In order to select the most appropriate steel to deal with pressurized hydrogen over long periods of time, the fatigue crack propagation rate of quenched and tempered CrMo and CrMoV steel grades was assessed by means of tests performed on thermally pre-charged specimens in a hydrogen reactor at 195 bar and 450ºC during 21 hours. Cylindrical samples were used to measure the hydrogen content and their desorption kinetics at room temperature and compact tensile specimens to determine the fatigue crack growth rate. Under the aforementioned pre-charging conditions, significant amounts of hydrogen were introduced, being much larger in the CrMoV steel grades, which also have much lower apparent diffusion coefficients, as precipitation of fine vanadium carbides during tempering provides strong hydrogen traps. Moreover, the fatigue crack growth rate increased significantly due to the presence of internal hydrogen in the CrMo grades for test frequencies lower than 10 Hz in comparison with tests performed in air. Furthermore, the presence of vanadium carbides in the CrMoV steel significantly improved fatigue crack growth performance, the effective hydrogen diffusion distance per cycle and the hydrogen concentration in the process zone ahead of the advancing crack being considerably reduced.

No abstract

No abstract

No abstract

The Miner number M, used as a tool for lifetime prediction of mechanical and structural components in most of the standards related to fatigue design, is generally accepted as representing a damage stage resulting from a linear progression of damage accumulation. Nonetheless, the fatigue and damage approach proposed by Castillo and Fernández-Canteli, permits us to reject this conventional cliché by relating M to the normalized variable V, which represents percentile curves in the S-N field unequivocally associated to probability of failure. This approach, allowing a probabilistic interpretation of the Miner rule, can be applied to fatigue design of mechanical and structural components subjected to variable amplitude loading. The results of an extensive test program on concrete specimens under compressive constant and load spectra, carried out elsewhere, are used. A parallel calculation of the normalized variable V and the Miner number M is performed throughout the damage progression due to loading allowing probabilities of failure to be assigned to any value of the current Miner number. It is found that significant probabilities of failure, say P=0.05, are attained for even low values of M, thus evidencing the necessity of a new definition of the safety coefficient of structural or machine components when the Miner rule is considered. The experimental and analytical probability distributions of the resulting Miner numbers are compared and discussed, the latter still providing a nonconservative prediction in spite of the enhancement. A possible correction is analyzed.