Integrating machine learning (ML) techniques as an enabler for sustainable and smart sensor-assisted machining of heat-resistant superalloys (HRSA)

    Project Details

    Description

    The project aims to integrate machine learning (ML) techniques as an enabler for sustainable and smart sensor-assisted machining of heat-resistant superalloys (HRSA) complying with Industry 4.0 (smart technologies) and Industry 5.0 (human-centricity). Traditionally, machining of HRSA is performed using a direct closed feed loop approach which involves conducting a large and repetitive number of experiments to improve the quality of the machined part. HRSA possesses high hardness and strength which generates extreme temperatures at the cutting zone that negatively affects their surface microstructure. Synthetic fluids are a preferred option in high heat machining operations but can be harmful to humans and the environment since they require proper handling and disposal. An alternative is to use minimum quantity lubrication (MQL) and compressed cold air-cooling using vortex tube airgun, which are environmentally friendly cooling techniques. In addition, the use of next generation bio-based coolants has proven its effectiveness with HRSA over traditional coolants in terms of environmental and economic performance, but some limitations still exist especially in high-speed machining. The direct closed feed loop approach coupled with the use of synthetic fluids can be time consuming, costly, and is sensitive to human error. Employing state-of-the-art ML techniques (i.e., deep learning and Kernel-based methods) can analyse big data from machining and determine certain trends that are not apparent to humans to improve manufacturing efficiency and productivity with minimal human intervention. This work aims to provide a big data-driven approach enabling ML as a tool to overcome the limitations of employing sustainable elements in a machining process (i.e., coolants and cooling technologies). The effect of the proposed machining and cooling conditions on the surface finish and, hence, the mechanical performance of the machined parts will be investigated through a series of cyclic loading tests.

    Key findings

    Difficult to cut alloys, machining, machine learning, sustainable machining.
    StatusFinished
    Effective start/end date1/05/231/05/24

    Fingerprint

    Explore the research topics touched on by this project. These labels are generated based on the underlying awards/grants. Together they form a unique fingerprint.