Electrical and mechanical response of spin-coated PVDF-TrFE formed from MIBK and DMSO solvents under external excitations

Flexible polymer piezoelectric materials are of interest in electro-mechanical applications; however, the mechanical and electrical properties of these materials can be affected by their formation. In this study vibrational analysis was performed to explore the mechanical and electrical properties of piezoelectric PVDF-TrFE polymer using external excitations. Frequency response study of cantilever cantilevers was used to explore the natural mechanical response of a flexible KAPTON cantilever coated with PVDF-TrFE of different concentrations, and with different solvent formations. It is found that the mechanical response of the PVDF-TrFE coated KAPTON cantilever has reduced vibration amplitude when a combined Methyl Isobutyl Ketone (MIBK) and Dimethyl Sulfoxide (DMSO) is used, but it is not affected with only MIBK is used. When using a combined MIBK and DMSO solvent, the 1.25%w/v PVDF-TrFE material was easily mixed, and it possessed a higher piezoelectric electrical response to the MIBK solutions. The piezoelectrical coefficients were also calculated by applying a dynamic force to the cantilevers and found that q31 of cantilever coated with the PVDF formed using the mixture of DMSO and MIBK solution is higher than that formed using MIBK.


Introduction
Electroactive polymers (EAPs) are substances that, when stimulated by an outside electrical field or triggering force, can alter the shape or conduct electricity.Polyvinylidene fluoride (PVDF), a form of EAP, is utilized in a variety of microelectronics applications, including sensors, actuators, energy harvesting, and biometric robots.It exhibits piezo, pyro, and ferro-electric characteristics.In comparison to pure PVDF, the PVDF-TrFE copolymer exhibits stronger crystallinity and a greater piezoelectric response.The copolymer has an extended 'all-trans' shape even without physical techniques like stretching or annealing because the TrFE adds more fluorine atoms to the polymer chain and inhibits the molecular chains from adopting a particular conformation.A polar crystalline phase and spontaneous polarization are present in copolymers that include more than 17 mol% TrFE [1].
The commonly used piezoelectric materials include piezoelectric ceramics (such as PZT), perovskite structured PZT, and polymer like PVDF.On the macro scale, PZT is a suitable candidate due to its large output and high electromechanical coupling [2][3][4], but they are not compatible with flexible structures or with polymer properties.In contrast, PVDF and its co-polymers are advantageous due to their flexible nature, chemically inertness, toughness, stability under sunlight, and biocompatibility [5,6].Research on polymer-based devices is expanding rapidly because their Young's modulus is half of the traditional silicon-based materials, which increases surface stress sensitivity if the structure is long and thin [7].PVDF-TrFE and its co-polymers employed on cantilevers have been widely used for piezoelectric and pryo-electrical sensors, and also have a large potential in low-cost fabrication [8][9][10][11][12].Polymer piezoelectric materials coating structures are employed in many applications such as medical devices, environmental monitoring, chemical devices and in security tools [13][14][15][16][17][18].mechanical and electrical properties was also studied.Also, DMSO is also known to enhance piezoelectricity, and so the performance difference of solution containing DMSO is compared with solutions containing only MIBK.The mechanics of the PVDF-TrFE polymer were explored using flexible KAPTON cantilever cantilevers with mechanical vibration studies using a ground base shaker and measuring the vibrations.The piezoelectric response of the spin-coated PVDF-TrFE polymer formulations were explored on stiff aluminum cantilever cantilevers driven into vibration on the ground base shaker.Finally, the piezoelectric response of the spin coated PVDF-TrFE polymer is compared with commercially available piezoelectric materials.

PVDF-TrFE solution and sample preparation
Spin-coated PVDF-TrFE polymer formulations were explored by mixing solid powder of PVDF-TrFE material at various concentrations with MIBK solvent, and with DMSO together with MIBK solvent.The mixture is stirred until a clear solution forms.Instead of being suspended in the solvents, the polymer powder was totally dissolved.For both mechanical and electrical piezoelectric measurements, four mixture concentrations were studied.The different formulations used PVDF-TrFE powder dissolved in the solvent solution to enable spin coating.The powder was purchased from PolyK Labs, molar ratio 75/25 which has the highest β-phase so, and no poling and stretching is required to enable the piezoelectricity β-phase.Two different solvents were used to explore their effect.The solvent MIBK was selected as it is recommended by PolyK Labs, due to its ability to suitably dissolve PVDF-TrFE power.MIBK allows for longer solution stability when stored compared to DMF which was used by [6].The second formulation used a combination of MIBK and DMSO, as the addition of DMSO is known to give the solution a lower viscosity due to the presence of sulfur [1].For this study, DMSO was added to MIBK with the ratio of 80% DMSO and 20% MIBK.
Sample preparation for the experiments were as follows.First, the cantilevers were cleaned IPA and acetone.The spin coating process requires the PVDF-TrFE powder to be completely dissolved into a clear solution.Four solutions were prepared.Solution A was formed using MIBK and DMSO with 0.137 g of PVDF-TrFE powder in 10 ml of solvent, forming a 1.25%w/v concentration.The PVDF-TrFE powder was found to completely dissolve after 5 min of hand stirring.Solution B, C, and D were formed using MIBK with PVDF-TrFE powder mixed into concentrations of 1.25%, 2%, and 3% respectively, however, solution mixing was more difficult.
Solution B was formed using MIBK with 0.137 g of PVDF-TrFE powder in 10 ml of solvent, forming a 1.25% w/v concentration.The powder did not dissolve using only hand stirring.Magnetic stirring was then used.It was found that at room temperature (23 °C) after 20 min of magnetic stirring the PVDF-TrFE powder did not dissolve.Therefore, the temperature was elevated by putting the beaker with the mixture on a hotplate.The hot plate temperature was increased by 2 °C in steps while stirring the solution for 10 min at each step.If the solution was not fully dissolved, then it was further increased another 2 °C for 10 min for mixing.When a temperature of 42 °C was reached, the solution dissolved to a fully clear solution.Solution A was visibly less viscous than solution B, due to the presence of sulfur in DMSO.Solution C and D were also prepared using MIBK but were at higher concentrations.Solution C was at 2% (0.30 g of PVDF-TrFE powder in 15 ml of solvent), and solution D was at 3% (0.598 g of PVDF-TrFE powder in 20 ml of solvent).The amount of solvent is higher in C and D solutions because during stirring the solvents evaporate and results in low solution volume.The preparations of clear solutions C and D took 1 h and 2.5 h respectively, at 300 rpm constant magnetic stirring at room temperature.
Fourier transform infrared spectroscopy (FT-IR) was used to examine the absorption of infrared energy to determine the molecular vibrational mechanics of a material's crystal structure.The inbound electromagnetic field from the infrared radiation (IR) source interacts with the molecular bonding of the PVDF-TrFE based solution when the electric field component of infrared radiation (IR) and the molecule vibration are In table 1, the beta-phase peaks of PVDF-TrFE/MIBK solution shows CF 2 stretching at 840 cm −1 , 1278 cm −1 and 1290 cm −1 and CH 3 stretching at 1405 cm −1 , 1434 cm −1 .While the PVDF-TrFE/DMSO +MIBK solution shows its beta-phase peaks at 1405 cm −1 and 1434 cm −1 with strong CH 3 stretching.The mixture with MIBK+DMSO solution showed higher absorbance at 1405 cm −1 and at 1434 cm −1 as compared to only the MIBK solution, which is expected based on the 20% MIBK and 80% DMSO solvent concentrations.
The higher absorbance of the DMSO+MIBK solution at 1405 cm −1 and 1434 cm −1 with respect to MIBK solution is associated due to the CH 3 stretching bonds surrounding a highly electronegative S=O functional group in DMSO.Also, the molecular dipole moment of DMSO functional group (S=O) is higher than the functional group of MIBK (C=O).This makes DMSO more efficient than MIBK which results in enhanced charge transfer at 1405 cm −1 and 1434 cm −1 .In result, DMSO is responsible for increasing the piezoelectricity of the solution.The piezoelectric comparison will be discussed in a later section 6.The solutions were spin coated on the cantilever materials at 500 rpm for 200 s.For all solutions, after spin coating a two-step annealing process was used.First, for 10 min the sample is annealed in an oven at 50 °C to form a stable crystalline film.This is followed by a 10-minute cooling time at room temperature.Second, the sample is again annealed in the oven at 112 °C for one hour.This two-step process results in crystallization of the ferro-electric gamma-phase during the first step, and the second step enhances the density of PVDF film [33].In this way long chain formation is possible.The Curie temperature for PVDF-TrFE with 75/25 molar ratio is 112 °C-121 °C [34].Therefore, the annealing temperature was kept at 112 °C to avoid changing the material properties.After annealing, the thickness and surface roughness of the polymer films were measured using a KLA Alpha-Step 500 profiler, and the results are shown in table 2.
Table 3 highlights this work and contrasts it with other works in literature.The annealing temperatures in the other works differed, as the Curie temperatures change when the molar concentrations of PVDF-TrFE change.

Test setup for vibration testing
The test setup consists of a ground base shaker.Cantilever structures were explored as substrates for the mechanical and piezoelectric experimental testing of the PVDF-TrFE material.For rigid substrate testing an aluminum cantilever of thickness 1.54 mm was used, while for flexible substrate testing a KAPTON-HN (Dupont) flexible cantilever of thickness 127 μm was used (127 μm is the maximum available thickness of the HN family).A ground base shaker was used to drive vibration on the attached PVDF-TrFE coated cantilever cantilevers at various vibration frequencies and accelerations.The shaker system consists of 'U' shaped trunnion support base that supports the electrodynamic shaker body.The trunnion allows the shaker to be configured in different skewed orientations and angles for excitation.Changing the input force is important for exciting structures with vertical and lateral modes that are highly uncoupled.The cantilevers cantilever (either aluminum or KAPTON) were rectangular dimension of 6 ×2 cm, and mounted using a clamp to the shake platform, which covered 1 cm of the cantilever, leaving a free size of 5 ×2 cm.
A Polytec scanning vibrometer controller (OFV 5000) was used to monitor the driven vibration characteristics.It employs a Polytec (OFV-505) sensor head with a He-Ne laser of wavelength of 633 nm at a vertical distance of 51.5 cm.The laser spot is located approximately 1 to 2 cm along the length of the cantilever from the clamp point.The system model and test setup are shown in figure 2.

Mechanical experimental results for flexible Kapton cantilevers
Dynamic response vibration testing is widely used to investigate vibration-based sensors.The response to a vibrational force is of interest instead of actual vibration level, when exploring dynamic properties of a structure.Vibration testing describes the vibrational modes of a structure at resonant modes.To perform this test on the device samples, the KAPTON cantilever is clamped and mounted on ground-based shaker (as shown in figure 2), which drives the structure at selected gravitational 'g' values.For this flexible structure, the light weight and stress in the polymer layer can result in undesired vibrations (noise) and reduced amplitude which can change the desired output.Many groups have studied this issue using analytical modeling based on steady state forced analysis, transient response, and natural vibrations, etc on piezoelectric flexible structures to understand the vibration problem [2].However, due to undesired noise, it is hard to filter the actual data from the noise.In order to see the natural response of the structure, the vibration test is analyzed using frequency response function (FRF).The FRF analysis determines the resonant frequencies of the vibrating cantilever.A random excitation from 0 Hz to 1500 Hz is employed in vibration testing.A random signal has a continuous spectrum of varying amplitudes.In the output, we get the multiple resonance frequencies, this is also referred to as spectral density, which includes the noise-estimation based on measured data.In practical terms, it is not possible to get a noise-free estimation response.In the present case the data is averaged, with the average number set as 10.
Testing was first conducted with unit base excitation on KAPTON cantilevers coated with solutions A and B and the parameters for data acquisitions which include the reference value of attached accelerometer (Reference actual sensitivity9.07mV/g), frequency resolution (0.5 Hz), sampling frequency (8192 Hz), actual sensitivity of laser sensor (1000 mV/V) and the Fourier transform (FFT) setup parameters were kept constant.For each complete measurement the number of averages were set as 10.  with solutions A and B were closely matched for the first mode but changed at higher modes.A noticeable point is that the measured motion amplitude of the cantilever coated with solution B (MIBK solvent only) is higher than for the cantilever coated with solution A (MIBK+DMSO solvents).This is due to the different solvent solutions used, which results in the cantilever coated with solution A having reduced amplitude.The cantilever coated with solution B shows more flexibility and greater amplitude up to 224.60 Hz and then drops in amplitude if the frequency is further increased [35].Another reason that the solution A coated cantilever has less amplitude, is that after the annealing the solution A coated cantilever shows visible bending, as shown in figures 4(a) and (b).It should be mentioned that the cantilever coated with solution B did not exhibit visible bending in its structure.The noticeable difference between the FRF results of the cantilevers coated with the above-mentioned solutions is the shift in frequency values.
The same vibration test was repeated with an uncoated KAPTON cantilever, in order to confirm if these polymer coatings affect the KAPTON cantilever's own natural frequency and amplitude.We can see in figure 3(a) a higher amplitude for the first mode of the uncoated KAPTON cantilever compared to the cantilever coated with solution B, and it is more than double that of cantilever coated with solution A. Figure 3(a) shows that the uncoated KAPTON cantilever has its highest frequency at 458 Hz with only 4 modes.
This vibration test was also repeated with 2%w/v (sample C) and 3%w/v (sample D) PVDF-TrFE concentrations in MIBK solvent shown in figure 3(b).These results are summarized in table 4. We can in general see that comparing the MIBK samples B, C, and D, that for vibration modes f 3 , f 5 , and f 7 , the frequencies are shifting higher as polymer concentration is increased.Compared with sample A, we see in general that this MIBK+DMSO solvent mixture has lower vibration mode frequencies.
The experimental parameters for all the measurements were the same, the shift in frequencies is not only due to the material properties and thickness of the films.There are other factors that may also affect for example, different mass and stress distribution, noise in the near vicinity, air flow in the room.

Modal analysis using simulations
At higher frequency modes, there can exist locations along the structure that have minimal motion.For the above vibration measurements, the laser vibrometer was positioned so as to not be at one of these locations.A modal analysis was done with MATLAB using the Euler-Bernoulli equation of forced vibration shown in equation (2), which shows the force excitation f(x, t) a shear force acting vertical to the longitudinal axis of the cantilever, w(x, t) is the lateral displacement, and causing bending moment M(x, t) in equation (3).Equation (3) shows the relationship between the bending moment, shear force, and force excitation [36].
where E is the Young's Modulus (2.07 GPa), ρ is the density (1.42 ×10 6 g m −3 ), I is the moment of inertia (6.75 ×10 −11 kg•m 2 ), and A is the cross-sectional area of the flexible cantilever.The natural frequencies in forced vibration solution are determined by superposition principle.In that case the deflection can be expressed as equation (4), where ( ) W x n is the nth normal mode, ( ) q t n is the generalized coordinate in the nth mode, and equation (5) represents the modes of the uniform cantilever.
Equation (2) was used to simulate a KAPTON cantilever of identical geometry as the tested structures, in this case, a uniform cantilever of length x = 5 cm clamped at x = 0.After solving for the general solution, we get the W(x) which determines the modes along the length of the flexible cantilever.Figure 5 shows the first five natural frequencies or modes.The approximate position of laser vibrometer is shown that is used in our experiment to measure cantilever motion.This location was selected so as to not be at any location with minimal modal vibration amplitude.

Output piezoelectric voltage test
Piezoelectric output measurements were studied using a rigid aluminum cantilever of dimensions 6 × 2 cm, because a rigid device would provide more consistent measurements for driven vibration tests.
Studies with the 6 × 2 cm rigid cantilever aluminum cantilever required the PVDF-TrFE piezoelectric material to be placed between two electrodes, in order to measure the piezoelectric voltage signal.The aluminum cantilever itself is employed as the bottom electrode.This is then coated with the PVDF-TrFE piezoelectric material.Next, an insulation layer of a photoresist (AZ 1518) is deposited using spin coated method to avoid the possible contact of top electrode with bottom electrode, because one polymer layer deposition may have voids.The speed for AZ 1518 was set at 1000 rpm and the output thickness was approximately 4 μm.This thickness is enough to avoid the electrical shorting of the top and bottom electrodes when depositing the top electrode.Finally, a small patch of silver paint is applied using brush as a top electrode and allowed to dry at room temperature.
Electrical measurements were performed with rigid aluminum cantilever spin-coated with PVDF-TrFE solutions A, C and D. The cantilevers were clamped on a ground base shaker and excited at 100 Hz because there are minimum mechanical constraints at low frequencies, and this is also due to fact that the calculated resonance frequency of 6 cm×2 cm cantilever is 992.56Hz.Table 5 shows the voltage amplitudes, electric field (V/t) and the sensitivity (E/g) for the cantilevers.Electric field is found by dividing the voltage by the thickness of each structure's polymer film.For sensitivity, the electric field is divided by the measured base excitation 'g' of each structure.Similarly, there exists a small difference in g between measurements, and so this is corrected for in the sensitivity calculation.
We can see that the measured voltages for all the devices are similar, but due to the different materials thicknesses we get different electric fields.The Al cantilever coated with solution A gives an electric field more than twice that of Al cantilevers coated with solution C and D because of the thickness difference.The sensitivity of cantilever coated with solution A is higher than cantilever coated with solutions C and D.

Piezoelectric constants measurement
The purpose of this experimental investigation is to measure the piezoelectric constants of the PVDF-TrFE film.In this study, the piezoelectric electrical constants were calculated by applying dynamic force to a cantilever.The setup is shown in figure 6.The cantilever of area 0.0011 m 2 was clamped on both sides and a force sensor was attached, which is then attached to the base shaker.The direction of the base shaker is set horizontally to produce horizontal directional vibrations to calculate the strain (S 11 ) in one direction.The cantilever beam was mechanically excited at 100 Hz, and the output piezoelectric voltage (V ) and force (F) were computed.The aluminum beam's elastic modulus is 70.3 GPa.The dynamic force was 56 N on the cantilever beam.In equation (6), the strain (S 11 ) was determined using the elastic modulus formula.Since the polymer film was formed on a quite thick piece of aluminum (1.56 mm), there will always be same strain (S 11 ) on the thin polymer materials formed from the MIBK and MIBK+DMSO solutions.Here, S 11 represents unidirectional strain.After excitation, we get the same output voltage (V 3 ) of 0.01 V for both structures.We know that the electric field intensity E 3 is the ratio of voltage over thickness, so we obtain the electric field intensity for the structures coated with DMSO+MIBK solution and only MIBK solution as 3.96 10 V m 4 / and 1.05 10 V m 4 / respectively.From the piezoelectric constitutive equation, we know the stress-voltage form shown in equation (7).
In the absence of external applied charge, D will be zero, and so in terms of the setup configuration the equation (7) will become equation (8).
where E 3 is the electric field in the direction of polar axis and q 31 is the piezoelectric coupling coefficient from the stress-voltage form, where applied stress (1applied stress) is perpendicular to the output voltage (3polar axis) and S 11 is the strain.Solving equation (8) using the values for E 3 (found to be 3.96 ×10 4 V/m for DMSO+MIBK solution, and 1.05 × 10 4 V/m for MIBK solution), we obtain = q 5.5 10 V m 31 10 / for the material formed from the DMSO+MIBK solution, and = q 1.45 10 V m 31 10 / for the material formed from the MIBK solution.For both cantilevers, the top electrode area on the cantilever was 3.48 ×10 −5 m 2 , and the capacitance was subsequently measured by the capacitance analyzer meter from Jacksking (Model# M 6013) to be 0.18 nF.The charge density D 3 (where D 3 = Q/A electrode ) is calculated to be 5.17 × 10 −8 C m −2 , and the total accumulated charge Q (where Q = CV 3 ) equals 1.8 ×10 −12 C. The piezoelectric coupling coefficient 'e' may then be determined using equation (9) using the stress-charge form of the piezoelectric constitutive equation.
Since e E .S = 0, as no external electric field was applied, e 31 can be calculated using equation (10).In summary, table 6 shows the measured piezoelectric coefficients using the dynamic force test.We can see that there is a difference in the electric fields of the two devices due to different thicknesses.This results in different values of q 31 .From these results, we can also confirm that the solution of DMSO+MIBK gives a higher voltage coefficient than the MIBK based solution.

Contrast of PVDF-TrFE material with PVDF based commercial sensor
The output voltage and sensitivity of the above discussed experimental devices was compared with commercial LDT0-028 K vibration sensor to compare their behavior.The flexible commercial sensor was purchased from TE Connectivity (Model # LDT0-028 K) and is made of a 0.125 mm polyester substrate, coated with a 28 μm thick piezoelectric PVDF polymer film with screen-printed Ag-ink electrodes (figure 7).The properties of the commercial LDT0-028 K vibration sensor are mentioned in table 7.
The exploration is based on the independent responses from both of the devices and does not count the materials itself.Because the commercial LDT0-028 K vibration sensor employs PVDF while the lab configured devices have PVDF-TrFE as a piezoelectric material and also the dimensions are not equal.The fabrication process of the lab configured devices are already explained in the above section (figure 8 (a)), while the commercial LDT0-028 K vibration sensor is attached on aluminum cantilever having same dimensions (6 ×2 cm) with the help of conductive glue (figure 8 (b)).The experiment parameters were kept the same for both devices.Vibration measurements were done by mounting the aluminum cantilever on to the ground base shaker, and excited at 100 Hz in similar manner discussed in table 5.
The voltage amplitudes shown in table 8 are the contrast between the cantilevers coated with solution A, D, and the commercial LDT0-028 K vibration sensor.We can see that the voltage at 100 Hz is similar to the commercial LDT0-028 K vibration sensor, but the sensitivity is very low.In summary, the lab configured devices do not have an advantage over the commercial LDT0-028 K vibration sensor in sensor voltage produced during excitation, however, their material sensitivity is higher.

Conclusion
In this study we have investigated the mechanical and electrical response of spin-coated PVDF-TrFE coatings formed using two different solvent solution compositions (MIBK only, MIBK+DMSO), and with different concentrations of PVDF-TrFE (1.25%, 2%, and 3% w/v).It was observed that the choice of solvent and solution concentration has an effect on the PVDF-TrFE material's mechanical and electrical properties.Experiments showed that employing only the MIBK solvent had little effect on the mechanical properties of the PVDF-TrFE material.By contrast, employing the combined MIBK+DMSO solvent resulted in a noticeable change in the PVDF-TrFE material's mechanical properties, resulting in the coated flexible KAPTON cantilever having reduced vibration amplitude.The measured piezoelectric voltage response of the PVDF-TrFE material showed that at 100 Hz, sensitivity (E/g) of the cantilever coated with solution A (MIBK+DMSO solvents) is more than twice than the devices formed using only MIBK solvent.The piezoelectric electrical coefficients were also calculated and found that the cantilever coated with the mixture of DMSO and MIBK solution gives high voltage piezoelectric coefficient (q 31 ) as compared to the cantilever coated with only MIBK solution.This means that DMSO is more efficient than MIBK due to its polar nature.The piezoelectric response of the experimental device was checked with a commercially available LDT0-028 K vibration sensor.It was observed that the output voltage is the same, but sensitivity of the lab configured devices is higher.

Figure 1 .
Figure 1.The FTIR results for the solutions A, B, C and D and the dashed lines are the positions for the β-phase peaks.

Figure 2 .
Figure 2. The schematic of the test setup.
The frequency response function (FRF) results are shown in figure 3(a).For the cantilever coated with solution A, we can see 7 frequency modes ranging from 11.85 Hz to 688.75 Hz.It should be noted that sample A has two frequency values 209.37 Hz and 231.87 Hz which are close together, this may be due to extra vector directions at that resonance.For the comparison study, only 209.37 Hz is considered because it has apparently higher amplitude.For the cantilever coated with solution B, we can see 5 modes of frequency ranging from 11.91 Hz to 720.50 Hz.The frequencies of cantilevers coated

Figure 3 .
Figure 3. Vibration response for KAPTON coated cantilevers at unit base excitation.(a) Response for uncoated KAPTON, and KAPTON coated with solutions A and B. (b) Response for KAPTON coated with solutions C (2%) and D (3%).(c) Zoom in on figure 2(a) for frequencies 60 to 90 Hz.(d) Zoom in on figure 2(a) for frequencies 60 to 100 Hz.

Figure 4 .
Figure 4. (a) KAPTON coated with sample A showing bending after the annealing.(b) Bending of KAPTON cantilever when clamped.

Figure 5 .
Figure 5. First five modes over the length of the flexible cantilever.

Figure 6 .
Figure 6.The experimental setup for dynamic force test to calculate the force on cantilever.

Table 1 .
FTIR of solutions A, B, C and D. The position of β-phase peaks with the absorption values.The absorption values have arbitrary units.

Table 3 .
Comparison with spin-coated PVDF-TrFE material in literature.

Table 4 .
Natural frequencies of Kapton cantilever at different polymer concentrations randomly excited from 0-1500 Hz at unit base excitation.

Table 5 .
Measured voltages under base excitation at f = 100 Hz.

Table 8 .
Measured voltages under base excitation at f = 100 Hz.