The aim of this article is to explore the dynamic compaction characteristics of stone mastic asphalt (SMA) and the friction–lubrication effect of internal particles during the superpave gyratory compaction (SGC) process. Firstly, a calculated method for the compaction degree of an asphalt mixture in the gyratory compaction process was defined based on the multiphase granular volume method. Secondly, the gyratory compaction curves of asphalt mixtures were taken based on this calculation method of compaction degree. The dynamic change law of each compaction index (compaction, percentage of air voids, compaction energy index, etc.) during the compaction process was analysed. Finally, the effects of different initial compaction temperatures and different asphalt content on the friction–lubrication effect and compaction characteristics of asphalt mixtures were studied. Research shows that it is reasonable to define the compaction degree by the ratio of the apparent density of the asphalt mixture to the maximum theoretical density of the asphalt mixture during gyratory compaction. The dynamic prediction equations of the compaction degree K and the compaction energy index CEI with the amount of compaction were established, and could effectively predict the compaction degree, percentage of air voids and compaction energy index CEI. The compaction process of the asphalt mixture needed to go through three phases, including periods of rapid growth, slow growth, and stabilisation, and the compaction degree increased by about 10%, 5%, and 1%, in that order, finally tending towards a stable value. The effect of the initial compaction temperature on the forming compaction degree of the asphalt mixture is significant; therefore, it should be controlled strictly in the compaction construction of asphalt mixtures. When the initial compaction temperature of SMA-13 is about 170 °C, the compaction effect is optimal, and the effect of the increase in the amount of compaction at a later stage on the increase in the compaction degree of the asphalt mixture is very low. With the optimal asphalt content, the friction–lubrication effect between the asphalt and aggregate particles is optimal, because it can effectively form an asphalt film, reducing the frictional resistance of the particles moving each other during the compaction process, and the voids will be embedded and filled with each other, finally producing the best compaction result.

To investigate the dynamic evolution of macro- and fine-scale characteristics during the compaction process of Stone Mastic Asphalt (SMA-13), the following methodology was employed. First, the compaction characteristics were analyzed based on the Marshall compaction test, and an exponential regression model for compaction was established. A compaction coefficient was proposed to evaluate the ease of compaction of SMA-13. Second, Marshall specimens subjected to different number of blows (5, 15, 25, 50, 75, 100, and 125 times per face) were scanned using a Computed Tomography (CT) scanner, and image processing techniques were applied to precisely extract void characteristics (void number, void area, and equivalent void diameter) under various compaction states. Third, the evolution of void characteristics in SMA-13 during the compaction process was analyzed, and correlations with macroscopic compaction properties were established. The experimental results demonstrated that SMA-13 achieved optimal compaction at an asphalt content of 5.9% and an initial compaction temperature of 170 °C. The established compaction numerical regression model can effectively characterize the compaction characteristics of SMA-13. A higher compaction coefficient indicated easier compaction and better compaction performance. During compaction of SMA-13, the void number and void area exhibited exponential decay, and voids with equivalent diameters of 1–7 mm gradually decreased, showing a non-proportional linear decay in their distribution. In contrast, voids with equivalent diameters of 0–1 mm increased during compaction, and they are the dominant component of the void structure in SMA-13. During compaction, the void ratio of SMA-13 gradually decreased along the direction of height, and the distribution of void ratio was “great at both ends and small in the middle”. The void ratio at 5–55 mm decreased from about 10% to about 0%, and the void ratio distribution was relatively uniform. The void ratio at the bottom 0–5 mm and top > 50 mm was large and unevenly distributed, mostly between 10% and 40%.