Modeling the relationship between hardness and spray cooling parameters for pinion shafts using a neuro-fuzzy model strategy

Abstract The production of pinion shafts having a predefined hardness distribution faces the challenge of analytically describing the relationship between pinion shafts hardness and spray cooling parameters. The present work attempts to establish this type of mutual relationship based on radial basis function (RBF) neural networks regarding precision forged and spray-cooled pinion shafts. A method for increasing the ability of generalising RBF-Networks via fuzzy inputs is discussed. The results indicate that the relationship between the pinion shaft's hardness and spray cooling has been obtained by using the proposed network. The optimum process parameters can be selected to achieve the desired hardness profile on the basis of the predicted results.