Please use this identifier to cite or link to this item: https://scidar.kg.ac.rs/handle/123456789/11056
Title: Determination of tribological properties of aluminum cylinder by application of Taguchi method and ANN-based model
Authors: Milojević, Saša
Stojanovic, Blaza
Issue Date: 2018
Abstract: © 2018, The Brazilian Society of Mechanical Sciences and Engineering. Energy losses due to friction and wear in reciprocating machines could be potentially reduced by applying new surface, materials and lubrication technologies for friction reduction and wear protection. In this paper, the tribological properties of ferrous-based reinforcements are tested and compared with aluminum alloy (EN AlSi10Mg) as a base material for cylinder of air compressor. The ball-on-plate CSM tribometer is used to carry out these tests under dry sliding conditions and constant sliding distance, for three different values of sliding speed and normal load. The wear factor is analyzed by using Taguchi method as well as artificial neural network-based model, with the aim of finding the optimal parameters. The result of signal-to-noise ratio and analysis of variance shows that the best tribological properties were achieved with reinforcements. Material has the greatest impact on the wear factor (35.54%), followed by load (22.16%) and sliding speed (6.01%). A good superposition was reached of the results obtained by the Taguchi method with the results of the artificial neural network-based model. According to the analysis of surface micrographs, it can close that the bonding material is the most dominant mechanism of wear for both tested materials.
URI: https://scidar.kg.ac.rs/handle/123456789/11056
Type: article
DOI: 10.1007/s40430-018-1495-8
ISSN: 1678-5878
SCOPUS: 2-s2.0-85057053262
Appears in Collections:Faculty of Engineering, Kragujevac

Page views(s)

151

Downloads(s)

24

Files in This Item:
File Description SizeFormat 
10.1007-s40430-018-1495-8.pdf2.05 MBAdobe PDFThumbnail
View/Open


This item is licensed under a Creative Commons License Creative Commons