Intelligent Reliability Analysis Using MATLAB and AI: Perform Failure Analysis and Reliability Engineering using MATLAB and Artificial Intelligence (English Edition)

CHAPTER 1

Reliability Fundamentals

Introduction

The reliability theory was introduced during the World War II, and it plays a critical role in systems design and development. A valuable contribution was made to the development of the theory of reliability by the mathematicians like Gnedenkov, Belgayev, Solovyeu, Polovoko, Barlow, and Proschan.

Structure

In this chapter, we will discuss the following topics:

Importance of reliability and life analysis

Faults and failures

Various configuration of reliability system

Series, parallel, and hybrid arrangement

Reliable data acquisition in wireless sensor networks

Objectives

After studying this chapter, students should be able to understand the basic concepts of reliability, application area, and its performance parameters. The need of reliability for electrical and electronics components and devices is specified in this chapter. It would help the students to differentiate between faults, defects, and failures. The fundamental techniques and methods for reliability assessment and condition monitoring are discussed.

Reliability

In the modern era of integration, millions of transistors and other electronic components are connected on a single chip. As the number of devices enhance, reliability and validity become a challenging and critical issue. In a series connection, if any one of the components fails or degrades its performance, the complete system shuts down immediately. So, reliability analysis of all the individual components are equally important and necessary for the long term quality performance of the complete device or system.

The efficiency of the system which is performing a specific task, is described by the terms such as reliability, survivability. Reliability means the ability of the system to perform its intended function satisfactorily. The term reliability was defined by the Advisory Group on Reliability of Electronic Equipment (AGREE) as the probability of a product performing its intended function satisfactorily under given condition for a specified period of time. The reliability evaluation techniques were adopted mostly in military applications and aerospace industry for the improvement of quality. The improvement of the effectiveness of components of various kinds has received special attention concerning with numerous problems because of the advanced technology. Sometimes effectiveness is also referred to as quality in literature.

The concept of survivability is understood as the ability of the system to preserve the properties to serve its purpose under adverse conditions (viz, explosions, fine, inundation, and so on). The reliability of a system was determined by the properties of the system, namely, trouble proofness, reparability, and longevity. Trouble proofness is the property of a system to preserve its capability in the duration of a definite time under normal conditions. The prevention, detection, and elimination of failure are useful for improvement of the system that is called repairability. Longevity is the ability for a prolonged operation with the necessary technical maintenance including various kinds of repairs. Maintainability is defined as the probability that failed equipment is restored to operable condition in a specified time (known as down time). Availability is the measure of performance of repairable equipment. It is the combination of reliability and maintainability.

Failure

The word failure plays an important role in the context of reliability theory. The term failure is defined as the termination of the ability of an item to perform its intended function. The notion of failure is a useful characteristic of reliability analysis because it is mainly responsible for various numerical criteria of reliability analysis. Failures are classified into the following different ways – inherent weakness failure, sudden failure, gradual failure, catastrophic failure, and degradation failure. Some of the causes for failures of component in the system are poor designing of the components, lack of experience, poor maintenance policies, wrong manufacturing techniques, and human errors.

Role of probability laws in reliability theory

The reliability analysis of a system is based precisely on defined concepts like reliability function R (t), expected life E (T), hazard function h (t) and failure rate λ (t). A population of identical systems, which operate under identical conditions, fail at different duration of times and the failure phenomenon can only be described in probability terms. Thus, the reliability definition is based on the concept of probability theory. In practice, reliability evaluation is associated with some parameters which are described by probability distributions. The most useful continuous distributions in the theory of reliability are the exponential, weibull, gamma, normal, and log normal distributions, and the two most important discrete distributions are binomial and poisson.

Hazard rate

In the reliability theory, identifying failure model is an art. The concept of hazard rate is effectively used in reliability analysis to identify the failure distributions. For instance, if the hazard rate is more or less constant, then it would represent that time to failure follows exponential distribution. The exponential distribution is widely used in the theory of reliability. On the other hand, it is interesting to note that exponential distribution would arise as a particular case of weibull and gamma distributions which are most commonly used in life testing and reliability estimation.

Weibull distribution

Weibull distribution was named after the Swedish scientist Weibull, who proposed it for the first time in 1939 in connection with his studies on strength of materials. Weibull established that the distribution is also useful in describing the aging effect or wear-out failures KAO, proposing it as a failure model for vacuum tube failures, and Leiblin and Zelen used it for ball bearing failures. Mann et al has shown the number of situations where this model is applicable for other types of failure data. Normal distribution is also applicable as a failure model in the context of reliability. Davis observed that the failure data is fit to normal distribution in the case of incandescent lamps; Bozavsky also described the use of the normal distribution in reliability