14 reliability analysis（PDF 15页）.pdf

Reliability Analysis Total Quality Management - University of Michigan 1 Pat Hammett 1 Reliability Analysis 2 Reliability n Reliability Defined: Reliability is the probability that a system or product will perform in a satisfactory manner for a given period of time when operated under specified conditions. n Reliability Analysis: Analysis of system failures and repairs to: n increase the design life of a product. n reduce the likelihood of failures. n reduce manufacturing process downtime. n eliminate safety concerns.Reliability Analysis Total Quality Management - University of Michigan 2 Pat Hammett 3 Reliability and Six Sigma Projects n Consider the following defects: n High warranty costs due to excessive replacement of certain product components. n Late delivery of products due to excessive machine downtime. n An effective method to examine the effects of these defects is Reliability Analysis. 4 Product Types / Failure Conditions 1.Non-repairable items {replace} (e.g., light bulb) n interested in probability of first and only failure (hazard rate) n Analyze MTTF (mean-time-to-failure) 2.Repairable items (e.g., copy machine) n interested in probability that a failure will occur over some time period n Analyze MTBF (mean-time-between-failures)Reliability Analysis Total Quality Management - University of Michigan 3 Pat Hammett 5 Reliability Assessment n To describe failure patterns in order to predict future issues, we need statistical distributions. n Common distributions for non-repairable devices: n Weibull n Exponential* is special case where failure rate is constant. n Log-normal n Normal* n Common distributions for repairable devices: n Poisson n Nonhomogenous Poisson Process (NHPP) 6 Failure Pattern for Product Life Cycle - Bathtub Curve Bathtub Curve l,