Failures and Reliability Adam Adgar School of Computing and Technology

Failure ► Definition: Non-conformance to some defined performance criterion. ► Other terms Defect Malfunction Failure Fault Reject

Failure Classifications ► Cause Production related Stress related Misuse Inherent Wearout Maintenance induced ► Suddenness Immediate Gradual ► Degree Catastrophic Intermittent Partial ► Definition Applicable to specification Not applicable ► Result Critical Major Minor

Reliability ► Definition: Probability that an item performs its intended function for a stated period of time under specified operating conditions ► Manufacturing processes, machinery, systems are becoming more complex ► Reliability of whole process is dependent on all of the facilities, machines, components involved ► Reliability depends on design, operations, maintenance ► Generally defined as the ability to perform as expected over time (process, machine, product, person…)

Reliability ► Reliability of systems depends on many components ► Important concept because it is considered in several important issues Probability of failure System criticality Appropriate maintenance strategy

Maintainability ► Probability that a system can be retained in, or one that has failed can be restored to, operating condition in a specified amount of time. ► Maintainability is the totality of design factors allowing maintenance to be accomplished easily ► Design Issues Access of parts for repair Modular construction and standardization Diagnostic repair procedures and expert systems

Reliability Measurement ► Failure rate ► Mean time to failure MTTF ► Mean time between failures ► Mean time to repair MTTR number of systems failed number of operational hours number of systems failed number of systems tested FR (%) = FR (N) = operating time for system number of systems failed MTBF =

MTTF Information Bearing Number Millions of Revolutions Thirty Identical 6309 Deep Groove Ball Bearings Run to Fatigue Failure Under Test l Load Conditions From:Ball and Roller Bearings: Theory, Design, & Application, Eschmann, et al John Wiley & Sons. 1985

Example 1 ► Situation A company is interested in learning more about the reliability of one type of its production machines. The company tests 12 machines over a period of 1 week. During the test, one machine failed at 20 hours, another after 25 hours and another after 35 hours. The company operates day shift only (Mon-Fri) with a typical shift being 8 hours long. ► Required Failure rate as a percentage  FR(%) Number of failures during a period of time  FR(N) Mean time between failures  MTBF Estimated number of failures per shift given there are 100 production machines in total

Failure Rate Curve Infant Mortality Period Normal Life Period Wear- Out Period

Average Failure Rate

System in Series Component 1 XY Component 2 Component n

System in Parallel Component 2 Component 1 Component n XY

Simple Example ► Series R 1 = 99% R 2 = 95% R s = 0.99 x 0.95 R s = R s = % ► Parallel R 1 = 99% R 2 = 95% R s = 1 - (1-0.99) x (1-0.95) R s = 1 – (0.01 x 0.05) R s = R s = % R1R1 X R2R2 Y R1R1 X R2R2 Y

Average Component Reliability Number of components System Reliability %

Consequences Failure ► Production or Mission Impact Quantity Quality ► Environmental ► Health ► Safety ► Life Cycle Cost ► Morale

Example 3 ► Situation A consulting company has a thermographic camera which it uses to perform machinery inspections for its clients. The company estimates the cost associated to lost time for each breakdown of the camera is £500. Based on data extracted from the CMMS over the last 2½ years the table opposite was compiled The camera manufacture has offered a monthly service contract for £900. Past data has shown this to reduce breakdowns to only one every two months. ► Required Determine whether the company should purchase the service contract No. breakdowns No. per month