Electrical Discharge Machining of Titanium Alloys: A Comprehensive Review of Process Variants, Sustainable Dielectrics, Intelligent Optimization, and Biomedical–Aerospace Applications
TL;DR
The application of machine learning frameworks including artificial neural networks, genetic algorithms, support vector machines, XGBoost, and particle swarm optimization has transformed parameter optimisation, enabling accurate prediction of MRR, tool wear rate (TWR), surface roughness, and recast layer thickness.
Abstract
Titanium alloys remain among the most challenging materials to process by conventional machining due to their low thermal conductivity, high melting point, elevated chemical reactivity, and strong tendency toward tool adhesion. Electrical discharge machining (EDM) continues to be a pivotal non conventional manufacturing technology for shaping these alloys into complex geometries demanded by aerospace, biomedical, and automotive industries. This review extends and comprehensively updates earlier systematic analyses by integrating literature published between 2020 and 2026, covering wire EDM (WEDM), die sinking EDM, EDM drilling, and hybrid EDM configurations. A principal focus is placed on recent advances in dielectric fluid formulations including sustainable and bio derived alternatives such as refined, bleached, and deodorised (RBD) palm oil, vegetable oils, biosilica activated dielectrics, and water in oil nanoemulsions which simultaneously improve material removal rate (MRR) and surface integrity while reducing environmental impact. The role of novel electrode materials, including 3D printed aluminium alloy electrodes and cryogenically treated tool materials, is critically evaluated alongside powder mixed EDM (PMEDM). Hybrid processes combining ultrasonic vibration, laser pre drilling, and magnetic field assistance are assessed for their contributions to recast layer reduction and improved hole geometry. Crucially, the application of machine learning frameworks including artificial neural networks, genetic algorithms, support vector machines, XGBoost, and particle swarm optimization has transformed parameter optimisation, enabling accurate prediction of MRR, tool wear rate (TWR), surface roughness, and recast layer thickness. Long term corrosion behaviour and biointegration potential of EDM processed titanium in simulated body fluid are also reviewed. Four extensive summary tables synthesising experimental data across 120+ cited studies are provided as a structured reference for researchers and industry practitioners.