AN APPARATUS AND PROCEDURE FOR STRESS CORROSION TESTING

Abstract

The design and operation of a large capacity, stress corrosion testing apparatus of the alternate immersion type for use in aluminium alloy development research is described. The testing procedure was established by investigation of the important controllable experimental variables, of which the following are discussed: (i) specimen preparation by etching in sodium hydroxide, decreasing in carbon tetrachloride, and oxide stripping with phosphoric acid inhibited by potassium dichromate; (ii) the shape of the specimen with reference to the effect of corrosion at edges as a source of erratic variation; (iii) the electrolyte concentration in both sodium chloride and hydrogen peroxide; (iv) the ratio of specimen surface to volume of electrolyte. A statistical analysis is made of the established routine procedure, and the components of variation contributed by the material and by the experimental procedure are separated and evaluated as a function of the magnitude of the corrosion. The sample size required to yield results reproducible within ± 10% is calculated for two levels of probability, P = 0.9 and P = 0.99, for various degrees of corrosion. A table of typical results is given for different alloys in various electrolytes.