FM相関法による分散波の群速度測定 : FM信号を利用した相関法による水中音響計測[II]

抄録

When the medium of propagation is dispersive, the pulse method is used for the measurement of the group velocity because the pulse is propagated by energy-transmission, while the standing wave method is used for the measurement of the phase velocity. The correlation technique with some advantages has also been reported to be useful for analyzing propagation of energy. In this case a band-limited white Gaussian noise or AM signal where the Gaussian noise is modulated by M-sequence signal is applied as the measuring signal, and this method can be qualitatively explained as a system applying the envelope detection of the noise signal provided with information of the group propagation. In this paper we examine the use of FM signal which is effective when the mechanical Q of the transducer has such a high value as in under water acoustics, as reported in a preceding paper in detail, because its spectrum can be changed to compensate for the influence of the mechanical Q by using the random signal named binary Poisson process as a modulating signal. However, as the FM signal has a constant amplitude, it is inevitable to adopt a different method from the envelope detection mentioned above and so the system applying the heterodyne detection is investigated in Section 2 (Fig. 2). The following facts are recognized: (1) It is theoretically analyzed that the FM correlation technique applying the heterodyne detection enables us to measure the group velocity, and it is confirmed experimentally that the measurement of the group velocity in a fluid cylinder with a free boundary is consistent with the theoretical curve (Fig. 8). (2) Additionally, even if a signal of a high frequency range is applied, a correlator operating in a low frequency range can be used by changing the frequency of the local oscillator. In Section 3, we investigate the polarity coincidence correlation function in which a nonlinear operator is added. The results obtained are as follows: (3) If the arriving time interval of each receiving signal propagating through multiple sound paths is longer than a certain value, the group-transmission time can be measured discriminately from the cross-correlation curve as in the case without a hard limiter. This is also confirmed experimentally by measuring the sound propagation through a brass pipe (Fig. 5). (4) But, as a strong signal suppresses a weak signal in this case, the amplitude of each receiving signal is not in proportion to the corresponding correlation coefficient (Fig. 7). Accordingly we cannot directly measure the relative intensity of each receiving signal from the cross-correlation coefficient.