Random decision forest based sustainable green machining using Citrullus lanatus extract as bio-cutting fluid

Abstract

The presently employed synthetic cutting fluids in machining applications lead to skin irritations, hazardous aerosols, ingestion etc., to the machine tool operators. Accomplishing sustainable green machining through bio-cutting fluids, is the key motivating factor behind this work, even though the phenomenon of heat generation is inevitable in the machining process. In the present investigation, Citrullus lanatus extract is chosen as a potential candidate for bio-cutting fluid, since it offers multiple benefits such as increased tool life, cleaner production atmosphere, efficacious heat dissipation, healthy operating conditions, etc. Hybrid LM0-6SiCp-4Grp composite is selected for study since; it has potential applications in marine, aerospace, defence and automotive sectors. Experimentation is carried out with Citrullus lanatus extract as bio-cutting fluid in turning of hybrid LM0-6SiCp-4Grp composite under different processing conditions. The presence of green solution based on Citrullus lanatus extract curtails the rise of temperature significantly during turning operation. Higher magnitude of temperature decrease is observed when the percentage of the concentration of Citrullus lanatus extract is increased, irrespective of depth of turning. Bio-nature and cooling qualities of this green-cutting fluid can contribute its share for green machining and sustainable turning environment. The investigations are further progressed using machine learning algorithms based on the different categories of logical regressions implemented through the Python program. Random forest regression shows better results with 99.8% prediction accuracy compared to other machine learning approaches and the decision tree is constructed through the respective regression model. Decision tree has the ability to handle process parameters systematically and predicts the implications of Citrullus lanatus extract in decreasing the temperature at machine-tool interface during machining operation. This tree structured analysis provides a better inference that delivers flexibility in turning parameters and opens up wider options for operation. Thus, decision tree approach ensures optimized usage of production resources and increases the production capacity through judicial selection of cooling approaches associated with the turning process.