Experimental optimisation of the tribological behaviour of Al/SiC/Gr hybrid composites based on Taguchi’s method and artificial neural network
| Autor: | Ilija Bobić, Aleksandar Vencl, Slavica Miladinović, Blaža Stojanović, Jasmina Skerlić |
|---|---|
| Rok vydání: | 2018 |
| Předmět: |
wear
analysis of variance Materials science friction Alloy Composite number Aerospace Engineering chemistry.chemical_element 02 engineering and technology engineering.material A356 lubricated sliding Industrial and Manufacturing Engineering chemistry.chemical_compound Taguchi methods 0203 mechanical engineering Aluminium Silicon carbide Graphite Composite material Mechanical Engineering Applied Mathematics General Engineering Tribology 021001 nanoscience & nanotechnology compocasting 020303 mechanical engineering & transports chemistry Automotive Engineering engineering Taguchi method hybrid composites 0210 nano-technology artificial neural network Tribometer |
| Zdroj: | Journal of the Brazilian Society of Mechanical Sciences and Engineering |
| ISSN: | 1806-3691 1678-5878 |
| DOI: | 10.1007/s40430-018-1237-y |
| Popis: | This paper presents the investigation of tribological behaviour of aluminium hybrid composites with Al–Si alloy A356 matrix, reinforced with 10 wt% silicon carbide and 0, 1 and 3 wt% graphite (Gr) with the application of Taguchi’s method. Tribological investigations were realized on block-on-disc tribometer under lubricated sliding conditions, at three sliding speeds (0.25, 0.5 and 1 m/s), three normal loads (40, 80 and 120 N) and at sliding distance of 2400 m. Wear rate and coefficient of friction were measured within the research. Analysis of the results was conducted using ANOVA technique, and it showed that the smallest values of wear and friction are observed for hybrid composite containing 3 wt% Gr. The prediction of wear rate and coefficient of friction was performed with the use of artificial neural network (ANN). After training of the ANN, the regression coefficient was obtained and it was equal to 0.98905 for the network with architecture 3-20-30-2. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|
Full text from SpringerLink