K. CROES, "Statistical techniques for planning type I singly censored reliability experiments with two stress factors", Phd Thesis, Limburg University Center, 1999.
Copyright - [Précédente] [Première page] [Suivante] - Home
These : [THESE126]
Titre : K. CROES, Statistical techniques for planning type I singly censored reliability experiments with two stress factors, Phd Thesis, Limburg University Center, 1999.
Cité dans : [DIV366] Recherche sur l'auteur Kristof CROES, octobre 2002.
Auteur : Kristof Croes
Infos : Ph. D. Thesis
Source : Limburg University Center.
Date : 1999
Infos : en octobre 2002 : Software Responsible Reliability and Statistics, XPEQT, Transportstraat 1, B-3980 Tessenderlo
Lien : mailto:krc@xpeqt.com
Tel. : +32(0)13/611625
Fax. : +32(0)13/611642
Mobile : +32(0)485/556204
Web : http://www.xpeqt.com
Abstract :
[1] : [PAP158] -------
Lien : private/PhD_Croes_content.pdf - Content, 11 Ko, 3 pages.
Lien : private/PhD_Croes_1.pdf - Chapter 1, 156 Ko, 27 pages.
Lien : private/PhD_Croes_2.pdf - Chapter 2, 220 Ko, 29 pages.
Lien : private/PhD_Croes_references.pdf - References, 28 Ko, 6 pages.
1. Contents
Preface 1
1. Introduction 3
1.1. Reliability in microelectronics 3
1.1.1. The general structure of an IC 3
1.1.2. Miniaturization and its consequence with respect to reliability 4
1.1.3. Importance of reliability and today’s reliability requirements 4
1.1.4. Main causes of failure of some typical components of an IC 5
1.2. Reliability experiment 10
1.2.1. Basic steps for performing reliability experiments 10
1.2.2. Examples of characteristics of interest, stress factors and lifetime models 11
1.2.3. Most common test strategies for on-chip interconnects 15
1.3. Statistical analysis 18
1.3.1. Notations and assumptions concerning the setup of a reliability experiment 18
1.3.2. Example for an electromigration experiment 20
1.3.3. Failure time distributions 20
1.3.4. Parameter estimation 23
1.3.5. The choice of a suitable reliability measure 27
1.4. Definition of the problem 28
2. A new technique for planning type I singly censored reliability
experiments with two stress factors 31
2.1. Literature 31
2.1.1. Brief survey of literature 31
2.1.2. Major shortcomings of current techniques 33
2.2. The concept of an asymptotic variance 34
2.2.1. Concept of an asymptotic variance 34
2.2.2. Definitions and notation 35
2.2.3. Linear form of Black’s equation 36
2.2.4. Fisher information matrix of a type I singly censored reliability experiment
with two stress factors 38
2.2.5. EAV of several important reliability parameters 41
2.3. Planning experiments making use of the EAV 43
2.3.1. Literature 44
2.3.2. Parameters defining an experimental plan 44
2.3.3. Definition of the Optimum Plan Function (OPF) 50
2.3.4. Minimizing the OPF 53.Contents ii
2.4. Examples 54
2.4.1. Example 1 54
2.4.2. Example 2 64
2.4.3. Example 3 67
2.5. Conclusions 68
3. A closer look at the shortcomings of the proposed technique 71
3.1. A closer look at the second example of chapter 2 71
3.2. Forcing the planned experiments to stress at three different stress levels 74
3.2.1. Three different stress levels for the first stress factor 74
3.2.2. Three different stress levels for the second stress factor 77
3.2.3. Three different stress levels for both stress factors 79
3.3. Sensitivity of the new plans to misspecified values of the model parameters 81
3.4. Forcing the planned experiments to have equal stress times at the two
lowest stress levels 84
3.5. Conclusions 86
4. Comments on the assumptions made by the proposed technique 89
4.1. Monte Carlo simulation of reliability experiments 89
4.2. Objective method for making the distinction between the lognormal and the
Weibull distribution 91
4.2.1. Basic statistical definitions 91
4.2.2. Introduction of the problem 92
4.2.3. Method 93
4.2.4. Illustrative example 104
4.2.5. Comparison with the literature 106
4.2.6. Conclusions and advantages of the new method 108
4.3. The influence of gradients in the applied stress 108
4.3.1. Introduction of the simulations 109
4.3.2. Influence on the estimates of the median life h 112
4.3.3. Influence on the estimates of the activation energy Ea 114
4.3.4. Influence on the estimates of the dispersion parameter s 114
4.3.5. Influence on the estimates of low percentiles 116
4.4. Conclusions 119
5. Conclusions 121
Publication list 123.Contents iii
References 125
Appendix A. Terminology and abbreviations 131
Appendix B. Notation list 133
Appendix C. The number of random P -vectors for the determination of
MEAS(i) in the OPF 137
Appendix D. Approximate confidence intervals of the parameters
h hh hH, n, Ea, s ss s and p
N t 141
Samenvatting 143
Mise à jour le lundi 10 avril 2023 à 19 h 01 - E-mail : thierry.lequeu@gmail.com
Cette page a été produite par le programme TXT2HTM.EXE, version 10.7.3 du 27 décembre 2018.
Les informations contenues dans cette page sont à usage strict de Thierry LEQUEU et ne doivent être utilisées ou copiées par un tiers.
Powered by www.google.fr, www.e-kart.fr, l'atelier d'Aurélie - Coiffure mixte et barbier, La Boutique Kit Elec Shop and www.lequeu.fr.
