Topic : Reliability Models

Subject Code : HMMS18005

Subject Name : Statistical Quality Control and
Reliability
Delivered by : Dr. A. Rathi
Introduction of Reliability:
• Due to the rapid advancement of science and technology and to meet the ever
increasing needs of the society, industries are increasingly introducing automation
for producing goods ranging from the simplest to highly complex systems.
• However all processes and products are prone to failure. For example, a television
remote control may stop functioning or a malfunction may occur in any bother
household appliance; an automobile engine starter may fail, an aeroplane may
crash due to the failure of some of its components credit card transactions may
fail, and so on.
• We can think of many more similar examples from our daily life. So we would
like to understand failures, present them in a different manner and see how we can
try to prevent the interruption due to failure.
• Reliability is nothing but a science of better presentation of failure. For example, the
statement that 20% students of the class failed the test can be better presented as 80%
students of the class passed the test.
In other words,
• All systems are made up of parts and components.
• The failure of a part or component may lead to the failure of the whole system in some
cases and may not in others depending on the functional relationship of the components.
• In any case each component contributes to the system reliability (or unreliability) and it is
worthwhile to express the system reliability interms of component reliability.
• This necessitates the careful analysis of functions of all components and their failure
models.
What is reliability?

The quality of being trustworthy (truthful) or of performing consistently

well.

Eg: The measurement is considered reliable.

When we measure the temperature of a liquid sample several times
under identical conditions. The thermometer displays the same
temperature every time, so the results are reliable.
Why it is important to study?

Reliability is important because it determines the value of a

psychological test or study. If test results remain consistent when
researchers conduct a study, its reliability ensures value to the field of
psychology and other areas in which it has relevance, such as education
or business.
Use of reliability in day to day life.

Reliability engineering can be applied to many business functions,

from design to maintenance.
This type of engineering has been referred to as operations or activities
that produce goods in the manufacturing world to minimize equipment
downtime, lower maintenance costs, and avoid business operations
interruption.
Reliability may be defined in several ways:
• The idea that an item is fit for a purpose with respect to time. In the
most discrete and practical sense: "Items that do not fail in use are
reliable" and "Items that do fail in use are not reliable”.
• Reliability of a device is denoted as and defined as R (t)
• R(t) = P (T > t), t > 0. Where “T” is a continuous random variable
representing the life length of the device.
Definition: Reliability is defined as the probability that a product, system, or service will
perform its intended function adequately for a specified period of time, or will operate in a
defined environment without failure. (i.e) Reliability of a component is the probability that it
performs its intended function adequately for a specified period of time under the given
operating conditions.
Probability:
when a number of identical components operate under similar conditions, the failure
times of components generally vary from component to component and we cannot predict
their failure times exactly in advance. But we can describe the phenomenon of failure in
probabilistic terms.
For example, suppose we put identical components in operation at time t = 0 under
identical conditions. Suppose after t units of time, we have
(t) = number of components that are operating at time t
(t) = number of components that are failed at time t
Then Reliability R(t) at time t is defined as R(t) =
Thus, reliability of a component / system at time t is the probability that it performs
its function without failure. Being a probability, its value lies between 0 and 1. From
equation (1), we can see that,
If (t) = 0 , R(t) = 0 and
If (t) = , R(t) = 1
By definition (t) satisfies the following relation:
0 ≤ (t) ≤ for any time interval [0,t]
=> [on dividing by ]
=> 0 ≤ R(t) ≤ 1
Suppose that the reliability of the refrigerator of a particular company at 10 years is
R(t) = 0.9452. then this can be interpreted as follows: out of 10000 refrigerators of
this particular company, approximately 9542 refrigerators worked without failure
for 10 years.
Intended Function: The intended function of a product is the function/
job/ work which is expected to perform when we put it in an operation
under stated conditions.
For example, if a pump is designed to deliver at least 300 gallons
of water per minute, the intended function of the pump is to deliver 300
gallons or more water per minute.
Adequately: A product is said to perform adequately if it is performing
its intended function as expected.
For example, if the intended function of a water pump is to
deliver 300 gallons or more water and it is delivering 300 gallons or
more water per minute, it means that it is performing adequately. But if
the pump delivers less than 300 gallons of water per minute we say that
it has failed or not performing adequately.
Time: The reliability can be meaningful only if it is related to a time
interval or period of time.
For example, we can say that the reliability of a component is 0.98
for a mission time of 100 hours. But the statement ‘Reliability of a
component is 0.95’ is meaningless since the time interval is unknown. The
Reliability ‘t’ can be defined as follows:
R(t) =
In the above definition, time ‘t’ means the interval [0,t]
R(0) = 1 since no component fails at t = 0 and R(t →
Since no component can survive forever.
Reliability Function:
Let the random variable, called time to failure or failure time of the component / system be
denoted by T. Then by definition, reliability [R(t)] at time t, of the component/system is given by,
R(t) = Probability that the time to failure of the component / system > t
= P[T > t] ≥ 0 (Probability of any event A always satisfies 0 ≤ P(A) ≤ 1
The Reliability of a component / system when expressed in terms of time t is known as reliability
function and is generally denoted by R(t)
Let us consider two particular cases of the Reliability function:
when t = 0 and when t →
R(0) = P[T > 0] = 1 ( We assume that at t = 0, the component system is
performing its intended function)
R(t →) = P[T > t(t →)] =0 ( No component / system can perform its
intended function forever)
The Probability of survival of a component decreases as the life time of the component increases.
Ultimately this probability will approach zero, since no component can perform its intended forever.
Mean Time to Failure (MTTF):
we often interested in knowing the mean time to failure of a component
rather than the complete failure details. This parameter will be assumed to be
the same for all the components which are identical in the design and operate
under identical conditions.
If we have life – tests information on a population of N items with
failure times then the MTTF is defined as,
MTTF =
However, if a component is described by its reliability function and hazard
model, then the MTTF is given by the mathematical expectation of the
random variable T describing the time to failure of the component.
Therefore, MTTF = E[T] =
f(t) = = -
Hence, MTTF = -
= -t R(t) (limits 0 - ) + =
MTTF = E[T] = =
The MTTF can also be computed using the Laplace transformation of R(t), i.e.
MTTF = =
However,
= R(s)
Where R(s) is the Laplace transform of R(t),
Thus, MTTF = R(s)
Instantaneous Failure Rate or Hazard Rate:

• Hazard measures the conditional probability of a failure given

the system is currently working
• The failure density measures the overall speed of failures
• The hazard rate measures the dynamic speed of failures
• To understand the hazard function we need to review
conditional probability and conditional density functions
Time Dependent Hazard Models:

• Constant Hazard Model

• Linear Hazard Model
• Nonlinear Hazard Model
• Gamma Model:
Time Dependent Hazard Models:
Constant Hazard Model:
The constant Hazard Model takes the form, z(t) = 𝛌
Where 𝛌 is a constant and is independent of time. This characteristics is being exhibited
by many products, particularly, electronic components. This model has been used for
many years in reliability studies. An item with constant hazard rate will have the
following reliability and associated functions.
f(t) = 𝛌
R(t) =
F(t) = 1-
The mean time to failure of the item is,
MTTF = =
Linear Hazard Model:
Many components that are under mechanical stresses fail due to wear out or
deterioration. The hazard rate of such components increasing with time. The linear hazard
model is the simplest time – dependent model and has the form,
z(t) = bt, t > 0
Where b is constant.
R(t) = exp [- ]
= exp (-
f(t) = bt exp (-
It may be noticed that f(t) is a Rayleigh – Density function. The Mean time to failure is
given by,
MTTF = dt
Nonlinear Hazard Model:
The hazard rate is not always a linearly increasing function of time. It is therefore
useful to have a more general form of the hazard model:
Z(t) =
Where a and b are constants.
This gives as,
R(t) = exp[]
f(t) = aexp[]
This general form is known as the Weibull model and generates a wide range of curves
for various sets of a and b. This model includes the previously discussed models, i.e. b =
0 represents the constant – hazard model and b = 1 represents linearly increasing hazard
model.
Gamma Model:
Another two – parameter model which is applied in reliability work is the
gamma model. The hazard model and associated functions are,
Z(t) =
R(t) =

where a is a positive integer and 𝛌 is a positive constant.

f(t) =

MTTF =
For a > 1, z(t) increases and for a = 1, it becomes a constant – hazard rate
model. The above functions are applicable to a component which is replaced (a-1)
times by identical components with practically zero replacement time. In effect a
total of a components were used in sequence to accomplish the task. The case a = 1
represents that there is only one component.
Stress – Dependent Hazard Models:
Basically, the reliability of an item is defined under stated operating environmental
conditions. This implies that any change in these conditions can affect the failure rate of
the item and consequently its reliability. The failure rate of almost all components is stress
– dependent. A component can be influenced by more than one kind of stress. For such
cases, a power function model of the form
h(t) = z(t)
Can be postulated for each equipment, where are positive constants, are stress ratios for
two different kinds of stresses and z(t) is the failure rate at rated stress conditions. For
example, for an electrical component, the stress may be,
= and =
These models are used for accelerated testing of components. Experience shows that the
parameter ‘a’ takes a range of values varying from 1 to 8 depending upon the type of
component. In addition to stress factor, there can be other factors, such as the complexity
factor, application factor, quality factor, etc., which should be used for the correct
estimation of failure rates.
Derivation of Reliability Function using Markov Model:
This method of computing the reliability of a component can be used when we
know the hazard rate of the component. It is assumed that the component has only two
states:
State 0: the component is good
State 1: the component is failed

fail and reach state 1 in the time interval 𝜟t is z(t)𝜟t,

If the component is in state 0 at time t, then the probability that the component will

where z(t) = the hazard rate of the component

Let (t) = the probability that the component will not fail in time t
(t) = the probability that the component will fail in time t
The probability that the component will stay in state 0 at time t+ 𝜟t is given by
(t +𝜟t) = (t) – z(t) 𝜟t (t)
When the component fails, the probability of its remaining in state 1 ( failed state) is unity

at time t+𝜟t is,

because the component is not repaired. The probability of the component being in state 1

(t +𝜟t) = (t) + z(t) 𝜟t (t)

The probability equations can be rewritten as,
= -z(t) (t)

As 𝜟t → 0, we have
= z(t) (t)

-z(t)(t)
The solution of this equation yields
log = - +
= exp[- + ]
= exp[ - ]
Where are constants of integration.
Initially when t = 0 the component is in state 0 and therefore
=1=
Thus the reliability of the component is,
R(t) = = exp[- ]
• Example:
The following table shows a status report of a set of 1000 components
which were put to operation simultaneously
Operating 0 100 200 300 400 500 600 700 800 900 1000
time
(Hours)

No of 1000 895 810 730 660 600 545 495 450 410 373
surviving
components

Estimate the hazard rates and density function for the above data.
System Reliability Models:

• The main concern of a systems engineer is to estimate various reliability parameters

of systems he is dealing with.
• The systems may vary from simple (consisting of one or two components) to
complex (involving thousands of components).
• One approach for analyzing such systems is to decompose then into subsystems of
convenient size each representing a specific function, reliabilities of subsystems are
then estimated and combined to determine the reliability of the entire system using
certain probability laws.
• This approach requires a complete knowledge of the physical structure of the system
and the nature of its functions sufficiently well to determine the behavior of the
system in the event of failure of a subsystem. The subsystem may consist of one or
more components whose reliabilities are known.
System:
A group of devices or artificial objects or an organization
forming a network especially for distributing something or
serving a common purpose.

Example:
• a telephone system
• a heating system
• a highway system
• a computer system
Techniques of Reliability Evaluation:

Techniques of Reliability evaluations of a system depend upon

the configuration of its components.
In practice, several types of configuration are used. We shall
discuss the reliability evaluation techniques for the following:
(1) Series configuration
(2) Parallel configuration
(3) k out of m configuration
Series Configuration:
Suppose a system has n components. From a reliability point of
view, the components of a system are said to be in series configuration
if they are connected in such a way that,
• For the successful operation of the system of all n components must
perform their intended function, and
• only one needs to fail for system failure
Example: The string of lights (bulbs or LEDs) used in festivals and
weddings is a good example of a series system.
The system performs only when all bulbs /LEDs are working and only
one of them needs to fail for the system to fail.
• Consider a system having a total of n components.
• The successful operation of the system depends on the proper
operation or all the n – components, then we say that the system
configuration is a series or chain type.
• The information at the IN end will reach the OUT end only if all the n
components function satisfactorily.
• Many complex systems can be reduced to such a simple structure.

IN 1 2 n OUT
Reliability of Series System:
A Series system is nothing but a system, all of whose components
are in series configuration.
Let denote the event that the component ‘i’ is good. ( i.e.
functions satisfactorily) / performs its intended function successfully,
where i = 1, 2, 3, ….., n and the event that the component ‘t’ is bad /
performs its intended function not successfully.
If R denotes the reliability of the series system, then from the
definitions of reliability and series system, we have,
The event representing system success is then the intersection of . The
reliability of the system is the probability of this event and is given by
R = ()
 we take the intersection of events because reliability is a
probability and the definition of series system implies that for the
successful operation of the system, all components must work.
If Events are not independent (i.e. dependent) then we apply conditional
probability and get,
R=(
means the probability of the event on the condition that has occurred.
However, we have assumed that the components are independent and
therefore
R = ()
This can be evaluated by using failure events also. In this case
R = 1 – (Probability of the system failure)
The system fails if any of the components fail and therefore
R = 1- (
The reader can easily verify by using the relationship
)+)=1
The time dependent reliability function is,
R(t) = (t)
Where is the probability that the component ‘i’ is good at time t. If time to
failure of components are exponentially distributed, then in this case
= exp(-t)
and R(t) = ] = exp [ -t]
Mean time to failure of the system is,
MTTF = dt = =
Where is the mean life of the component i.
For any general hazard model,
R(t) = exp[-]
For series systems with components having constant hazard rates, the
following are true:
(i) The system failure – rate is the sum of the failure rates of
individual components.

This means that the system failure is also constant.

Example 1: Evaluate the reliability of the series system having the reliability
block diagram shown in the following figure for a mission 500 hours, where
the given reliabilities of the components are for a mission of 500 hours.
Assume that the failure of any component does not affect the functioning of
the other components.
Input 0.60 0.75 0.95 0.80 Output
Solution:
We know that the reliability of a series system is equal to the product of the
individual reliabilities of the components. Therefore, the required reliability
of the series system is,
R = 0.60 x 0.75 x 0.95 x 0.80 for a mission of 500 hours
= 0.342 for a mission of 500 hours
Limitations of the series system:
• In the previous example it is noted that the reliability of the system
is less than that of the worst component.
• This problem arises due to limitations of the series system. One of
the main limitations of the series system is that reliability of the
series system can never be better than the reliability of the worst
component in the system.
• The second major limitation of the series system is that reliability of
the system decreases as the number of components in the system
increases.
Example 2: A system has three components connected in series having reliabilities 0.40,
0.70, 0.80 respectively, for a mission of 400 hours. What is the percentage increase in the
reliability of the system in each of the following cases?
(i) Reliability of the first component is increased by 0.1 and that of the second and third
components remains the same.
(ii) Reliability of the second component is increased by 0.1 and that of the first and third
components remains the same.
(iii) Reliability of the third component is increased by 0.1 and that of the first and second
components remains the same.
Solution: first of all, we calculate the reliability of the system as per the given
reliabilities of the components.
R = 0.4 x 0.7 x 0.8 for a mission of 400 hours
R = 0.224 for a mission of 400 hours
The reliabilities of the system and percentage increase in the reliability of the system in
the cases (i), (ii) and (iii) are shown in columns 5 and 6 respectively of below table.
Reliability of the Reliability of Reliability of Reliability of Percentage Increase in the
First the Second the Third the system Reliability of the System
Component Component Component (6)
(1) (2) (3) (4) (5)
0.4 0.7 0.8 0.224 -
(i) 0.5 0.7 0.8 0.280 x 100 = 25%
(ii) 0.4 0.8 0.8 0.256 x 100 = 14.29%

(iii) 0.4 0.7 0.9 0.252 x 100 = 12.5%

It is noted that the improvement in the reliability of the system (in percentage) is higher
when reliability of the weakest component is increased by 0.1in comparison with the
cases when reliabilities of the other two components are increased one at a time by the
same amount (0.1).
• This suggests that if a system engineer wants to improve the
reliability of the series system, he / she should concentrate on the
improvement of the reliability of the poorest component.
• You may have heard a saying that a chain is only as strong as its
weakest link. This saying best fits the series system.
• So we can say that the unit (component / subsystem), which is the
poorest in the system from a reliability point of view, dictates the
reliability of a series system.
Example 3: Consider a component having reliability 0.95 for a mission of
100 hours. Evaluate the reliability of a series system in the cases having
2, 3, 4, …..,10 identical components having reliability 0.95 for the
same mission of 100 hours.
Solution: If the single component of a system has reliability 0.95 for a
mission of 100 hours, then the system will also have the same reliability for
a mission of 100 hours.
But the reliability of the series system for a mission of 100 hours will vary
if it has 2, 3, 4, ….,10 such identical components connected in series from a
reliability point of view.
In fact, the reliability of the system decreases with an increase in the
number of its components.
Number of Components Reliability of Series System % Decrease in the Reliability of
the System Compared to the
System having Single Component
1 0.95 -
2 = 0.9025 x 100 = 5%
3 = 0.8574 x 100 = 9.75%

4 = 0.8145 x 100 = 14.26%

5 = 0.7738 x 100 = 18.55%
6 = 0.7351 x 100 = 22.62%
7 = 0.6983 x 100 = 26.49%
8 = 0.6634 x 100 = 30.17%
9 = 0.6302 x 100 = 33.66%
10 = 0.5987 x 100 = 36.98%
The system mean life (T) and mean life of components ( are related by

Another important characteristics of a series system is that its reliability is always worse than the poorest
component and this is a point of concern for reliability engineers.
Example:
A system having three subsystems with reliability of 0.6, 0.9 and 0.8 will have a reliability of
0.6 x 0.9 x 0.8 = 0.432
Which is much less than 0.6. This illustrates the effect of an increase in the number of components on the
system reliability. Thus the component – minimization technique is one of the useful means of achieving
higher reliability of systems.
There are situations where the components are identical and for such cases
R=
For high – reliability components ( Say < 0.2)
R ≈ 1-nq, nq <1
Systems with Parallel Components:

• A parallel system is nothing but a system, all of whose components are

in parallel configuration.
• A System with m components is known as an m-unit parallel system if
and only if the successful functioning of any one of the components
leads to the system success.
• The components are connected across each other and there are m
parallel paths between the IN end the OUT end and the existence of
any one of them is sufficient to transmit information from the IN end
to the OUT end.
 1

.
.
.
m

Parallel System
The actual physical connectivity of the components in the system may
vary. But always remember the following two characteristics of a
parallel system:
(i) only one component needs to work successfully for the successful
operation of the system, and
(ii) All the components of the system must fail for system failure.

Let be the event that component i, performs its intended function

successfully, where i = 1, 2,3,…,n. Let denote the event complementary to
the event
Let denote the reliability of the component i, where i = 1,2,3,….,n for a
mission of t units of time.
If R denotes the Reliability of the parallel system for the mission of t
units of time, then by definition of reliability and parallel system,
the system fails only when all units fail. The system reliability is,
R = (U U…. U)
We take the union of events because the definition of reliability implies
that reliability is a probability and the definition of parallel system
implies that for the successful operation of the parallel system only one
components needs to work.
We Know that, P(A)+P(=1
Here, A = U U…. U
R = 1 - P( (By De Morgan’s Law)
R = 1 - P()
If events , are not independent, then on applying conditional probability we get,
R = 1 - P() PP
But if , are independent then are independent. Then we get,
R = 1 - P()
R = 1 – (1-P())
Since ), therefore
R = 1 – (1-)
or R = 1 -
Thus, the reliability of the parallel system is equal to one minus the
product of un reliabilities of the elementary components of the system.

In particular, if the parallel system has only two components, then

R = 1 – (1-) (1-)
And if the parallel system has three components, then
R = 1 – (1-) (1-)(1- )

Let us consider some examples for evaluating the reliability of parallel

systems.
Example:
Evaluate the reliability of the parallel system, which has the reliability block
diagram shown in the below diagram for a mission of 90 days. The reliabilities
of the components given in below diagram are for a mission of 90 days. Assume
that the failure of any component does not affect the functioning of the other
components. 0.60

0.75

INPUT OUTPUT
0.95

0.80
In this case, = 0.60, = 0.75, = 0.95 and = 0.80

We know that the reliability of a parallel system (R) having n components

with reliabilities , i = 1,2,3,….,n, is given by
R = 1 – (1-)

Therefore, R = 1 – (1-
R = 1 – (1-0.60)(1-0.75)(1-0.95)(1-0.80)
R = 1-(0.40 x 0.25 x 0.05 x 0.20) = 1- 0.001
R = 0.999
• In this example, the reliability of the system is very good compared to
the series system.
• The Reliability of the system is even better than the reliability of the
best component of the system.
• For the successful operation of the parallel system only one
components needs to work.
• So this improvement in the system reliability comes at the cot of an
initial investment of three additional components.
• Thus we have to balance the cost and reliability of the system.
• whether to give importance to the cost or to reliability depends upon the
severity of consequence of system failure.
• For example, there should be at least one additional parallel engine in the
aeroplane, whatever be the initial cost that the airline has to pay. So in this
case reliability is more important than cost.
• But in case of the component, say Television, the importance may be given to
cost instead of reliability because failure of a TV has little consequence
compared to crash of an aeroplane, which may result in loss of lives of
hundreds of people.
Let us see the effect of a component’s reliability on the reliability of the
parallel system with the help of the following example.
Example:
A System has three components connected in parallel from a reliability point
of view having reliabilities 0.20, 0.40, 0.50 respectively, for a mission of 400
hours. What is the percentage increase in the reliability of the system in each
of the following cases?
(i) Reliability of the first component is increased by 0.1 and that of the
second and third components remains the same.
(ii) Reliability of the second component is increased by 0.1 and that of the
first and third components remains the same.
(iii) Reliability of the third component is increased by 0.1 and that of the first
and second components remains the same.
Solution:
We first calculate the reliability of the system as per the given reliabilities of
the components.
R = 1 – (1- Where = 0.2, = 0.4, = 0.5
R = 1 – (1-0.20)(1-0.4)(1-0.5)
= 1 – (0.80 x 0.6 x 0.5)
= 1 – 0.240
R = 0.760
Now, reliabilities of the system and percentage increase in the reliability of
the system for the cases (i), (ii) and (iii) are shown in the column 5 and 6
respectively in the following Table.
 Reliability of the I Reliability of the Reliability of the Reliability of Percentage Increase in the
Component II Component III Component the System Reliability of the System
(1) (2) (3) (4) (5) (6)

0.2 0.4 0.5 0.760 -

(i) 0.3 0.4 0.5 0.790 x 100 = 3.95%

(ii) 0.2 0.5 0.5 0.800 x 100 = 5.26%

(iii) 0.2 0.4 0.6 0.808 x 100 = 6.32%

• From the column 6 it is noted that the improvement in the reliability of the
system (in percentage) is higher when reliability of the best component is
increased by 0.1 as compared to the cases when reliabilities of the other
two components are increased one at a time by the same amount (0.1).
• This suggests that if a system engineer wants to improve the reliability of a
parallel system, he / she should concentrate on the improvement of the
reliability of the best component.
• Thus, the best component dictates the reliability of the parallel system.
• Recall that the poorest component dictates the reliability of the series
system.
Effect of number of components on the reliability of a parallel system:
Example:
Consider a component having reliability 0.75 for a mission of 100 hours.
Evaluate the reliability of a parallel system when the system has 2,3, 4, …., 10
identical components for the same mission of 100 hours.
Solution:
If a system has a single component of reliability 0.75 for a mission of 100
hours., it will also have the same reliability for a mission of 100 hours.
But the reliability of the parallel system for a mission of 100 hours will vary in
the case having 2, 3, 4 … 10 identical components connected in parallel from
reliability point of view.
In fact, it increases with the number of components as the calculations given the
below problem:
Calculation of Reliability as Number of components varies from 1 to 10:
Number of Reliability of Parallel System % Increase in the Reliability of the
Components System Compared to the System
having Single Component
1 0.75 -
2 1-=0.9375 x 100 = 25%
3 1- = 0.9844 x 100 = 31.25%
4 1- = 0.9961 x 100 = 32.81%
5 1-=0.9990 x 100 = 33.20%
6 1-= 0.999756 x 100 = 33.30%
7 1-= 0.99993 x 100 = 33.33%
8 1- = 0.999985 x 100 = 33.33%
9 1- = 0.999996 x 100 = 33.33%
10 1-= 0.999999 x 100 = 33.33%
Thank You..!