Deformation prediction and analysis of underground mining during stacking of dry gangue in open-pit based on response surface methodology

Abstract

Deformation prediction and the analysis of underground goaf are important to the safe and efficient recovery of residual ore when shifting from open-pit mining to underground mining. To address the comprehensive problem of stability in the double mined-out area of the Tong-Lv-Shan (TLS) mine, which employed the dry stacked gangue technology, this paper applies the function fitting theory and a regression analysis method to screen the sensitive interval of four influencing factors based on single-factor experiments and the numerical simulation software FLAC^3D. The influencing factors of the TLS mine consist of the column thickness (d), gob area span (D), boundary pillar thickness (h) and height of tailing gangue (H). The fitting degree between the four factors and the displacement of the gob roof (W) is reasonable because the correlation coefficient (R²) is greater than 0.9701. After establishing 29 groups that satisfy the principles of Box-Behnken design (BBD), the dry gangue tailings process was re-simulated for the selected sensitive interval. Using a combination of an analysis of variance (ANOVA), regression equations and a significance analysis, the prediction results of the response surface methodology (RSM) show that the significant degree for the stability of the mined-out area for the factors satisfies the relationship of h>D>d>H. The importance of the four factors cannot be disregarded in a comparison of the prediction results of the engineering test stope in the TLS mine. By comparing the data of monitoring points and function prediction, the proposed method has shown promising results, and the prediction accuracy of RSM model is acceptable. The relative errors of the two test stopes are 1.67% and 3.85%, respectively, which yield satisfactory reliability and reference values for the mines.

摘要

科学分析露天坑下采空区稳定性是实现地下残矿安全高效回采的关键。 在对铜绿山矿地下残矿进行回采的同时, 利用上部露天坑回填干堆尾砂, 为研究该复杂条件下的采场变形问题, 采用单因素分析和 FLAC^3D 数值软件, 运用函数拟合和回归分析将间柱宽度 d、 空区跨度 D、 境界顶柱厚度 h、 尾砂堆高 H 等 4 个影响因素对空区位移的影响进行敏感区间筛选, 结果表明 4 个因子与位移量 W 的拟合度较高, 相关系数 R²≥0.9701。 建立 29 组满足 Box-Behnken 中心组合设计原则的响应面数组, 对敏感区间下的尾砂堆积过程进行模拟, 并结合方差、 回归拟合及显著性分析, 得到影响采空区稳定性的显著程度依次为 h>D>d&H, 且 4 个均为不可忽视的重要因素。 最后, 将预测模型对工程实验采场变形情况进行预测分析及误差验证。 验证结果表明, 函数预测较好, 相对误差分别达到 1.67%、 3.85%, 具有较强的参考价值和应用可靠性。