Accelerated Degradation Tests Planning With Competing Failure Modes
read more
Citations
Reliability Modeling and Analysis of Load-Sharing Systems With Continuously Degrading Components
Optimal inspection and replacement policy based on experimental degradation data with covariates
Maintenance policy for a system with a weighted linear combination of degradation processes
Utilizing experimental degradation data for warranty cost optimization under imperfect repair
Optimal condition-based maintenance policy with delay for systems subject to competing failures under continuous monitoring
References
Accelerated Testing: Statistical Models, Test Plans, and Data Analyses
Accelerated Testing: Statistical Models, Test Plans, and Data Analyses
Stochastic modelling and analysis of degradation for highly reliable products
Accelerated degradation tests: modeling and analysis
Degradation Data Analysis Using Wiener Processes With Measurement Errors
Related Papers (5)
RUL Prediction of Deteriorating Products Using an Adaptive Wiener Process Model
Stochastic modelling and analysis of degradation for highly reliable products
Frequently Asked Questions (17)
Q2. What future works have the authors mentioned in the paper "Accelerated degradation tests planning with competing failure modes" ?
The authors use three common optimality criteria to optimize the ADT plans. The authors also study different rules of three-level compromise plans, and they suggest 10 % or 20 % allocation rule to the middle stress in such problems. In addition, models incorporating multiple dependent shocks and degradation failures could be adopted in the ADT planning framework, where copula function can be applied in model construction, and revealing asymptotic properties of unknown parameters are challenging and of interest to investigate.
Q3. How many test units fail under the traumatic stress?
For the traumatic shocks, the authors assume that at the use condition, only 0.5% of the test units are expected to fail within the test duration 1500 hrs, and 30% of the test units fail under 173℃.
Q4. What is the main reason for failure of a product?
performance degradation is due to the natural aging and usage of a product, and if the performance degrades to an unsatisfactory level, the product is deemed failed although it may still work.
Q5. Why have stochastic models been widely used to model degradations?
Stochastic models have been widely used to model degradations because of their clear physical explanations and tractable mathematical properties [28].
Q6. Why is the variability of the two additional parameters significantly smaller than the degradation failures?
The reason behind this is that by considering shock failures, the variability of the two additional parameters contributes significantly to the sum of variance of all parameters because the information on shock failures is much less than degradation failures.
Q7. How many test units are there for the experiment?
The authors assume that there are 200 test units for the experiment and the maximum test duration is 1500 hours with the inspection interval 75 hours, i.e., ᵃ� = 20, Δᵅ� = 75.
Q8. Why is the degradation measure “censored” by a shock failure?
Due to the presence of shock failures, some degradation measurements for a certain test unit cannot be obtained, i.e., the degradation measures are “censored” by a shock failure.
Q9. Why do the authors change to see the variation in the optimal plans under [C2?
Because ᵅ� does not influence the optimal plans under [C1] and [C2], the authors change ᵅ� from 0.1 to 0.9 to see the variation in optimal plans under [C3].
Q10. Why is the optimal plan not explored at the maximum stress?
This is because the presence of shock failures prevents the optimal plan from exploring more at the maximum stress due to the risk of information loss caused by shocks.
Q11. What is the way to plan accelerated tests?
Neglect of any failure mode may significantly influence the optimality of reliability test plans and therefore the prediction accuracy of lifetime, thus it is necessary to consider multiple failure modes when planning accelerated tests.
Q12. What is the optimal stress for PF0?
For [C3], it is interesting to observe that when PF0 is large enough, such as 1%, the optimal stress lower stress increases as PF1 increases.
Q13. Why does the intercept parameter of shock models affect the optimal plan?
This is due to the fact that [C3] concerns more on the accuracy of extrapolation of the lifetime, and the intercept parameter of shock models determine the shock failure rate under use condition, which is of great importance when test planners try to predict the field lifetime.
Q14. What is the time to a shock failure?
Denote the time to a shock failure as ᵃ� , of which PDF and CDF are given byThe lifetime of the product, denoted by ᵃ� , is determined by either degradation failure or shock failure times, whichever comes first, i.e., ᵃ� = min{ᵃ� , ᵃ� }.
Q15. What is the reason why the contour plots in Section 4.1 have supported that the optimal plans?
the contour plots in Section 4.1 have supported that the optimal plans are unique within the feasible range of decision variables.
Q16. How many test units survive to the end of the test?
As is shown there are only three test units that survive to the end of the test and provide full degradation information during the test.
Q17. Why is the test plan relatively robust?
Due to the fact that the test plannersmostly wish to estimate lower percentiles, the test plan is relatively robust with respect to the change in ᵅ�.