- 3 - 257 



case aiscusstd by Taylor(l'), this aiffracted wave is absent, and tension, havin:) once set In, persists 

 indefinitely,. It is clear that such a theory underestimates damage, and this cin easily Be seen 

 to be true even if we issume that the plate leaves the water as soon as tension sets In. 



If water cannot stind tension, then tne plate will leave the water as soon as the pressure 

 drops to zero, furthermore the water will cavitate as the tension-pulse set up by reflection at the 

 plcite travels outwards. Initially, the cavitated water will all be moving slower than the plate, 

 but as the plate is brought to rest by the action of elastic and plastic stresses the water eventually 

 catches up. For a very large plate it is clear that damaje will be enhanced by this effect, and 

 this is the case which is discussed below. Fox and Rollo have pointed out that, if cavitation occurs 

 at a finite tension, the contribution to damage will vanish discont inuously if distance and weight 

 of charge are changed continuously and bring the peak explosion pressure Below a certain value. 

 For a finite plate conditions will jump discont inuously from the "cavitation and follow-up" to tfte 

 "diffracted wave" regimes in the same sort of way, and in such a case it Is at prssent an open 

 question which mechanism will be th» more damaging. In any case, It seems clear that diffraction 

 will predominete for i small plate, and cavit=ition for a large one, but it may be a matter of some 

 difficulty Ic deal with the intermediate case, where both mechanisms are probably occurring tcjether, 

 cavitation at the centre, and diffraction near the edges. 



Ext>erimental evide nce. 



It is perhips appropriate to mention at this point that the existence of cavitation in the 

 water under such conditions has been established photographically beyond all possible doubt by 

 Wright, Campbell and Senior in this country. U.S. Reports U.E. 18 and 19 are also relevant. In 

 the latter papers it is also shown that cavitation is just prevented from occurring if conditions are 

 such that the diffraction wave can reacn the centre of the plate at the moment when the total pressure 

 would otherwise have dropped to zero. Work on the motion of plates due to an explosion which also 

 gives evidence that cavitation occurs is in hand at Road Research Laboratory (8), at the Taylor Model 

 Basin(9) and at Admiralty Undex Works. This work is in qualitative agreement with theory, But until 

 a method of allowing for edge effects has been worked out a quantitative comparison is not possible. 



Little or no experimental evidence on the behaviour of sea-water under high rates of change 

 of tension seems to be available. It Is probably quite different from that of air-free water, which 

 is what is usially studied in the laboratory. Even the static tension which purs water will stand 

 does not seem to be known with any certainty. 



Theory for an infinite flats . 



(l) Assumpt ions made . 



As stated above, we shall be dealing with an infinite plate, and we shall also consider the 

 case of a plane wave at nornel incidence, so that conditions become effectively one-dimensional. We 

 alio assume that:- 



(a) The water cavitates at, or soon after, the instant at which the pressure drops 

 to zero. 



(b) When the water catches up the plate after cavitation, it Is brought to rest 

 (relative to the plate) and does not rebound, so that a layer of "reconstituted" 

 water gradually builds up on the plate. This water is treated as If it were 

 incompressible. 



Regarding assumption (a), there is considerable conflict of evidence on what the facts really 

 are, and it is suspected that more than one parameter may in fact be involved. For example, it nay 

 wtll be that the rate of change of tension is important, as well as the tension itself. 



Regarding assumption (b), the occurrences at the plate are probably very complex. Air-bubbles 

 may persist so that cavitation docs not disappear at once and pressure waves may be set up in the 

 "reconstituted" water as the thickness of tne layer grows. As the layer of "reconstituted" water 

 turns out to be quite thin, any waves in this layer would probably soon Be averaged out by reflection 

 at the plate and at the surface. Very recently(9), seme of tne observed "kinks" in the deflection- 

 time curves of diao^'rdjms have Been attributed to these waves. 



(2) 



