| Abstract: | A milling cutter instrumented with a three-component accelerometer is investigated as a sensor of dynamic cutting forces. Two major causes of measurement errors are considered. These causes are: (1) the flexible mode vibrations, and (2) inertial and viscous forces associated with the “rigid body” motion of the spindle. A self-tuning filter consisting of two subsystems is applied to attenuate these errors. The first subsystem converts accelerations from the rotating spindle into stationary coordinates. It also analyses the corrupted signal and calculates an optimal filter structure and the settings for the actual operating conditions. This information is utilized by the second subsystem, a digitally programmable filter, which performs signals correction in real time. Two examples are presented to illustrate performance of the proposed “natural” sensor. In the first example, a periodical force applied from an exciter is reconstructed from the accelerations measured during spindle rotation. The second example deals with estimation of a force impulse generated by means of an impact hammer. |
