Normal 2000

During the last decade novel computational methods have been introduced in a number of sectors of the engineering science. After proving their abilities in pattern recognition, data clustering, signal processing etc, artificial neural networks (ANN) and genetic algorithms (GA) have been initially used in engineering as tools for optimization of design methods. Later on, fuzzy logic methods have been used to model various highly complex and nonlinear systems based on a set of sample data and fuzzy “if-then rules”. A fuzzy inference system can model the qualitative aspects of human knowledge without employing any quantitative analyses. In the present work an Adaptive neuro-fuzzy inference system (ANFIS) is used in order to model the fatigue behaviour of composite laminates under different loading conditions. Several parameters that affect the predicting accuracy of the proposed technique are examined. Fatigue behaviour of all material configurations and loading patterns is adequately modelled by the aforementioned modelling technique, by using only a limited portion of the available experimental data.