Ph.D. Tezi Görüntüleme | |||||||||||||||||||||
|
|
||||||||||||||||||||
| Summary: In this study, lateral buckling of I beams having different supports and loads with lateral and torsional bracings and ideal spring constants of bracings were investigated and Artificial Neural Networks (ANN) method was applied to the lateral buckling load calculation of lateral bracing I beams by using obtained results. Thin-walled I beams are structural elements with very small torsional rigidities although their flexural rigidities are sufficiently big. Due to these characteristics, thin-walled I beams encounter a lateral buckling problem. The load applied to the beam cross-section and the support condition of the beam significantly affect lateral buckling load.Artificial intelligence method, which appeared in the second half of the XXth century and on which a great number of people study nowadays, was successfully used in the analyses of many problems and turned to be an alternative to classical programming. By reason of revealing the proved theories of ANN and other logical programming methods, the subject aroused numerous scientists’ interest and was applied to various problems. This study consists of three main chapters. Literature review, some brief introductory information about lateral buckling behaviour and ANN was given in the first chapter. In the second chapter, the studies carried out, findings and observation were presented. Firstly, the lateral buckling behaviour was analysed by the help of BASP program based on finite element model according to the load and support condition applied to thin-walled I beams. Secondly, the effect of lateral and torsional bracings to lateral buckling behaviour of I beams was studied by the help of BASP program, in this way determination of ideal spring constants of bracings was aimed. And finally, ANN approach was investigated according to cross-section properties of beam, beam length and spring constant and location of bracing in lateral buckling problems of thin-walled I beams with lateral bracing. In solution with ANN, computer programming based on supervised learning rule algorithm was used. It is seen that ANN produced appropriate solutions in lateral buckling problem. In the last chapter of the study, the results obtained and recommendations were summarised.Key Words: Lateral Buckling, Lateral Bracing, Torsional Bracing, BASP, Artificial Neural Networks
| |||||||||||||||||||||