Optimization of Steel Truss Girders in Pedestrian Bridges - Using Genetic Algorithm
Publicerad
Författare
Typ
Examensarbete för masterexamen
Master's Thesis
Master's Thesis
Modellbyggare
Tidskriftstitel
ISSN
Volymtitel
Utgivare
Sammanfattning
The optimization of any structure is an iterative process, where the Structural Engi neer evaluates how certain changes in one or more design variables affect the outcome
of the design. With increased number of variables, the design space increases expo nentially and it is not feasible to remodel the structure each time and analyze these
combinations to choose the best solutions. One way to overcome this problem is by parametric design, where the structure is programmed and governed by a number
of parameters so that any change in these parameters would result in a change in this structure. Although this approach saves time when it comes to the modelingprocess,
the designer still needs to run the analysis each time and document theresults to select the best set of parameters that fulfills the goals of the design. To
overcome this limitation, the implementation of genetic algorithms allow for a fasterexploration in the design space to find the most optimal solution.
In this thesis, the steel truss-girders used in single-span pedestrian bridges are stud ied. A parametric design model is developed to automatically produce these trusses
based on a set of parameters that govern the topology of the truss. Furthermore, afinite element analysis is conducted and the necessary design according to Eurocode
and the regulations of the Swedish Traffic Administration is done. Finally, a ge netic algorithm optimization tool is used to minimize the weight of the truss while
fulfilling all the checks.The study indicates that truss toplogies with bent upper chord have lower weights and this weight difference is increasing with longer spans. Furthermore, it is seen
that the diagonals and the upper chord stands for about 70% of the total weight ofthe truss and yet not fully utilized due to their contribution for the capacity against
global buckling which is the most critical check in the design. In addition, usinghigher steel grade resulted in lighter trusses. However, this effect decreased with
longer spans. Global buckling is more critical for trusses with higher steel grades,because the members are only available in cold-formed cross-sections.
Beskrivning
Ämne/nyckelord
Steel truss pedestrian bridge, Parametric design, Genetic algorithm, Python, Truss optimization, Topology Optimization