Practical Reliability Engineering - From Planning to Test Data Analysis and Extrapolation (RQ25-31-0)

SynopsisReliability is crucial to the success of product sales in today's competitive market. Reliability test is now a common routine to most of the manufacturing. However, the analysis of reliability test data is not straight forward since it involves advanced statistics. There are many existing software available for reliability data analysis, but most of them have inherent assumptions that are suitable for system reliability rather than component reliability. Also, without understanding the impact of these assumptions, analysis results could be invalid.

This course provides a basic understanding of reliability data analysis, and the various type of reliability data as well as their respective analysis methodology. The planning of accelerated life testing will also be presented, and the extrapolation of the accelerated test data to normal operating conditions will be discussed in details. Misconception in accelerated life testing and their extrapolation will be discussed.

What previous participants say about this course
Answers to the question 'what did you like most about the course'

• "Clear presentation and willingness of Dr Tan to discuss topics outside of course contents" - 21 Jun 06
• "Sample size software, time to failure calculation" - 21 Jun 06
• "Instructor has vast knowledge on the subject matter" - 21 Jun 06
• "Challenging the industry standard" - 21 Jun 06
• "Explanation of the reliability concept & application of the equation" - 21 Jun 06
• "Knowledgable instructor" - 21 Jun 06
• "Contents of the course & detailed explaination by the lecturer" - 21 May 07
• "The conceptual description and explaination provided has help me to understand the reliability theory, calculations & applications" - 21 May 07

What You Will Learn

• Reliability concepts and types of reliability data
• Non-parametric analysis of reliability data and its limitation
• Parametric analysis of reliability data and its usefulness for reliability improvement
• Probability plotting in reliability data analysis and its precautious
• Planning of life tests for products
• Various accelerated test models and their use in extrapolation of test data for reliability prediction.

Who Should Attend

• R&D engineers
• Reliability engineers
• Process engineers

PrerequisiteKnowledge of basic statistics.

Course MethodologyThis course is presented classroom style, with case studies to illustrate the concepts taught.

Course Duration3 days, 9am - 5pm

Course StructureModule 1 - Reliability concepts and reliability data
This module introduces some of the basic concepts of product reliability. Section 1 explains the relationship between quality and reliability and outlines how statistical studies are used to obtain information that can be used to assess and improve product reliability. Section 2 explains, in general terms, important qualitative aspects of statistical models that are used to describe populations and processes in reliability applications. Section 3 emphasizes the important distinction between studies focusing on data from repairable systems and nonrepairable units. Section 4 describes a general strategy for exploring, analyzing, and drawing conclusions from reliability data.

Module 2 - Models and Censoring for Failure time data
This module introduces basic concepts of modeling failure-time processes. The basic relationships among cumulative distributions, densities, survival, hazard, and quantile functions for modeling of continuous failure-time processes are explained. The module also explains briefly the importance of censoring, censoring mechanisms, and important assumptions about censoring mechanisms, needed for proper application of the methodology in reliability data analysis.

Module 3 - Non-parameteric estimation
The nonparametric (model-free) estimates described in this module are used as a tool for reliability data analysis. Section 1 starts with a simple method that applies to problems with complete data or single censoring. Section 2 explains the basic ideas of statistical inference and introduces the ideas behind the use of confidence intervals. Confidence intervals for complete data or single censoring are described in Section 3. The methods are generalized to the commonly encountered multiple censoring in Section 4. Simultaneous confidence bands (used for helping to choose a model) are presented in Section 5.

Module 4 - Parametric Distributions
This module introduces some basic ideas of parametric modeling and the most important parametric distributions. Parametric distributions are used extensively in reliability data analysis. Section 1 explains some of the basic concepts and motivation for using parametric models. Section 2 describes important functions of parameters like failure probabilities and distribution quantiles. Section 3 introduces the important location-scale family of distributions. Section 4 gives detailed information on these and the important log-location-scale distributions.

Module 5 - Probability Plotting
This module presents the important topic of probability plotting. Section 1 explains the basic concepts of probability plotting. Section 2 explains aspects of the practical application of probability plots, including the use of simulation to help interpret such plots.

Module 6 - Planning life tests
This module provides tools for evaluating and controlling estimation precision for a life test when censored data are expected. Section 1 introduces the basic ideas of test planning and uses simulation to illustrate and explain the effect that sample size has on sampling variability. Section 2 shows how to compute approximate sampling variability directly. Sections 3 and 4 show how to find the sample size needed to control sampling variability (or precision) and illustrate the ideas for the normal and exponential distribution. Section 5 applies these methods to problems involving Type I censored data with the Weibull and lognormal distributions. Section 6 describes methods for planning a test to demonstrate conformance with a specified reliability standard.

Module 7 - Accelerated test models This module describes models used for accelerated tests and introduces concepts of physics of failure. Section 1 motivates and describes the general methods for accelerating reliability tests. Sections 2, 3, and 4 describe, respectively, use-rate, temperature, and voltage acceleration. Section 5 describes some models with a combination of accelerating variables. Section 6 describes some other common forms of accelerated testing, and Section 7 described some potential pitfalls of accelerated testing.

Module 8 - Hand-on for practical reliability problems