Design and optimization of steel portal frames according to Eurocode using genetic algorithm

University essay from Lunds universitet/Avdelningen för Konstruktionsteknik; Lunds universitet/Väg- och vattenbyggnad (CI)

Author: Ghassan Numan; [2012]

Keywords: Technology and Engineering;

Abstract: One of the essential engineer’s jobs is to achieve the most economical technical solutions. Weight optimization is important since it provides a structure that can carry the applied loads in addition to fulfilling the structural requirements. In this project, a Matlab-algorithm has been developed to find the optimum design of steel portal frames according to “Eurocode 3: Design of steel structures” with regard to the weight. To get the final design of the frame, some inputs have to be specified by the user such as the material, the coordinates, the loads and the distribution pattern of the bracings. The algorithm goes through three essential steps before getting the optimal frame: - Analyzing the frame with help of CALFEM-toolbox: In this step, the frame is geometric nonlinearly analyzed due to certain load combination according to the Ultimate Limit State (ULS) and Serviceability Limit State (SLS) criterion. The internal forces and the displacement established and the axial force diagram, shear force diagram, bending moment diagram and the deformed shape of the frame to be calculated and plotted. - Checking the capacity of the frame according to Eurocode: Some constraints with regard to EC should not be violated. These checks may refer to the frame capacity, checking the capacity against the risk of buckling and the deformations which should not be exceed the Serviceability Limit State (SLS) limitations. - Finding the optimal design of the frame using Genetic Algorithm optimization method: Genetic Algorithm (GA) is an iterative searching method based on the evolution of species’ principle. The algorithm repeat the two steps above every iteration cycle trying to find the best design that has the minimum weight without violating the limitations. Since it is an iterative process, the time required to find the optimum design depends on some factors such as the speed of the computer, the number of variables the size of frame mesh etc. The project ends with some testing examples in which a frame with width span of 20 m, and a height of 6.5 m and uniformly distributed loads as snow loads are acting on the roof of the frame. In these examples, the algorithm has been tested to compare the design that have been got in case of fully braced frame, unbraced and by letting the algorithm to find the optimal number and position of the bracings along the frame. Another comparison has been done to see the difference of the design in case of absence /existing of the deformation limitations.

  AT THIS PAGE YOU CAN DOWNLOAD THE WHOLE ESSAY. (follow the link to the next page)