Simulation-based investigation of the reuse of unbroken specimens in a staircase procedure: Accuracy of the determination of fatigue properties

Bayesian analysis of fatigue data with multi-load-level damage accumulation: The benefits of rerun specimens

“…It was found that the benefits of rerun specimen data were substantial in comparison to mere virgin specimen tests. Similar conclusions were drawn in the recent simulation study by Roué et al [29], where censoring of fatigue limit (strength) was utilised and damage accumulation was not considered in the data analysis. The situation, in this case, is quite identical, given that the inference results presented here imply that significant damage accumulation on runout levels was unlikely.…”

“…Hence, for these data sets, the observations (fail/runout) could be considered censoring in terms of fatigue limit, and when the runout specimens are retested, interval censoring is obtained. Roué et al [29] state that if the fatigue damage was estimated, the uncertainties could be further reduced, and more information could be obtained from the specimens. The authors of this paper would also like to emphasise that this is the fundamental concept and novelty behind the statistical approach presented here.…”

“…In a very recent publication, Roué et al [29] present a simulation study on different approaches to analysing fatigue data from staircase testing. The study is very interesting, especially as the simulations consider the fatigue damage accumulation when reinserted specimens are utilised in the testing procedure.…”

Experimental strain‐based vibration control to obtain the fatigue strain limit by the staircase method

“…Object 2: For the second object, the hypothesis is that the experimental failure probability reflects the real FSL distribution. Roué et al [34] and Müller et al [32] concluded that the fatigue limit estimated from the staircase test represents the median threshold with a probability of failure of 50%. The object aims to minimize MSE of kernel-estimated CDF and the failure probability in each experimental strain level.…”