As repeated pre-stressing in air during the first stage, either overstressing at stress level 0.5 or 1kg/mm2 over the endurance limit with the cyclic repetition of 3.4×103 to 1.02×106, or understressing at stress level 1kg/mm2 below the endurance limit cycling from 1×106 to 7×107 was chosen, and followingly the specimen was exposed to air-saturated 5%H2SO4 at 25°C to which the repeated stress equivalent to prior endurance limit was applied during the second stage, until it was led to failure.
In order to give prior corrosion, rotating specimen under stress-free condition was exposed to air-saturated 5% H2SO4 at 25°C for 1h to 100h as the first stage. After having drained out the acid after prior corrosion, the specimen was cleaned and dried, then was subjected to repeated stress in air as the second stage, and was led to failure.
Secondary corrosion fatigue life in the second stage was slightly increased by primary overstressing in the first stage, but hardly any influence was given on secondary corrosion endurance by primary understressing in air. It was derived that, in view of these facts, endurance property conjoined with corrosion action was essentialy controlled by corrosion resisting property of metal within a given environment.
Exposure to the aqueous acid solution as the first stage, gave no practical deterioration to the specimen in the following test in the second stage if repeated stress applied in this stage did not exceed prior endurance limit. However, if the repeated stress applied in this stage was beyond prior endurance limit, serious decrease of the fatigue strength at N cycles was exhibited by short time of prior corrosion, for some austenitic stainless steel, and the reverse was exhibited for some other austenitic stainless steel.
It has been understood that this different tendency was presumably controlled by retained ductility of austenitic stainless steels, though corrosion resistance might be primary cause.