Reliability Maintenance Engineering 2 - 2 Reliability Techniques
Post on 25-Jan-2015
785 Views
Preview:
DESCRIPTION
Transcript
Reliability Engineering
Fred Schenkelbergfms@fmsreliability.com
RELIABILITY TECHNIQUES TO IMPROVE PERFORMANCE
Day 2 Session 2
Objectives
• Examining opportunities and reliability improvement techniques for robust design– RCM, FMEA, RCA, 6 step and FRETT
• Developing corrective actions and determining effectiveness
• Implementing the growth curve technique• Fault and failure forecasting
FMEA
• Organized risk analysis– Severity– Occurrence– Detection
• Use in Design process
RCA
• Root Cause Analysis– Determine fundamentals– Initiating event– Contributing events
• Use in design process
6 Steps+1
FRETT
Cause categories• design errors• material defects• fabrication and processing
errors• assembly and installation
deficiencies• maintenance-related or
procedural errors• unintended operating
conditions• operator error
Failure mode sets
• Force
• Reactive environment
• Temperature
• Time
Discussion & Questions
Corrective Action
• Plan Do Check Act process
• Experiments and measures
• Verify results
• Document results
Plan
Do
Check
Act
An Example
• Define problem
• Plan response
• Do experiment
• Monitor results
• Implement fix
Monitor Effectiveness
• What do you measure?
• What can you measure?
• Does absence of failure mean it’s fixed?– What is expected failure
rate?– What is probability of
failure?
Monitoring
• How to determine monitoring plan?
• Sample size?
• Duration?
• Focus on failure mechanism
Discussion & Questions
Example
Repair rate over time
• Trend plots– MCF – cumulative
failures over time– Inter arrival times– Reciprocals of inter
arrival times
• Duane Plot
• Sample Data• 1500 hour test• With each failure RCA
and improvements implemented
Observed failure times5, 40, 43, 175, 389, 712, 747, 795, 1299, 1478 hrs
Cumulative failure vs. time (MCF)
• Failure count vs time
• Straight line (roughly) means steady failure rate.
• Curve down – improvement over time
• Curve up – increasing problems over time
NIST Engineering Statistics Handbook 8.2.2.3
Inter arrival failure times
• Plot waiting time between failures
• Trend up – improvement
• Trend down – degradation
• Straight – no change over time
Reciprocal Inter arrival times
• Plot failure rate estimates since last failure
• Trend up – • degradation• Trend down –
improvement• Straight – no change
over time
Duane Plot
Time Cum MTBF
5 5
40 20
43 14.3
175 43.75
389 77.8
712 118.67
747 106.7
795 99.4
1299 144.3
1478 147.8
Discussion & Questions
Predictions & Forecasts
Risks
• Might just be wrong
• May miss failure mechanisms due to masking
• Unable to predict everything
Benefits
• Awareness of probability of failure
• Awareness of what to expect to fail
• Maintenance and logistics planning
Predictions & Forecasts
• Empirical models– Extension of field data– Extensions of
experiments– Extensions of vendor
data
• First principle models– Difficult to create– Does it apply in this
situation
Cautions
• Check assumptions• Check sensitivity
• Verify models and fits• Be conservative
The further from reality the more risk of being
wrong exists
Monitor and Adjust
• All models are wrong, some are useful
• Continue to refine and challenge any model
• RCA with focus on failure mechanisms is key to success
Discussion & Questions
Summary
• Examining opportunities and reliability improvement techniques for robust design– RCM, FMEA, RCA, 6 step and
FRETT• Developing corrective
actions and determining effectiveness
• Implementing the growth curve technique
• Fault and failure forecasting
Reliability techniques to improve performance
top related