PREDICTION OF THE WEAR CHARACTERISTICS OF ZA-27/SiC NANOCOMPOSITES USING THE ARTIFICIAL NEURAL NETWORK

Abstract

The zinc-aluminium casting alloy ZA-27 is well-established and is a frequently used material for plain bearing sleeves. It has good physical, mechanical and tribological properties. Its tribological properties can be improved further by adding hard ceramic particles to the alloy. The tested nanocomposites were produced by the compocasting process with mechanical alloying preprocessing (ball milling). Three different amounts of SiC nanoparticles, with the same average size of 50 nm, were used as reinforcement, i.e. 0.2, 0.3 and 0.5 wt. %. Tests were performed on a block-on- disc tribometer (line contact) under lubricated sliding conditions, at two sliding speeds (0.25 and 1 m/s), two normal loads (40 and 100 N) and a sliding distance of 1000 m. The prediction of wear rate was performed with the use of an artificial neural network (ANN). After training the ANN with architecture 3-4-1, the regression coefficient for the network was 0.99973. The experimental values and values obtained by applying the Taguchi method were compared with the predicted values, showing that ANN is more efficient in predicting wear.