LECTURE NOTES ON EEE 502: RELIABILITY AND MAINTENABILITY OF SYSTEMS

MAINTENANCE AND MAINTAINABILITY

1.1. MAINTAINABILITY
1.2. Certain Terms and Terminologies Relating to Maintainability
(i). Maintenance action
(ii). MTTR
(iii). Maintenance action (repair) rate:
(iv). Maintenance time constraint: (tm)
(v). Maintainability equation:
(vi). Utilization factor (U)
(vii). Repairability
(viii). Maintenance Action Time
(ix). Active Repair Time
(x). Logistics
(xi). Administrative Time
(xii). Availability (A)
o Steady State
o Instantaneous
o Mission
o Series-parallel Availability
o Unavailability
1.3. Analysis of System Availability
1.3.1. Steady State Availability
1.3.2. Instantaneous Availability
1.3.3. Mission Availability
1.3.4. Availability of Items in Series and in Parallel
1.4. Importance of Maintainability in Relation to Reliability
(i). Electric Power System
(ii). Consumer electronics – Television Set
(iii). Industrial process control system
(iv). Military Ballistic Missile
2. MAINTENANCE
2.1. Classification of Maintenance
A. Planned
• Preventive
o Routine
o Replacement
o Identification of Dormant failures
o Identification of Degradation failures
• Corrective

1
 o Short down
o Breakdown
B. Unplanned
• Emergency
2.2. Factors to consider in designing a maintenance policy
(i) Operational requirements
(ii) Equipment/system characteristics
(iii) Tools and test equipment
(iv) Spares provisioning
(v) Maintenance instructions and manuals
(vi) Logistics
3. TESTING METHODS
3.1. Categories of Testing
3.2. Prototype testing
3.3. Pre-production (qualification) testing
o Environmental
o Reliability
o Maintainability
o Packaging and Transport
o Physical and Electrical Characteristics
o Ergonomic
3.4. Production testing
3.5. Reliability Demonstration and Acceptance
3.6. Accelerated Life Testing
3.7. Relationship Between Testing and Inspection
3.8. Relationship Between Quality and Reliability
3.9. Sampling Plans (Approaches for Testing Small and Large Batch Quantities)

2
 1.1. MAINTAINABILITY
In a layman’s language, maintainability is the ease and speed with which a failed item can be
repaired and brought back into service. Perhaps at this stage, we may consider a simple
example in order to illustrate the concept of equipment/system maintainability. Suppose there
are two Television (TV) sets from different manufacturers, one of them has a control knob
outside the back cover to adjust the TV set when pictures are rolling vertically on the screen
while the other TV set has a similar control knob inside the back cover. Now, we take it that a
TV set is faulty (i.e. it has failed) if it has pictures rolling vertically on its screen and it is to be
repaired and brought back into normal service. At this juncture, it is not difficult to be
convinced that the TV set with the external knob will not require opening of the back cover
before reaching the knob and carrying out appropriate adjustment manually. Consequently, it
can be repaired in this case with greater ease and speed than the other TV set with internal
control knob. Therefore, we can say that the TV set with external knob has a higher
maintainability. In a more general case, we must consider many repair factors in an equipment
(or a TV set, as in our example) before we conclude whether or not one equipment has a lower
or higher maintainability than another similar equipment.
However, maintainability has a precise definition which is stated as follows:
Maintainability is the probability that a unit or system will be restored to operational efficiency
within a given period of time when the maintenance action is performed in accordance with
prescribed procedures. From the definition, it is clear that maintainability can be quantified
and can assume a dimensionless numerical figure between 0 and 1, except if expressed in
percentage form. However, it is important that maintainability figure should be kept as high as
possible if optimum use is to be derived from a unit or system. At this stage, it is pertinent to
mention that the achievement of high maintainability is dependent on some factors; such as
type of system location, efficiency and skill of the maintenance crew, and provision of spare
parts etc. Furthermore, we should note from the definition of maintainability the emphasis on
two very important phrases, viz; carrying out ‘maintenance action in accordance with
prescribed procedures’, and within ‘given time’. The simple explanation of these phrases is that
if we decide to put two skillful and efficient men on a repair job instead of one, we should
expect to increase the probability that the job will be completed within “a given time” if they do
the job as technically expected. The aim of maintainability is to design and develop
systems/equipment which can be maintained in the least time, at the least cost and with a
minimum expenditure of supporting resources (such as manpower, spare parts, tools and test
equipment, etc) without adversely affecting the equipment performance or safety
characteristics.

1.2 Certain Terms and Terminologies Relating to Maintainability

In order to quantify our study of maintainability, we should introduce some
terms/terminologies, which will be defined immediately below.
(i) Maintenance action: the prescribed operation to repair a failed item.
(ii) Mean time to repair (MTTR): is the mean of the time required to perform maintenance
actions. The MTTR can be developed at the design stage of a system/equipment by way of
analyzing the cogent maintenance tasks which will be required to correct a fault. A typical
sequence of maintenance tasks comprises fault localization, fault isolation, fault correction,
involving removal, replacement or reassembly, and alignment or adjustment, and checkout.

3
 The mean time anticipated for these operations is obtained and the MTTR is evaluated by
using the relationship:
∑ 𝑛𝑖 𝜆𝑖 𝑡𝑚𝑖
𝑀𝑇𝑇𝑅 = (1.1)
∑ 𝑛𝑖 𝜆 𝑖
A worked example is shown in Fig. 1.1 to further illustrate this point.

Component Qty Failure 𝑛 𝑖 𝜆𝑖 Average maintenance action time(hours)

Type 𝑛𝑖 rate Localization Isolation Disassembly Replacement Alignment 𝑡𝑚 𝑛𝑖 𝜆𝑖 𝑡𝑚
𝜆𝑖
Resistor 21 0.06 1.26 0.01 0.1 0.015 0.025 0.001 0.151 0.19026
Capacitor 12 0.04 0.48 0.12 0.1 0.02 0.02 0.001 0.261 0.12528
Variable
Capacitor 2 0.005 0.01 0.15 0.2 0.03 0.05 0.001 0.431 0.00431
Transistor 4 0.05 0.2 0.05 0.1 0.02 0.05 0.001 0.221 0.0442
Power
transistor 5 0.09 0.45 0.1 0.15 0.04 0.08 0.001 0.371 0.16695
Diode 4 0.05 0.2 0.06 0.2 0.02 0.05 0.001 0.331 0.0662
Zener Diode 5 0.01 0.05 0.05 0.15 0.02 0.04 0.001 0.261 0.01305
Transformer 3 0.02 0.06 0.04 0.08 0.06 0.06 0.001 0.241 0.01446
Transformer 2 0.03 0.06 0.08 0.04 0.05 0.05 0.001 0.221 0.01326
Loudspeaker 5 0.005 0.025 0.04 0.05 0.02 0.05 0.001 0.161 0.004025
Inductor 2 0.07 0.14 0.05 0.08 0.05 0.08 0.08 0.34 0.0476
∑ 𝑛𝑖 𝜆𝑖 = 2.935 2.99 0.689595
Fig. 1.1
∑ 𝑛𝑖 𝜆𝑖 = 2.935

∑ 𝑛𝑖 𝜆𝑖 𝑡𝑚 = 0.689595
∑ 𝑛𝑖 𝜆𝑖 𝑡𝑚 0.689595
𝑀𝑇𝑇𝑅 = = = 0.235 ℎ𝑜𝑢𝑟𝑠
∑ 𝑛𝑖 𝜆 𝑖 2.935
(iii) Maintenance action (repair) rate: (𝜇) is the number of maintenance actions that can be
carried out on a particular item per hour or average number of items (equipment) which can be
restored to normal working condition per hour, assuming that there is no waiting between
repair jobs. Maintenance action rate (u) is the reciprocal of the MTTR,
1
𝜇= (1.2)
𝑀𝑇𝑇𝑅
(iv) Maintenance time constraint: (tm) is the permissible repair time.
(v) Maintainability equation: The probability of completing one or more maintenance actions is
expressed through the maintainability equation as:
−𝑡𝑚
𝑀 = 1 − 𝑒 −𝜇𝑡𝑛 = 1 − exp ( ) (1.3)
𝑀𝑇𝑇𝑅

Example 1.1
The average time to repair any fault in a given system is 1 hour. Determine maintainability of
the system for a time of 21/2 hours.

4
 Solution
From equation (1.3)
𝑀 = 1 − 𝑒 −2.5/1
= 1 − 0.082 = 0.918
Comment: The probability, M of the system being returned to a working condition within
21/2 hours is 0.918 (or 91.8%).
Note:
From the foregoing, it is clear that the prediction of maintainability involves establishing a value
for the MTTR of a system. Although an accurate assessment of MTTR is a difficult task, however
a designer can aim at a low rate value of MTTR by considering such factor as accessibility of
components, built-in fault units etc.
(vi) Utilization factor (U): Utilization factor of a unit or system is defined as the ratio of the
operating time (top) to the sum of the maintenance time, (tm), idle time (tid), (which may
occur between completion of maintenance and use due to administrative reasons) and the
operating time. Mathematically, the utilization factor, U can be expressed as
𝑡𝑜𝑝
𝑈= (1.4)
𝑡𝑚 + 𝑡𝑖𝑑 + 𝑡𝑜𝑝
(vii) Repairability: Repairability is the probability that a unit or system will be restored to
operational efficiency within a given ‘active repair time’, if specified maintenance is
performed on it. Meanwhile, it should be noted that active repair time is the time during
which technicians are actually working on repairs, adjustment etc.
(viii) Maintenance action time: Maintenance action may consume time due to the following
activities which are associated with it:
(a) Reporting failure to the appropriate staff;
(b) The maintenance team or a maintenance technician collects the repair tools and possibly
some spares he guesses may be needed;
(c) Sending for additional spares, replacement, etc, if the need arises;
(d) Trying to verify the precise way in which the item fails to function and consequently
deciding to disassemble some parts prior to the proper commencement of fault
diagnosis;
(e) Locating the actual fault;
(f) Carrying out repairs or replacing a faulty unit with a good one and making any
adjustments, if necessary, and
(g) Testing the item after repairs to ensure satisfactory performance.

NOTE
From the foregoing it is clear that the ‘maintenance action time’ is the sum total of the times
spent on all the activities listed (a – g) above.
(ix) Active repair time: this is the time during which actual repair work; adjustments etc are
carried out by the maintenance technician(s). This period is covered by activities (d) – (g).
(x) Logistics time: This is part of the total repair time and represents the time taken to obtain
all the repair tools, materials and spares etc, and the movement of repair crew to the failed
item. It is covered by activities (b) and (c).
(xi) Administrative time: This is the time which is not specifically allocated to either active
repair or logistics. This includes the time lost due to the non-availability of maintenance
5
 staff, partly as a result of meal breaks (i.e. tea break or lunch time) or because the
maintenance technician needed for the repair job is yet to complete a previous job etc.
This time may be covered by the activities (a) and (e).
(xii) Availability (A): Availability is the probability that an item will perform its required function
at a stated instant of time or over a stated period of time. From Equation (1.4), if the idle
time is equal to zero (i.e. 𝑡𝑖𝑑 = 0) and the maintenance time becomes as small as possible,
then utilization factor will approach its maximum value, and can now be called availability
of a unit or system. Mathematically, this can be expressed as:
𝑡𝑜𝑝
𝐴 = 𝑈𝑚𝑎𝑥 = (1.5)
𝑡𝑚(𝑚𝑖𝑛) + 𝑡𝑜𝑝
On the other hand, we know from our previous knowledge of MTBF and
MTTF that the following assumptions are true:
(i) 𝑡𝑜𝑝 = MTBF (mean time between failures)
(ii) 𝑡𝑚(𝑚𝑖𝑛) = MTTR (mean time to repair)
Using these assumptions in equation (1.5) above,
𝑀𝑇𝐵𝐹
𝐴= (1.6)
𝑀𝑇𝐵𝐹 + 𝑀𝑇𝑇𝑅
If the availability of equipment is given as 0.90, this means that the equipment is working
satisfactorily for 90 percent of the time, and under repair, for the remaining 10 percent. The
concept of availability is further illustrated through worked examples shown immediately
below.
Example 1.2
An electrical generating set designed for continuous operation fails twice in a period of 123
days. The total time for repairs during the period is 3 days. Determine the following
parameters;
(a) MTBF (days)
(b) MTTR (days)
(c) Availability
Solution
(a) Total operating time = (123 – 3) days = 120 days
Total number of failures = 2
 MTBF = 120/2 = 60 days

(b)
𝑡𝑜𝑡𝑎𝑙 𝑡𝑖𝑚𝑒 𝑓𝑜𝑟 𝑟𝑒𝑝𝑎𝑖𝑟𝑠 3
𝑀𝑇𝑇𝑅 = = 𝑑𝑎𝑦
𝑡𝑜𝑡𝑎𝑙 𝑛𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝑓𝑎𝑖𝑙𝑢𝑟𝑒𝑠 2
(c)
𝑀𝑇𝐵𝐹 60
𝐴𝑣𝑎𝑖𝑙𝑎𝑏𝑖𝑙𝑖𝑡𝑦 = = × 100% = 97.6%
𝑀𝑇𝐵𝐹 + 𝑀𝑇𝑇𝑅 60 + 3/2

Example 1.3
The ratio between maintenance time and operating time of a complex system is 1:10. Calculate
the ratio between maximum time the system can be allowed to stand idle and the operating
time to achieve a maximum utilization of 0.8.
Solution
6
From equation (1.4)
𝑡𝑜𝑝 1
𝑈= = 0.8 = 𝑡 𝑡𝑖𝑑
𝑡𝑚 + 𝑡𝑖𝑑 + 𝑡𝑜𝑝 𝑚
𝑡𝑜𝑝 + 𝑡𝑜𝑝 + 1
Using the data given,
1
0.8 =
1 𝑡𝑖𝑑
+ +1
10 𝑡𝑜𝑝
𝑡𝑖𝑑
= 0.15 𝑜𝑟 0.15: 1
𝑡𝑜𝑝
1.3. Analysis of System Availability
In this section, we are going to discuss on different types of availability, and how to determine
the availability of a complex system consisting of sub-units connected either in series or parallel
configuration. There are three forms of system availability, namely: steady-state availability,
instantaneous availability, and mission availability.
1.3.1 Steady-state Availability (Ass)
Steady-state availability is the proportion of time that a system is availability for use when the
overall period is of considerable duration. Since availability of a repairable system is a function
of its failure rate, , and of its repair rate, , therefore Equation (1.6) can be put in another
form to read
1
𝑀𝑇𝐵𝐹 𝜇
𝐴𝑠𝑠 = = 𝜆 = (1.7)
𝑀𝑇𝐵𝐹 + 𝑀𝑇𝑇𝑅 1 1 𝜆 + 𝜇
+𝜇
𝜆
On the other hand, we can express system unavailability as:
Steady-state unavailability,
𝜆
̅̅̅̅̅
𝐴𝑠𝑠 = 1 − 𝐴𝑠𝑠 = (1.8)
𝜆+𝜇
Note
Example 1.2 previously discussed fits the situation of steady-state availability. The reader may
wish to have another look at that example. In the example, the values of MTBF and MTTR
involved are large enough for our purpose.

1.3.2 Instantaneous Availability

Instantaneous availability is the probability that a system will be available at any instant of time,
t. It is interest to note that instantaneous availability is sometimes called point availability. The
steady-state availability may be 0.5 as previously explained, but if we ask whether the
equipment will be available at, say, 12 noon in another day, the answer may be negative,
probably because it is under repairs. Consequently, at that particular moment in time, (12
noon), the instantaneous or point availability will be zero. Instantaneous availability can be
expressed mathematically as
𝜇 𝜆
− 𝑒𝑥𝑝[(𝜆 + 𝜇)𝑡] (1.9)
𝜆+𝜇 𝜆+𝜇
Similarly, instantaneous Unavailability 𝐴⃖ can be expressed as:
𝜆 𝜆
𝐴⃖ = 1 − 𝐴 = − 𝑒𝑥𝑝[−(𝜆 + 𝜇)𝑡] (1.10)
𝜆+𝜇 𝜆+𝜇
As time t becomes large, Equations (1.9) and (1.10) will become
7
 𝜇 𝜆
𝑎𝑛𝑑 𝑟𝑒𝑠𝑝𝑒𝑐𝑡𝑖𝑣𝑒𝑙𝑦
𝜆+𝜇 𝜆+𝜇
It should be noted that if t is large, Equation (1.10) becomes
𝜇
𝐴=
𝜆+𝜇
which is the same as Equation (1.8). In other words, instantaneous availability becomes equal
to steady-state availability when time value is large (i.e. when 𝑡 → ∞).
1.3.3 Mission Availability
Mission availability is more appropriate to military situation where no one expects that an
expensive military hardware will continuously function whenever required. Instead, one is
more concerned with the availability of an equipment/system when it is required for a
particular military mission. Thereafter, it may not function again. Hence, mission availability of
an equipment/system can be broadly defined as the probability that a mission of given duration
will be successfully completed within a given time.
Now, let us consider an example of a practical military situation Fig. 1.2 shows a simplified block
diagram of a missile launching and guidance system. The system is expected to guide the
motion of the missile launched into the air until it hits and destroys the target aircraft. And so,
the mission of the missile is to destroy the ‘enemy’ target aircraft. If the mission is successful
we say that the system is ‘mission-available’. In quantitative terms, mission availability can be
assessed. It normally has a numerical value between 0 and 1. However, after the
accomplishment of the said mission the system may or may not function again.

Fig. 1.2: Simplified block diagram of a missile launching and guidance system.

1.3.4 Availability of Items in Series and Parallel Combinations

Availabilities of sub-units can be combined in the same way as reliabilities. The availability of a
system, As consisting of n sub-units in series with availabilities 𝐴1 , 𝐴2 , … 𝐴𝑛 respectively is given
by:
𝐴𝑠 = 𝐴1 𝐴2 … 𝐴𝑛 (1.11)
8
Similarly, if the items are connected in parallel, the availability 𝐴𝑃 is given by:
𝐴𝑃 = 1 − [(1 − 𝐴1 )(1 − 𝐴2 ) … (1 − 𝐴𝑛 )] (1.12)
As an illustration, consider a system made up of two series sub-units with individual failure
rates of 1, 2 and repair rates of µ1 and 2 respectively.
𝜇1 𝜇2
𝐴1 = 𝑎𝑛𝑑 𝐴2 =
𝜇1 + 𝜆1 𝜇2 + 𝜆2
By substituting these values of A1 and A2 into Equation (1.11), we have
𝜇1 𝜇2
𝐴𝑠𝑠 =
(𝜇1 + 𝜆1 )(𝜇2 + 𝜆2 )
However, if 𝜇1 = 𝜇2 = 𝜇 and 𝜆1 = 𝜆1 = 𝜆 then
𝜇2
𝐴𝑠𝑠 =
(𝜇 + 𝜆)2
Suppose we consider a duplex system (two sub-units connected in parallel) with an individual
sub-unit failure rate given by 𝜆 and the repair rate given by 𝜇. From Equation (1.13), the
steady-state availability is
𝜇 𝜇 2 𝜇 2 − 2𝜇𝜆
𝐴𝑠𝑠 = 1 − (1 − 𝐴𝑖 )2 = 2𝐴𝑖 − 𝐴𝑖 2 = 2 ( )−( ) =
𝜆+𝜇 𝜆+𝜇 (𝜆 + 𝜇)2
A worked example is given immediately below to further explain how the availability of a
complex system can be calculated.
Example 1.4
A communication system comprises four component modules connected in series. Given that

UNIT MTBF (Hours) MTTR (Hours)

A 250 5
B 200 10
C 650 6
D 800 8

(i) Calculate the availability of the system,

(ii) Calculate the availability of the system if two of the unit C modules are connected in
parallel.
Solution:
(i) The availabilities of the units must first be determined. Hence, availability of unit

𝑀𝑇𝐵𝐹 250
𝑢𝑛𝑖𝑡 𝐴 = = = 0.98
𝑀𝑇𝐵𝐹 + 𝑀𝑇𝑇𝑅 250 + 5
200
𝑆𝑖𝑚𝑖𝑙𝑎𝑟𝑙𝑦, 𝑎𝑣𝑎𝑖𝑙𝑎𝑏𝑖𝑙𝑖𝑡𝑦 𝑜𝑓 𝑢𝑛𝑖𝑡 𝐵 = 210 = 0.95
650
𝐴𝑣𝑎𝑖𝑙𝑎𝑏𝑖𝑙𝑖𝑡𝑦 𝑜𝑓 𝑢𝑛𝑖𝑡 𝐶 = 656 = 0.99
800
𝐴𝑣𝑎𝑖𝑙𝑎𝑏𝑖𝑙𝑖𝑡𝑦 𝑜𝑓 𝑢𝑛𝑖𝑡 𝐷 = 808 = 0.99
Since the units are connected in series, the overall system availability is the product of
individual availability of the units given by, Equation (1.11).
𝑇ℎ𝑒𝑟𝑒𝑓𝑜𝑟𝑒 𝑜𝑣𝑒𝑟𝑎𝑙𝑙 𝑠𝑦𝑠𝑡𝑒𝑚 𝑎𝑣𝑎𝑖𝑙𝑎𝑏𝑖𝑙𝑖𝑡𝑦 = 0.98 × 0.95 × 0.99 × 0.99 = 0.91

9
 (ii) If two of the unit C modules are connected in parallel, then then whole system
arrangement will be configured as shown in Fig. 1.3.

A B D

C
Fig. 1.3
Using Equation (1.12) the resultant availability 𝐴𝐶𝑃 of two unit C modules connected in parallel
is given by
𝐴𝐶𝑃 = 1 − (1 − 𝐴𝐶 )2
Where 𝐴𝐶 = availability of one unit C,
From the above calculation, 𝐴𝐶 = 0.99
𝐻𝑒𝑛𝑐𝑒, 𝐴𝐶𝑃 = 1 − (1 − 0.99)2 = 0.999
Using equation (1.12) the overall system availability
= 0.98 × 0.95 × 0.999 × 0.99 = 0.922
̅
(xiii) Unavailability (𝐴): Unavailability of an equipment or system due to breakdown is the
complement of availability.
𝑖. 𝑒. 𝐴̅ = 1 − 𝐴
Multiplying both sides of Equation (1.6) by -1 and adding 1 to both sides of the result, we have
𝑀𝑇𝐵𝐹 𝑀𝑇𝐵𝐹 + 𝑀𝑇𝐵𝐹 + 𝑀𝑇𝑇𝑅
𝐴̅ = 1 − =
𝑀𝑇𝐵𝐹 + 𝑀𝑇𝑇𝑅 𝑀𝑇𝐵𝐹 + 𝑀𝑇𝑇𝑅

𝑀𝑇𝑇𝑅
̅=
𝑇ℎ𝑒𝑟𝑒𝑓𝑜𝑟𝑒 𝑈𝑛𝑎𝑣𝑎𝑖𝑙𝑎𝑏𝑖𝑙𝑖𝑡𝑦 𝐴
𝑀𝑇𝐵𝐹 + 𝑀𝑇𝑇𝑅

1.4 Importance of Maintainability in Relation to Reliability

From the proceeding sections in this chapter, it has been made clear that reliability and
maintainability combine together to give availability. Essentially, the key factors influencing
availability are mean time between failure (MTBF), mean time to repair (MTTR), and
performance before failure and after repair. For example, a highly reliable system with a long
MTTR may have lower availability than a less reliable system that is easily and quickly
maintained or repaired. On the other hand, the user may be willing to sacrifice either reliability
or maintainability to achieve higher performance (as in some military weapon systems).
Performance may be sacrificed, by contrast, to increase reliability and minimize MTTR.
As an illustration we shall consider immediately below some systems with highlights on their
availability, reliability, and maintainability as they affect user’s needs. The systems we shall
consider are an electric power system, consumer electronics systems, (a Television set), an
industrial control system (a chemical process control) and a military system (the ballistic
missile).
An Electric Power System
The consumer expects electric power supply all the time for his needs. In some cases, a
prolonged power failure may cause immeasurable hazard and tragedy too obvious for
mentioning. In fact, industrial users may suffer immense financial loss.
10
Power is made available to consumers almost all the time. Availability is kept close to 100
percent especially in developed nations through modestly reliable but highly redundant and
constantly maintained equipment. Redundant generating systems and power sharing among
different geographical regions provide back-up power to compensate for various overload
situations. The approximate useful life span of an electric power system is greater than 15
years.
Consumer electronics – Television Set
Based on research studies, the approximate useful life span of a TV set is between 7 to 10
years, if used approximately for about 2000 hours per year. Manufacturers aim for very long
MTBF and high availability for the first few years of ownership, in order to avoid warranted
repairs and against losing customers to market competitors.
An Industrial Process-control System
The electronic process-control system must regulate the chemical reaction to ensure precise
control of a product’s composition and this process normally runs 24 hours every day, without
breakdown or malfunctioning. Any downtime whatsoever or a slight deviation from the desired
product can cause enormous financial loss. Reliability, in this case, can be measured in terms of
the fraction of time that a processing unit produces products of a specified composition. Any
failure of the automatic control system triggers an immediate switch-over to standby systems
so that continuously good performance is maintained. The mechanical units of the plant may
require scheduled maintenance. Consequently, high reliability and short MTTR are crucial since
even minimal downtime can lead to a ruined batch of product and other forms of losses.
A Military ballistic missile
The user, the military Airforce command, expects the ballistic missile to be in operational
readiness up to the time it is launched. When launched, the missile must operate very reliably
and guide itself accurately to a prescribed target against all odds. Based on the mode and
circumstances of its operation the missile is expected to have relatively short MTBF and MTTR
as trade-offs in favour of high performance and reliability which is almost 100 percent.
2. Maintenance
Maintenance is a combination of any actions carried out to retain an item in or restore it to
normal operational standard.
The principal objectives of maintenance can be stated as follows:
(a) To extend the useful life span of an item;
(b) To increase availability of installed equipment;
(c) To ensure operational readiness of all equipment required particularly for emergency
purpose;
(d) To ensure the safety of people (personnel) using such equipment;
(e) To avoid exorbitant expenses on the repair of equipment, which might occur if the same
equipment was not maintained and allowed to fail
Generally, there are two types of maintenance, namely: Preventive maintenance and
Corrective maintenance. A detailed classification is given below.
2.1. Classification of Maintenance
Fig. 2.1 shows the classification of maintenance. The classification shown in the diagram is
clear and self-explanatory to a great extent. At this juncture one must sound a note of warning
that the classification given in Fig. 2.1 is generally broad and this should not be taken to mean
that all forms of maintenance will neatly fit into this classification. There are some types of
11
maintenance which will be mentioned and discussed in this section which cannot be so readily
classified. These include fixed time maintenance, condition-based maintenance, replacement
and operate-to-failure maintenance.
It is pertinent to discuss briefly each and all forms of maintenance shown in the diagram above
and possibly give an example of equipment or system to which they are applicable.

Fig. 2.1 Classification of maintenance

Planned Maintenance
Planned maintenance is the type of maintenance work organized and carried out with
forethought, control and records. Planned maintenance is not a specific type of maintenance,
but represents the application of maintenance tackled in a scientific manner. This covers the
comprehensive planning of the maintenance function and includes the whole range of
maintenance such as fixed-time maintenance, planned replacement, preventive, operate-to-
failure maintenance and condition-based maintenance, etc.
Fixed-time maintenance is the individual or group replacement or repair of items after the
expiration of a fixed period (e.g. changing the oil filter in a car after 2000 km, changing the
engine oil in an electric generator after, say, 500 running hours) before failure occurs.
Operate-to-failure maintenance involves only such maintenance tasks as normal lubrication and
adjustments of parts but no other predetermined action is taken to prevent failure.
Planned replacement can be very suitably applied in situations when an equipment operates as
an individual unit e.g. machine tools, mobile electric generators etc. Where a policy of planned
replacement is practiced the pivot of the planning is to ensure the money is available for the
purchase of new equipment and the provision of spare parts.
In condition-based maintenance, it is the actual current condition of the equipment/system
which determines the form and frequency of maintenance. Condition monitoring of equipment
can be carried out by using human senses (visual inspection, hearing, touch and smell), portable
test instruments and fixed monitoring systems such as simple remote alarm system,
minicomputers, XY plotters, etc.

12
The fundamental basis of any planned maintenance system hinges on deciding in advance the
following points:
(i) The individual items of equipment to be maintained;
(ii) The form, method and details of the maintenance scheme for each item;
(iii) The tools, test gears, replacement spares and time needed for carrying out a maintenance
job;
(iv) How frequency maintenance operations must be carried out;
(v) The method of administering the system; and
(vi) The method of analyzing the results.
The details required for (ii), (iii) and (iv) are likely to be found in the manufacturer’s manual,
normally supplied with the equipment). In order to apply the above listed points practically the
following elements are required:
(i) An inventory of all equipment to be maintained;
(ii) A complete list containing all the individual tasks that must be carried out on each item of
equipment;
(iii) A calendar of events, indicating the sequence of maintenance which must be carried out;
(iv) A method of monitoring and ensuring that the work listed in the programme is carried out;
and
(v) A method of recording the results and assessing the effectiveness of the overall
maintenance.
The benefits derivable from planned maintenance include the following:
(i) Fewer breakdowns of regular and correct maintenance operations are carried out;
(ii) Appreciable reduction of downtime since spares and equipment demands are known in
advance and are available immediately on request;
(iii) It ensures regular, simple servicing which is cheaper than sudden expensive stop-gap
repairs; and
(iv) It guarantees greater and more effective labour utilization because it is planned ahead and
is made to spread evenly throughout the year, thus reducing widely fluctuating demands
upon the labour force.

Planned Preventive Maintenance

This is a form of planned maintenance in which components or units of certain kinds are
removed and replaced by new components at specified interval of time appreciably shorter
than the mean life of the component in order to prevent failure of an equipment/system.
Preventive maintenance increases reliability by predicting failures and preventing them from
occurring. An attempt to introduce preventive maintenance for random failures can only lead
to a reduction in equipment reliability. When a component or unit has a known rate of wear,
the task is either to measure the amount of wear, (for example, the length of the carbon
brushes in a dynamo) or to keep an account of wears in terms of other related quantities.
Preventive maintenance is therefore, intended to provide checks to cover the wear-out period
of a component’s life, including detecting the onset of this period, as well as replacing worn-out
part.
Fig. 2.1 shows that planned preventive maintenance can be carried out either if the
equipment/system concerned is in service or out of service, leading consequently into running
preventive maintenance and shut-down preventive maintenance respectively. Regardless of

13
this clarification, there are four specific types of preventive maintenance, viz: routine
maintenance, preventive replacement, identification of dormant failures, and identification of
degradation conditions.
(i) Routine preventive maintenance
This covers such operation as clearing, lubricating and adjusting at specified intervals of time.
There is the need for care in carrying out this type of maintenance because more failures may
be caused than prevented.
(ii) Preventive Replacement
Preventive replacement is based on the policy of replacing items after a particular period of use
even if they have not failed. In that case, it is imperative to have knowledge of the wear-out
distribution of the particular items concerned in the equipment. Preventive replacement is
characteristic of wastages because many items may be replaced before they actually fail. On
the other hand, the great benefit of avoiding equipment failure can be readily achieved. The
cost of equipment failure and its attendant inconvenience may be far outweighing the cost of
the preventive replacement of certain items.
(iii) Identification of Dormant Failures
Some items fail in an equipment without necessarily causing the equipment itself to fail or
malfunction immediately. However, a failed item in equipment will definitely increase the
stresses on the remaining items and hence accelerate equipment failure. Consequently,
occasional checks for dormant faults are practicable although it may be expensive. A very
useful example is the indicator lamps in a motor car. They are there to signify the occurrence of
a fault, which if not rectified immediately may lead to greater problems. A burnt-out lamp
must be replaced immediately. A careful motorist must always check out for the proper
functioning of the indicators in his/her vehicle.
(iv) Identification of degradation condition:
In systems where extremely high reliability is expected, such as in aircrafts, it is often necessary
to carry out measurements of some parameters for the purpose of identifying parts which have
degraded or drifted and which might be liable to become a failure in the near future. In such
cases, preventive replacement may become inevitable.

Planned Corrective Maintenance

Planned corrective maintenance is normally used when an equipment/system has already failed
and it is to be repaired and restored back to service. From Fig. 2.1, corrective maintenance can
either be of two main types, namely: shut-down maintenance and breakdown maintenance.
In the case of shut-down maintenance, maintenance operations are normally carried out when
the equipment/system is out of service. This type of maintenance is usually carried out during
slack time in the case of systems which are running continuously but have other redundant
channels. A channel that has become faulty may be automatically withdrawn or shut-down
from service while a redundant channel is switched on in its place. Whereas, in a situation
where redundant channels are non-existent, shut-down maintenance is carried out during the
period when the equipment is not operating, particularly if the department or the organization
concerned is on a single shift working.
On the contrary, breakdown maintenance is often initiated by a sudden, unexpected failure of
an equipment/system. When this occurs, all available resources are urgently utilized to restore
the equipment to its normal operational standard in the shortest possible time. Furthermore,

14
breakdown maintenance is normally worked out after an equipment/system has been
considered in advance and provision made by planning the method of repair; tools and spares
requirements and possible replacements.
Unplanned Maintenance
Unplanned maintenance is characteristic of maintenance work not initially organized and not
carried out with any forethought, control and records. Whenever a system/equipment fails to
work as intended, then repairs and adjustments must be carried out to bring it back into
service. Since we can never foresee exactly when a particular type of failure will occur we
cannot plan for it in advance.
Unplanned Emergency Maintenance
Unplanned emergency maintenance is the type which necessitates a maintenance which has
not been foreseen and to which no advance thought has been given. This involves a failure that
is thought could not possibly happen. In a situation of emergency maintenance, all available
resources are urgently utilized to restore the equipment to its normal operational standard in
the shortest possible time. For example, a radio broadcast transmitter may suddenly cease to
transmit messages to public listeners probably due to a sudden failure of one of its
components. In this case, the maintenance crew will swing into action immediately in order to
rectify the fault and ensure it is put back into service without much delay. If a standby
transmitter is available it will be put immediately to use while repair work progresses on the
faulty one. Unfortunately, many industries/establishments in developing countries rely on this
type of maintenance, leaving things to chance because of lack of financial resources or foresight
to embark on planned maintenance. This is why many industrial machinery and equipment
pack up, leading occasionally to the closure of the industry itself. That is the bane of industrial
undevelopment.
2.2. Factors to Consider in Deciding a Maintenance Policy
In the preceding section, we have discussed different forms of maintenance, any or a
combination of which may be adopted for an equipment/system maintenance. Therefore it is
pertinent to discuss here those factors which should always be considered in choosing a
maintenance policy. In fact, the choice of a maintenance policy invariably leads to the decision
on the choice of a maintenance method. Such factors will include:
(vii) Operational requirements
(viii) Equipment/system characteristics
(ix) Tools and test equipment
(x) Spares provisioning
(xi) Maintenance instructions and manuals
(xii) Logistics
Now, we shall throw some light on the effect which each factor has on a maintenance
policy and how they assist in achieving good maintainability.
Operational Requirements
By operational requirement we mean the function equipment has to perform and the
conditions under which it operates. It is perhaps easy to see the considerable difference
between maintaining a communication satellite launched and working in outer space several
thousands of kilometers from the earth surface, and a domestic radio receiver used in a living
room. In line with the operational requirement of the satellite, once launched, it is kept
15
working continuously in the outer space for several years without direct contact with human
beings for any maintenance operation. One must therefore be prepared to pay heavily for this
by improving reliability in order to eliminate the possibility of a breakdown and maintenance.
Conversely, a domestic radio receiver is in an environment where spare parts are readily
available, and less reliability will be accepted.
Equipment characteristics
Equipment characteristics encompass such things as electrical and mechanical features which
the equipment must satisfy in line with its operational requirement. The latter includes factors
as reliability complexity, safety precautions, and the degree of heat generation and evacuation,
as in thermionic circuits. The complexity of an electronic circuit is manifested by the number of
components and the interconnections among them. It is easy to see that the greater the
complexity of a circuit, the more difficult it is to isolate faults, and the more tedious a
maintenance operation. However, if the task is difficult then, there is the need for increase in
demand for well-trained maintenance staff, or for other aids to the maintenance task.
Mechanical feature of equipment also affects the maintenance task mainly through
requirements for manual skills, involving dismantling and assembling one thing or the other.
Some of these tasks may be generally delicate since they may have to be carried out in confined
or narrow spaces within equipment.
Tools and Test Equipment
The provision of appropriate tools and test equipment influence the maintainability of
equipment considerably. However, the following points must be considered when specifying
tools and test equipment. These are: the test gear must be easy to set up and operate. It must
also be more reliable than the equipment/system for which it is designed. Finally, it must be
maintainable and available for purchase for as long as the system for which it is made exists. A
great deal can be done to achieve rapid maintenance by the use of built-in test equipment
(BITE) or indicators for immediate visual detection of fault.
Maintenance Personnel
The following points concerning maintenance personnel vis-à-vis their influence on the
maintainability of equipment must be considered:
- Level of training given
- Skill level employed
- Motivation
- Quantity and deployment of maintenance personnel.
Definitely, a complex system may involve a wide range of maintenance and hence a good level
of training will be required for the maintenance personnel. Ability in carrying out corrective
maintenance is achieved by a combination of theoretical knowledge and practical diagnostic
skills. Consequently, a very effective maintenance technician should have a balance of
theoretical knowledge and practical skill commensurate to the complexity and diversity of
equipment or systems for which he is engaged. An ill-motivated maintenance technician, no
matter the level of training given him and skill acquired, may be less productive. Staff
motivation is closely associated with a number of factors such as staff promotion,
accommodation, condition of service etc. In fact, an ill-motivated technician can possibly
sabotage the maintenance effort (policy) of his employer; although if he does that, it will be at
the risk of losing his job. The MTTR may be reduced by deploying two technicians where
16
necessary to a maintenance work instead of one, especially where complicated maintenance
operation is involved.
Maintenance instructions and manuals
The provision of a well-explained and carefully illustrated maintenance manual and instructions
is a very important tool in achieving good maintainability. Maintenance instructions and
manuals, containing a logical sequence of tests are normally used as an aid to fault diagnosis.
They may also contain information on safety hazards and precautions, spare requirements and
replacements, procedure for failure reporting, and details of preventive maintenance.
Spares Provisioning
Based on the information concerning failure rate of some components, provision could be
made for spares. Faulty equipment without a provision for urgently needed spare parts can be
temporarily forgotten altogether! No complete maintenance job can be achieved without
spares. However, for spares provisioning, allowance must be made for damages during
transportation and storage. The main considerations of spares provisioning are: failure rate-
which determines essentially the quantity of spares needed, cost of each spare which also
assists in fixing the quantity of spares to be maintained, and the supply time after ordering. If
the spares can be obtained easily and locally without much delay, there may be no sense in
massive stock-piling of spares.
Logistics
Logistics deals with time and effort involved in transporting equipment, spares and manpower
to the venue where maintenance action is needed. The aspects of what to centralize or
decentralize should be well appreciated particularly in relation to the nature and location of
equipment/system to be maintained. As a guide especially where small MTTR is vital, it is
advisable to decentralize:
(i) Small tools and standard items that may be frequently required for maintenance
operations;
(ii) Spares stocking;
(iii) Fragile test gear.
The best maintenance policy can be obtained as a result of the optimum combination of the
contributing factors discussed earlier. It is important for the technician to realize and
appreciate that the maintenance policy he is following evolves as a result of a balance between
these factors. Consequently, the recently discussed factors contribute immensely towards the
achievement of good maintainability.

3 Testing Methods
3.1 Categories of Testing
It is no simple matter trying to classify testing. This is because the criteria for the classification
are not so distinct and are therefore not free from causing possible overlapping. Nonetheless,
three stages/categories of testing can be identified in an ideal manufacturing process. The
three categories of testing are called prototype testing, pre-production (qualification) testing
and production testing. Apart from these, there exists reliability demonstration and acceptance
testing which of course, may not be a mandatory process after the production of an item. At
this stage, it suffices to say that this type of testing is based purely on an agreement between
the manufacturer and the customer. More light will be shed on it later on in this chapter.
17
Finally, there is also what is known as ‘accelerated life testing’ which actually does not stand on
its own as a category of testing but associated with any of the three categories of testing earlier
mentioned. Further explanation on accelerated life testing will be provided later on.
Meanwhile, we shall examine in some detail all the types of testing already referred to.
3.2 Prototype Testing
Before a design goes into production a prototype or few prototypes are first produced and are
thoroughly tested to prove that a design will meet the specification. If the tests reveal any
inadequacies in either performance or any other parameters, the design must be modified and
the testes repeated.
It is very desirable that prototypes are tested under the correct environmental conditions by
competent staff who are independent of the designers. Besides, there is the need for care not
to allow every shortcoming to be glossed over with the hope that it will be corrected at the pre-
production or production stage. At times, prototype testing might be made the responsibility
of competent staff in the reliability or the quality control department. Since the designer
should be free from the responsibility of testing the prototype of his own design, therefore, the
designer must be completely answerable for its performance and reliability. If tests reveal
some shortcomings, it is the designer’s responsibility to correct them. As a result of prototype
testing three types of errors may be revealed. They are:
(i) Design errors – here the prototype might have been produced according to drawing, but
still it does not work satisfactorily.
(ii) Manufacturing errors – where the prototype has not been made correctly, in line with the
design.
(iii) Parts faults – where individual components parts of the prototype are found to be
defective.

3.3 Pre-production (Qualification) Testing

The aim of pre-production (qualification) testing is to ensure that a production model fulfills all
the technical requirements stated in the Engineering specification. Pre-production testing is
normally done before bulk manufacture of the item is embarked upon in order to ensure that
the specifications can be met under normal manufacturing conditions and such tests, as usual,
will be performed by an independent department such as the quality control department. As a
result, pre-production testing will encompass performance test (at defined limits and margins),
environmental test, reliability test, maintainability test, packaging and transport test (involving
vibration and shock tests), and physical characteristics test (to consider the positioning of
operating knobs and switches etc. vis-à-vis the convenience of the operator). It will be quite
useful to examine in some details some of the important tests already mentioned under pre-
production testing.
Environmental Tests
This is the type of test normally carried out on the production model under simulated
environmental conditions. The test can, of course, be done in actual working environment if it
is so available. In the final analysis one is interested in proving by this test that the product will
function according to laid down specification for a sustained period and is free from damage
and degradation by specified extremes of its environment. In practice, test equipment may
include ovens which can simulate various combinations of temperature and humidity,
refrigeration plant and equipment to simulate various combinations of vibration, shock etc.
18
 Reliability Tests
The type of reliability test which is normally carried out on pre-production model concerns
mainly reliability measurement. Reliability measurements are normally embarked upon when it
is necessary to either estimate a reliability figure itself or determine the failure rate/MTBF. We
may wish to recall that both failure rate and MTBF are directly related to reliability as fully
discussed under reliability.
Maintainability Test
Maintainability test will be necessary in checking the relative speed which it takes the same
repair to be carried out on two alternative designs. This needs further explanation. If, for
example, we have two different designs of a particular item we will realize that each design will
imply different repair speed, depending on the nature of the design. This parameter is
particularly important in view of the desired availability of the product.
Packaging and Transport Test
It may make little sense in spending a lot in design and manufacture of a product without
adequate consideration for packaging and transport hazards. This explains why packaging and
transport test is essential. In a nutshell, this test is aimed at identifying potential causes of
defects during packaging, storage and transport. This test may involve vibration and shock
tests and also storage test.
Physical and Electrical Characteristics Test
Physical characteristics test is aimed at determining the physical dimensions and weight of a
product whereas electrical characteristics tests sets to determine the range of operating mains
power supply voltage or d.c. voltage, and frequency. Some devices are designed to operate on
a.c. 100-110V, 115 to 127V, 2000220V 230-250V, 50/60 HZ, etc. Power (mains) consumption
may also be determined in such a test.
Ergonomics Test
This test is aimed at determining the effects of interface of a pre-production model with
operators and maintenance personnel. Such test may reveal possible frequency of operator
mistakes or the time required by the operator to carry out an operation on an equipment. It
should be mentioned that the complexity of an equipment design, number and location of
operating switches and knobs constitute an important factor in determining an operator’s
convenience and accuracy in carrying out his job.

3.4 Production Testing

Production testing is designed to verify the conformity of a finished product/equipment with
specification. This test may involve some reliability proving and determination of possible
causes of early or burn-in failure. At this stage, failures may be attributed to component defect,
production methods etc. Design related problems may also arise but it is expected to have
been detected and corrected either at the prototype or the pre-production stage. Since tests
carried out during the pre-production stage embrace virtually all parts of the engineering
specification and take considerable length of time to complete, therefore, it will be up to the
engineering, inspection and/or quality departments to decide which important parts of the
specifications to check during and after production.

19
 3.5 Reliability Demonstration and Acceptance Testing
As mentioned earlier in Section 3.1, there exists what is called reliability demonstration and
acceptance testing associated with the post-production stage. It should be reiterated that this
is not a mandatory exercise, and when it happens it will be an affairs between the
manufacturer and the customer.
Reliability demonstration test is a test which is normally carried out when a manufacturer,
having produced certain equipment, wishes to demonstrate to a customer that its reliability is
at least as good as he claims it to be. Conversely, a customer may receive a finished product
untested and he decides to test it himself to verify whether or not the product satisfies the
reliability figure as stated in the specification. In this case the customer will be said to be doing
what is called a reliability acceptance test.

3.6 Accelerated Life Testing

Once a practical design has been developed, reviewed, and approved, a number of prototypes
systems are produced. These are normally subjected to exhaustive accelerated testing.
Accelerated life tests are usually employed in order to reduce the very long test times which are
often necessary. The underlying idea is that the normal life cycle of the item concerned should
be compressed into a very much shorter time for the purpose of the test. The intention is to
compress the effects of many years of normal use into a few weeks by operating the unit at
higher than normal uses. If properly carried, out, accelerated test results can be shown to
correlate reasonably well with actual reliability results under normal conditions. Temperature
and humidity combinations, shock and vibration are among the environments to which the
devices are exposed. It, therefore, follows that accelerated life test cannot be easily separated
from environmental testing, already discussed in section 3.3.

Accelerated life tests can take the following two main forms:
(i) Tests which may be carried out under normal use conditions except that what would in
practice be intermittent use now becomes a continuous use; and
(ii) Tests where the normal conditions of use are intensified, although it may be under
intermittent use.
Suppose we want to test a batch of microprocessor chips. In this case, the test engineer
operates the microprocessor beyond the normal limits specified for voltage, temperature, and
clock rate to test for operating margins and to expose latent failures. The stresses caused by
testing the device at excessive conditions generate information which are helpful in predicting
reliability over the devices’ lifetime. However, this type of test must always be carried out with
considerable caution.

3.7 Relationship Between Testing and Inspection

Various types of tests that can be carried out at different production stages have been
thoroughly discussed. However, there is a tendency to confuse inspection with testing unless
an effort is made to differentiate them. Testing normally takes place at the conclusion of each
production stage, after prototype production, after pre-production stage and after final
production stage. The attributes that may be tested during each stage have been exhaustively
discussed.

20
However, the inspection process may take place before, during and after each production
stage. Inspection is usually directed at detecting defective items. An item is said to be
defective if it fails to conform with the pre-stated specification. In any industrial process there
are a number of places where inspection can and should be done. It can take place at the input
(when inspection of raw materials or output of previous stage is involved), at the output (to
reduce the risk of bad quality being passed on and causing customers’ complaints), and finally it
may involve inspection of the process itself, sometimes at several points, and checking up how
it is working. The aims of inspection are mainly two-fold. Firstly, it may provide information
that may necessitate the adjustment of the production process in order to reduce the amount
of bad production. Secondly, it may be necessary to sort out the bad production and return
items for re-processing.
Ideally, inspection should extend to each produced item for the entire attribute. It is in general
impracticable due to the following reasons:-
(i) The inspection test may be destructive. An example is testing which involves bending
strength of transistor leads.
(ii) Inspection may take too much time and may make the product uneconomical.
(iii) Insufficiency of inspection personnel that may be needed to carry out inspection of
individual item.
(iv) Individual inspection of all produced items may even fail to detect all the defective ones,
more so if the border line cases are large. This may occur if there is either a calibration
problem with the testing equipment or operators’ tiredness.

3.8 Relationship Between Quality and Reliability

Quality can be defined as conformity to specification or fitness for purpose, and it can be sub-
divided roughly into two:
(i) Physical features: Provides information as to whether an item had dimension within
acceptable limits, and has a satisfactory appearance.
(ii) Performance: Provides information on whether an item works correctly. This is an
attribute which may not be easy to observe in figures.
Reliability was defined earlier as the probability that an item will perform a required function,
under stated conditions for a stated period of time. One the other hand, as we have just seen
performance is also an aspect of quality; they are related somehow although they are not
synonymous.
It may be argued that quality is a static feature and reliability a dynamic one. For example,
electronic equipment may be of high quality, but its reliability is measurable only in use.
Quality control can be seen as a major subset of reliability engineering. It encompasses such
activities as testing and selection of materials used in the manufacturing process. It also has to
do with the design and control of processing and assembling parts and equipment. Finally, it
also includes testing and inspection. In contrast, reliability tends to focus on design. Reliability
is concerned with the probability of failure at some instant in time, and, therefore, has its own
mathematics which is different from that in use by quality personnel.

3.9 Sampling Plans (Approaches for Testing Small and Large Batch Quantities)
Concept of Sampling Plan

21
We have discussed in Section 3.7 why it may be unreasonable to carry out inspection test on
individually produced items (especially when the quantity involved is large) unless absolutely
defect-free items are needed and we are ready to pay the cost. In practice, instead of
embarking on inspection testing of each item, some sort of acceptance sampling plan is always
adopted. The underlying idea is to reject the whole lot of the produced items based on
information received from a sample. As an example, if a batch size ‘N’ is taken for inspection a
random sampling is employed to collect ‘n’ samples for test. These samples are then tested
according to the prescribed and pre-decided sampling plans. The main aim of the sampling
plan is to provide criteria for deciding the lot acceptability in the light of the number found
defective from the sample. The point will be made clearer in the succeeding sub-stations
treated immediately below.
Single Sampling Plans
The single sampling plan is concerned with the selection of a sample of a given size (say, 50
resistors) from a lot of a known size (say, 1000 resistors) and check against its known
specifications may be, its resistance value, tolerance etc). The allowable number of defective
items is specified in the sampling plan depending on the risks both the customer and producer
are ready to accept. If the number of defective items (resistors) is less than or equal to that
agreed in the sampling plan, then, the lot is accepted (i.e. all the 1000 resistors will be assumed
alright), otherwise the lot is rejected.
For example, we can specify that a batch of 1000 transistors is to be inspected at 1 percent
A.Q.L. (Acceptable Quality Level). (A.Q.L is defined as the maximum percent of defectives
which can be considered satisfactory as a process average for the purpose of sampling
inspection). We may take a sample of ninety, and accept the batch if the number of defectives
does not exceed say, two.
Double Sampling Plans
In a double sampling plan, there are two sample sizes, and two allowed numbers of defectives.
An example of a double sampling plan may follow the pattern given below:

First sample =120, Accept number = 1, Reject number = 4

Second sample = 120, Accept number = 4, Reject number = 6

This means that a sample of one hundred and twenty should be taken and inspected. If the
number of defective items discovered does not exceed one, accept the batch on the first
sample. It will then be needless to take the second sample. However, if the number of
defective items is four or more reject the batch on the first sample. If the number of defective
items is more than one, but less than four, take a second sample of one hundred and twenty.
Now if the number of defectives from the second sample does not exceed four, accept the
batch, otherwise reject it.
Multiple and Sequential Sampling Plans
By extending the principle of double sampling plans to reach any desired number of stages then
we shall have what are called multiple sampling plans. In this case, the number of defective
items for each sample is normally specified in the sampling plan. It is worthy of note that as the
number of samples required to arrive at a decision increases, so the size of each sample must
22
decrease. Ultimately, a limit is reached with what is called ‘sequential sampling’, where the
sample size is one. The results are normally plotted on a graph, and the method involves taking
a sample of one, testing it and plotting the result. This procedure is then repeated, taking
successive samples of one, until the graph contains enough information on which to arrive at a
decision.

23