Reliability Maintenance Engineering Day 1 session 5 Measuring Reliability
Three day live course focused on reliability engineering for maintenance programs. Introductory material and discussion ranging from basic tools and techniques for data analysis to considerations when building or improving a program.
3. Objectives
• Reliability and the bathtub curve
• Calculating PDF plot
• What to do if you don’t have data at failure
mode level
• Measuring reliability of Line Replaceable Units
• Developing the next generation reliability
growth analysis
4.
5. Early Life Failures
Decreasing hazard rate
Typical failure causes
• Manufacturing defect
• Shipping damage
• Installation damage
6. Useful Life
Constant hazard rate
Typical failure causes
• Overstress
• Random events
Note: rarely occurs in real
life – may be useful if
change in hazard rate is
small enough
7. Wear out
Increasing hazard rate
Typical failure causes
• Material wear
• Rust/Oxidation
• Creep/Crack growth
• Embrittlement
10. Probability Distribution Plot
• Like a histogram
• Construction notes
• Continuous distributions
• Discrete distributions
f (x)dx = Pr a £ X £ b[ ]a
b
ò
f x( )= Pr X = x[ ]
11.
12. How to use PDF plot
• Probability plots
• How many will survive
• How many will fail
15. Data and Level
• Failure mechanism level
• Subunit level (LRU)
• Unit Level
• System level
16. Plots and levels
• CDF with system and
subunit information
• Continuous Distributions
• Discrete Distributions
F x( )= f m( )dm
-¥
¥
ò
F x( )= f i( )
i=0
x
å
23. Cumulative Failures over time
• During development or
improvement projects
• Plot total failures over
time (more later with
Duane and related
plots)
25. Summary
• Reliability and the bathtub
curve
• Calculating PDF plot
• What to do if you don’t
have data at failure mode
level
• Measuring reliability of
Line Replaceable Units
• Developing the next
generation reliability
growth analysis
Measuring Reliability to improve
Safety and Availability
Editor's Notes
Bathtub curve
Balance between investment and value
PDF Plots
The formula for the probability density function of the general Weibull distribution iswhere is the shape parameter, is the location parameter and is the scale parameter. The case where = 0 and = 1 is called the standard Weibull distribution. The case where = 0 is called the 2-parameter Weibull distribution. The equation for the standard Weibull distribution reduces to