Definition of Reliability

Reliability is defined as………

The Probability that a system or component

will perform its required function under stated
condition for a specified period of time
 Definition of Reliability –Probability
• Reliability is first and foremost definition as “
Probability “
• Numeric value between “0” and “1”
• The ability to express the reliability in numeric terms
allow for direct comparison between different
design alternatives of products
• Reliability 0.97 indicates that on average out of 100
item , 97 are performing their intended function
for a given period of time under given operating
condition
Definition of Reliability –Operating Condition
• Another aspect of the definition is “operating
condition “, which involves type and amount of usage
and the environment in which product is used
• Operating condition consideration like loading,
temperature , storage temperature , altitude ,
humidity , chemical exposure , shock , vibration ,
transportation handling etc
• Reliability often include extreme condition in
addition to ambient environment
 Definition of Reliability –time
• Final aspect of Definition is “ time “
• All product must operate for their lifetime
• Reliability is evaluated for a given frame of
time
• Device having 0.97 reliability with operating
time 1000 hr is inferior then component with
same reliability with 5000hr operating time
Mathematical Basic of Reliability

• Mathematics of the reliability is based on

Probability theory which deals with uncertainty
of random events
• In most reliability evaluation we deal on
quantitative measures such the time to failure of
the part, Where the random variable is time (t)
• Because the product can be found failed at any
time( time-space) after time (t=0).
• Ex. Any value between (0-5 ) In this case random
variable is continuous
Mathematical Basic of Reliability
• All random events ( including failures) have an
underlying probability function
• Given Continuous random variable
– Probability density function (pdf) as f(t)
– Cumulative distribution function (cdf) as F(t)
• Pdf and cdf gives complete description of
probability distribution of continuous random
variable
Probability distribution
• In practical situation data is collected to analyze
the behavior of system or components
• Data collection
– Data from actual operation / Series of experiment on
selected samples
– Data from Data bank : No failures for discrete
system , Time to failure , repair time for continuous
• Data processing
– Frequency distribution
– Probability distribution
State variable
The state of the item at time t may be described by the state
variable X ( t )

The state variable of a non repairable item is illustrated in

above figure will generally be a random variable.
Time to failure
• Time to failure of an item mean the time elapsing from when
the item is put into operation until it fails for the first time.
• We set t = 0 as the starting point. At least to some extent the
time to failure is subject to chance variations. It is therefore
natural to interpret the time to failure as a random variable, T .
• Different time concepts may be used, like
❑ Calendar time
❑ Operational time
❑ Number of kilometers driven by a car
❑ Number of cycles for a periodically working item
❑ Number of times a switch is operated
❑ Number of rotations of a bearing
In most applications we will assume that the time to failure T is
a continuous random variable (Discrete variables may be
approximated by a continuous variable)
Reliability Measures

❑ Mean time to failure (MTTF)

❑ Number of failures per time unit (failure rate)
❑ The probability that the item does not fail in a
time interval (0, t] (survival probability)
❑The probability that the item is able to function
at time t (availability at time t)
Physical Approach
Time Dependent
Reliability Measures
In this chapter we introduce the following measures:
❑The reliability (survivor) function R(t)
❑ The failure rate function z(t)
❑ The mean time to failure (MTTF)
❑ The mean residual life (MRL)
of a single item that is not repaired when it fails.
Life Distributions
The following life distributions are available in the
literature
❑ The exponential distribution
❑ The gamma distribution
❑ The Weibull distribution
❑ The normal distribution
❑ The lognormal distribution
❑ The Birnbaum-Saunders distribution
❑ The inverse Gaussian distributions
In addition ,three discrete distributions:
❑ The binomial distribution
❑ The Poisson distribution
❑ The geometric distribution
Distribution function
• Assuming that the time to failure T is continuously
distributed with probability density function( Failure
PDF ) f ( t ).
• Distribution function F(t) (Failure distribution
function ) denotes Probability that the item will fail
within time interval (0,t).
• The Distribution function of F(t) is Probability that the
item will fail within time interval (0,t).
Failure Probability density function
• The Failure probability density function (pdf) is defined as

• When Δt is small then

• When we are standing at time t and ask : what is the probability that
the item will fail in interval ( t, t+Δt) . The answer is approximately
f(t).Δt ( Product of Failure Density function and time )
 Failure Distribution function

The area under pdf –curve ( f(t)) is always 1 and

The area under the cdf –curve to the left of t is equal to F(t)

The area under cdf-curve between t1 and t2 is

 Reliability Function

R(t)= the Probability that the item will not fail in (0,t)
R (t) = the probability that the item will survive at least to
time t
R (t) = is also called the survivor function of the item
Failure rate function

• The probability that an item will fail in time interval

(t,t+Δt) when we know the item is functioning at
time t is ( Conditional probability )

• Failure Rate function can be obtained by dividing

above probability by length of time interval Δt and
letting Δt→0

For Small Δt
Failure rate function
• When we follow an item from time 0 and note
that it is still functioning at time t, the
probability that the item will fail during a short
interval of length Δt after time t is z(t) ·Δt
• The failure rate function is a “property” of the
item and is sometimes called the force of
mortality (FOM) of the item.
Some formulas
 Mean Time to failure
R(t)= 1-F(t)
dF(t)/dt= -R’(t) = f(t)

Since f(t) =-R’(t)

By partial Integration

If MTTF<∞ , it can be
shown that

in that case
 Mean Time to failure
• The expected value of the time to failure of
non repairable system is called mean time to
failure (MTTF)
• If equipment is repairable then expected value
of the time between repaired failure is
commonly described as Mean time between
failure (MTBF)
Mean residual life
• Consider an item that is put into operation at time t=0
and is still functioning at time t . The probability that
the item of age t survives an additional interval of
length x is

• R(x|t) is called the conditional survivor function of

the item at age t. The mean residual time (MRL)of the
item at age t is
 MTTF Basic Concepts
• MTTF definition
• The MTTF of the components or systems is the mean of the distribution of
times accumulated by identically operated components or systems
before they fail
• Note
– Once failed the component ( System ) get removed : There is no
repair
– Mean time between failure ( MTBF) applies to the component ( System
) which undergo repeated cycle of failure and repair
– MTBF and MTTF are often used interchangeably for non repairable
component
• How is MTTF is determined
– Test N identical components
– Record time to failure T1 , T2 , T3 …………..TK ( K<N )
– Estimated MTTF = total component –hour generated by all component whether they
failed or not / number of failed components K
• Example
– Test 3 devices : Test runs for 300 Hrs : 2 failure in 150 hr and 200hr
– MTTF = 300+150+200/2 = 325hrs
Distribution Transformer with failure pattern as shown below
 Exponential distribution
• Consider an item that is put into operation at time t =
0. Assume that the time to failure T of the item has
probability density function (pdf)

• This distribution is called the exponential distribution

with parameter λ, and we sometimes write T ~ exp(λ).
• The survivor function of the item is
 Exponential distribution
• The failure rate function is

• The failure rate function is hence constant and

independent of time. Consider the conditional
survivor function
 Exponential distribution
• A new item, and a used item (that is still functioning),
will therefore have the same probability of surviving a
time interval of length t.
• A used item is therefore stochastically as good as new
• So for estimation of the reliability function, the mean
time to failure , it is sufficient to collect data on the no
of hours of the observed time in operation and no of
failures.
• The age of item is of no interest in this connection
• Commonly used life distribution
 Failure Probability in a Time Interval
• A priori probability of failure in time interval t, Q(t)=1-e-λt
• A priori Probability:
– Probability calculated by logically examining existing
information
• A posteriori Probability :
– Conditional Probability that is assigned after relevant information
taken in account
• The probability of failure in the next interval t actually depends
conditionally upon its behavior preceding that interval . That is
it cant fail in that interval if it is already failed prior to that
interval
• It is, therefore, required to determine (a posteriori ) Probability
of a component failing in interval t given that it has survived
prior to that interval
Posteriori Failure
• Probabilty of Component failing during t given that it has
surrvied upto T Qc(t)=P(A|B)
Event
A failure Between T , T+t
B Surviving Upto T
Posteriori Failure
• Reliability evaluation in useful life of component is
relatively simple as exponential distribution
function is applicable
 Poisson Distribution
• The Poisson distribution is a discrete probability
distribution for the counts of events that occur randomly
in a given interval of time (or space).
– If we let X = The number of events in a given interval.
– Then, if the mean number of events per interval is λ
• The probability of observing x events in a given interval
is given by
 Poisson Distribution
• Can be used to evaluate the probability of an
isolated event occurring a specific no of times
in given time interval
• example no of failure , no of lighting stroke
in given time
• Requirements
– Events must be random
– Failure rate (Hazard Rate ) must be constant
• Only applies useful life period of the system
Poisson Distribution-Example
• If the Average no of cable fault per year per 10km of cable is
0.05. Evaluate probability the occurring of fault 0, 1,2 ,3 ,4 in
– 20 Year Period
– 40 Year Period
• Solution
• Failure rate λ=0.05f/year
• For 20 Year Period λt=0.05x20=1

For Twenty Year Period

No Of Failure 0 1 2 3 4
Probability 0.36778 0.36778 0.18389 0.061297 0.015324
 Poisson Distribution-Example

For 40 Year Period λt=0.05x 40=2

For Forty Year Period

No Of Failure 0 1 2 3 4
Probability 0.135335 0.270671 0.270671 0.180447 0.090224
Poisson Distribution-Example
Weibull distribution
• The time to failure T of an item is said to be
Weibull distributed with parameters α and λ[ T
~ Weibull(α, λ) ] if the distribution function is
given by

• The corresponding probability density function

(pdf) is
 Weibull distribution

Where λ is scale parameter α is

shape parameter

When α =1 weibull dis. Is equal to

exponential distribution
 Weibull distribution

Used to model life distribution , where the failure rate is

decreasing , increasing or constant
Example 2.1
Median and Mode
• The median life is defined by R(tm) =0.5
• The median divides the distribution in two halves. The
item will fail before time tm with 50% probability,
and will fail after time tm with 50% probability.
• The mode of a life distribution is the most likely
failure time, that is, the time tmode where the
probability density function f(t) attains its maximum