Since the ALVi;,' iull oont.-^in.i a si;i^-le uoiintci-i'tl ::e;'ja veld, 

 as veil iis v;eld3 for v'u four view ports .-.ind hatch a::d sinco tho hull 

 vill be hot forr.od in t-.c h3i;:iKphcrical pieces, the as-fabi'lcated hull 

 should perform at least as well at tl-,e other as-fabricated }:einispheres, 

 and probably as well as t!;e i-- tress relieved hemispheres of Fi^v.re 1. 

 Further it is possible, depending upon the titaniun alloy selected, 

 etc, that the hull vill '.:o sti-oss relieved. 



Thus the best ernneering jud^'ijent of the strength of the Titaniim 



hull for ALVIN, is that Ihe failure should be defined sornevhere betveen 



the limits of the strecf relieved and t],e r.achined curve. (rig 5). 



Further Figure 5 is bas?i on an assu'-ncd Z_S of l/3 in. vhich is probably 



a conservative value. !;r«-ever, it vas selected because i^ ij- felt that 



the distortions due to vilding the titaniiim sphere will be greater than 



the measured distcrtion;; obtained from welded Hi --103 spheres. Accordingly 



it is felt that the following minimiLm collapse pressvores and "^'■■^'j^ ratios 



should be attai:^able"": 



Yield Strength Goi::apsc Depth V^q -basic V>.^^- corrected Payload 

 psi Ft. increase 



■; V, 



>s 



100,000 18,!300 0.550 0.605 1975 



110,000 18^000 0.52/. 0.576 2300 



120,000 ISpOO 0.50;? 0.55? .'-^550 



"It should be noted tha' a similar technique was used to predict the 

 collapse depth for the ff-lOO ALVIN sphere. A depth of approximately 

 16,000 ft. was predictet based on a ip.f-r'j'cimuiii measured /\ of 0,050 in. 



3 



k 



