Application of machine learning on tool path optimisation and cooling lubricant in induction heating-assisted single point incremental sheet forming of Ti-6Al-4V sheets

Abstract

Induction heating-assisted single point incremental sheet forming was established for Ti-6Al-4V thin sheets at closed and above beta-transus temperature (980 °C). In order to eliminate geometric inaccuracy and adherence of lubricant on the surface caused by elevated temperature, a cooling lubricant system was designed for the forming tool to decrease the thermal expansion and friction. A radial basis function (RBF)-based tool path optimisation was developed to study the measured geometric accuracy, temperature, and forming force. By adjusting cooling lubricant control and integrating the RBF framework, the first optimised tool path was used to collect the results and to validate with the finite element (FE) model and theoretical geometric profiles. The output data were further studied by RBF and generate a second optimised tool path. The measured geometric coordinates revealed that the error percentage has been reduced to less than 5%. Further, the microstructure evolution analysed by scanning electron microscopy (SEM) indicated noticeable oxidation and alpha-layer for temperature around 1040 °C and the phenomenon was removed at temperature closed to 950 °C. The surface roughness and energy-dispersive X-ray analysis (EDX) revealed the optimised tool path distributed significant improvement in surface quality. The cooling lubricant system indicated optimal performance with RBF optimised tool path to support constant temperature and reduce friction and lubricant adherence on the surface.