Corrigendum to “Variability of the fatigue damage due to the randomness of a stationary vibration load” [Int. J. Fatigue 141 (2020) 105891]

The authors have recently discovered that a formula used in the article has been misinterpreted, due to an inattentive reading of Ref. [1]. The formula is that reported in Eq. (21) at Pag. 5: [Formula presented] in which [Formula presented] is an integer, [Formula presented] is the autocorrelation coefficient of the random process and [Formula presented] its time derivative. The quantity [Formula presented] represents the autocorrelation coefficient function of the fatigue damage, and it appears in the expressions used to compute both the variance [Formula presented] and coefficient of variation [Formula presented] of the fatigue damage [Formula presented] in a time-history [Formula presented] of length [Formula presented]. Although the formula above refers to an inverse slope [Formula presented] of the S-N curve, it has erroneously been applied in the article also to values [Formula presented] greater than two. Figs. 1 and 2 below, which refer to the same examples considered in the article, show that the use of the above formula also for [Formula presented] leads to an overestimate of [Formula presented] from time-domain simulations, with an error that increases with [Formula presented], but that reaches at most 12.8% for [Formula presented] (lightly damped linear oscillator). The authors would like to apologise for any inconvenience caused.