Genetic algorithm based optimum design of nonlinear planar steel frames with various semi- rigid connections

Abstract A genetic algorithm based optimum design method is presented for nonlinear multistorey steel frames with semi-rigid connections. The design algorithm obtains optimum frame by selecting appropriate sections from standard steel section tables while satisfying the serviceability and strength limitations specified in BS5950. The algorithm accounts for the effect of the flexibility of the connections and the geometric non-linearity of the members. The semi-rigid connections are modeled with the Frye–Morris polynomial model. The values of the coefficients, such as the diameter of bolts, the gauge and the geometric dimensions of angles used in the standardization constants are obtained by designing each connection in the frame during the optimum design cycles. The effective length factors for columns, which are flexibly connected to beams, are obtained from the solution of the nonlinear interaction equation. Several steel frames with different beam to column connections, such as extended end plate, and top and seat angle with and without web cleat, are designed using the algorithm. Each design is carried out twice, with and without considering the geometric non-linearity. Comparison of optimum frames has shown that consideration of geometric non-linearity results in greater economy. It is also noticed that taking the realistic behavior of beam–column connection into account produces more appropriate and in some cases even lighter designs.