A Study on the Formation Mechanism and Calculation Method of Surrounding Rock Pressure in Shallow-buried Loess Tunnel Considering the Influence of Vertical Joints

Vertical joints have significant effects on the stability of shallow-buried loess tunnel. Due to the presence of vertical joints distributed widely in loess, the formation mechanisms of surrounding rock pressure on shallow-buried loess tunnel and its computational method are far different from those in other types of soil tunnel. In this study, the distribution characteristics of surrounding rock pressure in shallow-buried loess tunnel were investigated, and discrete element software 3DEC was used to analyze the formation mechanism and development process of surrounding rock pressure in loess tunnel under the influence of joints. We correlated the stratum deformation characteristic with the frictional force caused by the mutual dislocation between joints, and revised the semi-empirical method (Chinese Code methods) in China. The results show that the measured data of radial surrounding rock pressure of shallow-buried loess tunnel are generally between 0-600 kPa, and 80% in the range of 0-200 kPa. The surrounding rock pressure increases with the burial depth, while the distribution of lateral pressure coefficient is relatively discrete. The existence of loess vertical joints (LVJs) aggravates the uneven distribution of the vertical pressure above the tunnel, and the reduction of the lateral earth pressure coefficient at rest under the influence of joints leads to an increase in pressure near the arch shoulder, these two constitute the essence of the joints action on the surrounding rock pressure. The surrounding rock pressure of the shallow-buried loess tunnel reaches the peak stress at about 0.5D from the tunnel face (D is the tunnel span), which is about 1.04-1.32 times of the initial stress. After the excavation of the upper step, the stress of the tunnel crown monitoring point decreased to about 60% of the initial stress. Compared with ignoring the influence of joints, the time of reaching the peak stress at the arch shoulder monitoring point is earlier, and the final stress release degree is smaller. The comparison of the analytical results (proposed method) with the numerical results and the field data exhibits good agreement, proving the proposed method's correctness.