normalized to give the weight fraction within the boundary layer displace- 

 ment thickness. It is evident that the very small concentration of only 

 2.9 ppm is sufficient to remove the laminar separation on the hemisphere, 

 and significant forward movement of transition on the Schiebe body is 

 accomplished, with somewhat larger concentrations. The main thrust of van 

 der Meulen's findings, then, appear to be confirmed, and the cavitation in- 

 ception inhibition on the hemisphere is traced to an overall viscous effect. 



Turbulence Levels 



We now, of necessity, touch briefly upon this ubiquitous standard 

 problem in experimental fluid mechanics; namely, the influence of free- 

 stream turbulence on flow past bodies. But, with the restricted emphasis 

 herein on cavitation inception, our concern is primarily to determine if 

 the stimulated transition causes a laminar separation to disappear or if 

 there is any direct effect on cavitation inception of bodies not having 

 such a separation. There are, of course, semi-empirical rules for the 

 turbulence effect (e.g., Hall and Gibbings 1972, Van Driest and Blumer 

 1963) and the turbulence-modified amplification method of Mack (1978). But, 

 as Reshotko (1976) has made clear in his review, the coupling between these 

 frees tr earn disturbances and boundary layer developments is unclear — 

 particularly for large disturbances that may bypass the linear growth pro- 

 cess. Thus, as in many areas of naval hydrodynamics, recourse must still 

 be made to the laboratory to study particular experimental situations. 



It has not been customary in much hydrodynamic work to quantify 

 turbulence levels — let alone their power spectra — although a few such data 

 on several water tunnels are now becoming available through the ITTC 

 reports (12th ITTC Cavitation Committee). Thus, it is no surprise that 

 there is no account of the effect of freestream turbulence on flow past 

 bodies of interest in naval hydrodynamic applications. One problem in 

 large hydrodynamic facilities has been the expense of the control of this 

 real fluid feature. For this purpose a water tunnel having a low level of 

 freestream turbulence to begin with is essential. The key features of one 

 such research facility are shown in Figure 28. Some features, particularly 



49 



