DESIGN FOR

RELIABILITY
Carandang, Monneth E.
Comia, Marc Edsyl R.
Gabog, Joshua D.
Mendoza, Hannah Kristina V.
Ramirez, Marites P.
Ramirez. Sharlene Mae
WHAT IS RELIABILITY?

The ability of a system to perform its intended

mission when operating for a designated period of
time, or through a planned mission scenario (or series
of scenarios), in a realistic operational environment.
The system, when operating in a true user’s
environment, is expected to be able to satisfy all of
the operational objectives desired and specified by
the customer.
 Probability
Satisfactory Performance
DEFINITION AND Time
Specified Operating Conditions
EXPLANATION
 THE RELIABILITY FUNCTION
R (t) = 1 - F(t)
Where:
R(t) = reliability function or aka (survival function)
F(t) = unreliability function
t = time
θ = Mean life
λ = instantaneous failure rate
M = mean time between failure
 THE FAILURE RATE
λ = number of failures/total operating hours

Example:
Failure rate may be expressed in terms of failures per hour, percentage of
failures per 1,000 hours, or failures per million hours. As an example, suppose
that 10 components were tested for 600 hours under specified operating
conditions. The components (which are not repairable) failed as follows:
component 1 failed after 75 hours, component 2 failed after 125 hours,
component 3 failed after 130 hours, component 4 failed after 325 hours, and
component 5 failed after 525 hours. Thus, there were five failures and the total
operating time was 4,180 hours.
 THE FAILURE RATE
Another example
The operating cycle for a given system is 169 hours. During that time, six
failures occur at the points indicated.A failure is defined as an instance when
the system is not operating within a specified set of parameters. Used the
figure below in computation:
Failure Rate (λ) = Number of failures / total mission time
Mission Time = Operating Time - Downtime
Where:
# of Failure = 6
Operating Time = 169
downtime = 2.1 + 7.1 + 4.2+ 1.8 + 3.5 + 8.3 = 27

λ = 0.04225

Mean time between Failure = 1 / 0.04225

A SYSTEM OPERATIONAL CYCLE. MTBF = 23.67 hrs
THE FAILURE RATE

TYPICAL FAILURE-RATE CURVE

RELATIONSHIPS.
THE FAILURE RATE

FAILURE-RATE CURVE WITH MAINTENANCE

(SOFTWAREAPPLICATION)
 THE COMPONENT RELATIONSHIP

SERIES NETWORKS

All components must operate in a satisfactory manner if the system is to function properly.
Assuming that a system includes subsystems A–C, the reliability of the system is the product
of the reliabilities for the individual subsystems expressed as

An electronic system includes a transmitter, a

receiver, and a power supply. The transmitter
reliability is 0.8521, the receiver reliability is 0.9712,
and the power supply reliability is 0.9357.
R = (0.8521)(0.9712)(0.9357) = 0.7743
 THE COMPONENT RELATIONSHIP

SERIES NETWORKS

If a series system is expected to operate for a

specified time period, its required overall reliability
can be derived.

OR CAN BE EXPRESSED AS
 THE COMPONENT RELATIONSHIP

SERIES NETWORKS

Suppose that a series system consists of four

subsystems and is expected to operate for 1,000
hours. The four subsystems have the following
MTBFs: subsystem A, MTBF = 6000 hrs; subsystem
C, MTBF = 10 500 hrs; subsystem D, MTBF = 3200
hrs. The objective is to determined the overall
reliability of the series network.

λ(a) = 1/6000 = 0.000167 failure/hour -(0.000797)(1,000)

λ(b) = 1/4500 = 0.000222 failure/hour R=e
λ(c) = 1/10500 = 0.000095 failure/hour
λ(d) = 1/3200 = 0.000313 failure/hour = 0.4507
 THE COMPONENT RELATIONSHIP

PARALLEL NETWORKS

one where several of the same components are in parallel and where all the components
must fail to cause total system failure. Assuming that components A and B are identical, the
system will function if either A or B, or both, are working.

R = RA + RB - (RA)(RB)
Consider next a network with three components in parallel. The network reliability is expressed as:
3
R = 1 - (1 - RA)(1 - RB)(1 - RC) OR R = 1 - (1 - R)
For a system with n identical components, the reliability is;
n
R = 1 - (1 - R)
 THE COMPONENT RELATIONSHIP

PARALLEL NETWORKS

A system includes two identical subsystems in

parallel and the reliability of each subsystem is
0.95.

R = 0.95 + 0.95 - (0.95)(0.95) = 0.9975

OR
2
R = 1 - (1 - 0.95) = 0.9975
 THE COMPONENT RELATIONSHIP

COMBINED SERIES–PARALLEL NETWORKS

Various levels of reliability can be achieved

through the application of a combination of
series and parallel networks.

Ra = RA(R B+ RC- RBRC )

 THE COMPONENT RELATIONSHIP

COMBINED SERIES–PARALLEL NETWORKS

R = [ 1 - (1 -R ) (1 - R )] [1 - (1 - R )(1 - R )]
b A B C D
 THE COMPONENT RELATIONSHIP
COMBINED SERIES–PARALLEL NETWORKS

Rc = [1 - (1 - RA)(1 - RB)(1 - RC)][RD ][RE+ R F- (R E)(RF )]

SHARLENE
RELIABILITY IN THE
SYSTEM LIFE CYCLE
 RELIABILITY IN THE SYSTEM
LIFE CYCLE
Reliability, as an inherent characteristic of design, must be addressed in the
overall systems engineering process beginning in the conceptual design phase.
Qualitative and quantitative reliability requirements are developed through
the accomplishment of feasibility analysis, the development of operational
requirements and the maintenance concept, and the identification and
prioritization of TPMs.
Figure 11. Reliability requirements in the system life cycle.
 SYSTEM REQUIREMENTS
Reliability requirements, specified both in quantitative and
qualitative terms, are defined within the context of the system
operational requirements and the maintenance concept.
This includes the following:
1. Definition of system performance and effectiveness factors, mission
profile(s), and system functional requirements.
2. Definition of the operational life cycle.
3. Definition of the environment in which the system is expected to operate
and be maintained.
4. Definition of the operational and supporting interfaces likely to impact
the system as it performs its mission(s) throughout its planned life cycle.
What reliability should the system have (exhibit) in
order to accomplish its intended mission
successfully, throughout the specified life cycle, and
in the environment defined?
 RELIABILITY MODELS
The system is defined, in “functional” terms, through accomplishment of the
functional analysis and the development of functional flow block diagrams.

Figure 12. Expanded reliability

model of system
Series–parallel relationships are established and developed further as
the design definition evolves.
The reliability model serves as a baseline for the identification of
possible weak areas and where possible design improvements can be
introduced.
 RELIABILITY ALLOCATION

Top-level reliability requirements are specified for the system, and these
requirements are then allocated to subsystem level, unit level, and down
to the level needed to provide a meaningful input to the design
 Referring to Figure 13, it is assumed that
at the system level there is an MTBF
requirement of 450 hours.

What should be specified as a reliability

requirement for Units A, B, and C?

Figure 13. Allocation of reliability requirements.

In response, the following steps are appropriate:
1. Identify the elements of the system where the design is known
and where reliability data are available or can be readily
assessed.
2. Identify the areas that are new and where design information is
not available.
 COMPONENT SELECTION AND
APPLICATION

The reliability of a system depends largely on the reliability of its component

parts, and the selection of parts must be compatible with the requirements of
the particular application of those parts.
The specific application of the component is of prime importance, particularly
when considering factors such as part tolerances and drift characteristics,
electrical and environmental stresses, and so on.
 Major emphasis in the design for reliability should consider
several factors:
1. The selection of standardized components and materials to
the greatest extent possible.
2. The test and evaluation of all components and materials
prior to design acceptance.

The challenge in design is to select and utilize only those

components that, when combined, are capable of meeting the
overall reliability requirements for the system.
 REDUNDANCY IN DESIGN
Under certain conditions in system design it may be necessary to
consider the use of redundancy to enhance system reliability by
providing two or more functional paths (or channels of operation) in
areas that are critical for successful mission accomplishment.
However, the application of redundancy per se will not necessarily
solve all problems but the use of redundancy may be the only solution
for reliability improvement in specific situations.
Form of Redundancy
1. Operating redundancy
2. Standby redundancy
Figure 14. Reliability block diagram illustrating
redundancy at system and subsystem levels.
 1. Subsystem A, B, and C are all operating.
2. Subsystems A and B are operating while
C is failed.
3. Subsystems A and C are operating while
B is failed.
4. Subsystems B and C are operating while
A is failed.
5. Subsystem A is operating while B and C
are failed.
6. Subsystem B is operating while A and C
are failed.
7. Subsystem C is operating while A and B
are failed.
Figure 9. Parallel network with three components.
8. Subsystems A, B, and C are all failed.
In the interest of simplicity, let R represent the reliability
and the unreliability. Then,
Sum of and represents all possible events,

It is assumed that the reliability of each block is 0.95.

Reliability of each of the individual
subsystems is as follows:
1. Subsystem G = 0.99
2. Subsystem F = 0.90
Figure 14. Reliability block diagram illustrating
3. Subsystem E = 0.93
redundancy at system and subsystem levels.
4. Subsystem D = 0.92
5. Subsystem C = 0.92
6. Subsystem B = 0.98
7. Subsystem A = 0.97
 The reliability of the redundant network
including subsystems C and D is

Reliability of each of the individual

subsystems is as follows: The reliability of the redundant network
1. Subsystem G = 0.99
including subsystems E and F is
2. Subsystem F = 0.90
3. Subsystem E = 0.93
4. Subsystem D = 0.92
5. Subsystem C = 0.92
6. Subsystem B = 0.98
7. Subsystem A = 0.97
 The reliability of the path is

Reliability of each of the individual

subsystems is as follows: The reliability of the combined network in
1. Subsystem G = 0.99
Figure 14 is
2. Subsystem F = 0.90
3. Subsystem E = 0.93
4. Subsystem D = 0.92
5. Subsystem C = 0.92
6. Subsystem B = 0.98
7. Subsystem A = 0.97
Figure 15. Standby redundant network.
 When determining the reliability of standby systems, the
Poisson distribution may be used.

- The probability of no failure

- The probability of one failure
When one operating subsystem and one standby subsystem are grouped
together, one must consider the probability that no failure or one failure will
occur.

where λt is the expected number of failures

If one operating subsystem and two standby subsystems are
grouped together, then the combined probability is
Suppose that one must determine the system reliability
for a configuration consisting of one operating
subsystem and one identical standby operating for a
period of 200 hours. This configuration is illustrated in
Figure 15, and it is assumed that the reliability of the
switch is 100%. The failure rate (λ) for each
subsystem is 0.002 failure per hour.
 To illustrate the difference between operating redundancy
and standby redundancy, assume that both of the
subsystems in Figure 15 are operating throughout the
mission.

As anticipated, the reliability of the standby system is higher

(0.9384) than the reliability of the system using operating
redundancy (0.8913).
 DESIGN REVIEW AND EVALUATION

The design configuration and the characteristics of the various system

components are evaluated in terms of compliance with the initially
specified reliability requirements for the system.
Some key areas of interest through the questions in a design review checklist:
1. Have reliability quantitative and qualitative requirements for the system been
adequately defined from the beginning?
2. Have these requirements been properly allocated to the various subsystems as
applicable?
 3. Are the reliability requirements realistic? Are they compatible
with other system requirements?
4. Has system design complexity been minimized?
5. Have system failure modes and effects been identified?
6. Are system, subsystem, unit, and component-part failure rates
known?
7. Are the failure characteristics known for each applicable
component part?
8. Has the system or product wearout period been defined?
9.Have component parts with excessive failure rates been
identified?
 10. Have all critical-useful-life items been identified and
eliminated where possible?
11. Have fail-safe characteristics been incorporated where
applicable?
12.Has the utilization of adjustable components been minimized
(if not eliminated)?
13. Have cooling provisions been incorporated in design “hot-spot”
areas?
14. Have all hazardous conditions been eliminated?
15. Have all system reliability requirements been met?
JOSHUA
RELIABILITY
ANALYSIS
METHODS
 Failure Mode, Effects, and Criticality
Analysis (FMECA)

The failure mode, effects, and criticality analysis (FMECA) is a design technique
that can be applied to identify and investigate potential system (product or process)
weaknesses. It includes the necessary steps for examining all ways in which a system
failure can occur, the potential effects of failure on system performance and safety,
and the seriousness of these effects.
1.Define system (product or process) requirements.
Describe the system in question, the expected outcomes, and the relevant technical
performance measures (TPMs). Figure 18 shows an example where the FMECA can be
applied to the manufacturing.

2. Accomplish functional analysis.

This involves defining the system in functional terms. A system may be broken down into
functional entities early in the life cycle and subsequently into a physical packaging
scheme

3. Accomplish requirements allocation.

This is a top-down breakout of system-level requirements.

4. Identify failure modes.

A “failure mode” is the manner in which a system element fails to accomplish its function. For
example, a switch may fail in an “open” position; a pipe may “rupture;” a given material may
“shear” because of stress; a document may fail to be delivered on time; and so on.

5. Determine causes of failure.

This involves analyzing the process or product to determine the actual cause(s) responsible for the
occurrence of a failure.

6. Determine the effects of failure.

Failures impact, often in multiple ways, the performance, and effectiveness of not only the
associated functional element but also the overall system.
7. Identify failure detection means.
For a process-oriented FMECA, it refers to the current process controls which may detect the
occurrence of failures or defects.

8. Rate failure mode severity.

This refers to the seriousness of the effect or impact of a particular failure.

9. Rate failure mode frequency.

Given that a function or physical component within the system may fail in a variety of ways, this
step addresses the frequency of occurrence of each individual failure mode.
10. Rate failure mode detection probability.
This pertains to the probability that process controls, design features/aids, verification procedures,
and so on, will detect potential failures in time to prevent a major system catastrophe.

11. Analyze failure mode criticality.

The objective is to consolidate the preceding information in an effort to delineate the more critical
aspects of system design.

RPN = ( severity rating)( frequency rating)( probability of detection rating)

12. Initiate recommendations for product/process improvement.
This pertains to the iterative process of identifying areas with high RPNs and evaluating the
causes, and the subsequent initiation of recommendations for product/process improvement.
 Fault-Tree Analysis (FTA)
The fault-tree analysis (FTA) is a deductive approach involving
the graphical enumeration and analysis of different ways in
which a particular failure can occur and the probability of its
occurrence. It may be applied during the early stages of design, is
oriented to specific failure modes, and is developed using a top-
down fault-tree structure. A separate fault tree is developed for
every critical failure mode.
 Stress–Strength Analysis
Of major concern in the design of the system, reliability is the stress
and strength characteristics of its components. Component parts are
designed and manufactured to operate in a specified manner when
utilized under nominal conditions. If additional stresses are imposed
because of electrical loads, temperature, vibration, shock, humidity,
and so on, then unexpected failures will occur and the reliability of the
system will be less than anticipated.
 A stress–strength analysis is often undertaken to evaluate the probability of identifying
a situation(s) where the value of stress is much larger than (or the strength much less than) the
nominal value. Such an analysis may be accomplished through the following steps:
1.For selected components, determine nominal stresses as a function of loads, temperature,
vibration, shock, physical properties, and time.
2.Identify factors affecting maximum stress, such as stress concentration factors, static and
dynamic load factors, stresses as a result of manufacturing and heat treating, environmental
stress factors, and so on.
3.Identify critical stress components and calculate critical mean stresses (e.g., maximum tensile
stress and shear stress).
4.Determine critical stress distributions for the specified useful life. Analyze the distribution
parameters and identify component safety margins. Applicable distributions may include
normal, Poisson, Gamma, Weibull, log-normal, or variations thereof.
5.For those components that are critical and where the design safety margins are inadequate,
corrective action must be initiated. This may constitute component-part substitution or a
complete redesign of the system element in question.
 Reliability Prediction
As engineering data become available, reliability
prediction is accomplished as a check on design in
terms of the system requirement and the factors
specified through allocation. The predicted values of R,
MTBF, and/or MTTF are compared against the
requirement, and areas of incompatibility are evaluated
for possible design improvement.

 Predictions are accomplished at different times in the system

development process and will vary somewhat depending on the type
of data available. Basic prediction techniques are summarized as
follows:
1.Prediction may be based on the analysis of similar equipment.
2.Prediction may be based on an estimate of active element groups
(AEG).
3.Prediction may be accomplished from an equipment parts count.
4.Prediction may be based on a stress analysis (discussed earlier).

 Reliability Growth Analysis

In response, a formalized plan needs to be implemented
such that reliability growth will occur in order to correct
the deficiency as early in the design process as
practicable, as compared to waiting until the system test
and evaluation phase and then discovering that the
design will not meet the specified requirements.

 RELIABILITY
TEST AND
EVALUATION
 Reliability Sequential Qualification Testing
Reliability qualification testing is conducted to provide an evaluation
of system development progress, as well as the assurance that specified
requirements have been met prior to proceeding to the next phase (i.e.,
the production or construction phase of life cycle). Initially, a reliability
MTBF is established for the system, followed by allocation and the
development of design criteria. System design is accomplished and
reliability analyses and predictions are accomplished to evaluate (on an
analytical basis) the design configuration relative to compliance with
system requirements.

 Reliability Acceptance Testing

As indicated, reliability testing may be accomplished as
part of qualification testing prior to commencing with full-
scale production (which has been discussed) and during full-
scale production on a sampling basis. To determine the
effects of the production process on system reliability, it
may be feasible to select a sample number of
equipment/units or products from each production lot.
 Reliability Life Testing
There are two basic forms of life testing in use: (1) life
tests based on a fixed-test time, and (2) life tests based on the
occurrence of a predetermined number of failures. The first
approach to life testing (based on time) assumes that a
computed fixed test time will be planned and a specified
quantity of failures will be predetermined. System
acceptance occurs if the actual number of failures during test
is equal to (or less than) the predetermined quantity of
failures at the end of the scheduled test time.
 Operational Reliability Assessment
The measure of system reliability has been based on a
combination of analytical studies, predictions, and the
demonstration of prototype models of various system
elements (e.g., equipment). The opportunity to observe the
system being utilized in a realistic user environment has not
been possible. Given the successful completion of all required
reliability qualification and acceptance tests, the next step is to
accomplish a true assessment of system reliability operating in
the user’s environment.