Finite weakest-link model of lifetime distribution of quasibrittle structures under fatigue loading

• Published in: Mathematics and Mechanics of Solids
• Main Authors: Le, Jia-Liang, Bažant, Zdeněk P.
• Format: Article in Journal/Newspaper
• Language: English
• Published: SAGE Publications 2013
• Subjects: Sea ice
• Online Access: http://dx.doi.org/10.1177/1081286513505463; http://journals.sagepub.com/doi/pdf/10.1177/1081286513505463; http://journals.sagepub.com/doi/full-xml/10.1177/1081286513505463

The design of various engineering structures, such as buildings, infrastructure, aircraft, ships, as well as microelectronic components and medical implants, must ensure an extremely low probability of failure during their service lifetime. Since such a low probability is beyond the means of histogram testing, we must rely on some physically based probabilistic model for the statistics of structural lifetime. Attention is focused on structures consisting of quasibrittle materials. These are brittle materials with inhomogeneities that are not negligible compared with the structure size, as exemplified by concrete, fiber composites, tough ceramics, rocks, sea ice, bone, wood, and many more at the micro- or nano-scale. This paper presents a finite weakest-link model of the fatigue lifetime of quasibrittle structures that fail at the fracture of one representative volume element (RVE). In this model, the probability distribution of critical stress amplitude is first derived by assuming a prescribed number of loading cycles and a fixed stress ratio. The probability distribution of fatigue lifetime is then deduced from the probability distribution of critical stress amplitude through the Paris law for fatigue crack growth. It is shown that the present theory matches well with the experimentally measured lifetime histograms of various engineering and dental ceramics, which systematically deviate from the two-parameter Weibull distribution. The theory indicates that the mean fatigue lifetime of quasibrittle structures must strongly depend on the structure size and geometry. Finally, the present model indicates that the probability distribution of fatigue lifetime can be determined from the mean size effect analysis.