OPTIMIZATION OF SOIL NAILED WALLS BASED ON THE GENETIC ALGORITHM

During recent decades, soil nailing, as a method for increasing the stability of slopes and excavations, has been of interest to engineers, due to its exibility, rapid construction, ease of construction and its economical aspects, compared with other conventional stabilization methods. In other word, soil nailing is a reinforcing method, using the shear strength of in-situ ground and the pullout resistance of soil nails inserted into the ground, and other word inter-act with the ground through grouting. Cost eectiveness is one of advantages of soil nailed walls. Hence, the optimized design of such systems should be undertaken with special care. In this research, decreasing the cost of construction of soil nailed walls, while ensuring their global, sliding and internal stability, is investigated as an optimization problem. A number of factors, such as the inclination, length, size and spacing of the nails, aect the stability of a slope with a specic geometry (i.e. height and slope angle). Furthermore, the in uences of these parameters on the safety factor vary with the shear strength parameters of soils. The abovementioned parameters cause this optimization problem to be multi-objective. Therefore, in order to arrive at a specic factor of safety (F.S.) for a particular slope, utilizing the minimum volume of material, a computer program is written in MATLAB language, in which the Genetic Algorithm is used for optimization. In order to examine the accuracy of the written program, a number of soil nailed wall design examples that have been presented in reliable publications, are redesigned by this program and it is shown that the last design results stand in noticeable savings in the use of the required material, while the desired safety factor is maintained at a constant. Finally, the in uence of some important cost parameters, such as the cost of drilling holes for nails, shotcreting and steel price, and the eect of each item on total cost, are analyzed using this program and compared with each other.