Project Details
Projekt
Identification of the nonlinear individual machine dynamic behavior of milling machines
Applicant
Professor Dr.-Ing. Jürgen Fleischer
Subject Area
Metal-Cutting and Abrasive Manufacturing Engineering
Term
since 2026
Project identifier
Deutsche Forschungsgemeinschaft (DFG) - Project number 546480484
The project investigates the identification of the nonlinear, individual machine dynamic behavior of milling machines. The aim is to develop a data-based modeling method based on experimentally recorded measurement data that enables precise mapping of the machine dynamic behavior of the machine structure as a function of external force excitations. The focus here is on the detection of nonlinearities, which have a significant influence on the accuracy of simulation-based stability analyses in milling processes. Modeling is based on measurements taken on the machine under consideration, which allows the individual machine dynamic behavior to be mapped without detailed prior knowledge of the machine structure, material properties, or controller parameters. To keep the amount of testing to a minimum, time-efficient measurement strategies and suitable force excitation profiles are being investigated. These are intended to ensure that the relevant machine dynamic behavior can be reliably recorded across the entire working space. Nonlinear system identification methods are used for modeling. These enable robust and data-efficient mapping of the nonlinear relationship between time-dependent force excitations at the machine TCP and the resulting vibrations of the machine structure. A particular focus is on analyzing the influence of different excitation characteristics (e.g., frequency, amplitude) on model quality and robustness, taking into account aleatory uncertainty. The identified models should be able to make direction-specific vibration predictions in the time domain for different TCP positions in the machine working space. To describe the dynamic behavior along a TCP path, interpolation across different model positions is planned. In a final validation, the extent to which the developed models can be transferred to real milling processes and the model quality that can be achieved under practical conditions will be investigated. The methodology developed in the subproject thus makes a significant contribution to the resource-saving and generalizable stability assessment of milling processes within the framework of process-informed models.
DFG Programme
Research Units
