Snap-Fıts Desıgn And Modelıng Wıth Artıfıcıal Neural Networks

Thesis Type: Postgraduate

Institution Of The Thesis: Gazi Üniversitesi, Fen Bilimleri Enstitüsü, Turkey

Approval Date: 2019




Snap-fits with easy assembly and disassembly are one of the most widely used and manufactured components in the industry, providing advantages in terms of ease of production. In the scope of this study, the stress, strain and deformations of the 6 types of snap-fits, C, I, L, S, T and V types, were investigated. This fasteners obtained by using 3 different materials such as PLA (Polylactic Acid), ABS (Acrylonitrile Butadiene Styrene) and PET-G (Polyethylene Terephthalate-Glycol) were dimensioned parametrically. Experimental sets were created by using the Taguchi method so that the number of experiments for each element was reduced to 54 experimental sets and the test was carried out. Analysis data using ANSYS software were analyzed with ANOVA analysis to determine whether or not to create an equation and effective parameters were determined. ANOVA analysis results were evaluated and modeled with Artificial Neural Network by using the data obtained with ANSYS writing. With the model created, an algorithm has been formed which has the ability to estimate the range of various sizes and types of materials