Automated Design of Steel Wide-Flanged Beam Floor Framing Systems Using a Genetic Algorithm

The design of steel floor framing systems involves an optimization problem. This problem involves choosing the structural elements used to construct the system so that strength, deflection, and performance constraints are met, while achieving the lowest cost possible for the structure. Traditionally, the design would be completed by an engineer using a trial-and-error, iterative procedure according to design specifications and guided by the experience of the designer. This process can be lengthy and is not guaranteed to yield an optimal solution to the problem. A better design tool is seen in the use of a genetic algorithm-based program that performs the design automatically. The genetic algorithm (GA) is a search technique modeled after the principles of survival of the fittest and adaptation. The GA does not explore every combination of discrete design variables, but rather uses a small population of individual solutions that improve as they adapt to fitness parameters over the course of 30 to 40 generations. The GA is a fast and efficient way to solve the floor framing problem which involves a large number of discrete design variables. A genetic algorithm-based program will be developed and implemented. It will have the ability to choose beam and girder sizes, deck and concrete slab characteristics, shear connector configurations, and floor panel layout based on simple user input. This user input involves floor panel dimensions as well as superimposed dead and live loading conditions. An output file is prepared by the GA for the program user detailing relevant data of least-cost solutions. Conclusions are made as to the influence of cost variations and floor panel geometry ratios on the most-economical solution. Suggestions are presented for building designers and researchers wishing to reduce the panel cost and improve vibration performance. Recommendations for future applications of the GA are provided along with suggestions for improving its performance