Optimum Design of Tunnel Supporting Systems Using Genetic Algorithm

Stability of underground spaces is one of the most important tasks in tunnel construction. With the passage of time, many support systems have been invented and evolved for tunnel stability. The support system in a rock tunnel can be one, or the combination of these options: rock reinforcement (i.e., rock dowels, rock bolts, rock anchors, etc.), steel ribs, precast concrete segments, lattice girders, cast in place concrete, shotcrete and etc. Today, rock bolts are very common in tunnel construction and they have various types and the methods of their installation are developed very much. One of their advantages is that they can be installed immediately after excavation. In conventional methods for rock bolt design, at first, arbitrary values for rock bolts diameter, rock bolts spacing in circumference and length of tunnel, and length of rock bolts are assumed. According to these values, the axial force in rock bolt and tunnel deformation are calculated and compared with available values. In this study, an algorithm for optimum design of rock bolt support has been introduced to minimize the cost of these supports. For rock bolt supporting system, diameter, length and arrangement of rock bolts were considered as design variables. The objective function is the cost minimization of supporting system. With respect to optimization problem Genetic Algorithm was used for minimizing objective function and finding optimum values for design variables. Finally, a case illustrates how the procedure was successfully used in the design of the tunnel and how much was saved with optimization of conventional method.