Chapter Five

Life Testing and Reliability

1
 Life Testing
• Life testing of products is the most essential
requirement in certain industries specially those
manufacturing electric tubes, razor blades etc.
• Life testing is performed so as to ensure whether
the product is as per specification or not. While
manufacturing it is essential to test the
component as well as the operations
• Life test ensures the manufacturer that the
product shall give good performance during its
guaranteed life.
2
 Classification of Life tests

a) Quality Control Life Test. These tests are for checking

the quality
b) Pilot-run Life Tests. The life tests are performed when
any change in the production is proposed, that is,
these are to check the proposed production.
c) Establishment Life Tests. These life tests are performed
whenever a new type of product is introduced.
d) Application Life Tests. When conditions for the use of
product are different than that for which it is generally
used, then these life tests are performed.
3
 Sample or Individual Life-Testing
• Whether a product is to be tested for life –test individually
or by sampling depends upon their importance.
• For very important products such as radar, transmitting
tubes where high reliability, greater stability and long life is
required, individual product must be tested for life-tests.
• Products of less importance such as electric bulb, fans,
electric press, radio receiving tubes etc. are put to sample
testing for life-tests.

4
 Failure Data Analysis
• When a component or a system does not
perform work satisfactorily, it is said to have
failed.
• The pattern of failure can be obtained from life
test results by testing a fairly large number of
items and observing the failure-rate
characteristics as a function of time.
• For example, tests were conducted on 1000
electric components. The number of
components that fail during each hourly interval
up to 5 hrs noted in the table given below. 5
Duration Number of Number of Failure Failure Rate Reliability
Failures Survivors Density

1 2 3 4 5 6

0-1 hrs 60 940 0.06 60/970 = 0.0619 0.94

1-2 hrs 40 900 0.04 40/920 = 0.043 0.90

2-3 hrs 30 870 0.03 30/885 = 0.034 0.87

3-4 hrs 40 830 0.04 40/850 =0.047 0.83

4-5 hrs 50 780 0.05 50/805 = 0.062 0.73

6
• From the table above, we see that:

• Failure density is the ratio of number of failures during a given

interval of time to the total number of items at the very
beginning of the test.
• Failure rate is the ratio of the number of failures during a
particular unit interval to the average means the average of
populations at the beginning and at the end of interval. Failure
rate is also known as hazard rate.
• Reliability is the ratio of survivors at any given time to the
total initial population. Reliability factors can also be called as
7
the probability of survival.
• Probability of Failure is the ratio of the number of
units failed (during a specified time ) to the total
population.
• In this example, probability of failure during the first
hour is 60/1000 = 0.06, and during first two hours is
(60+40)/1000 =0.01. therefore probability of failure
during first five hours is (60+40+30+40+50)/1000 =
220/1000 = 0.22. This can also be termed as
unreliability factor. Here it is important to mention
that Reliability factor + unreliability factor =1 8
 Mean Time To Failure (MTTF)
• As all the specimens do not fail at the same time,
they have different times to failure. Hence we are
required to calculate the mean time to failure of
all the specimens. If t1 is the time to failure for
the first and t2 is the time for the second
specimen, and tN is the time to failure for the
Nth specimen, the MTTF for N specimen will be

9
• Mean Time Between Failure (MTBF). It is the
average value of time intervals between
successive failure of a product or an equipment

Thus MTBF = equipment operating time/ No. of

observed Failures

10
 Reliability
• Reliability is a characteristic of an item, expressed by the
probability that the item will perform its required
function under given conditions for a stated time interval.
• From a qualitative point of view, reliability can be defined
as the ability of the item to remain functional.
• Quantitatively, reliability specifies the probability that no
operational interruptions will occur during a stated time
interval.
11
• This does not mean that redundant parts may not
fail such parts can fail and be repaired (without
operational interruption at item (system) level).
• The concept of reliability thus applies to
nonrepairable as well as to repairable items.
• To make sense reliability must be accompanied by
the definition of the required function, the
operating conditions, and the mission duration.
12
• Many formal definitions have been proposed
that are similar in their general intent but differ a
bit in their exact phrasing. Some of these are as
follows:
• Reliability is the probability of a device
performing its purpose adequately for the period
of the intended under the operating conditions
encountered.
13
• The reliability of a (system, device, etc.) is the
probability that it will give satisfactory
performance for a specified period of time under
specified operation condition.
• Reliability is the capability of an equipment not to
break down in operation. When an equipment
works well, and works whenever called upon to do
the job for which it was designed, such equipment
is said to be reliable. 14
• Reliability, therefore, provides a numerical measure
of “degree of excellence” through time.
• It is a facet of quality that, ideally, works at the
interface between design and specification.
• Experts in reliability are engaged in giving advice on
improving the “degree of excellence” by preventing
design errors to the extent that they are successful,
they are engaged in breaking down the barriers
between design and specification. 15
• In studies of human mortality, usually there is no
question of the moment of death of an individual.
• Similarly, in the types of mortality studies of
physical property made in connection with the
requirements of depreciation accounting, the
moment of retirement.
• In contrast, where the life of a manufactured
product is tested for purposes of acceptance
inspection. 16
• Elaborate test apparatus may be required to find
the exact moment at which the performance
ceases to be satisfactory.
• Moreover, specifications on what constitutes
satisfactory performance naturally depend on the
use to which the product is to be put.
• For example, a particular electronic device might
have a short life under one set of specifications
and a long life under another set. 17
• Although there are exceptions, when a simple
product fails (in the sense that its performance
no longer conforms to specification), the product
usually cannot be restored to its original
condition of giving satisfactory performance.
• Generally speaking, the failure of a component is
viewed as terminating its life.

18
• When a complex manufactured device fails, it
often may be possible to restore the device to its
original satisfactory performance, possible by the
replacement of one or more components.
• Consider that, say, 100 components are being
tested with a new component being substituted
whenever one fails.

19
• Even though the probability of a failure is constant
for any stated time interval, the actual number of
failures in successive equal time intervals will be
subject to chance fluctuations.
• Even though life may be renewed in this way, a
relevant question regarding any such complex
device is how long a time may be expected to
elapse between successive failures.
20
• Consider the following lot-by-lot acceptance sampling
plan.
• Select 22 items at random form a lot. Place these items
on test.
• Whenever an item fails, replace it with another item
selected at random from the lot.
• If the test continues for 500 h with not more than 2
failures, accept the lot. If 3 failures occur before the
500 h of testing, reject the lot and terminate the test.21
• Acceptance under this plan requires 22 (500) =
11,000 item hours of test with an acceptance
number of 2.
• Under the assumption that the probability of a
failure is the same for every item hours, the
calculation of the OC curve (Table 8.1) is the
same as if we had an ordinary single sampling
attributes plan with an n of 11,000 an c of 2.
22
• Our sampling plan has maximum of 500 h of
testing for the 22 items.
• However, it stipulated that testing should end
when the rejection number of 3 was reached. If
the third failure occurs after, say, 140h, no more
testing takes place; the decision on the lot is
reached with only 3,080 item hours of test rather
than with the stipulated maximum of 11,000 item
hours. 23