Abstract
A model for estimation of fatigue-life distribution function with respect to random character of the load process was presented in Part I. In this paper the extension of this model is proposed using a probabilistic interpretation of material characteristics. By means of a proper connection of the set of representing material parameters and loading characteristics the complete model of life estimation arises. This model enables computation of a fatigue-life distribution function (DF) respecting the stochastic nature of the interacting material and loading characteristics. A set of material parameters, entering the calculation in probabilistic form, is obtained by statistical treatment of the experimental dependences σ_a = f(2N_f) and ε_ap=f(2N_f), by means of the confidence and prediction intervals. A set of loading characteristics, representing the random process, is obtained by a proper segmentation of the load process record. Resultant DF allows to estimate the life for the required probability of failure occurrence (premature fatigue fracture) and can be widely used at design stage or in operation. DF enables to assess different combinations of materials and load conditions. The concrete procedure of its estimation is presented. DF obtained is in a good agreement with the experimental results.

Key words
fatigue-life estimation, life distribution function, fatigue-damage accumulation, hypothesis, hysteresis energy, random loading, probability of failure, reliability, mean stress, material properties inhomogeneity

Full text not available