Investigation of component-specific causality relationships in the dynamic behaviour of machine tools by identifying a modal fingerprint

• Applicant: Professor Dr.-Ing. Martin Dix
• Subject Area: Production Automation and Assembly Technology
• Term: since 2025
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 550608699

In order to minimise wear, failures and downtimes and at the same time make use of performance limits in dynamic behaviour, a large number of monitoring systems have been researched and developed in the field of machine tools. In the context of ensuring work accuracy, reliability and performance, assemblies such as the main spindle and feed drives in particular are receiving increased attention. This condition monitoring is largely based on signal analysis. Alternatively, there is model-based condition monitoring, which focuses on describing the machine and its conditions as a system and using measurement signals to determine the parameterisation of the models. Accordingly, the aim of the project is to determine recurring system descriptions in the form of modal parameters (natural frequencies, modal damping and eigenvectors) over the life cycle of machine tools. The identified modal parameters characterise the changes within the machine structure and allow the definition of a modal fingerprint. This fingerprint is dependent on the individuality of the machine itself (assembly, installation, design) as well as on the specific machine's life cycle events (collision, component wear, process specifications). These dependencies are investigated to assess the potential uses for structural monitoring based on changing modal parameters. The modal characterisation offers the possibility of a system-based description, which, by deriving causal relationships, goes far beyond the findings of current signal analyses in order to continuously analyse the dynamic system behaviour and thus be able to classify fault events. With the help of the natural frequencies and the natural modes, weak points and potentials for improvement can be easily recognised through their visualisation. On the other hand, these vibration responses and eigenvectors can be objectively evaluated and differentiated by criteria such as MAC, CoMAC, FRAC. The project investigates the limits within which normal time-variant events, such as position changes, can be separated from events such as wear and faults. A test cycle is to be developed that achieves adequate excitation and at the same time can be considered to be included in the life cycle of the machine. Furthermore, the question arises to what extent calculation models (state space descriptions, signal flow diagrams) and experimental descriptions (modal parameters, transfer functions) can be connected and how exchangeable they are. Furthermore, the limits being investigated to what extent events that do not occur directly near the sensor in their place of origin can nevertheless be recognised by changes in modal parameters.

• DFG Programme: Research Grants (Transfer Project)
• Application Partner: DECKEL MAHO Seebach GmbH; FunctionBay GmbH