Recent Observations on Cavitation and Cavitation Noise

This paper is primarily concerned with the acoustics of traveling bubble cavitation around foils or headforms. We begin with observations of individual bubbles and the acoustic signals they emit, our purpose being to identify areas of research which would enhance our tlriderstaridirig of the history of individual bubbles. Then we present some riumerical integrations of the Rayleigh/Plesset equation for the same flows. The comparison is encouraging in terms of future synthesis of the noise by ana l~ t ica l means. Finally, bubble interaction efFects which were omitted earlier are discussed and some recent anal) tical results including these effects are presented. Ilather t.han attempting a cornprehensive review of tho state of knowledge of cavitation noise, this paper will focus on several issues wllerc our understanding of the basic physical phenonlena is, a t best, quite limited. Our remarks will be confined t o the noise generated by bubble cavitatiort and we will not attempt to deal with the added corrtplicatior~s associated with fully or partially developed cavitatior~. 'l'he current state of knowledge of the noise generated by bubbly cavitation is thoroughly reviewed by Blake :1986] ant3 well represented by the proceedings of the two previous syrr~posiurns in this series, so it rnay be rriore 11sefu1 to focus on several key issues in order to identify areas which would benefit from further attention. T l ~ e ~rtost f:lndarnental approach to cavitation noise begins wit.11 tho nrlclci population of the incoming stream. By constructing the dynamics and acoustics for ea.ch ir~divid'lal size of nucleus, one should in theory be able to combine this information with the nuclei number distribution lo produce all the required information on cavitation noise levels and spectra. This, of course, assumes that bnhbles do not interact acoustically or hydrodynamically. Such interactions will be consideretl in a later section. For present purposes, however, bubble interactions will be neglected. Parenthetically we remark that, despite the fact that cavitation phenornena are now recognized to be initirnately related to the population of "cavitation" nuclei in the incornir~g liquid stream, it is st.ill too often the case that this distribution goes unmeasured. We must insist on this documentation for all cavitation experimeiits. Typical nuclei number distribution functions, N ( i to) where K,, is the nuclei radius in meters, a re shown in figure I ; N ( R o ) is defined such that the number of nuclei with sizes between Ro arid Ho + dlZo is N(Ro)dlCo. W e shall return a little later to discussiort of the effects of the nuclei number distribution. The present remarks will be confined to the cavitatior~ noise produced by bubble cavitation in floivs around bodies such as headforms or hydrofoils. Thus wt: shall be concerned with the behavior of cavitation bubbles in the presence of various flow phenornena such as pressure gradients, boundary layers, separation arid turbulence. A great deal of research has been dorie on the dynamics and acoustics of cavitation bubbles in qliiescerit liquid (Knapp, Daiiy and Ilamrnitt 119701). It is known, for example, from both experiments and analysis that .rvhen a bubble in a quiescent liquid collapses close to a solid t~ou~idary it microjet forms on the bubble surface furthest from the solid boundary and reaches very high velocities (Plesset and Chapman 11970j). 7'he current state of knowledge of this phenornena has recently been comprehensively reviewed by Blake and Gibson !1987]. Those authors reflect. a current body of opinion in which these microjets are belieled to be responsible for both the rnaterial damage and the noise createti by cavitation. Even for bubbles in a quiescent liquid, this view 1:laj. neeti to be modified in the light of the recent observatiorts by Kimoto 119871. Ile simultaneously took high speed motion pictures arid rrladt: local pressure ~neasurements on the surface beneath a collapsing cai itation bubble in it quiescent liquid and observed the instantaneous loading on the silrfacc!, ]lot orily a.s a result of the microjet, but also as a result o f the shoc:k wave generated when the remnant cloud of bublles collapses. It, is significant