Date of Award

Summer 2004

Document Type

Thesis - Restricted

Degree Name

Master of Science (MS)


Civil, Construction, and Environmental Engineering


Engineers consider straightforward calculations of capacity during the design process, and then subject the design to serviceability checks. For a steel framed floor system, these checks often include deflection and vibration. Floor systems have become lighter due to advances in material science and construction methods. While these advances help to reduce cost, they can result in structures that are susceptible to vibration problems that may cause discomfort for the building's occupants. The vibration response can be determined through numerous calculations that are heavily dependent on the entire design making design for vibration difficult. This thesis proposes a methodology for designing optimized systems utilizing a genetic algorithm (GA), which utilizes a search strategy that is modeled on the same mechanisms found in genetic evolution. The GA operates with a selected population of solutions. For each of these solutions, the capacity, cost and vibration response can be determined and used to evaluate each system's performance. Using a "survival of the fittest" methodology, the GA combines components of the individual systems to create a new generation (new population). This new population is evaluated, and the process cycles again. Through repetition, the optimal solution will present itself. Steel systems utilize a finite number of steel wide-flange sections, joists, decks and concrete thicknesses. These discrete design options are perfectly suited to the genetic algorithm. The GA has been used successfully in the past to obtain optimum or near optimum solutions to many design problems. The successful record of the GA with steel framed structures shows that the GA will be well suited to the problem of designing steel floor framing systems.



Restricted Access Item

Having trouble?