Practical Reliability Engineering, 5th Edition

3

Life Data Analysis and Probability Plotting

3.1 Introduction

It is frequently useful in reliability engineering to determine which distribution best fits a set of data and to derive estimates of the distribution parameters. The mathematical basis for the approaches to these problems was covered in Chapter 2.

3.1.1 General Approach to Life Data Analysis and Probability Plotting

The methods described in this chapter can be used to analyse any appropriate data, such as dimensional or parameter measurements. However, their use for analysing reliability time-to-failure (life data) will be emphasized.

General-purpose statistical software such as Minitab¯ includes capabilities for probability plotting. Since the Weibull distribution is the most commonly applied in reliability life data analysis, computer software packages have been developed specifically for this purpose, such as ReliaSoft Weibull++¯, SuperSMITH Weibull¯ and few others. This chapter makes use of ReliaSoft Weibull++ ¯ software to illustrate how to perform these tasks.

Note that probability plotting methods to derive time-to-failure distribution parameters are only applicable when the data are independently and identically distributed (IID). This is usually the case for non-repairable components and systems but may not be the case with failure data from repairable systems. The reason is that repaired systems can have secondary failures, which are dependent on the primary failures. Also due to successive repairs the population ...