Abstract: The dynamic derivation of dense core at the cutter bottom seriously affects the rock-breaking mechanism of disc cutter (hereinafter referred to as cutter), and then affects the tunneling efficiency of full face rock tunnel boring machine (TBM) to a certain extent. Referring to the H. Alehossein blunt wedge indenter rock penetration model, the derivation phenomenon of dense core and the complex stress field around the dense core area is investigted, and a constant cross-section flat edge disc cutter (hereinafter referred to as cutter) rock penetration model is established based on the cavity expansion theory and discrete modeling idea. Then, taking the vertical force and dimensionless average contact stress as the performance evaluation indices, and selecting the geometric dimension parameters of the blade section of the cutter ring and the physical and mechanical property parameters of the rock as the variables, the ten-factor and two-level L₁₂ (2¹¹) orthogonal tests are carried out. It is found that the blade width and penetration depth are the most sensitive factors. After that, the change characteristics of dimensionless contact stress at the cutter ring bottom are analyzed. It is found that because the dense core phenomenon is considered in the model, the stress of the rock under the cutter ring bottom is significantly greater than the uniaxial compressive strength, which is more in line with the engineering practice. Finally, two kinds of rock samples with different particle sizes are selected, four kinds of indenters are prepared, and the verification experiment of the theoretical model is carried out using the TRW-3000 rock breaking tester. The results show that both the cutter width and the penetration depth are sensitive factors for vertical force and dimensionless average contact stress. Because the dense core phenomenon is included in the assessment, the rock stress level at the cutter bottom is obviously greater than the uniaxial compressive strength under the intrusion of indenter. The maximum prediction error of the proposed model is less than 10%, which shows that it has good accuracy.