Practical Reliability Engineering 5th Edition by Patrick D. T. O’Connor, ISBN-13: 978-0470979815

Description

Practical Reliability Engineering 5th Edition by Patrick D. T. O’Connor, ISBN-13: 978-0470979815

[PDF eBook eTextbook] – Available Instantly

• Publisher: ‎ Wiley; 5th edition (January 30, 2012)
• Language: ‎ English
• 512 pages
• ISBN-10: ‎ 047097981X
• ISBN-13: ‎ 978-0470979815

Fully revised edition of the best-selling book, that presents a good balance of reliability mathematics, engineering methods and practices, providing an up-to-date overview for engineering students and reliability engineers alike.

With emphasis on practical aspects of engineering, this bestseller has gained worldwide recognition through progressive editions as the essential reliability textbook. This fifth edition retains the unique balanced mixture of reliability theory and applications, thoroughly updated with the latest industry best practices.

Practical Reliability Engineering fulfills the requirements of the Certified Reliability Engineer curriculum of the American Society for Quality (ASQ). Each chapter is supported by practice questions, and a solutions manual is available to course tutors via the companion website.

Enhanced coverage of mathematics of reliability, physics of failure, graphical and software methods of failure data analysis, reliability prediction and modelling, design for reliability and safety as well as management and economics of reliability programmes ensures continued relevance to all quality assurance and reliability courses.

Notable additions include:

• New chapters on applications of Monte Carlo simulation methods and reliability demonstration methods.
• Software applications of statistical methods, including probability plotting and a wider use of common software tools.
• More detailed descriptions of reliability prediction methods.
• Comprehensive treatment of accelerated test data analysis and warranty data analysis.
• Revised and expanded end-of-chapter tutorial sections to advance students’ practical knowledge.

The fifth edition will appeal to a wide range of readers from college students to seasoned engineering professionals involved in the design, development, manufacture and maintenance of reliable engineering products and systems.

Table of Contents:

Front Matter

Dedication

Preface to the First Edition

Preface to the Second Edition

Preface to the Third Edition

Preface to the Third Edition Revised

Preface to the Fourth Edition

Preface to the Fifth Edition

Acknowledgements

1 Introduction to Reliability Engineering

1.1 What is Reliability Engineering?

1.2 Why Teach Reliability Engineering?

Figure 1.1 Perception of risk.

1.3 Why Do Engineering Products Fail?

Figure 1.2 Load–strength – discrete values.

Figure 1.3 Load–strength – distributed values.

Figure 1.4 Load–strength – interfering distributions.

Figure 1.5 Time-dependent load and strength variation.

1.4 Probabilistic Reliability

1.5 Repairable and Non-Repairable Items

1.6 The Pattern of Failures with Time (Non-Repairable Items)

Figure 1.6 The ‘bathtub’ curve.

1.7 The Pattern of Failures with Time (Repairable Items)

1.8 The Development of Reliability Engineering

1.9 Courses, Conferences and Literature

1.10 Organizations Involved in Reliability Work

1.11 Reliability as an Effectiveness Parameter

1.12 Reliability Programme Activities

1.13 Reliability Economics and Management

Figure 1.7 Reliability and life cycle costs (traditional view).

Figure 1.8 Reliability/Quality and life cycle costs.

Figure 1.9 Reliability and life cycle costs (practical applications).

Questions

Bibliography

Periodic Publications

2 Reliability Mathematics

2.1 Introduction

2.2 Variation

2.2.1 A Cautionary Note

2.3 Probability Concepts

Figure 2.1 Samples with defectives (black squares).

2.4 Rules of Probability

Figure 2.2 Dual redundant system.

Example 2.1

Example 2.2

Example 2.3

Example 2.4

2.5 Continuous Variation

Figure 2.3 (a) Frequency histogram of a random sample, (b) frequency histogram of another random sample from the same population, (c) data of many samples shown with measurement intervals of 0.5.

Figure 2.4 Continuous probability distribution.

Figure 2.5 Measures of central tendency.

2.5.1 Measures of Central Tendency

2.5.2 Spread of a Distribution

2.5.3 The Cumulative Distribution Function

Figure 2.6 Typical cumulative distribution function (cdf).

2.5.4 Reliability and Hazard Functions

2.5.5 Calculating Reliability Using Microsoft Excel® Functions

Figure 2.7 Probability Density Function (pdf) and its application to reliability.

2.6 Continuous Distribution Functions

2.6.1 The Normal (or Gaussian) Distribution

Figure 2.8 The normal (Gaussian) distribution.

Example 2.5

Figure 2.9 (a) The pdf f(x) versus x; (b) the cdf F(x) versus x (see Example 2.5).

2.6.2 The Lognormal Distribution

2.6.3 The Exponential Distribution

2.6.4 The Gamma Distribution

2.6.5 The χ2 Distribution

2.6.6 The Weibull Distribution

2.6.7 The Extreme Value Distributions

Table 2.1 Sample data taken randomly from a common population.

Figure 2.10 Extreme value distributions.

2.6.7.1 Extreme Value Type I

2.6.7.2 Extreme Value Type II

2.6.7.3 Extreme Value Type III

2.6.7.4 The Extreme Value Distributions Related to Load and Strength

2.7 Summary of Continuous Statistical Distributions

2.8 Variation in Engineering

Figure 2.11 Shapes of common failure distributions, reliability and hazard rate functions (shown in relation to t).

2.8.1 Is the Variation Normal?

Figure 2.12 Curtailed normal distribution.

Figure 2.13 Effect of selection.

Figure 2.14 Skewed distribution.

Figure 2.15 Bi-modal distribution.

Figure 2.16 Four distributions with the same means and SDs.

2.8.2 Effects and Causes

2.8.3 Tails

2.9 Conclusions

2.10 Discrete Variation

2.10.1 The Binomial Distribution

Example 2.6

Example 2.7

2.10.2 The Poisson Distribution

Example 2.8

Example 2.9

2.11 Statistical Confidence

2.11.1 Confidence Limits on Continuous Variables

Example 2.10

Figure 2.17 Utilizing Excel’s Goal Seek to find Z-value corresponding to the 95% confidence interval.

Figure 2.18 Confidence levels for normal distribution.

2.12 Statistical Hypothesis Testing

2.12.1 Tests for Differences in Means (z Test)

Example 2.11

Example 2.12

2.12.2 Use of the z Test for Binomial Trials

Example 2.13

Table 2.2 Results for tests in Example 2.13.

2.12.3 χ2 Test for Significance

Example 2.14

2.12.4 Tests for Differences in Variances. Variance Ratio Test (F Test)

Table 2.3 Life test data on two items.

Example 2.15

2.13 Non-Parametric Inferential Methods

Table 2.4 Critical values of r for the sign test. Reproduced by permission of McGraw-Hill.

2.13.1 Comparison of Median Values

2.13.1.1 The Sign Test

Example 2.16

2.13.1.2 The Weighted Sign Test

2.13.1.3 Tests for Variance

2.13.1.4 Reliability Estimates

2.14 Goodness of Fit

2.14.1 The χ2 Goodness-of-Fit Test

Table 2.5 Data from an overstress life test of transistors.

Example 2.17

2.14.2 The Kolmogorov–Smirnov Test

Table 2.6 Failure data with ranked values of xi.

Example 2.18

2.15 Series of Events (Point Processes)

2.15.1 Trend Analysis (Time Series Analysis)

Figure 2.19 Arrival and interarrival values.

Example 2.19

2.15.2 Superimposed Processes

Figure 2.20 Rate of occurrence for superimposed processes.

2.16 Computer Software for Statistics

2.17 Practical Conclusions

Questions

Bibliography

Helpful introductory sources

More advanced works

3 Life Data Analysis and Probability Plotting

3.1 Introduction

3.1.1 General Approach to Life Data Analysis and Probability Plotting

3.1.2 Statistical Data Analysis Methods

Figure 3.1 Probability plotting alternatives in regards to the possible pdf of failure distribution.

3.2 Life Data Classification

Figure 3.2 Normal probability plot.

3.2.1 Complete Data

3.2.2 Censored Data

3.2.3 Right Censored (Suspended)

Figure 3.3 Complete data set.

Figure 3.4 Right censored data.

3.2.4 Interval Censored

Figure 3.5 Interval censored data.

Figure 3.6 Left censored data.

3.2.5 Left Censored

3.3 Ranking of Data

3.3.1 Concept of Ranking

3.3.2 Mean Rank

3.3.3 Median Rank

3.3.4 Cumulative Binomial Method for Median Ranks

Table 3.1 Median rank for the sample size of 5.

3.3.5 Algebraic Approximation of the Median Rank

3.3.6 Ranking Censored Data

3.4 Weibull Distribution

3.4.1 Two Parameter Weibull

3.4.2 Weibull Parameter Estimation and Probability Plotting

Figure 3.7 Weibull probability paper. Abscissa – ln t, Ordinate – ln ln

Example 3.1 Weibull Analysis using Rank Regression

Figure 3.8 Data plotted on Weibull paper for Example 3.1, β ≈ 2.0 and η ≈ 320.

Example 3.2 Calculating Adjusted Ranks

Table 3.2 Data summary and adjusted ranks calculation for Example 3.2.

3.4.3 Three Parameter Weibull

Figure 3.9 3-parameter Weibull distribution plotted with Weibull++®.

3.4.4 The Relationship of β-Parameter to Failure Rates and Bathtub Curve

Figure 3.10 Relationship between the bathtub curve and the Weibull slope β.

3.4.5 BX-Life

3.5 Computerized Data Analysis and Probability Plotting

3.5.1 Rank Regression on X

Figure 3.11 Minimizing distance in the X-direction.

3.5.2 Maximum Likelihood Estimation (MLE)

Example 3.3 Illustrating MLE Method on Exponential distribution

3.5.3 Recommendation on Using Rank Regression vs. MLE

Figure 3.12 Two-sided 90% confidence bounds.

3.6 Confidence Bounds for Life Data Analysis

Figure 3.13 One-sided confidence bounds.

Table 3.3 5 and 95% ranks for the sample size of 5.

3.6.1 Confidence Intervals for Weibull Data

3.6.2 Individual Parameter Bounds

Figure 3.14 Weibull++® two-sided 90% confidence bounds for Weibull distribution.

3.6.2.1 Fisher Matrix Bounds

3.6.2.2 Likelihood Ratio Bounds

3.6.2.3 Beta Binomial Bounds

3.6.2.4 Monte Carlo Confidence Bounds

3.6.2.5 Bayesian Confidence Bounds

Example 3.4 Manual Calculation of Confidence Bounds on the Weibull Parameter β

3.6.3 Alternative Methods for Calculating Confidence Bounds

Figure 3.15 Confidence limits for shape parameter β for different confidence values.

3.7 Choosing the Best Distribution and Assessing the Results

3.7.1 Goodness of a Distribution Fit

Figure 3.16 Weibull++® 90% confidence bounds on B10-life and Reliability (Example 3.1).

Figure 3.17 Weibull++® distribution ranking based on the goodness of fit (Rank Regression on X).

3.7.2 Mixed Distributions

Figure 3.18 Separate groups of data points.

Figure 3.19 Mixed Weibull distribution plotted with Weibull++®.

3.7.3 Engineering Approach to Finding Best Distribution

Example 3.5 Breaking Strength of a Wire

Table 3.4 Breaking strengths of 15 samples of wire of equal length.

Figure 3.20 Probability plot of the breaking strength (Weibull++®), Extreme value distribution.

3.8 Conclusions

Questions

Bibliography

4 Monte Carlo Simulation

4.1 Introduction

4.2 Monte Carlo Simulation Basics

4.3 Additional Statistical Distributions

Figure 4.1 Simplified Monte Carlo simulation procedure with y = f(x1, x2, xn).

4.3.1 Uniform Distribution

4.3.2 Triangular Distribution

Figure 4.2 (a) Rectangular and (b) Triangular distributions.

4.4 Sampling a Statistical Distribution

4.4.1 Generating Random Variables Using Excel Functions

Table 4.1 Statistical distributions sampling using Microsoft Excel®.

4.4.2 Number of Simulation Runs and the Accuracy of Results

Example 4.1

4.5 Basic Steps for Performing a Monte Carlo Simulation

Example 4.2 Calculating the Probability of Exceeding Yield Strength

Figure 4.3 Monte Carlo simulation process.

Figure 4.4 Monte Carlo Simulation using Microsoft Excel®.

4.6 Monte Carlo Method Summary

Table 4.2 Input variables generated by @Risk® for Example 4.2 (Reproduced by permission of Palisade Corporation).

Figure 4.5 Simulation results including the histogram and the best fit distribution for Example 4.2 using @Risk v.5.7.

Figure 4.6 Monte Carlo Simulation sensitivity analysis by @Risk®.

Questions

Bibliography

5 Load–Strength Interference

5.1 Introduction

5.2 Distributed Load and Strength

Figure 5.1 Distributed load and strength: (a) non-overlapping distributions, (b) overlapping distributions.

Figure 5.2 Effect of safety margin and loading roughness. Load L’ causes failure of a proportion of items indicated by the shaded area.

Figure 5.3 Truncation of strength distribution by screening.

5.3 Analysis of Load–Strength Interference

5.3.1 Normally Distributed Strength and Load

Example 5.1

5.3.2 Other Distributions of Load and Strength

5.4 Effect of Safety Margin and Loading Roughness on Reliability (Multiple Load Applications)

Figure 5.4 Failure probability–safety margin curves when both load and strength are normally distributed (for large n and n = 1) (Carter, 1997).

Figure 5.5 Characteristic regions of a typical failure probability–safety margin curve (Carter, 1997).

Figure 5.6 Failure probability–safety margin curves for asymmetric distributions (loading roughness = 0.3) (Carter, 1997).

Figure 5.7 Failure probability–safety margin curves for asymmetric distributions (loading roughness = 0.9) (Carter, 1997).

Example 5.2 (electronic)

Figure 5.8 Load data (sampled at 10 s intervals).

Table 5.1 Mean ranking of load test data.

Table 5.2 Failure data for 100 transistors.

Example 5.3 (mechanical fatigue)

Table 5.3 Maximum loads vs. percentages of the users applying those loads.

Table 5.4 Washing machine loads vs. percent of motors failing at 100 cycles.

Figure 5.9 Load-Strength distribution chart generated with Weibull++® for Example 5.3.

5.5 Practical Aspects

Questions

Bibliography

6 Reliability Prediction and Modelling

6.1 Introduction

6.2 Fundamental Limitations of Reliability Prediction

6.3 Standards Based Reliability Prediction

6.3.1 MIL-HDBK-217

6.3.2 Telcordia SR-332 (Formerly Bellcore)

6.3.3 IEC 62380 (Formerly RDF 2000)

6.3.4 NSWC-06/LE10

6.3.5 PRISM and 217Plus

6.3.6 China 299B (GJB/z 299B)

6.3.7 Other Standards

6.3.8 IEEE Standard 1413

6.3.9 Software Tools for Reliability Prediction

6.4 Other Methods for Reliability Predictions

6.4.1 Field Return Based Methods

6.4.2 Fusion of Field Data and Reliability Prediction Standards

6.4.3 Physics of Failure Methods

6.4.4 ‘Top Down’ Approach to Reliability Prediction

6.5 Practical Aspects

6.6 Systems Reliability Models

6.6.1 The Basic Series Reliability Model

Figure 6.1 Series System.

6.6.2 Active Redundancy

Figure 6.2 Dual redundant system.

6.6.3 m-out-of-n Redundancy

6.6.4 Standby Redundancy

Figure 6.3 Reliability block diagram for a missile system.

6.6.5 Further Redundancy Considerations

6.7 Availability of Repairable Systems

Figure 6.4 (a) Non-repairable system and (b) Repairable system.

Table 6.1 Reliability and availability for some systems configurations. (R. H. Myers, K. L. Wong and H. M. Gordy, Reliability Engineering for Electronic Systems, Copyright © 1964 John Wiley & Sons, Inc. Reprinted by permission of John Wiley & Sons, Inc.)

Example 6.1

Table 6.2 MTBR and replacement costs for the four modules.

Table 6.3 Cost per year of replacing the modules.

6.8 Modular Design

Example 6.2

6.9 Block Diagram Analysis

Figure 6.5 Block diagram decomposition.

Example 6.3

6.9.1 Cut and Tie Sets

Figure 6.6 (a) Cut sets and (b) tie sets.

Example 6.4

6.9.2 Common Mode Failures

Figure 6.7 Effect of common mode failure.

6.9.3 Enabling Events

6.9.4 Practical Aspects

6.10 Fault Tree Analysis (FTA)

Figure 6.8 Standard symbols used in fault tree analysis.

6.11 State-Space Analysis (Markov Analysis)

Figure 6.9 Reliability block diagram of engine.

Figure 6.10 FTA for engine (incomplete).

Example 6.5

Figure 6.11 Two-state Markov state transition diagram.

Figure 6.12 Tree diagram for Example 6.5.

6.11.1 Complex Systems

Figure 6.13 Transient availability of repaired system.

6.11.2 Continuous Markov Processes

Figure 6.14 State–space diagram for a single-component repairable system.

6.11.3 Limitations, Advantages and Applications of Markov Analysis

6.12 Petri Nets

Figure 6.15 Basic structures of logic relations for Petri nets.

6.12.1 Transformation between Fault Trees and Petri Nets

Figure 6.16 A fault tree.

Figure 6.17 Correlations between fault tree and Petri net.

Figure 6.18 The Petri net transformation of Figure 6.16.

6.12.2 Minimum Cut Sets

Figure 6.19 Minimum cut sets of Figure 6.18.

Figure 6.20 The absorption principle of equivalent Petri nets.

6.12.3 Marking Transfer

Example 6.6

Figure 6.21 Petri net diagram for an airbag controller with a detection system (Kleyner and Volovoi, 2008).

6.13 Reliability Apportionment

6.14 Conclusions

Questions

Figure 6.22 System XYZ block diagram.

Bibliography

7 Design for Reliability

7.1 Introduction

Figure 7.1 Cost of design change.

7.2 Design for Reliability Process

Figure 7.2 Design for reliability (DfR) activities flow.

7.3 Identify

7.3.1 Benchmarking

7.3.2 Environments

7.3.3 Environment Distribution

Example 7.1

Figure 7.3 Statistical distribution of the annual driving distances (per passenger car in Europe).

7.3.4 Quality Function Deployment (QFD)

Figure 7.4 Quality function deployment for electric motor design.

7.3.5 Programme Risk Assessment

7.4 Design

7.4.1 Computer-Aided Engineering

7.4.2 Failure Modes, Effects and Criticality Analysis (FMECA)

7.4.3 Steps in Performing an FMECA

Figure 7.5 FMEA worksheet for AIAG-3 method.

Figure 7.6 MIL-STD-1629 Method 102 worksheet for criticality analysis.

7.4.4 Uses for FMECA

7.4.5 FMECA Software Tools

Figure 7.7 Part of output listing from CARE®, the FMECA software.

7.4.6 Reliability Predictions for FMECA

7.4.7 Load-Strength Analysis

7.4.8 Hazard and Operability Study (HAZOPS)

Table 7.1 Load–strength analysis example.

7.4.9 Parts, Materials and Processes (PMP) Review

7.4.10 Non-Material Failure Modes

Table 7.2 HAZOPS on motion system (partial).

7.4.11 Critical Items List

7.4.12 Load Protection

7.4.13 Protection against Strength Degradation

7.4.14 Design Reviews

7.4.15 Design Review Based on Failure Modes (DRBFM)

7.4.16 Human Reliability

7.5 Analyse

7.5.1 Field Return and Warranty Data Analysis

7.6 Verify

7.6.1 Degradation Analysis

7.6.2 Configuration Control

7.7 Validate

7.8 Control

7.8.1 Design Analysis for Processes

7.8.2 Variation

Figure 7.8 Shaft–bore interference.

7.8.3 Process FMECA

7.8.4 ‘Poka Yoke’

7.8.5 Testability Analysis

7.8.6 Test Yield Analysis

7.8.7 Maintainability Analysis

7.9 Assessing the DfR Capability of an Organization

7.10 Summary

Questions

Bibliography

8 Reliability of Mechanical Components and Systems

8.1 Introduction

Figure 8.1 Material behaviour in tensile stress.

8.2 Mechanical Stress, Strength and Fracture

Figure 8.2 Stress–strain for different materials (generalized).

8.3 Fatigue

Figure 8.3 S–N curve.

Figure 8.4 Random overload.

Example 8.1

Figure 8.5 S–N diagram for the part in Example 8.1.

Figure 8.6 Strength deterioration with cyclic stress.

Figure 8.7 Typical fatigue failure (schematic).

8.3.1 Design against Fatigue

8.3.2 Maintenance of Fatigue-Prone Components

8.4 Creep

8.5 Wear

8.5.1 Wear Mechanisms

8.5.2 Methods of Wear Reduction

8.5.3 Maintenance of Systems Subject to Wear

8.6 Corrosion

8.7 Vibration and Shock

8.8 Temperature Effects

Figure 8.8 Waterfall plot.

8.8.1 Humidity and Condensation

8.9 Materials

8.9.1 Metal Alloys

8.9.2 Plastics, Rubbers

8.9.3 Ceramics

8.9.4 Composites, Adhesives

8.10 Components

8.11 Processes

8.11.1 Fasteners

8.11.2 Adhesives

8.11.3 Welding and Soldering

8.11.4 Seals

Questions

Bibliography

Fracture mechanics

Wear

Corrosion

Vibration and shock

Materials and components

9 Electronic Systems Reliability

9.1 Introduction

9.2 Reliability of Electronic Components

9.2.1 Stress Effects

9.2.1.1 Current

9.2.1.2 Voltage

Figure 9.1 Parameter drift.

9.2.1.3 Temperature

Figure 9.2 Temperature vs. reliability for electronic components.

9.2.1.4 Power

9.3 Component Types and Failure Mechanisms

9.3.1 Integrated Circuits (ICs)

9.3.1.1 Application-Specific ICs

9.3.1.2 Microelectronics Packaging

Figure 9.3 Examples of electronic components. (a) Leadless chip capacitor (b) Quad flat pack IC package (QFP) (courtesy DfR Solutions) (c) Ball grid array (BGA) IC package.

Figure 9.4 Five stacked die 4 GB flash memory (pyramid stacking with wire bond interconnects).

9.3.1.3 Hybrid /Microelectronic Packaging/Multichip Modules

Figure 9.5 Micro-hybrid.

9.3.1.4 Microelectronic Component Attachment

9.3.1.5 Microelectronic Device Failure Modes

9.3.1.6 Microelectronic Device Specifications

9.3.1.7 Microelectronic Device Screening

Figure 9.6 Typical failure density functions of electronic components when no component burn-in has been carried out.

Table 9.1 Microelectronic device screening requirementsa.

9.3.2 Other Electronic Components

9.3.2.1 Discrete Semiconductors

9.3.2.2 ‘Passive’ Components

9.3.2.3 Capacitors

9.3.2.4 Electro-Optical Components

9.3.2.5 Cables and Connectors

9.3.2.6 Insulation

9.3.3 Solder

9.3.3.1 Tin-Lead Solder

9.3.3.2 Lead-Free Solder

9.4 Summary of Device Failure Modes

Table 9.2 Device failure modes.

9.5 Circuit and System Aspects

9.5.1 Distortion and Jitter

9.5.2 Timing

9.5.3 Electromagnetic Interference and Compatibility

9.5.4 Intermittent Failures

9.5.5 Other Failure Causes

9.6 Reliability in Electronic System Design

9.6.1 Introduction

9.6.2 Transient Voltage Protection

Figure 9.7 Logic device protection. Diode D1 prevents the input voltage from rising above the power supply voltage. Capacitor C1 absorbs high frequency power supply transients.

Figure 9.8 Transistor protection. Resistor R1 limits the base current IB and capacitor C1 absorbs power supply high frequency transients.

9.6.3 Thermal Design

9.6.4 Stress Derating

Table 9.3 Device derating guidelines.

Figure 9.9 Temperature–power derating for transistors and diodes (typical).

9.6.5 Component Uprating

9.6.6 Electromagnetic Interference and Compatibility (EMI/EMC)

Figure 9.10 Digital circuit noise decoupling.

9.6.7 Redundancy

9.6.8 Design Simplification

9.6.9 Sneak Analysis

Figure 9.11 Sneak analysis basic patterns (hardware).

9.7 Parameter Variation and Tolerances

9.7.1 Introduction

Figure 9.12 Parameter distributions after selection.

9.7.2 Tolerance Design

9.7.3 Analysis Methods

9.7.3.1 Worst Case Analysis

9.7.3.2 The Transpose Circuit

9.7.3.3 Simulation

Figure 9.13 Transpose circuit.

Figure 9.14 Monte Carlo analysis of filter circuit.

9.8 Design for Production, Test and Maintenance

Questions

Bibliography

General

Components

Mechanical and thermal effects and design

EMI/EMC/ESD

Tolerance design and electronics testing

10 Software Reliability

10.1 Introduction

10.2 Software in Engineering Systems

Table 10.1 Comparison of Hardware and Software Reliability Characteristics.

10.3 Software Errors

10.3.1 Specification Errors

Figure 10.1 Voting redundant system.

10.3.2 Software System Design

10.3.3 Software Code Generation

10.4 Preventing Errors

10.4.1 Specification

10.5 Software Structure and Modularity

10.5.1 Structure

10.5.2 Modularity

Figure 10.2 Structured versus unstructured programming.

10.5.3 Requirements for Structured and Modular Programming

10.5.4 Software Re-Use

10.6 Programming Style

10.7 Fault Tolerance

10.8 Redundancy/Diversity

10.9 Languages

Figure 10.3 Fault tolerant algorithm.

10.10 Data Reliability

10.11 Software Checking

10.11.1 FMECA

10.11.2 Software Sneak Analysis

Figure 10.4 Software sneak patterns.

10.12 Software Testing

10.12.1 Managing Software Testing

10.13 Error Reporting

Figure 10.5 Software error reporting form.

10.14 Software Reliability Prediction and Measurement

10.14.1 Introduction

10.14.2 The Poisson Model (Time-Related)

10.14.3 The Musa Model

Example 10.1

10.14.4 The Jelinski–Moranda and Schick–Wolverton Models

10.14.5 Littlewood Models

Example 10.2

10.14.6 Point Process Analysis

10.15 Hardware/Software Interfaces

10.16 Conclusions

Figure 10.6 Software development for reliability.

Questions

Bibliography

11 Design of Experiments and Analysis of Variance

11.1 Introduction

11.2 Statistical Design of Experiments and Analysis of Variance

11.2.1 Analysis of Single Variables

Example 11.1

Table 11.1 Times to failure of 20 bearings.

Table 11.2 Values of for the data of Table 11.1.

Table 11.3 Values of for the data of Table 11.1.

Table 11.4 Values of WS for the data of Table 11.1.

Table 11.5 Sources of variance for the data in Table 11.1.

Table 11.6 Example 11.1 Minitab® solution.

Figure 11.1 Minitab® chart showing statistically significant difference between samples 1 and 4.

11.2.2 Analysis of Multiple Variables (Factorial Experiments)

Example 11.2

Table 11.7 Results of experiments on ‘O’ ring seals.

Table 11.8 The data of Table 11.7 after subtracting 100 from each datum.

Table 11.9 Analysis of variance table.

11.2.3 Non-Normally Distributed Variables

11.2.4 Two-Level Factorial Experiments

Example 11.3

Table 11.10 Results of a three-factor non-replicated experiment.

Table 11.11 Response table and interaction of effects A, B, C.

Table 11.12 Analysis of variance table.

11.2.5 Fractional Factorial Experiments

Figure 11.2 Main effects plots for Example 11.3 using Minitab®.

Table 11.13 The full design matrix for a 24 factorial experiment.

Table 11.14 Table 11.13 omitting rows where ABCD gives minus.

Table 11.15 Sixteenth fractional factorial layout for seven main effects.

11.3 Randomizing the Data

11.4 Engineering Interpretation of Results

Figure 11.3 Temperature–pressure interactions.

11.5 The Taguchi Method

Figure 11.4 Taguchi method (1).

Figure 11.5 Taguchi method (2).

Table 11.16 Results of Taguchi experiment on fuel system components (Example 11.4).

Example 11.4

Figure 11.6 Results of Taguchi experiment (Example 11.4).

11.6 Conclusions

Questions

Bibliography

12 Reliability Testing

12.1 Introduction

12.2 Planning Reliability Testing

12.2.1 Using Design Analysis Data

12.2.2 Considering Variability

12.2.3 Durability

12.3 Test Environments

Figure 12.1 Typical CERT environmental cycles: electronic equipment in a vehicle application.

Figure 12.2 CERT test facility

12.3.1 Vibration Testing

Figure 12.3 Road transport vibration levels.

12.3.2 Temperature Testing

12.3.3 Electromagnetic Compatibility (EMC) Testing

12.3.4 Other Environments

12.3.5 Customer Simulation Testing

12.4 Testing for Reliability and Durability: Accelerated Test

12.4.1 Test Development

Figure 12.4 Stress, strength and test failures (1).

Figure 12.5 Stress, strength and test failures (2).

Figure 12.6 Stress, strength and test failures (3): wearout failures.

12.4.2 Accelerated Test

Figure 12.7 Effect of accelerated test on the bathtub curve.

Figure 12.8 Stress ranges and types of failures.

12.4.3 Highly Accelerated Life Testing

12.4.4 Test Approach for Accelerated Test

12.4.5 HALT and Production Testing

Table 12.1 DoE/HALT Selection

12.4.6 DoE or HALT?

12.5 Test Planning

Figure 12.9 Example of a parallel test flow for an electronic device.

12.6 Failure Reporting, Analysis and Corrective Action Systems (FRACAS)

12.6.1 Failure Reporting

12.6.2 Corrective Action Effectiveness

Questions

Bibliography

13 Analysing Reliability Data

13.1 Introduction

13.2 Pareto Analysis

Figure 13.1 Pareto plot of failure data.

13.3 Accelerated Test Data Analysis

13.4 Acceleration Factor

Table 13.1 Acceleration factor relationship with reliability functions.

Example 13.1

13.5 Acceleration Models

13.5.1 Temperature and Humidity Acceleration Models

13.5.1.1 Arrhenius Model

Table 13.2 Commonly used activation energy values for different failure mechanisms.

13.5.1.2 Eyring Model

13.5.1.3 Peck Temperature-Humidity Model

13.5.1.4 Lawson Temperature-Humidity Model

13.5.1.5 Tin-Lead Solder

13.5.1.6 Lead-Free Solder

13.5.2 Voltage and Current Acceleration Models

13.5.3 Vibration Acceleration Models

Example 13.2

13.6 Field-Test Relationship

Example 13.3

13.7 Statistical Analysis of Accelerated Test Data

Table 13.3 Accelerated test results (Example 13.4).

Example 13.4

Figure 13.2 Weibull plot at three stress levels (Weibull++®)

Figure 13.3 Life vs. stress plot generated with ALTA® software

13.8 Reliability Analysis of Repairable Systems

13.8.1 Failure Rate of a Repairable System

Figure 13.4 Plotted data of Example 2.19.

13.8.2 Multisocket Systems

Figure 13.5 The failure pattern of a multisocket system.

Example 13.5 (Reprinted from Ascher and Feingold (1984) by courtesy of Marcel Dekker, Inc.)

13.9 CUSUM Charts

Figure 13.6 Bus engine failure data (a) First failures (191), (b) Second failures (105), (c) Third failures (101), (d) Fourth failures (96), (e) Fifth failures (94).

Table 13.4 Reliability test data Target = 95% (T).

Figure 13.7 (a) Run chart of data in Table 13.4. (b) CUSUM chart of data of Table 13.4.

13.10 Exploratory Data Analysis and Proportional Hazards Modelling

Figure 13.8 Time series chart: failure vs time (overhaul interval 1000 h).

13.11 Field and Warranty Data Analysis

13.11.1 Field and Warranty Data Considerations

Figure 13.9 Warranty claims root causes.

13.11.2 Warranty Data Formats

13.11.2.1 Individual Claims Data Format

13.11.2.2 Statistical Data Format

Table 13.5 Example of MIS (Month in Service) data (January – July 2011).

13.11.3 Warranty Data Processing

Table 13.6 Example of warranty data presented in the ‘Nevada’ (or the ‘Layer cake’) format.

Table 13.7 Cumulative percent failures for 6 months. Based on the data in Table 13.5.

Example 13.6

Questions

Figure 13.10 Weibull charts of the test results.

Bibliography

14 Reliability Demonstration and Growth

14.1 Introduction

14.2 Reliability Metrics

Example 14.1

14.3 Test to Success (Success Run Method)

14.3.1 Binomial Distribution Approach

Table 14.1 Required test sample sizes for reliability demonstration at 50 and 90% confidence.

14.3.2 Success Run Test with Undesirable Failures

14.4 Test to Failure Method

14.5 Extended Life Test

14.5.1 Parametric Binomial Method

Example 14.2

14.5.2 Limitations of the Parametric Binomial Model

14.6 Continuous Testing

Table 14.2 MTBF confidence limits.

Example 14.3

14.7 Degradation Analysis

Figure 14.1 Degradation analysis diagram.

14.8 Combining Results Using Bayesian Statistics

Example 14.4

14.9 Non-Parametric Methods

14.9.1 The C-Rank Method

Example 14.5

Figure 14.2 Solution to Example 14.2 using Weibull++® reliability demonstration calculator DRT

14.10 Reliability Demonstration Software

14.11 Practical Aspects of Reliability Demonstration

14.12 Standard Methods for Repairable Equipment

14.12.1 Probability Ratio Sequential Test (PRST) (US MIL-HDBK-781)

Figure 14.3 Typical probability ratio sequential test (PRST) plan.

14.12.2 Test Plans

Table 14.3 MIL-HDBK-781 PRST plans.

14.12.3 Statistical Basis for PRST Plans

Table 14.4 MIL-HDBK-781 fixed length test plans.

14.12.4 Operating Characteristic Curves and Expected T