622 - 2 - 



(ii) So^l 



Initial velocity at centre v^ ft. /sec. 



•lone factor n* 



When tnc Jome analysis methoa is aocliea the oeotn is calculateo by taking the acme height 

 at Jiff-r^nt acme raaii ana orooortioninj tnem to the dome height at the centre. This is done 

 at thr=- sm-11 time intervals from the first ao=e.,ranc2 of the don,e. It is founO that the calculate: 

 -oth varies both -long the raaius an; «ith time. The variation is an indication of the aistortion 

 of tl« aome ano as a measure of this the dome factor, n, has oeen oefined aS:- 



10" (rean aecth calcjlateO at a oome raoius equal to 3/2 0,. 



n =. g^ ° 



° m',n oeoth calculatea at a dome raoius equal to i ) 







In oractice It is found that the oart of the dome between radii equal to 0.5 and 1.5 x th? 

 occth of detonation is tn,^ most satisfactory for measurement and calculation. 



The factor n has ?. oositivc value when the calculatea deoth tends to increase outwnras from 

 the centre ana is associ'^tec with a lorn" which is r£latively high at the rim. When n is negative 

 the greatest aeoths arp calculatea at the inner rrii anJ the aome is relatively high near the centre. 



(iii) 



Time from detcnaticn to first acoearance 



of the clume through the dome t^ seconds 



height of dome when olume acoears h feet 



mean initial velocity of the olume _ _ c^^ 



Di scussion of results. 



The results ootainea from the analysi;. of the cine records are given in Table 1. The 

 range of deoths ccver^a in the results examined is equivalent to a variation of from ju st over 

 »n na,ri„ 7n foot with a ^00 lb. charje if the deoths are corrected in orooortion to \/ 



arly 70 f-:et with 



300 lb, charge if the deoths are corrected in orooortion to /charge weight. 



On the same C^sis of :t 300 lb. charge the variitions found in v^o are from under 2.000 to 

 6,000 square feet cer second for charges in deeo water and from U.OOO tc 8,CC0 square feet oc-r second 

 fcr charges ;n the bottom.. 



The bread relati;n f;una between Jjme velrcity anc olume velccity which was the first result 

 in this latest analysis is shown in Figure 1. Included in this figure are Kark VII decth charge 

 results described in the original reoort on the oome analysis method. It will De seen at once that 

 for all weights jnd deoths of charges, and for charges both on and off the bottom there is a general 

 tendency for v and v to increase together. In the range consiaerea v is on the average equal to 

 0.75 V although the scattering cf the ooints is equil tc a ± JOJ variati:;n en this. Despite the 

 wide variation there is clear evidence frrm the figure that a relation exists between the d^me and 

 the olume. 



The next steo in the analysis was to discover if the error found in the depth determination 

 were also related to the olume characteristics. The oercentage error in the oeoth calculation 

 was plotted on the basis of the plume-dome velocity ratio V/v^, but the points were found to be 

 widely scattered and no relationshio was aocarent. This suggested that if a relation existed it 

 involved a second factor. It has been observed that the domes were not always uniform in shape »nd 

 that in some instances large variations occurred in the depth calculated at different dome radii. 

 It seemed possible that these variations would th;n\selv=s influence the accuracy of the mean oeoth 

 figure. Eximples of the variation in ieoth calculatea across the 3ome raaius are shown in Figure 2. 

 The three results illustrateo are those which give the highest positive and negative values of the 

 Oome factor which were obtained in these tests and a zero value. 



In Figure 3a the percentage error in the depth calculated by the original dome analysis 

 method is plotted on the basis of the plume factor 7Sv^ and each result is labelleo with the value 

 of the dome factor n founo in its calculation. It will be seen that without the aome factor n 

 there is no relationship apparent but that with the oome factors shown the results at once align 

 themselves in banas across the figure. This is emohasisea by the cross lines which have been drawn 

 in the figure to correspond tc the average relation between percentage error, n and ^/v shown by 

 the results. 



This 



