http://reliabilitycalendar.org/Webinars/Chinese.htmlASQ Reliability DivisionChinese Webinar SeriesOne of the monthly webinarson topics of interest toreliability engineers.To view recorded webinar (available to ASQ ReliabilityDivision members only) visit asq.org/reliabilityTo sign up for the free and available to anyone livewebinars visit reliabilitycalendar.org and select ChineseWebinars to find links to register for upcoming events
ReliabilityReliability is the probability that an item willperform its intended functions withoutfailure for a specified time interval underdefined conditions.)(, tstRyReliabilit
Reliability EngineeringReliability Engineering provides the theoreticaland practical tools whereby product’sreliability (capability) can be specified,designed in, assessed, tested out,demonstrated (verified), and monitored witha desired confidence.
“Reliability Must Be Verified”Prior to acceptance into production / service, aproduct should be evaluated to checkwhether the RELIABILITY exhibited under useconditions exceeds the minimum acceptablerequirement.
Reliability Demonstration Test• Reliability Demonstration Test (RDT) isdesigned for claiming a specific reliabilityrequirement with a statistical confidence;e.g.,• The outcome of a RTD is a decision to acceptor reject it (the claim).CL@)(,0 GRtstR
Classic Reliability Demonstration Test• Reliability Requirement:• Outcome of n UUT tested under usecondition s(t) to time t0 has a BinomialDistribution,riinitRtR000 )()(1GRtstR )(,0
Practically, Zero Failure Test• Success Run• Test: n units all tested under use condition totime t0 without any failures, where• Classic statistical theory• Non-parametric)ln()1ln(GRCLn
Statistical InferenceTest Outcome:no failures aftert0 for all n testunits@ CLGRtstR )(, 00
Practical Issues• Besides a long test duration, a large sampleis usually required.ReliabilityGoalConfidenceLevelSampleSize0.90 90% 220.95 90% 450.95 95% 580.99 95% 298
Improved Techniques• Tested under use condition• With/without failures• Classic statistical theory• Non-parametricAcceleratedLife TestingBayesianDegradationTestingParametric(Distribution)
Better Strength, Better ReliabilityUse/TestStressA betterStrength means ahigher Reliability(at test condition).GR
Be Caution: There May Be a Gap …TestStressRequiredStrengthUseStressActuallyDemonstratedminimumStrengthSame reliability (required and demonstrated by test)Lower reliability may be seen in the field.
Test StressDemonstratedminimumStrengthReliability Demonstration“Demonstrated”minimumReliability undertest stress
UseStressDemonstrated thesame minimumStrength at the same CLOverload Success RunTestStressRequiredReliabilityTestTarget• Demonstrated the same reliability with a small sample• Same test duration• Require / apply the stress & strength knowledge)ln()1ln(TGTRCLn)ln()1ln(GRCLnGTG RRnnT
Summary• SSI concept may help you to plan and execute the ReliabilityTesting correctly.– Pay attention to and understand the effect of different between teststress and use stress.• SSI concept may help you to plan and design a more efficientReliability Testing .– Combining with life-stress relation and lifetime distribution, applyingSSI may further shorten test duration with a smaller sample.• More study is needed to incorporate the SSI into RDT planning.– Theoretical methodologies– Practical tools