Explanation of Impact Sensor Apparatus

The force sensor could be positioned at various locations and orientations within the finisher (Fig. 1).  In all cases, the 4 mm diameter AA6061-T6 aluminium impact cap was the only part of the sensor to contact the media through a hole in the housing assembly (Fig. 1).  The rest of the sensor and its mounting arm were connected to a brick wall and isolated from the impacting media by the housing assembly.

Fig. 1: Force sensor and housing schematic. Not to scale.

A piezoelectric force sensor (Dytran Instruments Inc.,Chatsworth,CA, model 1051V1) was used to measure normal forces. Sensor signals were recorded using an analog-to-digital converter (National Instruments, Austin, TX PCI-6071E), capable of sampling the output signal with 12-bit resolution every 1 μs; a  sampling frequency that is at least 10 times less than the shortest impact duration measured in the finisher. All electrical cables were shielded to minimize the electrical noise due to the finisher motor.  Utilizing the NI-DAQ v7 driver (National Instruments Corporation,Austin,TX,USA), a “C language” computer program was written to save all recorded data to a file.  All subsequent data analysis (see Section 2.4) was performed using Matlab v7 (The MathWorks, Inc.,Natick,MA,USA).

With the sensor cap covered, and the housing submerged in the finisher, it was found that the background vibration reaching the sensor through the wall together with residual electrical noise resulted in a background signal of less than 0.3 N. Typical contacts had force maxima in the range of 1 N to 25 N; therefore, the signal to noise ratio was acceptable.

Fig. 2(a) shows a typical force signal recorded over 5 s when the sensor was submerged in the vibratory finisher. Two contact types were identified in the force signals: impact and non-impact contacts (see Section 3.2), shown in different shades. There are relatively long periods during which there is no contact between the media and sensor.  Fig. 2(b) shows an expanded view of the signal between 0.8 and 1.5 s demonstrating that the contacts occur in bursts, generally every 21 ms, corresponding to the period of the finisher vibration at 47 Hz.

Fig. 2: (a) Typical signal recorded over 5 s when the sensor is submerged in thevibratory finisher.  The portion of the signal between 0.8 and 1.5 s is expanded in (b).  

%d bloggers like this: