Excitation without Proof Mass

135 The composite structure of sample D, with width 9 mm, length 1 cm, and total thickness of 196 m was weighed at 0.11g. This gives an average density of 6240 kgm 3 . Sample C and D series were found to have a thickness of 208 m and 192 m respectively. Both samples D and C series were assumed to have the same density. Since the natural frequency structure is inverse proportional to the length of the cantilever structure, therefore the Youn g’s modulus of the structure can be estimated by using equation 3-8 . The calculated Young’s modulus of sample C and D series are 3.78  10 10 Nm 2 and 1.17  10 10 Nm 2 respectively. The total Q -factor, Q T of the structure can be determined experimentally by exciting the free-standing structures over a range of frequencies close to the fundamental resonant frequency to determine the value of the full bandwidth at half maximum electrical output power, then substituting this value into equation 6-7. Figure 6-6 shows that the calculated values for Q T of the samples lie in the range 120 to 215, with the largest value associated with sample D3, which is roughly a square shape. Shorter or longer cantilever lengths do not appear to exhibit the same Q -factor as those having a square structure. This is because shorter or longer cantilever structures suffer losses at different rates and with different dominant factors. The energy dissipation losses at the support are dominant for a shorter structure [108], while air damping losses become dominant for longer cantilever structure [114]. With the measured Q-factor value, the total damping ratio for the samples was calculated to be in the range of 0.002 to 0.005. Figure 6-6: Q T as a function of cantilever length. 136

6.4.2 Excitation with Proof Mass

Attaching additional proof masses to the cantilever beam can further reduce the resonant frequency. As an example, the natural frequency of sample D5 is reduced from 505 Hz to 68 Hz with proof masses of 2.22 g. as shown in Figure 6-7. The measurement results show that the natural frequency of the structure is not affected by the distribution of the proof masses. Figure 6-7: Experimental results in agreement with theoretical calculation for resonant frequency as a function of mass for sample D5 with length 13.5 mm. The Q -factor of sample D5 was reduced from about 185 to about 30 when a proof mass of 2.2 g was attached as shown in Figure 6-8. The mechanical damping ratio obtained from the calculation by using equation 2-8 are in the range from 0.003 to 0.016 when a range of proof mass up to 2.2 g were attached to a cantilever of length 18 mm. The coupling factor appears to be increasing rather linearly with the proof mass from 0.06 to about 0.2, as shown in Figure 6-9. 137 Figure 6-8: Q T as a function of mass for sample D5 for four different proof mass distributions. Figure 6-9: Coupling factor as a function of mass attached to a cantilever with length 18 mm.

6.5 Electrical Characterisation

The same series of samples was used to investigate the electrical output performance from the piezoelectric cantilever structures. A modest electrical power output a few nano-watts was produced when the composite unimorph structure was operated in its bending mode. The output power is affected by the distance from the centroid of the piezoelectric material layer to the neutral axis of the composite cantilever, d . The samples used in the experiment have a d value of 6 µm, which was calculated from equation 3-22 by using the parameters in Table 6-1 and with the assumption that the