Abstract:To address the issue that the traditional ellipsoid model cannot accurately describe the complex morphology of the draw body (characterized by a coarse upper part and a fine lower part) during ore drawing in a single drift with a large-scale structure, this study established a large-scale physical ore drawing model to conduct single-drift ore drawing experiments, based on the predicted distribution of blasted fragment sizes of actual ore-rock samples from a mine. By arranging marker particles in layers and recording the ore drawing process, a butted model of double semi-ellipsoids of revolution was innovatively adopted to characterize the morphology of the draw body, and the Levenberg-Marquardt least squares algorithm was applied for high-precision fitting (with correlation coefficients generally > 95% ). The experimental results show that: ① The volume of the draw body (V) exhibits a power function relationship with the ore drawing height (h) and the drift opening width (w), expressed as V = 30. 42 × (h / w)2. 626 (R 2 = 0. 999 2), and the stability is poor when the relative height h / w < 4; ② The ore drawing height (h) shows strong linear relationships with the major semi-axis (a) and minor semi-axis (b) of the draw body, which are a = 0. 513 6h-5. 043 (R 2 = 0. 999 3) and b = 0. 090 8h + 3. 627 3 (R 2 = 0. 999 2) respectively; ③ The dilution rate (p) increases linearly with the drawn volume (V), following the relationship p = 8. 941 × 10-5 V-32. 38 (R 2 = 0. 913 3). Furthermore, a scale conversion method for on-site engineering application was proposed. Based on the geometric similarity criterion, the concept of dimensionless volume was introduced to realize the conversion of experimental data to the on-site scale, ensuring the effective application of theoretical results in practical scenarios. This study quantitatively reveals the morphological characteristics of the draw body and the intrinsic relationships between its key parameters under the condition of single drift with a large-scale structure. The established mathematical models can be directly used to predict the draw body volume and expected dilution rate under specific ore drawing heights and drift widths, providing an important theoretical basis and operational guidance for on-site precision control of ore drawing processes, optimization of ore recovery, and reduction of dilution.