The pressure is then built up incrementally 

 while the change in volume in the volume- 

 ters is recorded. At a pressure point where 

 critical behavior is predicted, the pressure 

 increments are decreased and the test pro- 

 ceeds very slowly while watching for the first 

 sign of a non-linear volume change. When 

 this occurs it is taken to be the onset of a 

 transition from elastic to plastic yield in the 

 hull structure. In the initial development of 

 this technique, strain gages are attached to 

 compare the test results, and to assist in 

 locating the local effects which caused test 

 termination. Without the assistance of strain 

 gages, it would be virtually impossible to tell 

 precisely where the critical stress occurred. 



For larger vehicles some measure of the 

 pressure hull behavior is attained through 

 model testing. In this procedure a scale 

 model is constructed in the same fashion and 

 of the same materials as the hull; the model 

 is then subjected to pressure testing in a 

 tank as if it were the full scale hull. Such 

 scale models are occasionally tested to fail- 

 ure as a means of verifying calculations. 



Pressure tanks themselves can be ex- 

 tremely complex, sensitive and potentially 

 dangerous. Mavor (32) reported a tank fail- 

 ure at 4,300 psi, with one of ALVIN's hulls 

 inside, which blew off a hatch but left the 

 windows undamaged and tight. 



Endo and Yamaguchi (33) present an excel- 

 lent description of pressure and materials 

 testing facilities at Mitsubishi Heavy Indus- 

 tries. The paper is not only a good summari- 

 zation of the devices available, but includes 

 requirements for deep submergence mate- 

 rials as well. 



CORROSION AND ITS 

 CONTROL 



As will become evident in later sections, a 

 great quantity of dissimilar metals are 

 joined and juxtaposed within the exostruc- 

 ture and the pressure hull. Corrosion protec- 

 tion and control is another concern of past 

 and present submersible builders. 



Corrosion control on submersibles which 

 are routinely launched/retrieved for each 

 dive is somewhat less complicated than those 

 continuously in the water — mainly because 

 vehicles removed from the sea may be 



washed with fresh water and will dry. On the 

 other hand, availability of components and 

 cost result in a situation where less than 

 optimum corrosion resistance and galvani- 

 cally incompatible materials must be used. 

 Likewise, cathodic protection as a general 

 method often is impractical and difficult ow- 

 ing to the geometrical complexity of compo- 

 nents. Though procedures for corrosion con- 

 trol vary from vehicle-to-vehicle, certain 

 problems are common to all. Consequently, 

 the procedures followed in the DEEPSTAR 

 submersibles fairly well represent a number 

 of common problems and solutions. Symonds 

 and Woodland (34) present the steps taken to 

 prevent corrosion on DEEPSTAR 20000 

 based on 4 years of operational experience 

 with DEEPSTAR 4000. These are summa- 

 rized below. 



Four areas of potential corrosion were rec- 

 ognized on the DS-20000: General corrosion, 

 galvanic corrosion, crevice corrosion and 

 stress corrosion. 



General Corrosion 



Two protective measures were foreseen to 

 prevent general corrosion: Painting and 

 cathodic protection. 



Painting: 



According to DS-20000''s specifications al- 

 most all metallic vehicle components would 

 be painted by a polyurethane paint system 

 known as Magna Laminar X-500. On the 

 pressure sphere, four layers are applied: 

 Wash primer, primer, primer surface and a 

 finish coat. On other components the primer- 

 surface layer is omitted. Where viewports, 

 hatch and electrical penetrators join the 

 hull, two priming layers of the Magna Lami- 

 nar X-500 are applied and, subsequently, sili- 

 cone grease is applied during assembly. 



Cathodic Protection: 



A comprehensive cathodic protection sys- 

 tem would be impractical owing to the com- 

 plex geometry which requires placement of a 

 large number of anodes at the sacrifice of 

 weight and space. Because of coating defects, 

 zinc anodes in the form of flexible steel-cored 

 line known as Diamond Line was proposed 

 because its flexibility permits adaptation to 

 a number of complex geometrical situations. 

 Lengths of Diamond Line would be attached 



275 



