Failure modelling in pin-loaded joints using an adaptive neuro-fuzzy approach
This data was imported from Scopus:
The aim of this study is to investigate the performance of Adaptive Network-based Fuzzy Inference System (ANFIS) for failure prediction in composite and aluminium joints. Accurate prediction of failure in pin-loaded joints is crucial to optimising the design of a structure. This paper proposes a new approach, which is derived from soft computing, to investigate failure load prediction in a CFRP (Carbon Fibre Reinforced Plastic) composite (and aluminium) plate with a circular hole, which is subjected to a tensile load by a pin. To evaluate the results of mentioned method, parametric studies were performed experimentally. The end distance to diameter (E/D) ratio was changed from 0.7 to 3. Data sets from 39 different carbon-fibre-reinforced plastics and 40 aluminium joints have been used to develop two ANFIS training models for the prediction of failure. It has been found that ANFIS can be trained to model tensile failure at least as well as the other classical classifier (C4.5) regarding the composite joints. However, it could outperform the classical methods for aluminium joints. The procedures developed in this work could be used in design with little further modification. © 2006 Civil-Comp Press.