposed on graded crushed stone by the wall and simulate the real CBR test environment. The micro-parameter calibration of the virtual test is carried out by comparing the indoor and virtual CBR specimens of a single particle size specimen and three groups of graded crushed stone specimens. The comparison result shows that the stress–strain characteristics of the graded crushed rock obtained by the discrete element simulation during the uniaxial penetration process have a high degree of similarity, which can verify the accuracy of the model establishment. With the increase in the penetration depth, the penetration force of the aggregates of various particle sizes gradually increases, and the penetration force and the penetration depth are basically linear, and when the particle size is greater than 9.5 mm, the increase in particle size has little effect on the CBR test results. Under the certain conditions, the contact stiffness of graded crushed stone particles with particle sizes of 4.75 mm, 9.5 mm, 13.2 mm, 16 mm, and 19 mm should be 0.88 × 10⁷ (N/m), 0.98 × 10⁷ (N/m), 1.10 × 10⁷ (N/m), 1.25 × 10⁷ (N/m), and 2.05 × 10⁷ (N/m), respectively. The recommended value of the contact stiffness of the small spherical particles increases with the increase in the particle size. This model can provide a basis for studying the micromechanical state of graded crushed stone and physical mechanics tests.

Smart aggregate (SA) is a multi-module integrated embedded monitoring sensor that has been employed in the construction, operation and maintenance of civil infrastructure. However, the differences in the shape and Yong’s modulus of the SA and the host matrix lead to the matching errors between the SA sensor readings and the actual in-situ stress values. A systematic analysis of how the modulus ratio of the SA to the host matrix (E _p/E _m), the length-to-height ratio of the sensor (L/H), and the non-homogeneity of the particle media interfere with the sensor monitoring data has been conducted in this paper using the finite element method and the discrete element method. In addition, the differences of the stress response from SA sensor under different load contact areas have been further investigated. Simultaneously, the numerical analysis has been validated by means of both theoretical derivations and laboratory tests. The analysis results show that E _p/E _m and L/H are the two key factors affecting the matching error. In particular, when E _p/E _m > 1, the matching error is positive and vice versa, while the L/H of the sensor is negatively correlated with the matching error. It is also noteworthy that the errors in the SA sensor monitoring data due to media inhomogeneity can only be negligible when the sensor diameter is larger than 20 times of the largest particle size in the host matrix. This study provides a theoretical guidance for optimizing the design of SA sensors and improving the accuracy of their measurement results.