Sigmoidal crack growth rate curve: statistical modelling and applications
Article first published online: 28 SEP 2012
© 2012 Wiley Publishing Ltd.
Fatigue & Fracture of Engineering Materials & Structures
Volume 36, Issue 4, pages 316–326, April 2013
How to Cite
PAOLINO, D. S. and CAVATORTA, M. P. (2013), Sigmoidal crack growth rate curve: statistical modelling and applications. Fatigue & Fracture of Engineering Materials & Structures, 36: 316–326. doi: 10.1111/ffe.12001
- Issue published online: 12 MAR 2013
- Article first published online: 28 SEP 2012
- Received in final form 19 July 2012
- crack length prediction;
- fatigue crack growth;
- life prediction;
- probabilistic model
The present paper proposes a statistical model for describing sigmoidal crack growth rate curves. Major novelties are: a) exploitation of the maximum likelihood principle for obtaining material estimates by pooling together experimental data belonging to the different crack propagation regions; b) a general formulation which allows to adopt different sigmoidal models and any kind of statistical distribution for the model variables; c) fatigue life predictions through numerical integration of analytical functions with no need of Monte Carlo simulations. Experimental data taken from NASGRO database are used to check the validity of the statistical model in estimating material parameters included in the crack growth NASGRO algorithm. Illustrative plots of number of cycles to failure and crack length after a given number of cycles are presented, showing good agreement between the proposed statistical model and NASGRO results.