ANN Samples Generation Using 2D Dynamic FEM for Predicting Machining Vibrations

Autor: Iulia-Maria Prodan, Tudor Deaconescu, Cosmin-Ioan Niță, Andrei-Ionuț Berariu
Rok vydání: 2020
Předmět:
Zdroj: Springer Proceedings in Physics ISBN: 9783030541354
DOI: 10.1007/978-3-030-54136-1_39
Popis: Cutting operations are difficult to predict in terms of dynamic behavior and prove to be equally difficult to control. The productivity and quality of the generated surface are influenced directly by phenomena like chatter, adhesion and wear, which left undetected can be destructive and cost-intensive. The paper presents and discusses a 2D finite element method that can be used to simulate and extract the cutting forces frequency components for various milling operations. The proposed method represents an effective approach to predicting such nonlinear behavior and entails unwrapping the analytical chip section and running of plane stress simulations using a linear kinematic trajectory. The results are rewrapped according to the geometry of the tool, which is used to generate the overall dynamic behavior in the machine coordinate system. The paper concludes with considerations concerning future directions in deep learning in milling operations and the necessary effort for creating high-end control algorithms.
Databáze: OpenAIRE