The objective of the grinding process is to obtain a better surface finish (roughness, accuracy, etc.) or a high material removal rate of the workpiece. These grinding process response variables depend largely on both the grinding conditions and the topography of the grinding wheels formed during dressing. In this regard, the paper focuses on determining the optimal dressing conditions for grinding wheels dressed with diamond roller dressers during rough and finish grinding. A new multi-objective optimization approach for determining dressing conditions has been proposed, which leads to Pareto-optimal solutions for increasing the grinding process production rate, reducing roughness, and increasing the accuracy of ground surfaces. To demonstrate the performance of this approach, one process of dressing electrocorundum grinding wheels with novel diamond rollers made of medium- and high-strength synthetic diamonds with mixed grain sizes has been selected. Empirical models have been developed to examine the production rate of the grinding process, the roughness of the ground surfaces, and the accuracy of the machined parts. Multi-objective optimization based on a genetic algorithm has been performed by applying two methods: determining an optimal compromise area for the dressing process conditions, and the weighted utility function method. The novelty of the implementation is due to the process of identifying the precise optimal dressing parameters specifically for the investigated mixed-grit diamond rollers. The optimization results show that rough grinding requires uni-directional dressing with a diamond roller AC32 at a feed rate of 1.4 mm/min, a dressing speed ratio of 0.8, a dress-out time of 1.0 s, and a grit size ratio of 2.56, ensuring a production rate of 875 mm3/min, a surface roughness Ra up to 1.25 µm, and a cylindricity deviation up to 12.5 µm. For fine grinding, optimal counter-directional dressing parameters include a feed rate of 0.26 mm/min, a speed ratio of 0.23, a dress-out time of 5.0 s, and a grit size ratio of 2.2, which guarantee a minimum surface roughness Ra of 0.38 µm, a cylindricity deviation of 8.1 µm, and a production rate of at least 600 mm3/min.
Александрова et al. (2026) studied this question.