Adaptierbare Parametrierung von Diagnosesystemen durch Verwendung digitaler Antriebssignale in der Prozessüberwachung

The presented thesis deals with the integration of monitoring and diagnostics applications into machine tool control. For this purpose, static and dynamic influences of a machine structure are analyzed, modeled and verified with an example process of a multi-blade milling cutter based on digital drive signals. Friction, inertia and the machine dynamic during the milling process are the major focuses which have a strong influence on the signal characteristics and therefore have to be identified. In order to realize a generic control integrated diagnostics, these influences have to be modeled and considered in-depth. A generic approach is developed based on the friction curve by Stribeck. The acceleration behavior is mainly characterized by the configuration of a machine axis cascaded controller and can be modeled by a simplified controller structure. Additionally, the characteristic of the machine structure during the machining process is studied. The dynamic transfer behavior between the axis drive shaft and the tool center point can be described based on metrological analysis and the controller characteristics. Using the results monitoring modules for overload and tool wear are developed. These modules are based on signals available in the numeric control structure. In order to monitor collision situations, a signal pattern is developed. This pattern defines the feed drive velocity area where a contact between the tool and the work piece is allowed. The pattern is based on the information about the tool and current velocities of all axes. Therefore, definitions of torque or force limits are unnecessary. In case of a collision, control internal safety mechanisms are triggered. Furthermore, TEAGER´S algorithm is applied on the digital drive signals to identify the signal amplitude. Since the process parameters often change during the life time of a milling tool, the identification of the current parameters is crucial. Hence, several approaches allowing the online identification of the feed per tooth and the radial depth are studied, presented and verified. Based on these investigations and on the identification of the signal amplitude, the tool wear state can be monitored. The presented methods are applied and verified on different machine tools. The results are verified and critically discussed at the end of this thesis