Reliability Introduction PDF

An Introduction to Reliability and
Life Distributions
Dr Jane Marshall
Product Excellence using 6 Sigma
Module
PEUSS 2011/2012 Reliability and Life distributions Page 1
Objectives of the session

• Probability distribution functions
• Life time distributions
• Fitting Reliability distributions using Hazard
Plotting
• Interpretation
1
Data types of interest
• Sample data from a population of items
• For example:
– 100 ipods put on test, 12 fail, analyse the times to
failure
– 1000 aircraft engine controllers operating in-service,
collect all the times to failure data and analyse
• Not only times but distance or cycles etc.
Histogram
Histogram of hours to failure
35 120.00%
30 100.00%
25
80.00%
Frequency
20
60.00%
15
40.00%
10
5 20.00%
0 .00%
9 5 75 5
8.
5 75 5 25 or
e
37 3. 12 .8 .2 .6
6. 46 1. 91 45 73 00 M
23 69 11 13 16
Hours to failure
Frequency Cumulative %
2
Probability distribution
Hours to failure
35
30
25
Frequency
20
15
10
5
0
-500 -5 0 500 1000 1500 2000 2500
Hours to failure
• The area under the curve is equal to 1

• The area under the curve between two values is the
probability
Failure Time distributions

• PDF (Probability density function)
• The CDF (Cumulative Distribution Function)
– The CDF gives the probability that a unit will fail before time t
or alternatively the proportion of units in the population that
will fail before time t.
• The Survival Function (sometimes known as
reliability function)
– Complement of the CDF.
• The Hazard Function
– Conditional probability of failing in the next small interval
given survival up to time t.
3
Probability density Function:
• PDF - Probability of falling between two values
1.2
1
PDF, f(t) (%)
0.8

t2
Frequency
0.6 P(t1<t<t2)= t1
f(t) dt
0.4
0.2
0
1 1 2 3 3 4 5 5 6 7 7 9
Value
Probability distributions
Hours to failure
0.5
Relative frequency
0.4
0.3
0.2
0.1
0
-500 -0.1 0 500 1000 1500 2000 2500
hours to failure
Probability of failure between 500 and 1000 hours is given by the area
4
Standard Normal distribution
m
-2s -1s +1s
+2s
68.27%
-3s +3s
95.45%
99.73%
Cumulative distribution function

• The CDF known as F(t)
1
1.2 t
1 F( t) = -f ( t ) dt
CDF, F(t) (%)
0.8
Frequency
0.6
F(t) gives the
0.4
0.2
probability that a
0 measured value will fall
1 1 2 3 3 4 5t 5 6 7 7 9 between -  and t
Value
Failure Function, F(t)
5
Cumulative distribution
Cumulative probabilty
1.2
cumulative probability
1
0.8
0.6
0.4
0.2
0
-500 -0.2 0 500 1000 1500 2000 2500
hours to failure
The probability of failure before 500 hours is 0.8

or 80% will have failed by 500hrs
Survival function
• The survival function or reliability function R(t)
1.2
1
(%)
0.8
Frequency
0.6
R(t) = 1 - F(t) and
R(t)
0.4 F(t) = 1 - R(t)

0.2
0
1 1 2 3 3 4 5t 5 6 7 7 9
Value
6
Survival Function
Survival Function
1
Probability of survival
0.8
0.6
0.4
0.2
0
-500 0 500 1000 1500 2000 2500
Hours to failure
The probability of surviving up to 500 hrs is 0.2

Or 20% have survived up to 500 hrs
Hazard function
• The Hazard function is defined as probability of
failure in next time interval given survival to time
t 1
1.2
Reliability Function R(t)
1
• h(t) = f ( t ) = f ( t )
(%)
1 - F (t )
0.8
R (t )
Frequency
0.6
h(t)
0.4
• Figure shows 0.2
increasing hazard 0
function 1 1 2Hazard
3 3 Function
4 5 5h(t)6 7 7 9
Value
7
Bath-tub curve
Hazard function
Useful Life
Time
Infant Wear Out
Mortality
Probability distributions
• Exponential distribution
• Weibull distribution
• Normal distribution
• Lognormal distribution
8
Exponential distribution
• Simplest of all life models
• One parameter, 
• PDF, f(t) = e- t
• CDF, F(t) = 1- e- t and R(t) = e- t
• Hazard function, h(t) =  i.e. constant
• MTBF = 1/  and failure rate = 
• 1/ is the 63rd percentile i.e. time at which 63%
of population will have failed
Exponential distribution
1.0
Probability
Hazard Function
0.5 0.160
0.155
0.0
Rate
0 10 20 30 40 50
Survival Function
0.150
0 10 20 30 40 50
9
Failure rate - example
• 10 components of a particular type in each PCB
• 5 PCBS in each unit
• 200 units in the field
• Total operating time to date for all units is 10,000 hours
• There have been 30 confirmed failures of this component
• The failure rate is given by:
– 30/5*200*10*10,000 = 0.000003 = 3 fpmh (failures per million hours)
• The MTTF is 1/0.000003 = 333,333
Example
• 100 units in the field
• Total operating hours is 30,000
• Number of confirmed failures is 60
• MTBF = 30,000*100/60 = 50,000
• Removal rate includes all units removed
regardless of whether they have failed
• Use 200 removals
• MTBR = 15000
10
Weibull distribution
• Most useful lifetime in reliability analysis
• 2 parameter Weibull
– Shape parameter - 
– Scale parameter - 
• When < 1 decreasing hazard function
• When > 1 increasing hazard function
• When =1 constant hazard function
•  is the characteristic life, 63rd percentile

t
  
 1   
PDF : f (t )  
t e


t
 
 
CDF : F (t )  1  e

t
 
 
Re liability : R (t )  e
11
• h(t) =  t-1

• When =1, h(t)= 1/ =  therefore =1/ 
• When >3.5 the distribution approximates to a
normal distribution
Three parameter Weibull

• A three parameter distribution can be used if
failures do not start at t=0, but after a finite
time . The parameter,  is called the failure-
free time or location parameter

 (t  ) 
 
  
F (t )  1  e
PEUSS 2011/2012 Reliability and Life distributions 24
12
Bath-tub curve and the Weibull
Hazard function
Useful Life
=1
<1
>1
Time
Infant Wear Out
Mortality
Normal Distribution
• Not used as often in reliability work
– Can represent severe wear-out mechanism
– Rapidly Increasing hazard function
• e.g.’s, filament bulbs, IC wire bonds
• Location parameter, m , is the mean

• Scale parameter, , is the standard deviation
• Lognormal more versatile, always positive
13
Fitting parametric distributions
Fitting parametric distributions

• Censoring
• Repaired and non repaired
• Probability plotting
• Hazard plotting
14
Censoring structures
• Complete data
• Single censored
– Units started together and data analysed before all units have
failed
– Right, interval and left
• Time censored
– Censoring time is fixed
• Failure censored
– Number of failures is fixed
• Multiply censored
– Different running times intermixed with failure times – field data
Complete Data
15
Right Censored data
Interval Censored
16
Left censored
Repaired and non-repaired data
• Non-repaired data when only one failure can

occur and interested in time to failure
• Repaired data when interested in the pattern of

times between failures
17
Probability plotting
Areas to be covered
• Introduction to probability plotting
• Assumptions
• How to do a Weibull plot
• Estimating the parameters
• Testing assumptions
• Examples
18
What is probability plotting?
• Graphical estimation method
• Based on cumulative distribution function CDF or F(t)
• Probability papers for parametric distributions, e.g.
Weibull
• Axis is transformed so that the true CDF plots as a
straight line
• If plotted data fits a straight line then the data fits the
appropriate distribution
• Parameter estimation
Assumptions
• Data must be independently identically
distributed (iid)
– No causal relationship between data items
– No trend in the time between failures
– All having the same distribution
• Non-repaired items
• Repaired items with no trend in the time
between failures
• Time to first failure of repaired items
19
Example of test for trend
• Machine H fails at the following running times
(hours):
– 15, 42, 74, 117, 168, 233, and 410
• Machine S fails at the following running times

(hours):
– 177, 242, 293, 336, 368, 395, and 410

Trend Analysis
machine H running times to failure
machine S running times to failure
8 8
7 7
order number
order number
6 6
5 5
4 4
3 3
2 2
1 1
0 0
0 100 200 300 400 500 0 100 200 300 400 500
machine time to failure machine time to failure
This system is getting better with This system is getting worse with time,
time, the failure times are getting the failure times are getting closer and
further and further apart closer together.
In neither case can Weibull analysis be used as there is

trend in the data.
20
Making a Weibull plot
Rank the data

• Probability graph papers are based on plots of
the variable against cumulative probability
• For n< 50 the cumulative percentage probability
is estimated using median ranks tables
• For n< 100 use benard’s approximation for the
median rank ri
ri = i - 0.3
n+0.4
Where i is the ith order value and n is the sample size
21
Example
Failure number (i) Ranked hrs at failure (ti) Median Rank from tables
Cumulative % Failed at ti - F(t)
1 300 6.7
2 410 16.2
3 500 25.9
4 600 35.5
5 660 45.2
6 750 54.8
7 825 64.5
8 900 74.1
9 1050 83.8
10 1200 93.3
• Plot times on x-axis

• Plot CDF on y-axis
• Fit line through the data
• Draw perpendicular line from
estimation point to the
fitted line.
• Read off the estimate of β
• η is the value given on
from the intersection line
22
Interpreting the plot
• If the data produced a straight line then:
– The data can be modelled by the Weibull distribution.
• If β<1 then data shows a decreasing hazard function
– e.g. Infant mortality, weak components, low quality
• If β=1 then data shows a constant hazard function
– e.g. useful life of product
• If β>1 then data shows a increasing hazard function
– e.g. wear-out, product reaching end of life
• η is the value in time by which 63.2% of all failures will
have occurred and is termed the characteristic life
Bath-tub curve and the Weibull

Hazard function
Useful Life
=1
<1
>1
Time
Infant Wear Out
Mortality
23
Interpreting the plot
• If the data did not produce a straight light then:
– There may be an amount of failure-free time
• This may appear concave when viewed from the bottom
right hand corner of the sheet
– There may be more than one failure mode present
• This may appear convex shape or cranked shape (also
known as dog-leg shape)
• In this case the data needs to separated into failures
associated with each failure mode using expert judgement
and analysed separately
Example: Poor fit due to 3 months

offset
Weibull Plot of Time-in-Service Weibull Plot of Time in service (Months)
2-Parameter Weibull - 95% CI 3-Parameter Weibull - 95% CI
Censoring Column in Censoring - ML Estimates Censoring Column in Censoring - ML Estimates
99 99
90 Table of Statistics 90 Table of Statistics
80 80
70 Shape 1.87010 70 Shape 0.873095
60 60
50 Scale 57.6561 50 Scale 297.337
40 40
30 30
20 Mean 51.1897 20
Thres 2.9997
10 10
5 5
Pe r c e n t
Pe r ce nt
3 3
2 2
1 1
0.01 0.01
0.1 1.0 10.0 100.0 0.0001 0.0010 0.0100 0.1000 1.0000 10.0000 100.00001000.000010000.0000
Time in service (Months) Time in service (Months) - Threshold
The same data plotted with a three-Parameter Weibull distribution shows a good
fit with 3 months offset (location – 2.99 months)
24
Example of two failure modes
Mode 2 Beta
= 11.9
Weibull CDF
Mode 1 Beta
= 0.75
Time to failure ( hours)

Adjusted rank for censored data
Reverse Median
Rank Time Status rank Adjusted rank rank
1 10 Suspension 8 Suspended...
2 30 Failure 7 [7 X 0 +(8+1)]/ (7+1) = 1,125 9,8 %
3 45 Suspension 6 Suspended…
4 49 Failure 5 [5 X 1,125 +(8+1)]/ (5+1) = 2,438 25,5 %
5 82 Failure 4 [4 X 2,438 +(8+1)]/ (4+1) = 3,750 41,1 %
6 90 Failure 3 [3 X 3,750 +(8+1)]/ (3+1) = 5,063 56,7 %
7 96 Failure 2 [2 X 5,063 +(8+1)]/ (2+1) = 6,375 72,3 %
8 100 Suspension 1 Suspended...
25
Weibull Analysis using
software tools
• Number of software packages that can do Weibull
plotting (and other distributions), these include:
– Minitab
– Relex
– WinSMITH
– Reliasoft
• Concentrate on getting good quality data, correct
assumptions and correct interpretation from the
software
Hazard Plotting
26
Contents
• Assumptions
• Fitting parametric distributions
• Estimating parameters
• Using results for decision making
Assumptions
• Non-repaired items
• Repaired items with no trend in the time
between failures
• Time to first failure of repaired items
• Individual failure modes from non-repaired items
• Can deal with censored data
in particular multiply censored data
27
Hazard Plotting
• Cumulative hazard function

t
H(t)= h(t) dt
0

t
H(t)= f(t) /1-F(t) dt
0
H(t)= -ln[1-F(t)]
• Relationship allows derivation of cumulative

hazard plotting paper
Weibull Hazard Plotting


• h(t) =  t-1 and H(t) = t
  ()
• If H is the cumulative hazard value then
Log t = 1 log H + log 

• Weibull hazard paper is log-log paper
• The slope is 1/  and when H=1, t= 
28
Hazard plotting procedure
• Tabulate times in order and rank
• Reverse rank
• For each failure, calculate the hazard interval
– hi = 1/ no of items remaining after previous
failure/censoring (i.e. 1/reverse rank)
• For each failure, calculate the cumulative hazard
function H
n
– H = h1 + h2 + ………. +
• Plot the cumulative hazard against life value
Example 1 – vehicle shock

absorbers Distance (km)
6700 F 17520 F
6950 17540
7820 17890
8790 18450
Distance to failure for 9120 F 18960
Shock absorbers 9660 18980
9820 19410
F denotes failure 11310 20100 F
11690 20100
11850 20150
11880 20320
12140 20900 F
12200 F 22700 F
12870 23490
13150 F 26510 F
13330 27410
13470 27490 F
14040 27890
14300 F 28100
29
Example 1 – vehicle shock
absorbers 20 19 17520 F 1/19 0.2385
Rank Reverse Distance Hazard Cumulative
rank (km) (1/rank) hazard 21 18 17540
1 38 6700 F 1/38 0.0263 22 17 17890
2 37 6950 23 16 18450
3 36 7820 24 15 18960
4 35 8790 25 14 18980
5 34 9120 F 1/34 0.0557 26 13 19410
6 33 9660 27 12 20100 F 1/12 0.3218
7 32 9820 28 11 20100
8 31 11310 29 10 20150
9 30 11690 30 9 20320
10 29 11850
31 8 20900 F 1/8 0.4468
11 28 11880
32 7 22700 F 1/7 0.5896
12 27 12140
13 26 12200 F 1/26 0.0942 33 6 23490
14 25 12870 34 5 26510 F 1/5 0.7896
15 24 13150 F 1/24 0.1359 35 4 27410
16 23 13330 36 3 27490 F 1/3 1.1229
17 22 13470 37 2 27890
18 21 14040 38 1 28100
19 20 14300 F 1/20 0.1859
Example 1
• Plot the data on log 2 cycle paper x log 2 cycle
paper
• Estimate Weibull shape parameter
• Estimate Weibull scale parameter
• Interpret results
30
Example 1
Cumulative hazard plot for shock absorbers on linear
paper
30000
25000
distance
20000
15000
10000
5000
0
0 0.2 0.4 0.6 0.8 1 1.2
Cumulative hazard
Example 1
Log cumulative hazard for shock absorbers
100000 = 2.6
= 28500km
log distance
10000
R2 = 0.98
1000
0.01 0.1 1 10
log hazard
31
Example 1
Since looking at one known failure mode use the estimated
parameters to fit to the distribution
Survival plot for vehicle shock absorbers with
Beta =2.6 and Eta=29000km
1.2
1
0.8
0.6
0.4
0.2
0
0 5000 10000 15000 20000 25000 30000 35000 40000
kilometers
Example 2:Hazard plot on

linear paper
Cumulative hazard plot for O ring failures
2000
1500
h o u rs
1000
500
0
0 1 2 3 4 5 6
cumulative hazard
32
Example 2: Hazard plot on log
paper
log cumulative hazard for O ring failures
10000 = 1.01
1000 = 360hrs
log hours
100
R2 = 0.98
10
1
0.01 0.1 1 10
log cumulative hazard
Example 2: Interpretation of
results
•  = 1.01 is approximately an exponential
distribution and constant failure rate
•  = 360 hrs = 1/ = Mean Time to Failure
• Calculating the MTTF from the data gives:
– Total hours/number of failures
– 26839/73 = 367 hrs
33
Example 2: Survival function
R(t) - Survival Function for O ring failures
1.2
1
0.8
0.6
0.4
0.2
0
0 500 1000 1500 2000
Hours to failure
Example 2: Failure Distribution
F(t) for O ring failures
1.2
Probability of failure
1
0.8
0.6
0.4
0.2
0
0 500 1000 1500 2000
Hours to failure
34
Example 3 : pumps
pump no age at failure failure mode
• Two dominant failure 1 1180 m
modes 2 6320 m
– Impeller failure (I) 3 1030 i
4 120 m
– Motor failure (m)
5 2800 i
6 970 i
7 2150 i
8 700 m
9 640 i
10 1600 i
11 520 m
12 1090 i
Example 3: ignoring failure

modes
log cumulative hazard for all failures
10000
1000
age
100
10
1
0.01 0.1 1 10
cum hazard
35
Example 3: impeller failure
log cumulative hazard plot for impeller failures
10000 = 1.95
1000 = 1900hrs
a ge a t fa ilur e
100
R2 = 0.93
10
1
0.1 1 10
cumulative hazard
Example 3: motor failure

log cumulative hazard plot for motor failures
10000 = 0.76
1000 = 3647hrs
a ge a t fa ilure
100 R2 = 0.978
10
1
0.01 0.1 1 10
cumulative hazard
36
Advantages of Cum Hazard
Plotting
• It is much easier to calculate plotting positions

for multiply censored data using cum hazard
plotting techniques.
• Linear graph paper can be used for exponential
data and log-log paper can be used for Weibull
data.
Disadvantages of Cum
Hazard Plotting
• It is less intuitively clear just what is being plotted.
– Cum percent failed (i.e., probability plots) is meaningful and
the resulting straight-line fit can be used to read off times
when desired percents of the population will have failed.
– Percent cumulative hazard increases beyond 100% and is
harder to interpret.
• Normal cum hazard plotting techniques require exact
times of failure and running times.
• With computer software for probability plotting, the
main advantage of cum hazard plotting goes away
37
Summary
• Important lifetime distributions
– Failure distribution (CDF), Survival function R(t) and
the hazard function h(t)
• Some parametric distributions
– Exponential, Weibull and Normal
• Weibull probability plotting
• Distribution fitting using hazard plotting
techniques
38

Reliability Introduction PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Reliability Introduction PDF

Uploaded by

Copyright:

Available Formats

An Introduction to Reliability and

PEUSS 2011/2012 Reliability and Life distributions Page 1

Objectives of the session

PEUSS 2011/2012 Reliability and Life distributions Page 2

PEUSS 2011/2012 Reliability and Life distributions Page 3

PEUSS 2011/2012 Reliability and Life distributions Page 4

• The area under the curve is equal to 1

Failure Time distributions

PEUSS 2011/2012 Reliability and Life distributions Page 7

PEUSS 2011/2012 Reliability and Life distributions Page 9

Cumulative distribution function

Failure Function, F(t)

PEUSS 2011/2012 Reliability and Life distributions Page 10

The probability of failure before 500 hours is 0.8

0.4 F(t) = 1 - R(t)

PEUSS 2011/2012 Reliability and Life distributions Page 12

The probability of surviving up to 500 hrs is 0.2

PEUSS 2011/2012 Reliability and Life distributions Page 14

PEUSS 2011/2012 Reliability and Life distributions Page 16

PEUSS 2011/2012 Reliability and Life distributions Page 17

PEUSS 2011/2012 Reliability and Life distributions Page 19

PEUSS 2011/2012 Reliability and Life distributions Page 20

PEUSS 2011/2012 Reliability and Life distributions Page 23

Three parameter Weibull

• Location parameter, m , is the mean

PEUSS 2011/2012 Reliability and Life distributions Page 27

Fitting parametric distributions

PEUSS 2011/2012 Reliability and Life distributions Page 28

PEUSS 2011/2012 Reliability and Life distributions Page 30

PEUSS 2011/2012 Reliability and Life distributions Page 31

PEUSS 2011/2012 Reliability and Life distributions Page 32

PEUSS 2011/2012 Reliability and Life distributions Page 33

Repaired and non-repaired data

• Non-repaired data when only one failure can

• Repaired data when interested in the pattern of

PEUSS 2011/2012 Reliability and Life distributions Page 34

PEUSS 2011/2012 Reliability and Life distributions Page 35

PEUSS 2011/2012 Reliability and Life distributions Page 36

– 15, 42, 74, 117, 168, 233, and 410

• Machine S fails at the following running times

– 177, 242, 293, 336, 368, 395, and 410

In neither case can Weibull analysis be used as there is

PEUSS 2011/2012 Reliability and Life distributions Page 41

Rank the data

PEUSS 2011/2012 Reliability and Life distributions Page 43

• Plot times on x-axis

PEUSS 2011/2012 Reliability and Life distributions Page 44

Bath-tub curve and the Weibull

Example: Poor fit due to 3 months

PEUSS 2011/2012 Reliability and Life distributions Page 48

Time to failure ( hours)

Adjusted rank for censored data

PEUSS 2011/2012 Reliability and Life distributions Page 50

PEUSS 2011/2012 Reliability and Life distributions Page 51

PEUSS 2011/2012 Reliability and Life distributions Page 52

PEUSS 2011/2012 Reliability and Life distributions Page 53

PEUSS 2011/2012 Reliability and Life distributions Page 54

• Relationship allows derivation of cumulative

PEUSS 2011/2012 Reliability and Life distributions Page 55

Weibull Hazard Plotting

• The slope is 1/  and when H=1, t= 

PEUSS 2011/2012 Reliability and Life distributions Page 56

Example 1 – vehicle shock