The work examined how the spatiotemporal evolution of the coal-rock interface morphology affected the top coal caving ratio. The geological conditions of the 3307 fully mechanized caving face at Tang'an Coal Mine served as the research context. PFC2D simulations were conducted to analyze the spatiotemporal evolution of coal-rock interface morphology during mining advancement. Simulations covered the caving intervals of 0.8, 1.6, and 2.4 m, with a mining height of 3 m and the top coal thickness of 3 m at a mining-to-caving ratio of 1:1. A quantitative relationship was established between the initial and final caving boundaries and coal-rock interface morphology. This revealed the intrinsic connections between the morphology, coal-rock movement, and the distribution of residual coal in the goaf. The influence of differences in coal-rock interface morphology on the top coal caving ratio was determined. The dynamic changes in the initial caving boundary (theta(1)) occurred over time and space for caving intervals of 0.8, 1.6, and 2.4 m at a mining-to-caving ratio of 1:1. The initial caving boundary (theta(1)) was categorized into the types A(I) similar to A(II), B-I similar to B-II,B- and C-I similar to C-II based on its differences. The initial caving boundary (theta(1)) variation pattern determined the final caving boundary (theta(2)) and corresponding coal-rock interface morphology. Initial caving boundary angle theta(1) for types I and II decreased when the coal caving interval increased. For the same interval, initial boundary theta(1) for types I and II followed the pattern: theta(I-1) > theta(II-1). The corresponding change pattern for final boundary theta(2) was as follows. When theta(1) > 90 degrees, theta(1) approximate to theta(2); when theta(1) < 90 degrees, theta(1) < theta(2). For initial boundary theta(1) > 90 degrees in A(I), A(II), and B-I, N-1 parallel to N-2, and the morphological differences between types A(I) and A(II) were minimal. Morphological changes were gradual, which resulted in minor fluctuations in the top coal caving volume. For initial boundary theta(1) < 90 degrees in B-II, C-I, and C-II, theta(1) < theta(2). The difference between types I-II of B-I similar to B-II and C-I similar to C-II became more pronounced as the interval length increased. The difference between the peak and trough corresponding to types I similar to II increased. The sinusoidal periodic change pattern in top coal caving volume became more evident, which decreased the top coal recovery rate. The minimal morphological differences and gentle spatiotemporal distribution of type-A(I) and -A(II) coal-rock interfaces offered new insights for intelligent coal caving control and applications in this mine or similar mines.
