Ground control remains a critical challenge in underground mining, where the structural integrity of stope boundaries directly influences safety and resource recovery. Traditional empirical and analytical support design methods often rely on over simplified assumptions, failing to capture the complex, non-linear behaviour of rock masses subjected to deep, high-stress environments or structural discontinuities. This paper presents an advanced numerical modelling framework aimed at optimizing stope support design by integrating high-fidelity simulations with site-specific geotechnical data. Utilizing a combination of three-dimensional continuum and discontinuum modelling techniques specifically Finite Element Method (FEM) and Discrete Element Method (DEM) this study simulates the progressive failure mechanisms of rock walls and backfill interactions under dynamic mining sequences.

The proposed methodology explicitly incorporates anisotropic rock mass properties, discrete fracture networks (DFNs), and transient stress redistributions induced by adjacent blasting and sequential extraction. By executing parametric sensitivity analyses, the models evaluate the performance of various ground support systems, including cable bolts, shotcrete, and rock bolts, under varying stress regimes. The optimization protocol utilizes a machine learning-driven response surface methodology to balance structural reliability against implementation costs, defining optimal bolt spacing, length, and anchoring strategies.

Results indicate that advanced modelling reduces conservative over-design by up to 22% while identifying localized damage zones that empirical methods routinely overlook. Specifically, the discontinuum models successfully predict key block raveling and wedge instability along dominant joint sets, allowing for proactively targeted reinforcement. Ultimately, this integrated modelling approach provides a robust, data-driven engineering tool that enhances underground safety, minimizes dilution, and optimizes support costs in complex geological settings.

Keywords: Stope support optimization, Advanced numerical modelling, Ground control, Discrete Element Method (DEM), Rock mass behaviour, Underground mining geomechanics