Genetic Algorithm Based Design of Bearing Thrust Block for Horizontally Bent Ductile Iron Pipes

The paper is aimed at optimizing bearing thrust block design for thrust restraint using Genetic Algorithm. The design was achieved by calculating the bearing Area from the design equation and then breaking the area in to the corresponding bearing width and height (b and h). The optimization was achieved by obtaining optimum values of the design variables (bearing width and height) that would give minimum volume of the truncated pyramid concrete block at the minimum bearing Area.Optimum values of the design variables obtained from genetic algorithm were successfully compared with the different values of the design variables obtained using the traditional Iterative approach. Five different iterates were compared with the Genetic Algorithm result for different soils. The volume was found to increase up to 95%, 90.97%, 74.55%, 59.72%, 33.77% and 119.05% for soft clay, silt, sandy silt, sand, sandy clay and hard clay respectively in contrast to Genetic Algorithm result. Optimum solution was found when the block height (h) varied between one and two times the block width (b) except for hard clay in which the block width was between one and two times the block height (h).