Tweet Share

• Send by email

  Send by email

  

  Numerical investigation of the interaction of parallel twin tunnels during excavation and support

  The construction of twin tunnels is a prevailing practice in either conventional or mechanized shield-driven tunnelling for security and functionality. Despite that, existing methodologies for tunnel design usually focus on single tunnels, neglecting the potential interaction between neighbouring tunnels. Such interaction can have significant effects, especially in closely spaced twin tunnels. This paper investigates the interaction between deep parallel twin tunnels in terms of stress, loads and internal forces on the primary support, examining a wide range of: (i) pillar width values (transversal distance between the axes of the adjacent tunnels), (ii) geotechnical conditions (strength and deformability parameters of rockmass), (iii) stiffness/rigidity of primary support, (iv) overburden height (depth) values, (v) longitudinal distance between the excavation faces of the adjacent tunnels and (vi) lateral earth pressure coefficient/geostatic stress ratio values. Particularly, the investigation focuses on the changes occurring in the shape and the values of the stress distribution, due to the influence of the aforementioned parameters and the interaction effect, leading to the change of loads and internal forces on primary support. The adequacy of the primary support measures is examined and evaluated with the use of capacity limit curves. An additional aspect of the investigation is the determination of the parameter causing the largest influence on the interaction effect and the determination of the critical value of the pillar width that limits the interaction effect. These issues are examined via 3D numerical analyses with the finite element code ABAQUS for circular tunnels excavated and supported with sequential method. The paper assesses the reliability and efficiency of equivalent simplified 2D numerical analyses for the under-investigation issue and the limitations on the applicability of such a method, due to the predominant three-dimensional nature of the problem. The results are presented in terms of time-history stress and normalized parameters that indicate the influence of the examined parameters on the interaction effect.

  Recipient :

  Name ^* :

  E-mail address ^* :

  ^* Required fields

  Send   or  Cancel

• Print