Abstract
Porous titanium is a new biomechanical implant material due to the good mechanical properties and biocompatibility. In order to meet the manufacturing needs of anisotropic porous titanium with complex micro-structures, the ultrasonic atomization with excellent cooling and lubrication effect is integrated with the micro-milling operation. Accurate prediction of cutting forces in the machining process is a principal factor as it exerts indispensable influence on the process efficiency and surface integrity. The prediction power of the traditional physics-based mechanistic cutting force model has been limited by the explicit relationship between force components and process variables based on the cutting mechanism and process geometries. In this paper, a comprehensive physics-guided intelligent system for stochastic cutting forces estimation is developed in the ultrasonic atomization-assisted micro-milling operation of porous titanium. The physics-based mechanistic cutting force model is proposed based on the determination of in-process random porosity of machined porous titanium with tool-workpiece contact. The influence of process uncertainty related to size effect and tool run-out is considered in the proposed physics-based mechanistic model. To overcome the poor nonlinear fitting and indeterminate cutting constants with numerous experiments, the machine learning algorithm has been merged into the physics-based mechanistic model to establish the physics-guided intelligent system of stochastic cutting force prediction. The predicted cutting force values of the proposed comprehensive physics-guided machine learning system are validated by the measured results with a wide variety of ultrasonic atomization-assisted micro-milling tests.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call