TA552 : Optimization of 3D Steel Moment frxames baxsed on Modified Reliability Using Genetic Algorithm
Thesis > Central Library of Shahrood University > Civil & Architectural Engineering > PhD > 2020
Authors:
Abstarct: In the last few decades, reducing the time and cost of designing structures and saving on materials has made the issues of structural optimization particularly important. In addition, the reliability of structural systems in recent years has been an issue that has been considered by designers. However, due to the time-consuming optimization baxsed on the reliability of the structural system, many researchers by using conservative and simplistic assumptions have estimated the probability failure of structures baxsed on the probability failure of members. Alternatively, the effect of axial force and bending moments interaction is ignored which has led to poor design and economic results.
The present thesis is baxsed on the development of optimization of three-dimensional steel moment frxames by considering the axial force and biaxial bending moments interaction in plastic hinge formation and tries to obtain new achievements with purposeful steps.
For this purpose, optimization of three-dimensional steel moment frxame using island genetic algorithm and reliability analysis of structural system using the proposed Modified Latin Hypercube Sampling, M-LHS, has been performed. In addition, in order to consider the interaction of the axial force and biaxial bending moments, solutions have been proposed and studied.
In the proposed M-LHS method, the arrangement of intervals is selected so that the largest volume of samples is in ranges with specified distances from the mean value. According to the results, it is suggested that about 70 to 80% of the samples be selected from these ranges.
The optimization results of the three-dimensional moment steel frxames showed that the assumption of plastic hinges in the members due to bending moment alone and regardless of the effect of axial force did not cause much difference with considering the interaction of axial force and biaxial bending moment. In this case, optimal weight of the structure has been increased by only about 4%, taking into account the interaction of internal forces. In addition, the optimal weight obtained from the proposed solution to consider the interaction of axial force and biaxial bending moment, compared to the method of elastic analysis with plastic failure criterion, has increased by about 2.1% in the three-story frxame and 3.3% in the six-story frxame. This could indicate that due to the compliance of the elastic analysis of the structure with the initial yield failure criterion in the proposed solution, the answers are more reliable and the need for time-consuming elasto-plastic analysis has been eliminated.
Keywords:
#Reliability-baxsed optimization #Probability failure of structural system #Interaction of the axial force and biaxial bending moments #Modified Latin Hypercube method #Island genetic algorithm #Three-dimensional steel moment frxames Keeping place: Central Library of Shahrood University
Visitor:
Visitor: