Abstract: This study addresses the abnormal failures, such as blade fractures and wear, encountered by disc cutters when excavating hard rock with high uniaxial compressive strength. It is crucial to understand how disc cutter geometric parameters affect excavation efficiency and load, especially when performing disc cutter excavation on hard rock. We propose a rock modeling approach that combines a cohesive model with a solid model, leveraging the characteristics of a cohesive element. The model follows the double linear constituted evolution criterion and the Benzeggagh–Kenane damage failure criteria. Calibration of the cohesive model’s microscopic parameters was based on the uniaxial compressive strength test and Brazilian splitting tests, using stress reduction rate curves and crack morphology from these experiments. We developed a simulation model for rock breaking during disc cutter excavation and validated it through linear cutting tests. The study examines linear cutting test results and the simulation of rock breaking during disc cutter excavation results. This study examines how disc cutter width affects rock breaking by considering load curves, rock breaking volume, and specific energy. Results indicate that the cohesive constitutive model effectively simulates rock crack initiation and propagation. The rock breaking process comprises three stages: elastic, crack initiation–propagation, and rock unloading. The crack initiation–propagation stage is critical for rock breaking efficiency. Results show that disc cutter load and rock break volume increase with increasing penetration, blade width, and cutter spacing. Specific energy initially decreases with blade width increase and then increases. Optimal excavation efficiency occurs with a blade width of 13 mm and a spacing of 80 mm, whereas a 7-mm blade fails to achieve complete rock fragmentation, resulting in low efficiency.