CRE (ASQ) key terms

The probability that an item performs a required function without failure for a stated period of time when used under stated operating and environmental conditions. It is expressed as a probability and always depends on defined conditions and duration.

The conditional probability that an item fails in a given short interval, given that it has survived to the start of that interval, expressed per unit time. Often denoted lambda and commonly stated in failures per million hours.

A graphical model of failure rate over product life showing three regions: infant mortality with decreasing rate, a useful life with roughly constant rate, and wearout with increasing rate. It guides screening and replacement decisions.

The early-life period in which a population experiences a declining failure rate caused by latent manufacturing or material defects. It is usually removed by burn-in or environmental stress screening before shipment.

The late-life region of the bathtub curve where failure rate rises as components degrade through cumulative damage such as fatigue, corrosion, or depletion. Wearout failures are time-dependent and can be predicted and prevented by scheduled replacement.

The average operating time until failure for a nonrepairable item or a one-shot device. For a constant failure rate it equals the reciprocal of the failure rate and represents the expected life.

The average operating time between successive failures of a repairable system. It combines operating time and the number of failures and is used to characterize repairable equipment availability and maintenance demand.

A disciplined program of design activities, including requirements allocation, derating, parts selection, and analysis, that builds reliability into a product before testing. It emphasizes preventing failure rather than discovering it after production.

Operating a component below its rated stress, such as voltage, temperature, or load, to reduce the probability of failure and extend life. Derating guidelines specify the fraction of rated stress permitted for each part type.

The process of apportioning a system-level reliability requirement among subsystems and components so that if each meets its assigned goal the system goal is achieved. Allocation may be based on complexity, criticality, or prior experience.

A logic model that represents how component successes combine for system success, using series, parallel, and k-out-of-n structures. It converts component reliabilities into an overall system reliability estimate.

A reliability structure in which every element must succeed for the system to succeed, so the system reliability equals the product of element reliabilities. Series systems are only as reliable as their weakest element.

A reliability structure in which the system succeeds if at least one element succeeds, so system failure requires all elements to fail. Added parallel paths increase reliability but add cost, weight, and complexity.

Providing more than one means of performing a function so that a single failure does not cause loss of the function. Redundancy may be active, standby, or k-out-of-n and is a primary architecture-level reliability technique.

The failure of multiple components or channels from a single shared cause, such as a common power supply, environmental event, or design error. It limits the benefit of redundancy and must be addressed through diversity and separation.

An analytical estimate of expected failure rate or reliability derived from parts data, stress levels, and environmental factors, often using handbook methods. It supports design trade studies and comparison, though it is not a substitute for demonstrated test evidence.

A preliminary reliability estimate that sums generic part failure rates multiplied by quality and environmental factors, assuming constant failure rate and series operation. It requires only a parts list and is used early in design.

The systematic improvement in reliability achieved by finding, analyzing, and correcting failure modes during development testing. Growth is tracked on plots of cumulative failure rate versus test time and modeled by standards such as Duane or AMSAA.

An iterative reliability growth process in which failures found in testing are analyzed for root cause, corrective actions are implemented, and testing resumes to verify improvement. It converts test failures into durable design improvements.

Testing at stress levels above normal use, such as elevated temperature or voltage, to provoke failures sooner and infer life at use conditions. It relies on a life-stress model such as Arrhenius or inverse power law.

A life-stress relationship in which reaction or failure rate increases exponentially with absolute temperature, used to extrapolate accelerated test results to use temperature. It requires knowledge of an activation energy for the failure mechanism.

A qualitative development test that applies progressively increasing stresses, such as temperature and vibration, to find operating and destruct limits and reveal design weaknesses. It is intended to improve design margins rather than to estimate field life.

A formal test with predefined accept and reject criteria used to verify statistically that a reliability requirement has been met. It typically assumes a distribution, a confidence level, and a producer or consumer risk.

A flexible life distribution with shape, scale, and location parameters that can model decreasing, constant, or increasing failure rates. Its shape parameter indicates whether failures are infant mortality, random, or wearout in nature.

The probability, expressed as a percentage, that a statistical interval or demonstration result correctly captures the true population parameter. Higher confidence requires more test time, more samples, or more failures allowed.

The probability that a failed item can be restored to a specified condition within a stated time when maintenance is performed under stated conditions with prescribed procedures and resources. It is a design characteristic measured by repair time distributions.

The average elapsed time to restore a failed item to operation, including diagnosis, access, repair, and verification. It is a key input to availability and is reduced by modular design, diagnostics, and accessible layout.

The fraction of time a system is capable of performing its function, combining reliability and maintainability. Inherent availability equals MTBF divided by the sum of MTBF and MTTR for constant-rate repairable systems.

Scheduled servicing, inspection, or replacement performed to prevent failure or slow degradation. Its interval should be set from wearout characteristics and cost tradeoffs rather than arbitrary schedules.

Condition-based maintenance in which monitored parameters such as vibration, temperature, or oil debris trigger maintenance action before functional failure. It targets the developing fault rather than a calendar interval.

A structured team method that identifies potential failure modes, their effects, and their causes, then ranks them by severity, occurrence, and detection to prioritize corrective action. It is performed early and updated as design matures.

A deductive top-down technique that models how combinations of basic events, linked by AND and OR gates, can cause a defined undesired top event. It supports probability estimation and identification of single points of failure.

A component or path whose failure alone causes loss of the entire system function. It is revealed by fault tree or dependency analysis and is normally eliminated through redundancy or design change.

A systematic investigation to determine the underlying physical, human, or organizational causes of a failure rather than merely its symptoms. Common tools include the five whys, cause-and-effect diagrams, and fault trees.

A reliability model in which failure occurs when the applied stress exceeds the item's strength, both treated as distributions. Increasing the margin between mean strength and mean stress reduces the overlap and the failure probability.

A screening process that operates items for a period before delivery to precipitate latent defects and remove infant mortality failures. Effective burn-in requires a stress and duration that trigger weak units without consuming much useful life.

A ranking of failure modes or items by the combined consequence and likelihood of failure, often extending FMEA into FMECA. It directs limited reliability and safety resources to the most consequential risks.