In the usual optical correlator, signals are stored in the from of transmittance variation on film. The use of film seriously limits the rapid processing of data since film filters cannot be adjusted to adapt to changes in target characteristics. The application of an ultrasonic light modulator (ULM) overcomes this limitation of the film filter:ULM can generate spatial amplitude and phase required for optimum filtering by the operation of electrical waveforms. A real-time optical correlator which used an ULM was introduced by Slobodin in 1963. In the Slobodin's configuration correlation is obtained by scanning an amplitude mask or replica of the signal with the first-order diffracted light of the ULM. The fixed reference mask should be changed in accordance with the incoming signal waveform. Some authors have suggested the use of a second ULM instead of the fixed mask. Atzeni and Pantani have recently reported some experiments using a dual-channel ULM inserted in a Slobodin's processor. In this correlator, the modulation factor of the ULM is small to eliminate higher-order diffracted light by a spatial filter. But it is possible to obtain a large modulation factor if differrant carrier frequencies are used. These carrier frequencies are set so as not to overlap the spectrum of the diffracted light. Fig. 5(a) shows the autocorrelation of rectangular pulses by which the same carrier frequencies are amplitude modulated. Using two different carrieres, the signal to noise ratio of the optically matched-filter could be improved as shown in Fig. 5(b). When the modulation factor of ULM is large, the distortion of the amplitude modulated light also becomes large, but that of intensity modulated light can be small. As the photodetector has quadratic characteristics it is very suitable for use in intensity modulated light as shown in Fig. 6. In these optical correlators the length of the ULM or the radius of the lens is finite, so that the duration time of the signal to be correlated is limited. When the reference signal has delay time, that of the signal particularly becomes short, because the two signals travel in the ultrasonic cells. Some experimental results are shown in Fig. 7.
Firstly, vertical and horizontal types of liquid atomization equipments were made, using 20 kHz ultrasonic vibration. As the ultrasonic sources some magneto-strictive ferrite transducers were used, attached with a stepped horn and a conical horn made of aluminum, and driven by a 300 Watt type ultrasonic generator (Figs. 1 and 2). According to the results of the atomization efficiency tests on these equipments, their atomization ability was above 20 liters per hour and it was not much affected by the frequency deviation of the ultrasonic oscillator (Figs. 4〜7). Thus it was confirmed that our atomization equipments may be applied to internal combustion engines, oil burners and so on. There have been many sugestions that the cause of the liquid atomization by ultrasonics is due to the unstable surface waves of large amplitude produced on the liquid surface. But there are also some who held the view that cavitation bubbles produced in the liquid partly cause the liquid atomization phenomena. In order to determine the cause of liquid atomization phenomena, a barium titanate transducer was attached to the top of the conical horn. When the test liquid was atomized on its surface, the output waveform of the barium titanate transducer was observed by an oscilloscope and analyzed by a frequency analyzer. Moreover, an aqueous solution of luminol was used as a test liquid and we examined whether the so-called sonoluminescence could be observed or not in the dark field. As a result we were able to confirm that cavitation bubbles were not produced in the liquid in our experimental conditions liquid atomization. Secondly, we attached a test piece, as shown in Fig. 12, to the top of the conical horn. Driving the transducer of the atomization equipment by means of ultrasonic pulse waves, we experimented in order to make the test liquid atomize (Fig. 11). The results obtained confirmed that when the frequencies of the control pulse wave were in the range from 60 Hz to 100 Hz, the efficiency of liquid atomization was much decreased (Fig. 14). In the case when the frequency of the control pulse was equal to 120 Hz, photographs of the surface waves of the liquid in the atomizing state ware taken (Fig. 15). These photographs show the fact that the standing waves were produced in the liquid surface on the ringed part of the test piece and that minute droplets of the liquid were driven off from the top of the waves. We rationalized these experimental results by means of an approximate theory and thus explained the reasons why such stationary, standing waves were produced.