Time-based subcycle formulation for fatigue crack growth under arbitrary random variable loadings

Yongming Liu, Karthik Rajan Venkatesan, Wei Zhang

Research output: Contribution to journalArticlepeer-review

13 Scopus citations


A time-based subcycle fatigue crack growth (FCG) formulation and validation are proposed to calculate the fatigue crack growth under general random variable amplitude loadings. The intrinsic difficulties of the classical cycle-based formulation for general random variable loadings are discussed first. Several typical spectrums that are not appropriate for cycle-based FCG are illustrated, such as the “Christmas tree” spectrums. A time-based subcycle formulation is then proposed to address this difficulty. The proposed model includes three major component: (1) a time-based crack growth kinetics function at the subcycle (time) scale; (2) an efficient crack tip opening displacement (CTOD) estimation method; (3) a crack tip plasticity zone tracking algorithms for crack opening level determination of a growing crack. Detailed derivation and calculation procedures are given. Following this, several numerical examples are illustrated for the proposed model under different loading spectrums for the crack growth and CTOD calculation. Randomly generated loading spectrums are used to illustrate the capability of the proposed method under arbitrary loadings. Next, in-house testing for “Christmas tree spectrum” and literature data on several representative variable loading spectrums are used for model validation. Finally, some conclusions and future work are drawn based on the proposed study.

Original languageEnglish (US)
Pages (from-to)1-18
Number of pages18
JournalEngineering Fracture Mechanics
StatePublished - Sep 2017


  • Fatigue crack growth
  • Random
  • Subcycle
  • Variable amplitude

ASJC Scopus subject areas

  • Materials Science(all)
  • Mechanics of Materials
  • Mechanical Engineering


Dive into the research topics of 'Time-based subcycle formulation for fatigue crack growth under arbitrary random variable loadings'. Together they form a unique fingerprint.

Cite this