Characterization and machine learning-based parameter estimation in MQL machining of a superalloy for developed green nano-metalworking fluids

摘要

Metalworking fluids (MWFs) are always been an integral part of a million-dollar global manufacturing industry. With a paradigm shift towards ecological sustainability and fossil-fuel preservation, commercial-grade petroleum-derived mineral oils are eventually substituted by environment-friendly bio-lubricants. However, these bio-lubricants fall short in terms of machining performance and are yet to be majorly explored and commercialized. In this regard, the present research focuses on the development and systematic assessment of vegetable-extracted edible oil-based Nanofluids as a potential replacement for the existing MWFs. Initially, the developed nano-MWFs are evaluated for physio-thermal, tribological, and mist flow characteristics. Later, the machining performance of the developed bio-lubricants is evaluated to understand whether the manufacturing requirements of a difficult-to-machine material are met in a MQL-based turning process. A comparative assessment of cutting temperature, surface quality, chips formation and tool wear is accomplished under different cutting environments during the MQL turning of a hard-to-machine Nimonic alloy. Finally, a machine learning-based prediction model has been proposed to identify the best among the developed nano-MWFs. This model integrates the experimental results of both characteristic properties and the machining responses. This work is the first of its kind with an extensive number of nano-MWF combinations being tested to study their effectiveness in countering the boundary conditions of high temperatures, friction and pressure at the chip-tool interface in MQL machining of a hard-to-machine metal. The proposed methodology helps researchers and industries to arrive at a conclusion on the best suitable nano-bio-lubricant for any given input setting.