Test Conditions: Significant Wave Height, 16 ft 

 Head Seas 

 Sliip Speed, About 6 knots 



Tfie sfiip pitclied at a rate of about 320 variations per liour. 

 The test sample consisted of 210 variations in pitch angle. 



Theoretical Rayleigh Distribution 

 E = 6.3 (degree )2 



Experimental Histogram 



Variation in Pitch Angle in degrees 

 Fig. 8 Distribution of Variation in Pitch Angle, Aircraft Carrier 



Table 5 Statistics of Log-Normal Distribution of Wave- 

 Height Observations in the North Atlantic Ocean 



Pefiod ol 

 Records 



Ocean 

 Station 



Latitude 

 deg.min 



Longitude 

 deg, min 



Number ol 

 Observations' 



Median Value ol 

 Sign. Wave Height 



Mean Value ol 



Logarithm ol 



Sign. Wave Height 



Variance {a') ol 



Logarithm at 



Sign. Wave Height 



1/49- 6/S4 



A 



62°0I)-N 



JFOO'W 



12,891 



6.34 



1.8472 



0.4524 



1/49-12/54 



B 



5S^9-N 



smo-w 



15,547 



6.59 





8859 



0.4434 



1/49-12/S4 



C 



S2°45'N 



35°30-* 



16,857 



6.75 





9100 



0.3763 



1/49-12/54 



D 



44°00-N 



41=00-1 



16,804 



6.26 





8338 



0.3843 



1/49-12/54 



E 



35°00-N 



48° 00-* 



16,777 



4.56 





5164 



0.3765 



1/47- 6/53 



1 



eroo'N 



15°20'W 



11,274 



7.36 





9960 



0.3747 



1/47- 6/53 



J 



52»30'N 



20° OC* 



12,016 



7.40 





0021 



0.3863 



1/49-12/53 



K 



45°0O'N 



16°0OTI 



11,182 



6.20 





8240 



0.3033 



1/49-12/53 



M 



66°00'N 



O2°0o'e 



14,321 



4.99 





6076 



0.2344 



1/49- 6/54 



H 



36°00-N 



70°00'« 



14,607 



5.08 





6251 



0.4237 



.E.ch * ■!=„ =h„.=»,lz.. . „. .... 











«II >.s..llhn .n: n.l.>.l loclthm.. 











pothesis that this sample was taken from a Ray- 

 leigh distribution. Consideration of the vari- 

 ances due to instrument errors would have re- 

 sulted in wider confidence limits, Table 2. 



As a further check on the validity of the hy- 

 pothesis, the chi-square test was applied to the 

 grouped data, in accordance with the method of 

 chapter 23, reference (12). This test gave a 

 rather low value of chi-square. Table 3, thus indi- 

 cating that there are no inconsistencies between 

 the test data and the hypothesis. 



Figs. 4 and 5 show the distribution for the poor- 

 est set of data obtained in the sea tests of the 

 Unimak. In this case also there are no signifi- 



cant differences between test data and the 

 hj^othetical Rayleigh distribution on the basis 

 of either the chi-square test or the variation in the 

 distribution of quantiles. 



Inasmuch as none of the more than one hundred 

 distributions of ship motion or huU stresses (longi- 

 tudinal, huU-girder bending moments) indicated 

 sigfnificant deviations from the hypothesis, this 

 hypothesis may be accepted confidently for ships 

 of the type tested. 



Thus far only the results of the Unimak sea 

 trials have been discussed. A similar analysis 

 also has been made of the results obtained from 

 sea tests (in the Atlantic Ocean) on a destroyer 



13 



