Data mining approach for selection of plateau honing parameters with dual objectives

摘要

Surface topography of cylinder liners plays an important role in terms of functionality of the engine. The surface is finished by honing process in various stages namely rough, finish and finally plateau honing to have proper plateau portion (bearing area) at the top to resist wear and grooves at desired honing angle to retain lubrication during engine operation. Selection of parameters and their levels in plateau honing operations to achieve required surface topography is found to be very difficult. For the first time, data mining approach is used in this investigation to select significant honing parameters and their levels considering two important bearing area parameters namely core roughness depth (R_k) and material portion in the valley zone (M_(12)) together. Data mining model was trained and tested using experimental data. The significant input plateau honing parameters have been identified and graphical representation of the decision tree has been also presented for easier understanding.