Chapter         12

Reliability and Maintenance

Engineering

12.1 Introduction
Reliability is the probability with which an item or system will remain functioning under stated
operational and environmental conditions for a specified period of time. Quantitatively, reliability
is the probability of success. There are two types of reliability, predicted reliability and assessed
reliability, which are very likely to differ. The predicted reliability is the estimated quantity based
on the known reliabilities of individual components. The assessed reliability is arrived at by
conducting reliability tests. A failure is said to occur with a product when the same stops to
function. Failure rate is the probability that a new item will fail only after a duration of time t
but before t + δt and is designated by λ(t). When the expected duration of service, t is taken as a
parameter, the quantity R(t) is called quantitative reliability, and is related to the failure rate by
the following relation.
dR(t ) dt
λ (t ) = −
R(t )
where R(t ) = p (τ > t )
Here, τ is the service life of the product without failure and p is the probability of service
life greater than the given time t.
Integrating and substituting the boundary condition that at the beginning of time, t = 0, the
reliability is equal to 1, and further assuming that the failure rate is independent of time, we
obtain the following exponential relation for reliability.
R(t ) = e − λt

where λ is the failure rate.

Reliability engineering is concerned with four key elements of this definition:
1. Reliability is a probability. This means that failure is treated as a random phenomenon:
it is a recurring event and the likelihood of failures varies over time according to the
given probability function.
2. Reliability is predicted on ‘intended function’. It means an operation without failure.
3. Reliability applies to a specified period of time. In practical terms, this means that a
system has a specified chance that it will operate without failure before time t. Reliability
engineering ensures that components and materials will meet the requirements during
the specified time.
 Reliability and Maintenance Engineering  237

4. Reliability is restricted to operation under stated conditions. This constraint is necessary

because it is impossible to design a system for unlimited conditions. The operating
environment must be addressed during design and testing.
The objectives of reliability engineering, in order of priority to minimize costs and generating
reliable products, are:
1. To apply engineering knowledge and specialist techniques to prevent or to reduce the
likelihood or frequency of failures.
2. To identify and correct the causes of failures that do occur, despite the efforts to prevent
them.
3. To determine ways of coping with failures that do occur, if their causes have not been
corrected.
4. To apply methods for estimating the likely reliability of new designs, and for analysing
reliability data.

12.2 Reliability curves

The continuous distributions used in reliability are exponential, normal and Weibull. The
frequency distributions of these probability distributions are shown in Figure 12.1. These are
functions of time.
Reliability can be calculated as,
t
Rt = 1 − ∫ f (t )dt .
0
The reliability curves for various distributions such as exponential normal and Weibull
distribution, respectively, are shown in Figure 12.2.
Relationship Among λ(t), f(t), F(t) and R(t)
After defining the failure rate, probability density function (PDF), cumulative density function
(CDF) and reliability, it is important to establish a relationship among them to find the individual
values. The failure rate is denoted by, λ(t) at time t is expressed as :
f (t ) f (t )
λ (t ) = = ;
1 − F (t ) R(t )
d
λ (t ) = − (ln R(t ))  (12.1)
dt
1 0.5 1 Weibull β
Normal 1 ⎛t − θ⎛
2 f(t) = αβ tβ − 1 e−α t
Exponential 1 −
2 ⎝ σ ⎝
β = 0.5
f(t) = e
σ √ 2π β=1 γ =0
f(t) f(t) = (1/θ )e−t/θ f(t) f(t) β = 3.5 α=1

0 t 0 t 0 t

Figure 12-1: Frequency distribution as a function of time

238  Industrial Engineering and Management

1 1 1
αt β
Rt = e−
t
Rt − t Rt Rt = 1 − ∫ f(t)dt Rt β = 3.5
θ 0
Rt = e β= 1

β = 0.5
0 t 0 t 0 t

Figure 12-2: Reliability as a function of time

The PDF, f(t) at time t is expressed as:

d d
f (t ) = F (t ) = − R(t )
dt dt
t

∫
− λ ( t ) dt
f (t ) = λ (t )e 0
(12.2)

The CDF, F(t) for time t is expressed as:
t

F (t ) = ∫ f (t )dt; F (t ) = 1 − R(t )
0
t

∫
− λ ( t ) dt
F (t ) = 1 − e 0
(12.3)

The reliability, R(t) for time t is expressed as:
∞

R(t) = 1 − F(t); R(t ) = ∫ f (t )dt

t
t

∫
− λ ( t ) dt
R(t ) = e 0
(12.4)

Failure Rate Curve: Failure rate curves for exponential, normal and Weibull distributions,
respectively, are shown in Figure 12.3.
f (t )
Failure rate, λ =
R(t )
where λ is the failure rate and θ is the mean time to failure (MTTF) or mean life.
2
1 1 −
1 ⎛ t−θ ⎛ 1
2⎝ σ ⎝ β −1
e λ = σβ t
λ= 2
∞ −
1 ⎛ t−θ ⎛
dt β = 3.5
2⎝ σ ⎝
λ ∫e β = 0.5
t
λ = 1/θ λ λ
1/θ β=1

0 t 0 t 0 t

Figure 12-3: Failure rate as a function of time

 Reliability and Maintenance Engineering  239

12.3 Failure Pattern

Product breakdowns often tend to follow a Poisson distribution, i.e. failures occur infrequently
during the normal life of a product. Equipment manufacturers have found that the failure during
the very early and late stages of product life often differs from the experienced during the normal
operating life. This difference in failure rates is depicted in the form of the ‘bathtub’ curve as
shown in Figure 12.4. Early failures (may be due to improper assembly or damage in shipment)
may tend to follow a negative exponential pattern. During the typical operating lifetime, failure
occurs on a rare-event basis, often described by a Poisson distribution. As components wear out
and fail, the products may follow a pattern described by a normal distribution.
The bathtub curve represents the failure rate of a population of statistically identical items
and it is the same as the life cycle of the product involving the debugging phase, chance-failure
phase and wear-out phase. For a constant failure rate (CFR) of a product that is irreparable, the
MTTF is defined as:
∞
1
MTTF = ∫ R(t ) dt =
0
λ
If an item is reparable, then the same quantity is termed as mean time between failures (MTBF).
Operating time Total time − Non-operating time TT − NOT
MTBF = = =
Number of failures Number of failures F
The reliability analysis is conducted in two stages, first at the design stage and then at the
production stage.
Failure rate may be expressed in the following two ways:
1. Percentage failure (FR per cent).
2. Number of failures per given operation time (FRn).

Number Number of failures Number of failures F

FR% = FR%of= failures ; FR n = Number; FR nof= failures = F =
Number testedNumber tested Operating time Operating − NOT TT − NOT
TTtime
where FR%where FR% isfailure
is percentage percentage
rate; failure
FR n is rate ; FR nofisfailure
number number perofgiven
failure per given
operation operation time;
time;
F is n umber of failures; TT is total time; NOT is non-operating
F is number of failures; TT is total time; NOT is non-operatingg time. g time.
Failure rate

Initial Chance Wearout

phase failure failure

Time

Figure 12-4: Bath-tub failure curve

240  Industrial Engineering and Management

Example 12.1: Twenty machines have been operated for 100 hours. One machine fails in
60 hours and another fails in 70 hours. What is the MTBF and what will be reliability at 500
hours and 900 hours?

Solution:
Eighteen of the machines ran for 100 hours, while one machine ran for 60 hours and another one
ran for 70 hours. Thus, total running time is 18 × 100 + 60 + 70 = 1930 hours.
Operating time 1930
MTBF = = = 965 hours per failure
Number of failures 2
1 1
Failure rate ( λ ) = = = 0.0010362 failure/hour
MTBF 965
R500 = e − λt = e −0.0010362 × 500 = 0.595
R900 = e − λt = e −0.0010362 × 900 = 0.394

Mean time to failure: The MTTF is one of the measures which characterize the failure
distribution for a component or system. It is defined by
∞
MTTF = E (T ) = ∫ tf (t )dt (12.5)
0

which is the mean or expected value of probability distribution defined by f(t). It is also expressed
in terms of reliability function as
∞
MTTF = ∫ R(t )dt (12.6)
 0

Variance: The variance, σ 2 is another measure, which is used to characterize the failure
distribution. It is defined as
∞
σ 2 = ∫ (t − MTTF) 2 f (t )dt (12.7)
0 
The variance represents the average squared deviation from the MTTF. It is a measure of spread
or dispersion of the failure times about the MTTF. It can also be written as
∞
σ 2 = ∫ t 2 f (t )dt − ( MTTF) 2 (12.8)
0 
The square root of the variance is the standard deviation.

Example 12.2: The time-to-failure PDF for a system is f(t) = 0.02; 0 ≤ t ≤ 50 days
Find
(a) Reliability for time t, R(t)
(b) The hazard rate function at time t, λ(t)
(c) The MTTF
(d) The standard deviation
 Reliability and Maintenance Engineering  241

Solution:
(a) The relation between reliability function and PDF is
∞

R(t ) = ∫ f (t )dt
t
Using the given f(t) value
∞ 50
R(t ) = ∫ (0.02)dt = ∫ (0.02)dt = 1 − 0.02t
t t

Therefore, R(t ) = 1 − 0.02t

(b) The relation for λ(t) in terms of reliability function and PDF is
f (t )
λ (t ) =
R(t ) , using the given f(t) value and the R(t) value derived in part (a),
0.02
λ (t ) =
1 − 0.02t
∞
(c) Using the relation, MTTF = ∫ R(t )dt
0
50
50
⎡ t2 ⎤
MTTF = ∫ (1 − 0.02t )dt = ⎢t − 0.02 ⎥ = 25 days
0 ⎣ 2 ⎦0
∞
(d) Using the relation, variance, σ 2 = ∫ t 2 f (t )dt − ( MTTF) 2
0
50
50
⎡ t3 ⎤
σ = ∫ t (0.02)dt − ( 25) = ⎢0.02 ⎥ − 625 = 208.33
2 2 2

0 ⎣ 3 ⎦0
Therefore, standard deviation, σ = 14.4 days.

Example 12.3: A component has the following liner hazard rate, where t is in years: λ(t) =
0.05t, t ≥ 0
(a) Find R(t) and determine the probability of a component failing within the two months
of its operation.
(b) What is the design life if a reliability of 0.96 is desired?

Solution:
(a) The relation between reliability function and failure rate function is
t

∫
− λ ( t ) dt
R(t ) = e 0
, substituting given failure rate function
t

∫
− ( 0.05t ) dt
2
R(t ) = e 0
= e −0.025t
2
R(t ) = e −0.025t
242  Industrial Engineering and Management

The probability of a component failing within the two months (2/12 year) of its
operation,
⎛ 1⎞ ⎛ 1⎞ ⎛ 1⎞
Pr ⎜ T ≤ ⎟ = 1 − Pr ⎜ T > ⎟ = 1 − R ⎜ ⎟
⎝ 6⎠ ⎝ 6⎠ ⎝ 6⎠
2
⎛ 1⎞
−0.025 ⎜ ⎟
⎝ 6⎠
= 1− e = 0.00069

(b) Design life if a reliability of 0.96 is desired, i.e., R(t) = 0.96

2 2
From part (a), R(t ) = e −0.025t , therefore 0.96 = e −0.025t

−1
t= log(0.96) = 1.277 years
0.025

12.4 Basic reliability models

In this section, some important failure patterns based on the exponential, Weibull and normal
probability distributions models are discussed.

12.4.1 CFR Model

The exponential probability distribution has a constant failure rate (CFR) and is called a CFR
model. It is one of the most common and important distributions in reliability engineering.
Failures due to completely random or chance events will follow this distribution. The failure
behaviours of most of the electronic components are represented by the exponential distribution.
The expressions for R(t), F(t), f(t), MTTF and σ 2 are derived in the case of the CFR model as:
For CFR model, let us assume λ (t ) = λ , t ≥ 0, λ > 0, then using the relation in Equations (12.3)
and (12.4), the expressions of R(t) and F(t) are derived and given in Equations (12.9) and (12.10).
t

∫
− λ ( t ) dt
R(t ) = e 0
= e − λt  (12.9)

and F (t ) = 1 − R(t ) = 1 − e − λt (12.10)



R(t) and F(t) curves are shown in Figures 12.5 and 12.6, respectively, for three different values
of λ in case of CFR model.
Using Equation (12.2), the expression for f(t) is derived as:
dR(t ) d(e − λt )
f (t ) = − =− = λ e − λt (12.11)
dt dt 
where λ is the distribution parameter.
Curves for f(t) are shown in Figure 12.7 for different values of λ.
 Reliability and Maintenance Engineering  243

Lambda = 4.0
1
Lambda = 1.0

1 .4
=0
da
Lam F(t) Lamb
R(t) bda
= 0.
4
Lamb
da = 1
Lambd .0
a=4

t t

Figure 12-5: Reliability, R(t) Figure 12-6: Cumulative distribution function, F(t)

Lambda = 1.0
Lambda = 3.0
Lambda = 5.0

f(t)

Figure 12-7: Probability density function

Using Equation (12.5), the expression for MTTF is derived as

∞ ∞
1
MTTF = ∫ R(t )dt = ∫ e − λt dt = (12.12)
λ
0 0 
Using Equation (12.8), the expression for variance is derived as
∞ 2
⎛1⎞ 1
σ 2 = ∫ t 2 λ e − λt dt − ⎜ ⎟ = 2 (12.13)
0 ⎝λ⎠ λ 
An important property of the CFR model is memorylessness. It means the time to failure of a
component is not dependent on how long the component has been operating. There is no ageing
or wear-out effect. The probability that the component will operate for the next 1000 hours is
the same regardless of whether the component is brand new and has been operated for so many
hours. This property is consistent with the completely random, and independent nature of the
failure process. For example, when external, random environmental stresses are the primary
cause of failures, the failure or operating history of the component will not be relevant.
244  Industrial Engineering and Management

Example 12.4: A component experiences chance (CFR) failures with an MTTF of 1000 hrs.
Find the following:
(a) The reliability for a 100 hr mission
(b) The design life for a 0.96 reliability
(c) The median time to failure

Solution:
(a) For CFR model,
1 1 − λt
Failure rate, λ = = per hour and R(t ) = e
MTTF 1000
⎛ 1 ⎞
−⎜ ⎟100
Therefore, R(100) = e ⎝ 1000 ⎠
= 0.904

(b) R(t ) = e − λt = 0.96, therefore

1
Design life, t = − log(0.96)
λ
1
t=− log(0.96) = 40.82 hours
⎛ 1 ⎞
⎜⎝ ⎟
1000 ⎠
(c) For the median life to failure, tmed
R(t med ) = e − λtd = 0.50
1
median life to failure, t med = − log(0.5)
λ
1
t med =− log(0.5) = 693.14 hours
⎛ 1 ⎞
⎜⎝ ⎟
1000 ⎠

12.4.2 Weibull Model

There are many situations in which the failure rate cannot be constant as in the case of CFR
model. In such cases, the Weibull model is more versatile and appropriate which can incorporate
the decreasing and increasing failure rate. The mechanical components, which fail due to the
gradual wear and tear, are modelled by Webull distribution.
The expressions for R(t), F(t), f(t), MTTF and σ 2 are derived in the case of the Weibull
failure model as:
For Weibull model, the failure rate λ (t ) is given by the relation
β −1
β⎛t⎞
λ (t ) = ⎜ ⎟ , θ > 0, β > 0 , t ≥ 0 (12.14)
θ ⎝θ ⎠ 
 Reliability and Maintenance Engineering  245

where β is the shape parameter and θ is the scale parameter. The values of the shape parameter,
β are < 1, =1 and >1 for decreasing, constant and increasing failure rates, respectively.
Using Equation (12.4), we have
t t β −1
β⎛t ⎞ β
∫
− λ ( t ) dt − ∫ θ ⎜⎝ θ ⎟⎠ dt ⎛t⎞
−⎜ ⎟
⎝θ ⎠
R(t ) = e 0
=e 0
=e (12.15)

The variation of R(t) with time t for different values of β and θ is shown in Figures 12.8 and 12.9,
respectively.

1 Beta = 0.5
Beta = 1.5
Beta = 3.0 1 Theta = 0.5
Theta = 2.0
R(t)

R(t)

t t

Figure 12-8: R(t) for different values of β Figure 12-9: R(t) for different values of θ

Similarly, using Equation (12.3), the expression for CDF is derived as:
β
⎛t⎞
−⎜ ⎟
⎝θ ⎠
F (t ) = 1 − R(t ) = 1 − e (12.16)

The variation of F(t) with time t for different values of β and θ is shown in Figures 12.10 and
12.11, respectively.

1 Theta = 0.5
Theta = 2.0

Beta = 0.5 F(t)

F(t) Beta = 1.5
Beta = 3.0

t t

Figure 12-10: F(t) for different values of β Figure 12-11: F(t) for different values of θ
246  Industrial Engineering and Management

Using Equation (12.2), the expression for f(t) is derived as:

β
⎛t⎞
−⎜ ⎟ β −1
β
⎛t⎞
dR(t ) d(e ⎝θ ⎠
) β ⎛t⎞ −⎜ ⎟
⎝θ ⎠
f (t ) = − =− = ⎜θ ⎟ (12.17)
e
dt dt  θ ⎝ ⎠
The variation of f(t) with time t for different values of β and θ is shown in Figures 12.12 and
12.13, respectively.

Beta = 0.5 Theta = 0.5

Beta = 1.5 Theta = 2.0
Beta = 3.0
f(t)
f(t)

t t

Figure 12-12: f(t) for different values of β     Figure 12-13: f(t) for different values of θ

Using Equation (12.6), we have

∞ ∞ ⎛t⎞
−⎜ ⎟ β ⎛ 1⎞
MTTF = ∫ R(t )dt = ∫ e ⎝θ ⎠
dt = θ Γ ⎜1 + ⎟ (12.18)
0 0 ⎝ β⎠
∞
Using Equation (12.8), σ 2 = ∫ t 2 f (t )dt − ( MTTF) 2
0
⎡ ⎛ t ⎞β ⎤ 2
β −1
∞
⎛ β ⎞⎛ t ⎞
⎢ − ⎜ ⎟ ⎥ dt
⎢⎣ ⎝ θ ⎠ ⎥⎦ ⎡ ⎛ 1 ⎞⎤
σ 2 = ∫t2 ⎜ ⎟⎜ ⎟ e − ⎢θ Γ ⎜1 + ⎟ ⎥
0 ⎝ θ ⎠⎝θ ⎠ ⎣ ⎝ β ⎠⎦
⎧⎪ ⎛ 2⎞ ⎛ ⎛ 1 ⎞ ⎞ ⎫⎪
2

= θ ⎨Γ ⎜ 1 + ⎟ − ⎜ Γ ⎜ 1 + ⎟ ⎟ ⎬
2
(12.19)
⎪⎩ ⎝ β ⎠ ⎝ ⎝ β ⎠ ⎠ ⎪⎭

The value of gamma function can be taken from the table provided in the ‘Appendix’ 2 of the
book.

Example 12.5: For a system having a Weibull failure distribution with a shape parameter of 1.5
and a scale parameter of 500 days, find the following:
(a) R (200 days)
(b) MTTF
(c) median time to failure, tmed
(d) The design life for a reliability of 0.96
(e) The standard deviation
 Reliability and Maintenance Engineering  247

Solution:
(a) For Weibull failure model,
β
⎛t⎞
−⎜ ⎟
⎝θ⎠
R(t ) = e , where β = 1.5, θ = 500 days
Therefore, for t = 200 days
1.5
⎛ 200 ⎞
−⎜
⎝ 500 ⎟⎠
R( 200) = e = 0.7765

(b) For Weibull model (having t0 = 0), the expression for MTTF is
⎛ 1⎞
MTTF = θ Γ ⎜1 + ⎟ , substituting the parameters values
⎝ β ⎠
⎛ 1⎞
MTTF = 500 Γ ⎜1 + ⎟ = 500 × 0.90167 = 450.835 days
⎝ 1.5 ⎠

(c) For median life to failure, tmed

β
⎛t ⎞
−⎜ d ⎟
⎝θ ⎠
R(t med ) = e = 0.50
1

Median life to failure, t med = θ ( − log(0.5)) β

1
t med = 500( − log(0.5))1.5 = 301.61 days
⎛t⎞
−⎜ ⎟ β
⎝θ⎠
(d) R(t ) = e = 0.96, therefore 1

Design life, t = θ ( − log(0.96)) β

1
t = 500( − log(0.96))1.5 = 59.28 days
(e) For Weibull model, the expression for variance is
⎧ 1 ⎞ ⎞ ⎫⎪
2
2 ⎪ ⎛ 2⎞ ⎛ ⎛
Variance = σ = θ ⎨Γ ⎜1 + ⎟ − ⎜ Γ ⎜1 + ⎟ ⎟ ⎬
2

β ⎠ ⎝ ⎝ β ⎠⎠ ⎪
⎩⎪ ⎝ ⎭
Substituting the parameter values,
⎧⎪ ⎛ 2⎞ ⎛ ⎛ 1 ⎞ ⎞ ⎫⎪
2

σ 2 = 500 2 ⎨Γ ⎜1 + ⎟ − ⎜ Γ ⎜1 + ⎟ ⎟ ⎬
⎝ 1.5 ⎠ ⎝ ⎝ 1.5 ⎠ ⎠ ⎪
⎩⎪ ⎭
{
= 500 2 × 1.18819 − 0.901672 = 93, 795.30 }
Therefore, standard deviation, σ = 306.26 days

12.4.3 Normal Model

The normal distribution is most widely used distribution for modelling the stochastic behaviour
of systems. In reliability practices, it is used to model fatigued and weared out phenomenon of
the components. The PDF, f(t) is a well-known bell-shaped curve as shown in Figure 12.14 and
it is expressed as:
248  Industrial Engineering and Management

Sigma = 0.5
Sigma = 1.0

f(t)

Figure 12-14: Probability density function, f(t) for different values

1 ( t − μ )2
1 −
f (t ) =e , where − ∞ < t < ∞ 
2 σ2
(12.20)
2πσ
where parameters µ and σ are the mean and standard deviation of the distribution, respectively.
∞ 1 ( t − μ )2
1 − dt
R(t ) = ∫ e 2 σ2
(12.21)
t 2πσ 
The closed-form solution of this integral cannot be obtained. It is calculated by numerical
solution. First, the PDF, f(t) expression is transformed into standard normal variable, z, function
T −μ 1 − z2 / 2
using transformation, z = . After transformation, f(t) reduces to ϕ ( z ) as ϕ ( z ) = e .
σ 2π
Then, the cumulative distribution function, F(t) is derived as
⎛T −μ t −μ ⎞
z
F (t ) = Pr (T ≤ t ) = Pr ⎜ ≤ = ψ ( z ) = ∫ ϕ ( z )dz (12.22)
⎝ σ σ ⎟⎠ −∞ 
This can be solved with the help of standard normal distribution tables.
Further, the reliability expression reduces to
⎛T −μ t −μ ⎞
z
R(t ) = Pr (T > t ) = Pr ⎜ > = 1 −ψ ( z ) = 1 − ∫ ϕ ( z )dz (12.23)
⎝ σ σ ⎟⎠ −∞ 
The variation of R(t) and F(t) with time t for different values of σ is shown in Figures 12.15 and
12.16, respectively.

12.5 Evaluation of System Reliability (Rs)

Most products are made up of a number of components. The reliability of each component and
the configuration of the systems consisting of these components determine the system reliability
or in other words the reliability of the product. Although the product design, manufacture and
 Reliability and Maintenance Engineering  249

1
1
Sigma = 0.5
Sigma = 1.0

Sigma = 0.5
F(t) Sigma = 1.0 R(t)

t t

Figure 12-15: Reliability, R(t) for different Figure 12-16: Cumulative distribution function, F(t) for
values of σ different values of σ

maintenance influence reliability, improving reliability is largely the domain of design. One
common approach for increasing the reliability of the system is through the redundancy in
design, which is usually achieved by placing components in parallel. As long as one component
operates, the system will work.

12.5.1 System Components in Series

When the components of a system are in series as shown in Figure 12.17, for working of the
system, each component must operate. Assuming that the components operate independently of
each other and there are n components in series with the reliability of the ith component being
denoted as Ri, the system reliability is given by,
n
Rs = R1 × R2 × R3 × .......... × Ri × .......... × Rn = ∏ Ri
i =1
For higher the value of n, reliability of the system will be lower. Even if each individual component
is over designed with higher reliability, the effect is nullified by the number of components in
series. Since manufacturing limitations and resource limitations restrict the maximum reliability
of any given component, any effort to redesign those results in lesser number of components in
the series is a viable alternative.
When the exponential model is used, the system reliability in terms of the failure rates of
individual components is given by

1 2 3

Figure 12-17: Components in series connection

250  Industrial Engineering and Management

⎛ ⎞
⎜ n ⎟
− ⎜⎜ ∑ λi ⎟⎟ t
−λ t −λ t −λ t −λ t −λ t
Rs = e 1 × e 2 × e 3 × .......... × e i × .......... × e n = e ⎝ i = 1 ⎠ 
⎜ ⎟
(12.24)

Thus, the failure rate of the system (equivalent failure rate) with an exponential model for
reliability (chance-failure phase of bathtub curve) is
n
λs = ∑ λ
i
i =1
If each component that fails is replaced immediately by another component that has the same
failure rate, the MTTF of the system is given by
1
MTTF =
n
∑ λ
i
i =1
If all components in the series have an identical failure rate, say λ, the MTTF of the system will
be a special case of the above.
1
MTTF =
nλ

12.5.2 System Components in Parallel

The components in parallel (as shown in Figure 12.18) are said to be redundant. The system
operates as long as at least one of the components operates and the system fails only if all
the components fail. Redundant components improve the system reliability and the braking
mechanism with dual subsystems in automobiles is an example.
The equations for calculating the system reliability, failure rate and MTTF for a system with
components in parallel are as follows.
For exponential failure model,
n ⎛ −λ t ⎞
Rs = 1 − ∏ ⎜1 − e i ⎟
⎜ ⎟
i = 1⎝ ⎠

Figure 12-18: Components in parallel connection

 Reliability and Maintenance Engineering  251

Rs = 1 − {(1 − R1 ) × (1 − R2 ) × (1 − R3 ) × .......... × (1 − Ri ) × .......... × (1 − Rn )}

n
= 1 − ∏ (1 − Ri ) (12.25)
i =1

In the special case of all components having the same failure rate,

( )
n
RS = 1 − 1 − e −λ t

and the MTTF is

1⎛ 1 1 1⎞
MTTF = 1 + + + ....... + ⎟
λ ⎜⎝ 2 3 n⎠

Example 12.6: Which system, (i) or (ii), has the highest reliability at the end of 90 operations
hours?
(i) Two CFR components in parallel, each having an MTTF of 900 hrs.
(ii) A Weibull component with a shape parameter 2 and a characteristic life of 10,000 hrs in
series with a CFR component with a failure rate of 0.00005 per hour.

Solution:
(i) Reliability for a system with two CFR components in parallel, each having an MTTF of
900 hrs.
Component reliability in CFR model is given by R(t ) = e − λt , where
1
Failure rate, λ = = 1 per hour
MTTF 900
⎛ 1 ⎞
−⎜ ⎟ 90
Therefore, R(100) = e − λt = e ⎝ 900 ⎠ = 0.9048.
System reliability in parallel configuration
R(t ) = 1 − (1 − R1 (t ))(1 − R2 (t )),
Since both components are identical, R1(90) = R2(90) = 0.9048
R(90) = 1 − (1 − 0.9048)(1 − 0.9048) = 0.9909
(ii) Reliability for a system with a Weibull component with a shape parameter 2 and a
characteristic life of 10,000 hrs in the series with a CFR component with a failure rate
of 0.00005. β
t ⎛ ⎞
−⎜ ⎟
⎝θ⎠
Weibull component reliability, R1 (t ) = e , where β = 2, θ = 10,000 hrs
2
⎛ 90 ⎞
−⎜
⎝ 10000 ⎟⎠
R1 (90) = e = 0.9999
252  Industrial Engineering and Management

Exponential component reliability, R2 (t ) = e − λt, where λ = 0.00005 per hour

R2 (90) = e − ( 0.00005)( 90 ) = 0.9955
System reliability in series configuration is given by R(t ) = R1 (t ) R2 (t ),
Therefore,
R(90) = 0.9999 × 0.9955 = 0.9954
So, system reliability is higher in system (ii).

12.5.3 System Components in Series and Parallel Mixed

A system may be called complex system, if some components are in series and others are in
parallel (as shown in Figure 12.19). However, the calculation of system reliability of complex
systems is complicated because by using a reduction process where each sub-system that is in
parallel is represented by the equivalent series element, the system reliability can be finally found
by the equations pertaining to the system with components in series.

2
1 4
3

Figure 12-19: Mixed connection (series and parallel)

Rs = R1 × {1 − (1 − R2 ) (1 − R3 )} × R4

Example 12.7: In a system, 10 components, each of reliability factor 0.95, are in series. What
is the overall reliability of the system? If these components are arranged in parallel connection
and have reliability factor of 0.3 each, find the reliability of the systems.

Solution:
In series connection:
n
Rs = R1 × R2 × R3 × .......... × Ri × .......... × Rn = ∏ Ri = (0.95) = 59.87%
10

i =1
In parallel connection:
{ }
Rs = 1 − (1 − R1 ) × (1 − R2 ) × (1 − R3 ) × .......... × (1 − Ri ) × .......... × (1 − Rn )
n
= 1 − ∏ (1 − Ri ) = 1 − (1 − 0.3)10 = 97.17%
i =1
Example 12.8: In a system, there are four components in parallel followed by three components
in series. The components in parallel have a reliability of 0.7 each and those in series have a
reliability of 0.8 each. Determine the reliability of the entire system.
 Reliability and Maintenance Engineering  253

Solution:
The overall reliability of the components in parallel connection is

RI = 1 − (1 − Rp ) = 1 − (1 − 0.7 ) = 0.99
4 4

The overall reliability of the components in series connection is

RII = Rs 3 = ( 0.8 ) = 0.512
3

RII = Rs 3 = ( 0.8 ) = 0.512

3

Reliability of the system, RS = RI × RII = 0.9919 × 0.512 = 0.5078 = 50.78

Reliability of the system, RS = RI × RII = 0.9919 × 0.512 = 0.5078 = 50.78%

Example 12.9: The length of time that a particular piece of equipment operates before failure
is a random variable with cumulative distribution function.
2.6
F (t ) = 1 − e −0.043.t
What is the probability that the equipment operates for more than 5 years without experiencing
failure?

Solution: The probability that the equipment operates for more than t years without experiencing
failure.
R = 1 − Probability that it will fall within period t

(
= 1 − F (t ) = 1 − 1 − e −0.043.t
2.6

)=e −0.043.t 2.6

Rs = e −0.043.(5) = 0.0593 = 5.993%

2.6

Example 12.10: The MTTF of engine, gearbox and tyre system of a vehicle are 280, 3200 and
140 hours, respectively. Determine the reliability of the vehicle at 30 hours of use. Determine
the time at which the reliability is 90 per cent.

1
Solution: We know that MTTF = and R = e − λt
λ
MTTFEngine = 280 hours, MTTFGearbox = 3200 hours, and MTTFTyresystem = 140 hours
Case I:
Rs = R1 × R2 × R3 = e − 30 280 × e − 30 3200 × e − 30 140 = 0.7183 = 71.83%
Case II:
Rs = e − t 280 × e − t 3200 × e − t 140 = e − t 0.0110 = 90 ⇒ t = 9.578 hours

Example 12.11: (a) For the network shown in Figure 12.20, deduce an expression for the
system reliability in terms of the component reliabilities. Assume that each component has a
reliability of R.
254  Industrial Engineering and Management

(b) Compute the system reliability if R = 0.98

2
1 R 4
R R
R
3 R

8
R R R

5 6 7

Figure 12-20: Network of components

Solution:
Given, R1 = R2 = R3 = R4 = R5 = R6 = R7 = R8 = R = 0.98
Components, R2 and R3, are in parallel. Let us replace them by an equivalent subsystem, then the
reliability of subsystem, R23, is
R23 = 1 − (1 − R2 ) (1 − R3 )
R23 = 1 − (1 − R ) = 2 R − R 2
2
Therefore,
Component, R1, subsystem, R23 and component, R4 are in series. Let us replace them by an
equivalent subsystem, R1234, then the reliability of subsystem, R1234, is
R1234 = R1 × R23 × R4 ,
Therefore, ( )
R1234 = R × 2R − R 2 × R = 2R3 − R 4
Components, R5, R6 and R7 are in series. Let us replace them by an equivalent subsystem, then
the reliability of the subsystem, R567, is
R567 = R5 × R6 × R7 ,
Therefore, R567 = R × R × R = R3
Subsystems, R1234 and subsystem, R567 are in parallel. Let us replace them by another equivalent
subsystem, R17, then the reliability of the subsystem, R17, is
R17 = 1 − (1 − R1234 ) (1 − R567 ) ,
Therefore, { (
R17 = 1 − 1 − 2R3 − R 4 )} {1 − R } 3
= 3R3 − R 4 − 2R6 + R7

Now, subsystems, R17 and component, R8, are in series. Let us replace them by another equivalent
subsystem, R18, then reliability of subsystem, R18, is
R18 = R17 × R8,
Therefore, R18 = 3R3 − R 4 − 2R6 + R 7 × R = 3R 4 − R5 − 2R 7 + R8 .
(b) Putting R = 0.98
R18 = 0.977
 Reliability and Maintenance Engineering  255

12.6 Improvement in Reliability of a System

The ways used to improve the reliability of a product or system are as follows.
1. Improve the design of components.
2. Simplify the design of the system.
3. Improve production techniques.
4. Improve quality control.
5. Test components and the system.
6. Install parallel systems.
7. Perform periodic preventive maintenance.

12.6.1 Redundancy
Redundancy is the duplication of critical components or functions of a system with the intention
of increasing reliability of the system. Redundancy does not just imply a duplication of hardware,
since it can also be implemented at the software level. However, to avoid common mode of
failures, redundant elements should be realized independently from each other. From an operating
point of view, redundancies are classified as active redundancy, warm and passive redundancies.
From system configuration point of view, the redundancies are classified as high-level and­
low-level redundancy. These redundancy types are elaborated in the following subsections.
Active redundancy: It is also known as parallel or hot redundancy. In this case, redundant
components are subjected from the beginning of the same load as operation elements. If any one
element fails, then remaining redundant components share the load of failed components. Since
all the components are active, the failure rate in the reserve state is the same as in the operation
state.
Warm redundancy: It is also known as lightly loaded redundancy. Redundant components are
subjected to a lower load until one of the operating components fails. Load sharing is possible in
this case. The failure rate in the reserve state is lower than in the operating state.
Passive redundancy: It is also known as standby or cold or unloaded redundancy. Redundant
components are subjected to no load until one of the operating components fails. In this case, no
load sharing is possible. The failure rate in the reserve state is assumed to be zero.
High-level redundancy: In high-level redundancy, parallel–series circuit is used. In this circuit,
the system is counted as success if any one of the parallel circuit arms is functioning, which in
turn is possible only if all the components of that particular arm are operating. It is a case of
redundancy at the subsystem level.
A typical example of a high-level redundancy is shown in Figure 12.21. This arrangement
can be converted into the simple parallel form by first evaluating the series reliability of each
path and then treating them as two components in parallel.
Low-level redundancy: In low-level redundancy, series–parallel circuit is used. As shown in
Figure 12.22, Unit A has two parallel redundant elements, Unit B has two parallel redundant
256  Industrial Engineering and Management

A B A B

A B A B

Figure 12-21: RBD for high-level redundancy Figure 12-22: RBD of low-level redundancy

elements. For Unit A to be successful, 1 or 2 must operate, for Unit B success, 3, 4 must operate
for system success. It is a case of redundancy at component level.
It can be proved that low-level redundancy is more reliable than the high-level redundancy.

12.7 Design guidelines for Reliability

1. Carry design–redesign iteration with the help of manufacturing and assembly engineers
until a final design that is most acceptable and efficient evolves. Once finalized, follow
the design in the subsequent steps of product life cycle.
2. Formulate detailed specifications as exactly as possible.
3. Give priority to just in time and lean manufacturing with minimum waiting times and
inventory.
4. Carry out the process planning activity, based on the final design, if possible, using an
efficient CAPP system, keeping the machine facilities in close reference.
5. Taking into account, while deciding the machining/forming parameters in process plan,
all the environmental and current status of machine tools instead of those quoted in
their catalogues. Catalogues give the specification of a new and unused machine tool,
whereas the machine tools in the shop floor may have depreciated.
6. Plan for on-line inspection methods and Non-Distructive Testing (NDT), wherever
appropriate and feasible.
7. Provide for condition monitoring so that the present data can help in the manufacture of
future products.
8. Ensure that the design data is clearly understandable and is widely displayed on the
shop floor so as to be visible to the operators.
9. Plan for future corrections based on the on-line data collected during the present
activities conducted on the product during the entire product life cycle.

12.8 Reliability Testing

The purpose of reliability testing is to discover potential problems with the design as early as
possible and, ultimately, provide confidence that the system meets its reliability requirements.
Reliability testing leads to reduced unscheduled downtime, increased production opportunity/
profitability and savings in repair cost and part inventory. It also facilitates the transition from
‘run to failure’ to pro-active maintenance. Some of the commonly used reliability testing
techniques are vibration analysis, balancing and alignment services, ultrasonic testing infrared
thermography. Reliability testing may be performed at several levels. Complex systems may be
tested at component, unit, assembly, subsystem and system levels.
 Reliability and Maintenance Engineering  257

12.8.1 Component Stress Testing

Stress testing is about simulating large workloads to see how the component performs under
peak usage conditions. With component stress testing, you isolate the constituent components
and services, figure out what functional and interface methods they expose. The idea here is to
stress each component in isolation far beyond what the normal application would experience.

12.8.2 Integration Stress Testing

After stressing each individual component, the entire system with all of its components and
supporting services is subjected to stress. Integration stress testing is largely concerned about
interactions with other services, processes and data structures from both internal components and
other external application services. Integration testing starts with just basic functional testing.
It is required to know the coded pathways and user scenarios; understand what the users are
trying to do, and identify all the ways the user goes through the application.

12.8.3 Accelerated Reliability Testing

Accelerated life tests are conducted on materials, components and manufacturing processes to
determine their useful life in the required product application. Their purpose is not to expose
defects, but to identify and quantify the failures and failure mechanisms which cause products
to wear out at the end of their useful life. Because of this, accelerated life tests must last long
enough to cause the samples under test to fail by wear out. The test time may typically vary from
a few weeks to a few months.
Accelerated life testing employs a variety of high stress test methods to shorten the life of a
product or quicken the degradation of the product’s performance. The goal is to efficiently obtain
performance data that, when properly analysed, yields reasonable estimates of the product’s life
or performance under normal operating conditions.

12.9 Maintainability
Maintenance management is the direction and organization of resources in order to control the
availability and performance of industrial plant to some specified level. Maintenance plays a
supporting role to keep equipments and machines operating effectively to carry out the required
production processes and to maintain quality, quantity and cost standards. Some important
definitions of maintenance are given below as:
Maintenance is that function of maintenance management that is concerned with the
day to day problem of keeping the physical plant in good operating condition.
 —Harold T Amrine, John A Ritchey
Maintenance relates to profitability through equipment output and equipment running
cost. Maintenance work rises the level of equipment performance and availability, but
at the same time it adds to running costs. The objective of an industrial maintenance
department should be the achievement of the optimum balance between these effects.
 —A Kelley, Harri S.M.J.
258  Industrial Engineering and Management

The ability of an item, under stated conditions of use, to be retained in, or restored to a state
in which it can perform its required function(s), when maintenance is performed under stated
conditions and using prescribed procedures and resources. Maintainability cab be expressed in
terms of ‘Mean Time To Repair’ (MTTR).
Maintainability criteria have two aspects.
1. The ability of an equipment to meet operational objective with a minimum expenditure
of maintenance effort under operational environmental conditions in which scheduled
and unscheduled maintenance is performed.
2. The probability that an item will be restored to specified conditions within a given
period of time when a maintenance action is performed in accordance with prescribed
procedure and resources.
It is the probability that a failed component or system will be restored or repaired to a specified
condition within a period of time when maintenance is performed in performed in accordance
with prescribed procedures.
If T is the random variable representing the repair time and t is the time in which the system
is repaired, then maintainability is defined as
M ( t ) = Pr (T ≤ t ) (12.26)

Let the repair time be exponentially distributed with parameter µ, then the repair-density
function is
g (t ) = μ e − μ t  (12.27)
Therefore,
t
M (t ) = Pr(T ≤ t ) = ∫ μ e − μt dt = 1 − e − μt (12.28)
0 
The expected value of repair time is called the mean time to repair (MTTR) and is given by
∞
MTTR = ∫ tg (t )dt (12.29)
0 
For an exponentially distributed repair time, this expression is simplified as
∞
1
MTTR = ∫ t μ e − μt dt = (12.30)
μ
0 

12.9.1 Factors Contributing to Maintainability at Design Stage

Some of the factors which contribute to maintainability of equipment are as follows:
1. Accessibility of assemblies and components
2. Standardization
3. Monitoring facilities
4. Procedures
 Reliability and Maintenance Engineering  259

5. Identification
6. Safety
7. Availability of literature
8. Training
One of the unforeseen benefits of design for maintainability is that an easily maintained design
will, more than likely, be easy to assemble and manufacture. The manufacturing engineer and
maintainability engineer on a concurrent engineering team have common issues, such as designs
that use common fasteners. This results in reduced amounts of different tool requirements for
maintenance or assembly and speeds both the assembly and disassembly processes.
An important element of maintainability is testability. Testability is a design characteristic
which allows the status of an item to be determined and the isolation of faults within the item to
be performed in a timely manner. The benefits include lowest programming cost, lowest testing
cost, longest test equipment life and highest product confidence.

12.9.2 Benefits of Maintainability

The major benefits of maintainability of a system are expressed in terms of manufacturing,
quality and support benefits as discussed below:

Manufacturing Benefits
1. Fewer false starts
2. Reduced labour costs
3. Reduced scrap costs
4. Faster production
5. Higher product quality

Quality Benefits
1. Fewer line stoppages
2. Lower inspection costs
3. Less MRB (Material Review Board) activity
4. Higher product quality

Support Benefits
1. Fewer maintenance actions
2. Less expensive maintenance actions
3. Shorter downtimes
4. Lower inventory costs
5. Lower support equipment accosts
6. Lower training costs
260  Industrial Engineering and Management

12.9.3 Human Factor Aspect in Design for Maintainability

The methodology of human factors engineering can be applied to systems design to minimize the time
and effort required to perform periodic preventive maintenance as well as unscheduled maintenance.
Field observation tools may be developed and field study may be conducted to ascertain the level of
effort required to maintain existing systems and to identify opportunities for system improvement.
The benefits of such field study aimed at design for maintainability include:
1. Greater task efficiency and reduced MTTR
2. Improved system performance and reliability
3. Increased process tool utilization
4. Fewer operational errors
5. Decreased stress in maintenance technicians
6. Reduced fatigue and incidence of cumulative trauma
7. Reduced training time and costs
8. Less reliance on maintenance manuals
9. Reduced installation and maintenance costs
10. Decreased maintenance staffing

12.10 Design for Maintainability

‘Maintenance is an activity to keep the resources in working condition or restore them to
operating status.’ Maintenance does not mean complete prevention of failure. Instead, it tries
to prevent failure and if failure occurs it restores the system very quickly and economically. To
improve the productivity and proper maintenance of tools, equipments and other resources are
required. There are a number of maintenance philosophies that are used for successful operation.
Predictive maintenance: It is the process of using data and statistical tools to determine when a
piece of equipment will fail and that piece of equipment or tools are changed before occurrence
of failure.
Preventive maintenance: It is the process of periodically performing activities such as lubrication
on the equipment to keep it running. It is the routine inspection and service activities designed to
detect potential failure conditions and make minor adjustments or repairs that will help prevent
major operating problems.
Breakdown maintenance: It is the repairs, often of an emergency nature and a cost premium of
facilities and equipment that have been used until they fail to operate. Breakdown maintenance
is likely to be sick leave, hospitalization or other healing period during which the employee is
incapable of performing at a normal level.
Total productive maintenance: It is based on philosophy of total quality management. Total
productive maintenance (TPM) is keeping the plant and equipment at its highest productive level
through co-operation of all areas of organization. The objective of TPM is to create a system in
which all maintenance activities can be planned and not interfere with the production process.
Surprise equipment breakdowns should not occur.
 Reliability and Maintenance Engineering  261

The overall goals of TPM are given below as:

1. Maintaining and improving equipment capacity.
2. Maintaining equipment for increasing the life.
3. Using support from all areas of operation.
4. Encouraging input from all employees.
5. Using teams for continuous improvement.

12.11 Maintenance Costs

Equipment breakdowns, idle workers and equipments, resulting in lost production time, delayed
schedules and expensive emergency repairs. These downtime costs usually exceed the preventive
maintenance costs for inspections, service and scheduled repairs up to an optimal point. Beyond
this optimal point, an increasingly higher level of preventive maintenance is not economically
justified, and the firm would be better off waiting for breakdowns to occur. Whereas the optimal
level of maintenance activity (M) is easily identified on a theoretical basis, in practice this
necessitates knowing a good deal about the various costs associated with both the preventive and
breakdown maintenance activities. This includes the knowledge of probability of breakdowns
and the amount of repair time required. Although these data are not always easily obtained, good
maintenance records will provide substantial help in estimating the probability distributions of
breakdown and repair times.

12.12 Availability
It is the probability that a system is available for use at a given time. It is a function of reliability
and maintainability. It is the operating time divided by total time, which is the available time per
day minus the planned downtime.
MTBF ∑ (T × F )
Availability = ; MTTR = .
MTBF + MTTR ∑F
where T = Repair time, F = Failure rate

Example 12.12: The following data was collected for an automobile:

MTBF = 500 hrs; Mean waiting for spares = 5 hrs
Mean time for repairs = 48 hrs; Mean administrative time = 2 hrs
Compute the availability of the automobile.

Solution:
Total mean down time = 5 + 48 + 2 = 55 hrs.
Using relation,
uptime
Availability =
uptime + downtime
262  Industrial Engineering and Management

500
Availability = = 0.90
500 + 55
The automobile would be available 90 per cent of the time.
There are few other related availability measures which are defined as:
1. Point availability, A(t) at time t is defined as:
2. Average availability over the interval [0,T] is defined as:
T
A(T ) = (1 / T ) ∫ A(t )dt
0

3. Mission or interval availability is defined as:

t
⎛ 1 ⎞2
At2 −t1 = ⎜ ⎟ ∫ A(t )dt
⎝ t 2 − t1 ⎠ t1
4. Steady-state availability is defined as:
A = lim A(t )
T →∞

Example 12.13: A plant runs in two shifts of 16 hours a day. During each shift of 8 hours, there
is 2 hrs of planned downtime. If machine is stopped each day on average of 110 min for set-up
and 75 min for breakdowns and repairs, find the availability of the equipment.

Solution:
MTBF Actual running time
Availability = =
MTBF + MTTR Planned running time
Planned running time = Total plant time − Planned downtime
= 16 − 2 × 2 = 12 hours
Actual running time = Planned running time − All other downtime
= 12 × 60 − (110 + 75) = 535 min
535
Availability = = 0.743»74.3 per cent
12 × 60

12.13 Serviceability
Service may be defined as the activity, including diagnosis, maintenance, repair and anything
else that affects the activity to keep the system functioning. Serviceability is a measure of the
following aspects:
1. How often the system needs servicing (the less often the better the serviceability)
2. How easy it is to service (the easier to service the better the serviceability)
3. How long it takes to service (shorter the time it takes to service better the serviceability)
4. How much the service costs (the lesser it costs to service the higher the serviceability)
 Reliability and Maintenance Engineering  263

Serviceability aspect is important because no product is 100 per cent reliable. So, serviceability
is not the same as reliability. Previously, many researchers have investigated on the aspect
of serviceability. Noteworthy contributions are that of Makino who proposed the Design For
Serviceability (DFS) and the work carried out at Ohio state university, which led to the method
of Design Compatibility Analysis (DCA). Based on a set of compatibility information, that is
good, poor and bad examples of design, DCA evaluates a candidate design with respect to its
specifications and constraints, provides an overall assessment, and suggests improvements. DCA
is effective in component selection, design for injection moulding, forging product and process
design, and serviceability design and personal computers. The methodology leads to computer
programs that incorporate current knowledge to improve future designs.

12.13.1 Design for Serviceability

Not taking serviceability concerns into consideration at the design stage can result in one or
more of the following problems:
1. The product cannot be repaired.
2. The cost of repair is very high.
3. Increased service and warranty expenditure.
4. Intangible effects on customer satisfaction.

12.13.2 Serviceability Variables

There are a few subcategories of the aspects of serviceability:
1. Target audience
2. Factors affecting serviceability
3. Subcomponents of serviceability
4. Repair trade-offs

12.13.3 Target Audience

The audiences for DFS are the inexperienced audience, the service engineer and the design
engineer. The inexperienced engineer can see how different designs affect the life-cycle cost of
a product by browsing through the DFS software. If the design experiences are gathered and
stored in a knowledge base supported by the software, this user can learn and keep at finger
tips the people’s design experiences. The service engineers need a tool to answer designers’
questions regarding the cost benefits of various designs and to defend their own ideas. The design
engineer is enabled by the knowledge of serviceability and a related software package to use
those concerns of serviceability at the early stages of design.

12.13.4 Major Factors Affecting Serviceability

Following is the list of factors that affect serviceability:
1. Reliability of components and subsystems
2. Labour cost
3. Inventory cost
264  Industrial Engineering and Management

4. Accessibility of components to be serviced

5. Availability of necessary parts, tools and anything else needed for service
6. Mechanic training
7. Customer preferences
8. Where the product is serviced
9. Length and coverage of warranty
10. Special tool, repair equipment and replacement part production
We know that Failure Modes and Effects Analysis (FMEA) is a tool to find out the mode and
criticality of failure. In the context of serviceability, a new term is defined, Service Mode Analysis
(SMA) which involves steps that are to be taken to correct, or prevent the various malfunctions
that a system experience in service and the issues associated.

12.13.5 Subcomponents of Serviceability

Following are the subcomponents of serviceability:
1. Diagnosability
2. Maintainability
3. Repairability
(a) Malfunction repairability
(b) Crash repairability

12.13.6 Repair Trade-offs

The repair trade-off is the ease of removal and decreased total manufacturing cost versus loss in
ability to make fine repairs and the cost of a module. If only one small and inexpensive component
costing less than one dollar fails, will it be necessary to remove and replace a 100 dollar module
that only takes 10 min to remove? Or just replace the inexpensive component, which takes 3 hrs
to access (the mechanic has to open up the modular part, which may be complicated and therefore
has an elaborate disassembly procedure)?
The decision depends on the relative costs of labour and the module in question. If a
component in the modular part of an automobile has a life of 1000 miles (needs replacement after
that) and rest all the components in that modular part have a life of 10,000 miles, then the entire
modular part may have to be replaced for every 1000 miles if the component is not separable
from the module or difficult to access inside the intricacy of the modular part. Therefore, such
aspects of serviceability related repair trade-off aspects should be taken into account at the design
stage itself. The permanency in the connection between the components of a modular part has a
tendency to increase the reliability of the part but this has to weigh against the increased module
cost, should a repair be necessitated.

12.14 Housekeeping and 5S Concepts

Housekeeping is a part of maintenance management that supports to maintain the workplace
properly. A Japanese concept widely known as 5S concept has been used to maintain the
workplace. The details of 5S concept are given below as:
 Reliability and Maintenance Engineering  265

1. Seiri: Seiri means sort the unneeded equipments, tools, furniture; unneeded items
on walls, bulletins; items blocking aisles or stacked in corners; unneeded inventory,
supplies, parts; safety hazards, those are not required at the workplace.
2. Seiton: Seiton means arrangement of equipment in proper order. Use the items and
then return to its place. It deals with items not in their correct places; correct places
not obvious; aisles, workstations, and equipment locations not indicated; items are not
putting away immediately after use, etc.
3. Seiso: Seiso means keep the workplace and equipments clean and dustless. It is concerned
with floors, walls, stairs, equipments and surfaces not cleaned; cleaning materials not
easily accessible; labels, signs broken or unclean; and other cleaning problems.
4. Seiketso: Seiketso means standardization. It is concerned with the necessary information
not visible; standards not known; checklists missing; quantities and limits not easily
recognizable; items cannot be located within 30 sec.
5. Shitsuke: Shitsuke means self-discipline. It is concerned with number of workers
without 5S training; number of daily 5S inspections not performed; number of personal
items not stored; number of times job aids not available or up to date.

 Summary
In this chapter, the reliability of a system has been explained and various frequency distributions
for reliability have been discussed in detail. Series parallel and mixes connection of components
and reliability calculation using these connections have been demonstrated. Maintainability and
its significance in manufacturing have been explained. Similarly, the concepts of availability,
serviceability and 5S have been emphasized. After completion of this chapter, a student will be
able to calculate reliability of a system, mean time to failure, MTBF and availability of an item.

Multiple-Choice Questions
1. The probability with which an item or system will remain functioning under stated operational and
environmental conditions for a specified period of time, is known as
(a) quality (b) responsiveness
(c) assurance (d) reliability
2. Reliability engineering is concerned with key elements
(a) intended function (b) specified period of time
(c) stated conditions (d) all the above
3. Which of the following is represented by Bathtub curve?
(a) failure rate (b) reliability
(c) availability (d) none of these
4. Mean time between failure is used for
(a) irrepairable items (b) repairable items
(c) both (a) and (b) (d) none of these
266  Industrial Engineering and Management

5. Exponential distribution is used for

(a) decreasing failure rate (b) increasing failure rate
(c) constant failure rate (d) none of these
6. Normal distribution is used for
(a) decreasing failure rate (b) increasing failure rate
(c) stochastic behaviour of systems (d) none of these
7. Weibull Distribution is used for
(a) decreasing failure rate (b) increasing failure rate
(c) constant failure rate (d) all the above
8. The mechanical components, which fail due to the gradual wear and tear, are best modelled by
(a) exponential Distribution (b) normal distribution
(c) Weibull distribution (d) none of these
9. In a system, of 10 components, each of reliability factor 0.90 are in series. What is the overall
reliability of the system?
(a) 0.95 (b) 0.75
(c) 0.55 (d) 0.35
10. In a system, of 10 components, each of reliability factor 0.90 are in parallel. What is the overall
reliability of the system?
(a) 0.99 (b) 0.89
(c) 0.79 (d) 0.59
11. In the case of active redundancy,
(a) the redundant components are subjected to a lower load until one of the operating components
fails.
(b) the redundant components are subjected from the beginning of the same load as operation elements.
(c) the redundant components are subjected to no load until one of the operating components fails.
(d) none of these
12. In the case of warm redundancy,
(a) The redundant components are subjected to a lower load until one of the operating components fails.
(b) The redundant components are subjected from the beginning of the same load as operation
elements.
(c) The redundant components are subjected to no load until one of the operating components fails.
(d) None of these
13. In the case of passive redundancy,
(a) the redundant components are subjected to a lower load until one of the operating components
fails.
(b) the redundant components are subjected from the beginning of the same load as operation elements.
(c) the redundant components are subjected to no load until one of the operating components fails.
(d) none of these
 Reliability and Maintenance Engineering  267

14. Under predictive maintenance,

(a) the components are changed periodically
(b) the components are changed after failure
(c) the components are changed based on prediction of failure
(d) none of these
15. Breakdown maintenance is used when
(a) cost of failure is more than the cost of maintenance
(b) cost of failure is equal to the cost of maintenance
(c) cost of failure is less than the cost of maintenance
(d) none of these

Answers
1. (d) 2. (d) 3. (a) 4. (b) 5. (c) 6. (c) 7. (d) 8. (c) 9. (d)
10. (a) 11. (b) 12. (a) 13. (c) 14. (c) 15. (c)

Review Questions
1. Define the term ‘reliability’. Discuss the objectives of reliability engineering.
2. Discuss the bathtub curve used for failure pattern.
3. What do you mean by MTBF and MTTF?
4. Explain the basic failure rate models: (a) Constant failure rate model, (b) Weibull model, (c) Normal
model.
5. Derive the expressions for reliability of a systems in (a) Series (b) Parallel.
6. What are the ways to improve the reliability of a system?
7. Write short notes on various types of redundancy used to improve the reliability.
8. Discuss the methods used for reliability testing.
9. What do you mean by maintainability? Discuss the factors contributing to the maintainability at
design stage.
10. Write short notes on (a) maintenance policy, (b) availability, (c) serviceability (d) 5s principles of
housekeeping.

Exercises
1. Fifty machines have been operated for 1000 hours. Ten machines fail in 640 hours and 20 machines
fail in 800 hours. What is MTBF and what will be the reliability at 1500 and 2000 hours?
2. The time-to-failure density function (PDF) for a system is
f(t) = 0.01 0 ≤ t ≤ 50 days
Find
(a) Reliability for time t, R(t)
(b) The hazard rate function at time t, λ(t)
268  Industrial Engineering and Management

(c) The MTTF

(d) The standard deviation
3. A component has the following linear hazard rate, where t is in years: λ(t) = 0.06t, t ≥ 0
(a) Find R(t) and determine the probability of a component failing within the two months of its
operation.
(b) What is the design life if a reliability of 0.98 is desired?
4. A component experiences chance (CFR) failures with an MTTF of 900 hrs. Find the following:
(a) The reliability for a 50 hr mission
(b) The design life for a 0.98 reliability
(c) The median time to failure
5. For a system having a Weibull failure distribution with a shape parameter of 0.5 and a scale parameter
of 200 days, find the following:
(a) R (100 days)
(b) MTTF
(c) median time to failure, tmed
(d) The design life for a reliability of 0.98
(e) The standard deviation
6. In a system, there are five components in parallel followed by four components in series. The
components in parallel have a reliability of 0.8 each and those in series have a reliability of 0.6 each.
Determine the reliability of the entire system.

 References and Further Readings

1. Agrawal, K. K. (1993), Reliability Engineering (Boston: Kluwer Academic Publishers).
2. Balaguruswami, E. (1984), Reliability Engineering (New Delhi: Tata McGraw-Hill).
3. Birolini, Alessandro (2010), Reliability Engineering: Theory and Practice (Berlin: Springer).
4. David J. Smith (2001), Reliability, Maintainability and Risk, 6th edition (Auckland: Butternorth
Heinemann).
5. Ebeling, Charles E. (2005), Reliability and Maintainability Engineering, 1st edition (New York:
Waveland Pr. Inc.).
6. Ebeling, Charles E. (2009), An Introduction to Reliability and Maintainability Engineering, 2nd ed
(Boston: Mcgraw-Hill).
7. Pham, Hoang (2003), Handbook of Reliability Engineering (New York: Springer).
8. Stapelberg, Rudolph Frederick (2009), Handbook of Reliability, Availability, Maintainability and
Safety in Engineering Design (London: Springer).