Neural Network Analysis for Erosion Wear of Nickel-Aluminide Coatings on Steel by Plasma Spraying

Abstract

In the present investigation plasma spray inter metallic coating of Nickel-aluminide was deposited on mild steel substrates. The response of plasma sprayed nickel-aluminide coatings to the impingement of such solid particles has been presented in this work. Nickel pre-mixed with alumina powder is deposited on mild steel substances by atmospheric plasma spraying at various operating power level. The coatings are subjected to erosion wear test. An erosion test setup developed in our laboratory is used to simulate real time erosive situations. Dry silica sand of average particle size 400 micron is used as the erodent. The erosion rate is calculated on the basis of 'coating mass losses. The erosion studies are made and different velocities and impingement angles. A computational technique (ANN analysis) is used to predict the rate of erosion wear under various operational conditions. This technique involves database training to predict property parameter evolutions in process having large number of interdependent variables. This paper presents the database construction, implementation protocol and also the set of predicted results related to the erosion wear rate of nickel-aluminide coating. It is shown that the erosion wear is strongly influenced by the angle of impact. The test is conducted at room temperature i.e.27°C and 60% RH. Ni3Al coatings deposited at different power levels (10, 12, 16, 20, 24 kW) are found to exhibit different wear rate under similar test conditions.