Measuring Reliability

Data collection
Servicing records,
warranty records,
repaired product records
spares used records
Data from development testing
Disposal records
Customer complaints
Customer reports and comments can also be
used to help complete a data set.
Insurance claims and coverage records

1
Reliability Metrics
Failure Rate, or its inverse
– Mean Time Between Failure: MTBF
MTTF – Mean time to failure for non-repaired
items
Removal Rate, or its inverse
– Mean Time Between Removal: MTBR
Time To Failure: TTF
Maintenance Free Operating Period: MFOP
Failure Free Operating Period: FFOP
Probability of failure

Histogram

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

𝑃(𝑡 < 𝑡 < 𝑡 = 𝑓 𝑡 𝑑𝑡

Probability distributions
Hours to failure

0.5
0.4
Relative frequency

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%

1 ( )
𝑓 𝑡 = 𝑒
𝜎 2𝜋

Cumulative distribution function

The CDF known as F(t)

𝐹 𝑡 =∫ 𝑓 𝑡 𝑑𝑡

F(t) gives the probability that a measured

value will fall between -  and t

5
 Cumulative distribution
Cumulative probabilty

1.2
cum ulative 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)

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

F(t) = 1 - R(t)

6
Survival Function
Survival Function

1
Probability of s urvival 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
ℎ 𝑡 = =
( ) ( )

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 = 
MTBF is the 63rd percentile i.e. time at
which 63% of population will have failed

Calculating the Failure Rate

Assuming the data fits an exponential

distribution
The equipment failure rate λ is estimated as:
– Total number of failures/total equipment hours
The component failure rate λ is estimated as:
– Total number of failures/total component hours

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 1,000 hours
There have been 30 confirmed failures of this component

The failure rate is given by:

– 30/(5*200*10*1,000) = 0.000003 = 3 fpmh (failures per million
hours)
The MTTF is 1/0.0000003 = 333,333

Example
• 100 units in the field
• 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 = 15,000

10
Why predict reliability
Forecasts for:
– Support costs;
– Spare requirements;
– Warranty costs;
– Marketability;
– Design alternatives;
– Design improvement;
– Lessons learned and design rules;
– Compliance.

Reliability data inputs

• Historical field failure data, however:
• No data is collected beyond warranty;
• Time to failure is not recorded or difficult to derive;
• Cause of failure is not determined;
• Symptom reported rather than root cause;
• Spread of customer base inconsistent;

• Supplier information

• Physics of failure models

• Test data

• Data handbook predictions

11
Failure rate prediction
When no data is available
Parts count analysis
Parts stress analysis
Assumes:
– Components are logically connected in
series
– Component failure rates are constant in time
i.e. exponential distribution
– Component failures are independent

Parts count method

(Mil Hdbk 217)
Overall equipment failure rate is
calculated using:

• 𝜆=∑ 𝑁𝜋 𝜆

• Where n = number of parts categories (e.g. electrolytic

capacitors etc)
• 𝑁 = quantity of th part
• 𝜋 = quality factor of th part
• 𝜆 = base failure rate of th part

12
Limitations and benefits
Benefits:
– Quick, little training required, minimum
information required, software available
Limitations:
– Assumes components cause units to fail, simple
structures only, cannot evaluate fault modes,
poor precision, handles constant failure rate only
Mil handbook 217 and industrial
counterparts
– Predict system failure rate

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

13
Weibull distribution

PDF:𝑓 𝑡 = t e

CDF:F 𝑡 = 1 − 𝑒

Reliability:𝑓 𝑡 = 𝑒

Weibull distribution
ℎ(𝑡) = 𝑡
When =1, h(t)= 1/ , therefore  =1/
When >3.5 the distribution approximates to
a normal distribution
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

14
 Bath-tub curve and the Weibull
Hazard function
Useful Life

=1
<1
>1

Time
Infant Wear Out
Mortality

Weibull Reliability Plot

15
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, s, is the standard
deviation
Lognormal more versatile, always positive

How can these distributions help in

reliability analysis?
Parameter estimation – in the case of Weibull -  and 
using probability plotting methods:
– 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
Investigate the pattern of failures over-time
To help identify causes of failure
Accelerated life test data analysis
Maintenance planning

16
Example of two failure modes

Mode 2
Beta =
11.9

Weibull CDF

Mode 1
Beta =
0.75

Time to failure ( hours)

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

17