9.2 The Necessity of Process Monitoring
Increasing demands on quality and higher time and cost pressures require more process security. This however has been made increasingly difficult by expanding process and component complexity [KLOC01].
An economical utilisation of automatic grinding processes presupposes their behaviour to be reproducible. Despite the increased expenditure on quality control and on technologically optimised process design, faulty parts can nevertheless still not be fully avoided in batch production.
Periodical and stochastic disturbances influence the grinding process, but the effects can not always be completely compensated. Accordingly, service life behaviour of the often very expensive grinding tools is also subject to unpredictable variations, which leads in automatic manufacturing processes to a conservative process design in order to assure the target quality level. Nonetheless, flawed machining sequences and workpieces can still be found again and again [TOEN88, WECK90].
Besides stock allowance fluctuations and hardness variation in the workpieces, other important disturbances in this context are varying grinding wheel properties as well as operator-related parameter modifications. One should not assume a consistent and predictable grinding wheel behaviour, necessitating a reliable and easily realisable monitoring of the service life end (Fig. 10-1).
Even today, there exist only insufficient measuring methods that would make possible a reliable online monitoring of the process and workpieces. The industrial operator is thus forced to control the output by means of random sampling. Such a procedure is however time-consuming and only reliable to a limited extent.
There is for this reason a constant need for new sensors and monitoring systems that can be used for preventative quality assurance. The goal is that workpiece faults are avoided by means of online monitoring of the process.
Various physical effects can be utilised for tool and process monitoring in grinding and dressing. Increasing grinding wheel wear for example can lead to higher machining forces. This offers the possibility of making use of those forces for the sake of process monitoring. This can be accomplished either directly, e. g. with the help of piezo-electric force sensors, or indirectly from the resulting extension of the machine components or the amount of motor current (Fig. 10-2).
•
 |
early dressing
• scrap parts during assembly
• high reject costs
Fig. 10-1 The necessity of process monitoring
An increase in tool wear leads however to a stronger dynamic stimulus of the workpiece/tool/machine-tool system as well. Vibrations can consequently also be exploited for process monitoring.
In the low-frequency range, these vibrations can be extracted from the force or extension signal. As a high-frequency oscillatory signal on the other hand, the acoustic emission (AE) generated during grinding and dressing is especially suitable. As opposed to force measurement, acoustic emission measurement is not subject to temperature drift and is also extremely sensitive in machining processes with very small chip cross-sectional areas. Although the AE signal is an indirect measure, it has the advantage over others, such as the motor current, of very small rise times [N. N.11, N. N.12].