In this report, we took notice of the degradation effect of the 28kHz ultrasonic cavitation produced in polymer solutions such as the polystyrene-benzene or polyvinyl pyrrolidonwater solution, and examined the effects of the static pressure and the excitation mode of a transducer to the polymer degradation. Ferrite transducers (28kHz) are attached to the 1/2 wavelength-resonant step horn of stainless steel, and they are excited by an RC-oscillator or a function generator and a broad band power amplifier. The sample volume of the liquid tested is about 18cm^3 and the temperature change during the experiment was held at ±1℃ by a cooling system that was set outside the pressure vessel. Using the polystyrene-benzene solutions (whose mean molecular weights are No. I=3. 23×10^5, No. II=1. 81×10^5, No. III=1. 76×10^5), we examined first the change of intrinsic viscosity with the ultrasonic irradiation time (Figs. 2 and 3), and obtained the results that the intrinsic viscosity decreases rapidly within 30 minutes of the irradition time and the rate of decreasing of viscosity has its maximum at a static pressure value of 9kg/cm^2, which is small under a high temperature in spite of the high intensity of ultrasound. Next, under the condition of constant irradiation time we examined the change of intrinsic viscosity with the static pressure value, and obtained the results that the polymer degradation effect of cavitation has its maximum under a static pressure value of several atmospheres and it is weakened above that pressure value, but is recognized up to the elevated pressure value of about 20kg/cm^2 in spite of using a relatively weak intensity of ultrasound (Figs. 4 and 5). As shown in Fig. 4, the rate of decrease of the intrinsic viscosity increases with the irradiation time and that a tendency to the static pressure value is nearly the same regardless of the irradiation time. As shown in Fig. 5, in case of the same temperature of experiment, the pressure value at which the polymer degradation effect of cavitation has its maximum increases with the intensity of the ultrasound. Under the conditions of the same intensity and temperature, we cannot obtain a sufficient viscosity decrease as the concentration of polymer solution becomes greater. Using the polyvinyl pyrrolidon-water solution (whose mean molecular weight is about 7. 0×10^5), we examined first the change of intrinsic viscosity with the ultrasonic irradiation time as a parameter of the ultrasonic intensity (Fig. 6), and obtained the results that the intrinsic viscosity monotonously decreases with the irradiation time and the rate of decrease of the intrinsic vicosity is very large in spite of a relatively weak intensity of ultrasound. Next, using a continuous irradiation and a burst pulse irradiation of ultrasound, we tried to examine the polymer degradation effect of cavitation. Under the atmospheric pressure we compared the degradation effect between the continuous irradiation and the burst pulse irradiation (burst time is 10msec and rest time is 20msec) in case of the same net irradiation time (Fig. 7). The results indicate that, in case of burst pulse irradiation, the intrinsic viscosity generally decreases more rapidly than in case of continuous irradiation, and the inclination is shown clearly under the condition of a short irradiation time. When the net irradiation time is 15 minutes and constant, we examined the change of intrinsic viscosity of the PVP-water solution in the temperature range of 20℃, 30℃ (Figs. 8〜10) and obtained the same results as mentioned above.
In a normal listening, a complex tone is perceived as a single tone with one difinite pitch corresponding to the pitch of the fundamental frequency. However, the pitch of a complex tone consisting of a small number of higher harmonics and without the fundamental is often ambiguous. An experiment was carried out in four subjects adjusted the frequency of pure tone to match the pitch of an amplitude-modulated (AM) tone, in order to study information and the mechanism for pitch perception of tones. It was found that several different pure tones were perceived to be equal to an AM tone in pitch (Fig. 3 and Fig. 4). The frequencies of the pure tones corresponded nearly to the constituent frequencies, the combination tone frequencies of the type 2f_1-f_2 and the subharmonics of them (Fig. 4). Considering the experimental result, the pitch perception mechanism was inferred as follows: There exist two kinds of informations as the cues for pitch perception. Place information are tonotopic organization on auditory neuron layers (Fig. 5), whereas temporal information is formed by time intervals of neuronal discharges (Fig. 6). Each information provides several different cues for pitch perception. According to the folded histogram (Fig. 9), the response of an auditory primary neuron to a pure tone stimulus shows a systematic locking to the cycle for frequencies up to some 4-5kHz. On the other hand, tone chroma changes regularly with one octave periodicity for frequencies up to some 4-5kHz. However, the regularity of the relationship between frequency and chroma disappears over that frequency (Fig. 8). These physiological and psychological knowledges suggest that tone chroma is based on the synchronization between tone stimulus waveform and time pattern of neuronal discharges. Then it is inferred that pitch in a narrow sense (tone height) is mainly given rise to by the place information and tone chroma is given rise to by the temporal information (Fig. 10). The reason why several different pure tones were matched to an AM tone in pitch perception can be interpreted that the subjects perceived pitches paying attention to specific cues among many different cues (Fig. 10).
It seems that there has not been any practical and convenient method by which two kinds of internal loss, magnetic loss and mechanical loss, can be measured and estimated for magnetostrictive transducers excited in large amplitude. In this paper, a new method is proposed which is composed of an analogy of the usual open-short circuit test generally used for power transformer loss measurement. The proposed idea is discussed and examined on 28kHz rectangular ferrite transducers excited in large amplitude. As the method analogous to the open circuit test, the transducer undertest is clamped by the magnetostrictive force of another transducer with the same specification bonded on the acoustic radiation surface. The magnetic loss is immediately measured from electric input power measurements on the transducer at various alternating driving magnetic flux amplitudes (Fig. 2 and Fig. 8). As the method analogous to the short circuit test, the transducer under test is excited at B-type resonance frequency with no acoustic loading. The mechanical loss is immediately measured from electric input power measurements at various vibrational amplitudes of the radiation surface (Fig. 3 and Fig. 10). Leakage impedance must be taken into account in the present measuring method. As a countermeasure for this problem, leakage inductance is compensated by a mutual inductor circuit (Fig. 4). This method is examined by setting a simple searching coil (Fig. 5 and Fig. 6). Excitation winding resistance, which forms a real part of leakage impedance, is immediately known from loss measurement of an air core winding with the same size, and the loss is to be subtracted from the data. The whole internal loss which arises in a practically operating transducer is estimated by the summation of two kinds of internal loss measured by the present measuring method. This is examined by comparing the summation with the actual total loss measured by introducing a dummy load test based on a vibro-meter method (Fig. 13). The result shows a good agreement in the range of vibrational velocity of the radiation surface up to 7cm/sec rms (Fig. 14).
In predicting and controlling the propagation of noise in the open air, it s very important to take account of the shape of a building that is the noise source, the cross sections of railway track and road, as well as barriers and surrounding buildings. Then, the theory of the free-field diffraction of a spherical sound wave by a thin half-plane is basically necessary. The early approximate diffraction theories of Fresnel and Kirchhoff are well known, where it is supposed that the linear dimensions of the opening are large compared with the wave-length and the diffracting plane is perfectly absorbent. The rigorous solution of a diffraction problem of a plane wave incident on a thin half-plane was first given by A. Sommerfeld. On the basis of his solution H. S. Carslaw solved the three-dimensional case of a spherical wave, and H. M. Macdonald gave a useful form of the exact solution. However, Maekawa's experimental curve is often used for the calculation of noise reduction of a barrier, which is obtained by experiment to satisfy the Kirchhoff's approximate condition. This paper makes clear the behavior of Macdonald's exact solution by a numerical calculation which seems to have been surprisingly unknown, and presents detailed data of experiments, many of which do not satisfy the Kirchhoff's approximate condition. The theoretical and experimental results are shown in Figs. 2-13 as follows: 1) According to the Macdonald's exact solution there is a contribution of image source to diffraction, and sound attenuation by a screen ATT. is given by three normalized distance parameters, namely, R_N, distance normalized by a half wave-length from the point of observation to the source; R^^_N, to the image source, and R_<1N>, over the edge of the half-plane to the source. 2) The Kirchhoff's approximate solution is in agreement with the expression where the contribution of image source is neglected and R_<1N>≃R_N in the exact solution. 3) All of the experimental results are in very good agreement with the exact solution, no matter where the sound source and the point of observation are located. 4) The Maekawa's experimental curve deviates largely from the experimental results and the exact theoretical results when either the source or the point of the observation is comparatively near the half-plane. 5) By the useful simple approximate expression of the exact solution which is accurately given by Bowman & Senior, the approximation error is less than 0. 5dB only if λ/4<R_1, here R_1 is the minimum distance from the point of observation over the edge of the half-plane to the source. Consequently, we can predict and control the propagation of noise in the open air very accurately by applying the Macdonald's exact solution or that approximate expression.