Dependence of the Impact Response of Polyvinylidene Fluoride Sensors on Their Supporting Materials’ Elasticity

2.1. Materials

The PVDF piezoelectric film was purchased from Jinzhou KeXin Eectronic Materials Co. Ltd, (Jinzhou, China). The purchased sheet had been already stretched and poled. The type of PVDF film was KXP02, the thickness was 100 μm. The polarized polymer material had a zigzag molecular structure which as sketched in Figure 1(a). The atoms of fluorine and hydrogen were separated in the two sides of the carbon chain. There were spontaneous dipoles in this structure. After polarization PVDF could exhibit piezoelectric properties, and the charge coefficient, d₃₁, d₃₂ and d₃₃ were 17 ± 1, 5∼6 and 21 ± 1 pico-Coulomb per Newton (pC/N), respectively.

Copper foil (DongGuan FeiYing Electronic Material Co. Ltd., DongGuan, China) was used as the electrode material. The thickness was 0.02–0.05 mm and there was a conducting resin on one side of the foil.

A charge amplifier was constructed to transform the sensing charge generated by PVDF into an output voltage signal. The principal circuit and a photo of the charge amplifier are given in Figure 1(c,d), respectively. Here CA 3140 was an integrated circuit operational amplifier (Harris Semiconductor Melbourne, Florida, USA), which was used to transfer the piezoelectric charges generated by the PVDF sensor. As shown in Figure 1(c), the impact sensing charges were input from the IN port; by the used of the parallel resistor-capacitor (200 MΩ and 100 pF) in the negative feedback loop of CA 3140, an impact-sensing voltage signal could be output from the OUT port. Furthermore, a low pass filter was used to screen out the ultimate voltage signal.

2.2. Experiments

The sensors were prepared based on PVDF films and copper foil with a sandwiched structure, as outlined in Figure 1(b). First two out-leads were soldered on tailored copper foils, and the Cu foils were transmitted and conglutinated onto the surfaces of PVDF film by the conducting resin on the copper foils. Second, the PVDF sensors were laid on different backstops with flat surfaces, which were desk, rubber and sponge. Third, a falling path was formed with hardboard to guarantee the impacting force happened exactly on the PVDF sensors. Here, two kinds of weights, which were 20 and 50 g, were tested; the weights fell down freely from three different heights, 10, 15, 20 cm.

2.3. Detection System and Equipments

The detection system was depicted in Figure 1(d) and the principal circuit for the charge amplifier was presented in Figure 1(c). The charges generated by the PVDF sensors' piezoelectricity were introduced into the charge amplifier, through the capacitor in Figure 1(c) the charges' variation of the PVDF sensor could be transformed and screened out as a voltage signal. Finally, the impacting sensitive voltage signals were displayed by a TDS2022B oscilloscope (Tektronix Inc. Beaverton, OR, USA), and the data were collected by a computer through a USB port and analyzed by Origin Lab 8.1 (Origin Lab Corporation, Northampton, MA, USA).

3.1. PVDF Sensors' Voltage Response Results When Supported by the Desk

The shock response of the PVDF sensor was measured on the desk; the temporal voltage response curves are presented in Figure 2(a). The falling impact conditions are adjusted according to Table 1. That means higher impacting energy could be obtained when the heavier object fell from the higher position. The impact energies were estimated and are listed in Table 1, by assuming all the mechanical potential energy was transformed to impact energy. To contrast their response characteristics under different conditions, the collected voltage signals were moved along the x-axis to aim at the same zero time.

The main explanation of the results is that the intensity of polarization in the PVDF film will be changed when it suffers a falling impact. To maintain the electric neutrality, it will release or capture charges from outside circuits as given in Figure 1(c). The quantity of these charges is measured and transformed into a voltage signal by the measuring circuits, given in Figure 1(c,d).

In Figure 2(b), the abscissa is the impact energy; the approximate values are given in Table 1. Similar curve shapes for these two analyzed items can be found in Figure 2(b). Since the area was the product of Vpp and the recovery time, so it was reasonable to deduce that the difference of the recovery time under different impacting conditions was tiny. Approximately calculations indicated that the recovery time for the desk supported sensor was about 0.15∼0.18 s.

By comparing the horizontally moved curves in Figure 2(a), it was found that: (1) under different impact conditions the positive peaks' amplitudes were not obviously changed, while the negative peaks were lowered, so the value of V_pp must be changed by the increasing impact energy, as shown by the right y-axis in Figure 2(b). (2) After the impact, the response voltages would return to their original value. In this progress there was a recovery time, a sample of which was depicted as a yellow double arrowhead line for the dashed black line in Figure 2(a), but it seemed that the variation of their recovery time was very small. Furthermore, V_pp and the integration areas under peaks were calculated by MATLAB, and their relation with the changing impacting energy was plotted in Figure 2(b). Here V_pp means the maximal output voltage minus the minimum output voltage, and integration areas were the areas under the curves from the highest to the lowest. These two items were plotted in a double- y-axis coordinates, as shown in Figure 2(b).

3.2. Results of the Elastically Supported PVDF Sensor

After changing the backstops, the experiments mentioned above were repeated on the 100 μm PVDF sensor used in Figure 2. Here two kinds of supporting materials were tested, which were sponge and rubber. The temporal response curves were plotted in Figure 3(a-c). The values of Vpp for all the figures in Figure 3(a-c), and Figure 2(a) are analyzed and plotted in Figure 3(d).

It was found from Figure 3(a), that the response curves' vibration of the sponge supported sensor was stronger than the rubber supported one under the same impact conditions. We deduced that the reason might lie in the elasticity of the supporting materials. The elasticity of sponge was bigger than that of rubber, so it offered a buffering action to PVDF film sensor and made it suffer a second elastic force after undergoing the impact, and this force could cause a second-piezoelectric-signal which was labeled in Figure 3(a,c), while for the rubber supported one, this signal was relatively small. This deduction could be supported by further comparison with Figure 2(a). On the hard material (the wooden desk) with almost no elasticity, no second-piezoelectric-signal was detected.

Another result that could be obtained in Figure 3(a–c) was that Vpp changed with the increasing impacting energy. Their relationship was proposed in Figure 3(d). The ascending curves in Figure 3(b) indicated that with the increase in impact energies, the values of Vpp were enhanced, but the slopes of these curves were obviously different. Desk-supported sensors exhibited smaller V_pp variation than rubber and sponge supported ones. For the same sensor (100 μm), the biggest curves' slope of Vpp vs. impacting energy was displayed by the sponge-supported one, the slope of which was 96.62 V/J., while for the same sensor, when it was supported by rubber, the slope was 82.255 V/J; and when it was supported by the desk the voltage response saturated quickly.

3.3. Discussion

According to the experiments mentioned above, PVDF sensors could respond to a suffered impact and transform the mechanical energy into an electronic signal, but there were differences in their response signals when the sensors' supporting materials were changed.

Firstly, the impact power could be expressed in the value of the peak-to-peak voltage, so the V_pp vs. impact energy curve was chosen as the response curve. When the impact energy was in the range of 0.02∼0.08 J, the bigger V_pp values could be transduced by a hard supported sensors (like the desk used in this work) than by the sponge or rubber supported ones, but the difference of impact energy could not be discriminated by the desk supported sensor, which was shown by the black and purple lines in Figure 3(d), and the quantitative detection could be realized better by the sponge supported sensor than the rub ber or desk supported one, whose sensitivities were given in above subsection.

Secondly, the recovery time between differently supported sensors also varied. The green and black curves in Figure 3(a–c) represent the sponge supported sensor's response signals under three different conditions. These solid curves were magnified in the time range of 0.41∼0.66 s here, as shown in Figure 3(a), to observe the sponge supported PVDF sensor's temporal response characteristics. It was deduced according to the black solid curve in Figure 4(a), that the sensor might endure a sympathetic vibration after it suffered a falling impact induced by a 50 g weight dropped from a 10 cm height, since the amplitude output voltage was enhanced after suffering the impact and could not recover. Furthermore, the response time data was extracted and plotted under recovery-time vs. impacting-energy coordinates, shown in Figure 4(b). As given in the explanation of Figure 2(a), the recovery time for desk supported sensors was about 0.15 to 0.18 s. So the characteristic curve about recovering time for the desk supported sensor was nearly flat as shown by the fitted curve in Figure 4(b), but for the sponge and rubber supported sensor the recovery time was up-and-down and no order could be found.

Thirdly, comparisons between Figure 2(a), and Figure 3(a–c) indicate that, for the same sensor (100 μm), when the supporting materials are changed from desk, rubber, to sponge, more obvious second piezoelectric signals could be observed. This phenomenon was coincident with the elasticity of these three materials.