Project Details
Projekt
FOR 639: Specific Predictive Maintenance of Machine Tools by Automated Condition Monitoring
Subject Area
Mechanical and Industrial Engineering
Term
from 2006 to 2013
Project identifier
Deutsche Forschungsgemeinschaft (DFG) - Project number 14394217
The paramount aim of this Research Unit is to minimise machine downtimes and unnecessary maintenance cycles, by which the economic efficiency of the manufacturing business can be considerably increased. This efficiency is achieved by adding new approaches of condition-oriented maintenance to existing maintenance systems based on statistical statements at a rough estimate. Resulting from this is the scientific task to establish the current condition of a machine and its components on the basis of measurable values. Since the condition itself cannot be directly measured, the obvious thing to do is to analyse the signals standardly available in a machine and the information that can be directly read from it. In modern machine tools such direct information is usually available in the form of motor temperature and recorded motor currents (feed drives and main spindles), speed (main spindle), and position (in the feed drives). Moreover, in many applications further sensors are used to determine additional information relevant with regard to control technology. This includes, for instance, the direct measurement of the position on the slide of an axis or the measurement of relative acceleration.
The aim is to develop a general methodology for the automated prediction of damage, with the help of which it is possible to reduce the costs to a minimum. The basis for the prediction of damage is to systematically analyse the control and drive signals of the machines considered, so that no additional costly monitoring sensors are necessary. Basing on statistical failure probabilities and an evaluation of the application-specific machine load, it is intended to infer the wearing state of the components and hence an imminent damage from changes in the signal with the help of static and dynamic models of all wearing components. If a machine breakdown is predicted reliably and in time, then automated plans for maintenance can be generated, enabling a production that is optimal with regard to time and costs.
The aim is to develop a general methodology for the automated prediction of damage, with the help of which it is possible to reduce the costs to a minimum. The basis for the prediction of damage is to systematically analyse the control and drive signals of the machines considered, so that no additional costly monitoring sensors are necessary. Basing on statistical failure probabilities and an evaluation of the application-specific machine load, it is intended to infer the wearing state of the components and hence an imminent damage from changes in the signal with the help of static and dynamic models of all wearing components. If a machine breakdown is predicted reliably and in time, then automated plans for maintenance can be generated, enabling a production that is optimal with regard to time and costs.
DFG Programme
Research Units
Projects
- Antriebsnahe Prozess- und Maschinendiagnose (Applicant Verl, Alexander )
- Entwicklung eines Modells zur optimalen, zustandsabhängigen Instandhaltungsplanung anhand technischer Diagnoseinformationen (Applicant Bertsche, Bernd )
- Experimentelle Analyse des Verschleißzustandes von Werkzeugmaschinen sowie deren konstruktive Optimierung (Applicant Heisel, Uwe )
- Steuerungsbasierte Maschinendiagnose (Applicant Klemm, Peter )
Spokesperson
Professor Dr.-Ing. Uwe Heisel
