Microbially Induced Calcite Precipitation (MICP) is an emerging technology for bioremediation and soil consolidation. However, its pore-scale mineralization process is governed by multiple complex factors (e.g., microbial species, urea concentration, and evolutionary distribution of pore vortex structures), exhibiting strong nonlinearity and spatiotemporal evolutionary uncertainties. Experimental investigations into the underlying microscale mechanisms remain challenging, primarily due to the difficulty in real-time monitoring of coupled reaction-flow processes and dynamic pore structure changes within porous media. To address this gap, a pore-scale multi-physics coupled numerical model for MICP reactive mass transport in evolving porous media has been established by integrating the Lattice Boltzmann Method (LBM), Eulerian Finite Element Method (FEM), and Cellular Automaton (CA). A Physics-Informed Neural Networks (PINNs) approach is incorporated to achieve high-spatial-resolution multi-resolution numerical simulations. Combined with GPU parallel acceleration technology, an in-house general multi-physics solver for three-dimensional (3D) coupled flow-reaction processes in complex porous media has been developed. This solver enables full-scale simulation of mineralization processes in 3D microfluidic chips and accurately reproduces the experimental phenomena observed in chip tests. Based on the validated model, the distribution characteristics of calcium carbonate precipitation and the effects of initial pore structures on its evolutionary process have been quantitatively analyzed, providing quantitative recommendations and predictive tools for optimizing bio-grouting strategies. Furthermore, the mechanism of vortex phenomena induced by dynamic pore structure evolution and their impacts on mineralization uniformity have been explored. Using the Liutex vortex identification method, quantitative analysis of vorticity evolution and distribution has been conducted to investigate the correlations among vorticity, calcium carbonate production, and solute mixing degree. This study preliminarily and quantitatively reveals the dynamic coupling mechanism between vortex flows and reaction processes in the pore-scale MICP system of porous media.