Numerical modeling of complex geological and geotechnical processes involving multiphysics coupling, fluid-solid interactions, and large deformations remains challenging. Traditional continuum-based methods (e.g., finite difference method) lack micro-scale descriptionsn, while existing DEM coupling methods (e.g., DEM-CFD, DEM-LBM) face trade-offs between computational cost and accuracy. To address this issue, A multi-field and fluid-solid coupling method for porous media based on DEM-PNM (pore network model) was proposed within the self-developed DEM platform (MatDEM)[1]. This method dynamically constructs the pore network through spatial contact relationships between discrete particles in the particle assembly, enabling simulation of low-speed Darcy flow in the pores. The pore pressure is determined by fluid density and temperature, while permeability depends on macroscopic calibration and contact relationships. A bidirectional coupling model is employed for the fluid-solid interaction. The method utilizes full matrix computation, significantly optimizing computational efficiency for large-scale simulations. This method has been tested and applied mainly in three aspects: 1. By establishing convection and diffusion equations for multi-physical field parameters in the pore network, the migration of temperature and pollutants in porous media were succesfully simulated[1]; 2. Fluid-solid coupling simulation: the hydraulic fracturing under a 20 MPa constraint was simulated, and the fracture development pattern closely matched real experiments[2], while also simulating fracture branching in the metamorphic rock layers produced by magma intrusion on a geological scale; 3. Fine particle movement under a wide gradation condition: In a coarse particle fluid domain, when fine particles migrate within the network formed by coarse particles, they are mainly influenced by hydrodynamics applied by the fluid in addition to collisions between particles. Based on this method, phenomena such as piping and sand injection were simulated[3].