Abstract

In the fields of engineering design and fatigue life prediction, handling limited sample data has consistently posed a challenge. A hybrid distribution characteristics of equivalent structural stress (HDC-ESS) method is presented for the design of fatigue life prediction models with small sample sizes. The method integrates the equivalent structural stress (ESS) method with the maximum likelihood estimation (MLE) method based on the idea of sample augmentation, and adds closed-loop reliability verification to ensure its robustness and correctness. By employing three distinct approaches, an analysis and comparison were conducted on the failure data of small specimen obtained from high-frequency fatigue tests of welded joints. The outcomes reveal that the HDC-ESS method, in comparison to the conventional direct method and the MLE-integrated equivalent structural stress approach, exhibited an enhancement of goodness of fit by 0.1866 and 0.1988, respectively. Furthermore, it exhibited remarkable performance in enhancing stability and optimizing sample quantity.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call