Transient high-frequency ultrasonic water atomization F. Barreras, H. Amaveda, A. Lozano Abstract An experimental study was performed to im- prove the understanding of the characteristics of ultrasonic water atomization when excited with waves in the MHz range. In the present experiments, small volumes of water were atomized, observing the temporal evolution of the process. Typical diameters of the resulting droplets are of the order of a few microns. To visualize them, images were acquired with very high magnification. Appropriate lenses were used to enable high resolution at a distance from the flow. Droplet size distributions were also calculated with a Malvern diffractometer. Droplet exit velocity was measured using particle image velocimetry. It was noticeable that, as the remaining liquid mass deposited over the ultrasonic transducer decreased, the atomization characteristics changed, and a second peak of larger droplets appeared in the size distribution function. This phenomenon is related to the change in the curvature of the liquid surface. Although results are not conclusive, it appears that, under the conditions in this study, some observations about droplet formation are better described by cavitation phenomena rather than by the simplified surface wave theory usually invoked to explain these processes. 1 Introduction Ultrasonic atomization is a very effective way of generating small droplets. Two approaches are common in this con- text: passing the flow across a standing ultrasonic wave (Bendig 1988) or depositing the liquid over an ultrasonic transducer. The second case originates a fine mist of droplets that are ejected from the transducer at very low velocity. This procedure is going to be studied in the present work and is used, for example, in many com- mercial humidifiers. The present work is subject to some specific charac- teristics. First, the atomization is restricted to water. Sec- ondly, the ultrasonic frequencies were selected in the MHz range, because interest was focused on micron-range droplets. Finally, in contrast to most experimental studies reported in the literature, for some measurements discrete water volumes were atomized without supplying a con- stant flow to keep a fixed liquid level over the transducer. This is why atomization in these experiments has been described as transient. All these conditions are represen- tative of those that would occur if an ultrasonic atomizer were used to administer by inhalation a metered dose of a drug diluted in an aqueous solution. The objective of this work is to study the temporal evolution of the process, and the characteristics of the generated spray. 2 Historical review The possibility of generating a cloud of droplets by means of ultrasonic waves was first reported by Wood and Lo- omis (1927). Two different mechanisms have since been invoked to explain the ultrasonic atomization, capillary waves, and cavitation. Scientific publications on the sub- ject can be divided into two groups, with different ap- proaches depending on which one of these mechanisms is considered to be responsible for the spray formation. However, the interaction between them and the limits in which one could predominate over the other depending on the different atomizing situations are still not clear. The first studies on stationary waves on the free surface of a liquid mass subjected to periodic vertical forcing were reported by Faraday (1831). In 1871, Kelvin derived the well-known equation for capillary waves k ¼ 2pr qf 2  1 = 3 ð1Þ where k is the wavelength, r is the surface tension coeffi- cient, q is the liquid density, and f is the frequency of the surface waves (Kelvin 1871). This research was continued by Rayleigh (1883), who modified Kelvin’s equation and derived the expression k ¼ 8pr qF 2  1 = 3 ð2Þ where F is the forcing sound frequency. It is to be noted that the relation f=F/2 was obtained empirically from experimental measurements. Received: 23 October 2001 / Accepted: 25 February 2002 Published online: 5 June 2002 Ó Springer-Verlag 2002 F. Barreras, A. Lozano (&) LITEC/CSIC, Maria de Luna, 10, 50018-Zaragoza, Spain E-mail: alozano@litec.csic.es H. Amaveda Universidad de Zaragoza, Maria de Luna, 3, 50018-Zaragoza, Spain This project was partially supported by the Diputacio ´n General de Arago ´n, under contract P104/97. The PIV analysis software was provided by Dr. Julio Soria from Monash University. Experiments in Fluids 33 (2002) 405–413 DOI 10.1007/s00348-002-0456-1 405