When conducting dynamic tests, when simulating shock effects, the measurement system - from the sensor to the data acquisition unit is subject to significant overloads. In this circuit, the accelerometer is the most vulnerable link. Therefore, piezoelectric accelerometers have become widely used in measuring shocks as more reliable and durable, but they also have a number of disadvantages. Thus, under broadband vibration effects, excitation in the region of mechanical resonance is observed. That is, the material of the sensor crystal may not have an excessive mechanical load, but generate a large amount of output charge due to the resonance of the sensor. This increases the electrical signal and can lead to saturation or, in many cases, damage to the signal shapers and amplifiers following the sensor. The result of such modes of operation is the loss of data due to the displacement of direct current over time. There is often a shift of the zero level of the output signal. When piezoelectric elements are in resonance, there may be a relative displacement of the sensing element. This state of overload leads to the appearance of parasitic output charges, which leads to a change in the generated charge in a short period of time. Due to mechanical overload, under the action of high-intensity shocks, physical destruction of the sensor is also possible. In many cases, the use of mechanical filters made of viscoelastic materials, which are placed between the object of measurement and the piezoelectric accelerometer (sensor), prevents the undesirable consequences of these shortcomings and improves the process of measuring shocks. The use of mechanical filters made of viscoelastic materials makes it possible to significantly reduce the amplitude of oscillations of the piezoelectric accelerometer (over -12 dB) and shift its resonant frequency towards higher frequencies. That is, the use of mechanical filters significantly reduces the mechanical load on the sensor, which makes the process of measuring vibration (especially shock) more predictable.
Read full abstract