Grinding performance of bearing steel using MQL under different dilutions and wheel cleaning for green manufacture

Abstract

Cutting fluids have a fundamental role in machining, especially in grinding, since they are responsible for lubrication and refrigeration of the cutting region. On the other hand, they may cause damage to the environment and to the health of the operators. Thermal damage, geometric inaccuracies and degradation of the surface integrity of the workpiece are diminished with the use of cutting fluids. While in the conventional application of cutting fluid a large amount of oil or emulsion of oil with water is used, in the minimum quantity lubrication (MQL) technique a small amount of oil is applied via a jet of compressed air reaching, in many cases, results comparable to the conventional method. However, the lower cooling capacity of MQL and the clogging of the wheel pores caused by the hot chips are obstacles to be overcome. This work evaluates the addition of water to the MQL method in the external cylindrical plunge grinding of AISI 52100 steel to improve the deficiencies of this technique. The results analyzed include surface roughness (Ra), workpiece surface damages, roundness deviation, grinding wheel wear, grinding power, acoustic emission, workpiece subsurface microhardness and microstructure for conventional method, MQL with pure oil (1:0) and MQL with water in 1:1, 1:3 and 1:5 oil-water proportions with and without grinding wheel cleaning (a jet of compressed air towards the wheel surface to remove the chips which clog the wheel pores). MQL with pure oil produced the worst results in this work, but dilution of oil in water benefited this lubricant-refrigerant technique. In addition, the diluted MQL 1:5 associated with wheel cleaning jet (WCJ) was the alternative method that most closely approximated the results of the conventional technique, indicating the potential for using MQL with WCJ widely in the industry. However, it is necessary to continue researching this technique in order it could even outperform the conventional method in all its output variables. (C) 2020 Elsevier Ltd. All rights reserved.