conditions, PH14-8Mo-SRH950 and 15-7AMV in the annealed, RH1150 and 

 RH950 conditions. 



Residual Stresses 



Residual stresses are those which are present in an alloy when it 

 has been subjected to certain fabricating treatments, such as welding. 

 Such stresses are triaxial rather than uniaxial as in the intentionally 

 stressed specimens and, in addition, their levels are indeterminable. 

 Residual multiaxial welding stresses were intentionally induced in pre- 

 cipitation hardening stainless steels by making a 3-inch diameter weld 

 in the center of the 6-inch by 12-inch specimens. These specimens were 

 then exposed in the "as welded" condition; i.e., they were not subse- 

 quently heated to low temperatures to relieve the internal stresses 

 created by the welding operation. The data on these specimens as well 

 as some with other fabricating stresses are given in Table 10. 



Specimens of AISI Type alloys 630-H925, 631-TH1050, 631-RH1050 and 

 632-RH1100 failed by stress corrosion cracking in seawater at the surface 

 and at depths of 2,500 and 6,000 feet. 



Precipitation hardening stainless steel 15-7AMV-RH1150 failed by 

 stress corrosion cracking at stresses imposed by the insulators holding 

 it in the rack which caused the specimen to bow slightly. 



Stress corrosion cracking of alloys 15-7AMV-RH1150 and 15-7AMV-RH950 

 was caused by stresses induced by unreamed, drilled 1/8-inch diameter 

 holes. 



SUMMARY 



The purpose of this investigation was to determine the effects of 

 deep ocean environments on the corrosion of stainless steels. To accom- 

 plish this a total of 1,750 specimens of 57 different stainless steels 

 were exposed in the seawater at the surface and at nominal depths of 

 2,500 and 6,000 feet for eight time periods varying from 123 to 1,064 

 days. 



It must be re-emphasized that corrosion rates for stainless steels 

 calculated from wieght loss determinations are very misleading because 

 most corrosion is manifested as the pitting, tunneling and crevice types, 

 all localized. With these types of corrosion most of the surface area 

 of a specimen is uncorroded, therefore corrosion rates based on uniform 

 removal of metal from the surface do not reflect the true condition of 

 the metal. Hence, using corrosion rates of stainless steels calculated 

 from weight loss data for design purposes usually results in disappoint- 

 ment and frustration for the designer becuase the equipment is rendered 

 useless in much less time than the expected life because of localized 

 attack. 



15 



