696 _ 2 _ 



Experimental results . 



The variation of amclitufle with time at Boints at the front eage ana at the centre of 

 the reference boara (ooints at norizontal aistancss of 52 feet anO 55 feet from the charge) are 

 given in Figures 2 and 5. 



It is seen that in all cases exceot with the charge 2 feet fleec, the oositlv? amolituae 

 was greatest for the second wave. This w=is oDserveO oreviously for 2-oz. ana 2-1 b. charges. 

 It is also to Oe noted that the maximum negative amolituae was equal to :r greater than the 

 maximum pcsitivf amolituce. 



The oerioaicity of the waves was about the same for all charge Ofoths. The period in 

 the region of tns first ana second waves was b;tween 2.5 ana 3 seconds and decreased for the 

 subsequent smaller w-ives. 



The velocity of the waves across the board was about 8 feet/second from which the wave 

 length of the main wave is estiitated to be about 20 to 25 feet. 



In Figure 4 oositive amolituoes of the second waves in feet oivicea oy the fourth root 

 of the charge weight M in lb. are shown as functions of charge deDth/(maximum bubble raoius). 

 Includeo in Figure u are previous results for 2-oz. and 3-lb. charges of =o1ar Ammon gelignite 

 and for 300-lb. T.N.T. deoth charges. The results confirm previous conclusions that the 

 optimum aoDth for wave aroauction was of the order cf 0.6 times the maximum bubble radius. 



In a previous enquiry it was considered that approximate scaling might obtain if the 

 scale of the model were maoe proportional to the fourth root of the charge weight provided the 

 ratio of deoth cf charge to maximum bubble radius were kept constant. The "olar Ammon gelignite 

 results in Figure 4 show that the wave amplitudes from the larger charges aid not increase as 

 much as expcctea accrraing t: their scale relr.t ionshi s, =1 result which suggests that the larger 

 charges may not have been so efficient in wave making. It is to be noted, however, that the 

 results using 300-lb. Charges were in fair agreement with the present tests. This may mean 

 that orovidsd large enough charges are used, thert is not much change in efficiency of wave 

 oroduction as the scale of the model is incro-issd. 



