Abstract: The large-scale application of gangue preparation filling materials can not only alleviate the environmental pressure caused by its accumulation, but also significantly reduce the mining filling cost. Optimizing the material structure through dense aggregate accumulation technology is the core strategy to promote the decarbonization and performance enhancement of filling materials. A new low-carbon paste filling material system was developed based on the synergistic effect of multi-stage gangue crushing process (coarse/fine double-granularity) and high mineralized mine water. A detailed modelling method based on Abaqus-Python is constructed to realize the parametric generation of 2D random polygon and 3D random concave-convex polyhedral aggregates, and the accuracy of the model is verified by combining Fuller's theoretical equations with the experimental grading curves (the optimal R²=0.985), and the leaching characteristics of heavy metals and inorganic salts of the filling materials were analyzed. The concept of paste thickness was introduced to reveal the volume effect of slurry-aggregate, establish the multi-scale synergistic design method of gangue-mining water, and prepare two types of typical filling bodies of C5/C10 to verify the reliability of the design. The mesoscale aggregate model could generate two-dimensional random polygon and three-dimensional random concave-convex aggregate models with arbitrary gradation, and specimen No.3 (7∶3 coarse to fine aggregate) had the highest percentage of aggregate by volume at 65.5%. The grading curve tended to be close to Fuller's ideal state when the ratio of coarse and fine aggregates was 7∶3 (No.3) and 6∶4 (No.4); the 28 d compressive strength of specimen No.3 reached 9.7 MPa, which is 38.20%−41.40% higher than that of specimens No.5 and No.6; leaching indexes of No.3 complied with the limits of GB 5085.3—2007 and GB/T 14848—2017. The optimum paste thickness of C5 and C10 slurry was 14−20 μm and 23−32 μm respectively, and the 28 d compressive strength meets the engineering standard. The optimization of aggregate proportion based on Python-Abaqus and the synergistic effect of mining water to prepare gangue-mining water filling material will provide a technical paradigm for the low-carbon preparation of solid-waste-based filling materials and the utilization of water resources in mines; and open up the idea of static numerical simulation of gangue-mining water filling materials in the later stage.