3S2 SECONDARY PRESSURE WAVES 



posite wave formed by the direct wave and its reflections, as sketched in 

 Fig. 9.5 for a gauge and charge position nearer the bottom than the sur- 

 face. If comparison with theory is to be made simpl}^, the direct wave 

 profile is to be preferred, although the composite wave corresponds to 

 the actual state of affairs for the particular positions of the point of ob- 

 servation and boundary surfaces. The durations of bubble pressures 

 are so long that the difference between the two profiles is appreciable in 

 many interesting or necessary situations, and some means of correction 

 for reflections is necessary in order to compare with theory. 



SURR\CE^>^^ 



Fig. 9.5 Effect of surface and bottom reflections on observed bubble pulse 



pressures. 



Rather rough corrections can be made using point-by-point calcu- 

 lations, starting at the first rise in pressure and adding positive or nega- 

 tive contributions at later times determined by the path differences. 

 The errors in this procedure are evidently emulative, but are usually not 

 serious up to the maximum. At later times, the errors rapidly become 

 worse if any boundary is close, and if cavitation occurs near the surface 

 no correction is possible. An expedient sometimes adopted is to meas- 

 ure to the peak and multiply impulse and energy flux density values by 

 two, which is roughly equivalent to theoretical assumptions. The cor- 

 rection process and estimation of the peak time introduce random errors 

 of the order of ten to fifteen per cent, being smaller for peak pressure. 



The errors in analysis and interpretation are in addition to experi- 

 mental errors arising from cable signal and incorrect distances. The 

 magnitude of errors from all known causes is of the order five to ten per 

 cent for peak pressure, ten to thirty per cent for impulse and energy, 

 being greater for large charge measurements and the ''direct" wave, 

 corrected parameters. These errors would occur to some extent in 

 measurements even if the phenomenon itself were perfectly reproducible. 

 This, however, is not the case, as the discussion of part C makes clear. 



C. Lack of reproducibility of pressures. Although measurements of 



