Tearing-fatigue interactions in 316L(N) austenitic stainless steel

This paper presents the results from a programme of tearing, fatigue and tearing-fatigue tests performed on specimens from a 316L(N) stainless steel plate. All tests were carried out at ambient temperature. The experimental results have been compared with assessments performed using current guidance within the R6 defect assessment method. The work has shown that there is some evidence that fatigue cycling modifies the JR-curve behaviour of this material. In most cases, the data lie approximately 20-30% above the base-line JR-curve. However, whilst there may be a modest influence of fatigue crack growth on the ductile tearing characteristics, it is difficult to separate this from experimental scatter. In tearing-fatigue tests performed at a stress ratio, R=0.2, ductile tearing reduces the fatigue crack growth rates by up to 50%. This is likely to result from the presence of a residual compressive zone at the crack-tip, and increased crack closure due to the irregular and non-matching fracture surfaces generated by the ductile crack growth mechanisms. For R=0.1 tearing-fatigue tests, fatigue crack growth rates are apparently enhanced by a factor up to of 10, particularly during the latter stages of the tests when ΔK>60 MPam. This is likely to result from: (i) loading being in the elastic-plastic regime where the J-integral (rather than K) characterises the crack-tip fields, (ii) increments of ductile tearing which may occur during each fatigue cycle, and (iii) crack blunting which reduces crack closure effects. For the R=0.2 tearing-fatigue tests, the linear summation approach described in R6 provides a consistently conservative prediction of ductile, fatigue and total crack growth during the tests. However, for the R=0.1 tearing-fatigue tests, the Paris law under-predicts fatigue crack growth rates. This may be corrected by using the Kaiser equation, which acknowledges loading in the elastic-plastic regime and incorporates incremental growth due to tearing as well as fatigue. R6 provides conservative predictions of instability for the CT specimen geometry tested in the current programme, both in terms of the critical crack growth and load required for instability to occur.