Preface

The second edition of this book includes three new chapters: Chapter 16 on experimental modal analysis (EMA) and modal parameter estimation, Chapter 17 on operational modal analysis (OMA), and Chapter 19 in which I have included several examples of how to apply the many techniques presented in the book, both on real data and on synthesized data, i.e., data generated by some numerical model. The latter chapter includes real data from some of my recent research that are also available through my website. It is my hope that these new chapters will make the book even more comprehensive for educators, students, and practitioners alike. In addition to these new chapters, I have also rewritten a few parts, notably the part about correlation function estimators (Sections 10.4.1 and 10.4.2), for which I have found in recent years, that it is easier for my students to understand if presented for correlation as a convolution. Chapter 10 now includes a section presenting a new framework for signal processing that I developed in my recent research. In chapter 19 is also included a presentation of a new way of implementing impact testing which offers new advantages not found in traditional measurement systems. This is also available in the accompanying toolbox, ABRAVIBE, which is available from my website www.abravibe.com, along with almost everything from this book, including a solutions manual for all problems in the book. The ABRAVIBE toolbox for MATLAB (or GNU Octave) is a very comprehensive toolbox with functionality for most of the techniques presented in this book, including rotating machinery analysis, spectrum analysis, EMA, and OMA. My YouTube channel also complements this book with many lectures.

The material in the first edition of this book, published in 2011, had been developing in my mind for more than 20 years of teaching at the time of its first publication. During these years, I had been teaching over 250 shortcourses for engineers in the industry on techniques for experimental noise and vibration analysis and also on how to use commercial measurement and analysis systems. In addition, in the late 1990s, I developed and taught three master's level courses in experimental analysis of vibrations at Blekinge Institute of Technology in Sweden.

Noise and vibration analysis is an interdisciplinary field, incorporating diverse subjects such as mechanical dynamics, sensor technology, statistics, and signal processing. Whereas there are many excellent and comprehensive books in each of these disciplines, there has been a lack of introductory material for the engineering student who first starts to make noise and/or vibration measurements, or the engineer who needs a reference in their daily life. In addition, there are few textbooks in this field presenting the techniques as they are actually used in practice. This book is an attempt to fill this void.

My aim for this book is that it may serve both as a course book and as supplementary reading in university courses, as well as providing a handbook for engineers or researchers who measure and analyze acoustic or vibration signals. The level of the book makes it appropriate both for undergraduate and graduate levels, with a proper selection of the content. In addition, the book should be a good reference for analysts who use experimental results and need to interpret such results. To satisfy these rather different purposes, for some of the topics in the book I have included more detail than would be necessary for an introductory text. To facilitate its use as a handbook, I have also included a short summary at the end of each chapter where some of the key points of the chapter are repeated.

This book contains background theory explaining the majority of analysis methods used in modern commercial software for noise and vibration measurement and analysis. It also includes a number of tools which are usually not found in commercial systems, but which are still useful for the practitioner. With modern computer‐based software, it is easy to export data to, e.g., MATLAB/Octave (see below), and apply the techniques there.

Since it is an introductory text, most of the content of this book is of course available in more specialized textbooks and scientific papers. A few parts, however, include some improvements of existing techniques. I will mention these points in the descriptions of the appropriate chapters below.

Signal analysis is traditionally a field within electrical engineering, whereas most engineers and students pursuing noise and vibration measurements are mechanical or civil engineers. The aim has therefore been to make the material accessible particularly to students and engineers of these latter disciplines. For this reason, I have included introductions to the Laplace and Fourier transforms – both essential tools for understanding, analyzing, and solving problems in dynamics. Electrical engineering students and practitioners should still find many of the topics in the book interesting.

Signal analysis is a subject which is best learned by practicing the theories (perhaps that is a universal truth for all areas?). I have therefore incorporated numerous examples using MATLAB or GNU Octave throughout the book. Further examples and an accompanying toolbox which can be used with either MATLAB or GNU Octave can be downloaded from my website. More information about this is located in Section 1.6. I strongly recommend the use of these tools as a complement to reading this book, regardless of whether you are a student, a researcher, or an industry practitioner.

Chapter 2 introduces dynamic signals and systems with the aim of being an introduction particularly for mechanical and civil engineering students. In this chapter, the classification of signals into periodic, random, and transient signals is introduced. The chapter also includes linear system theory and a comprehensive introduction to the Laplace and Fourier transforms, both important tools for understanding and analyzing dynamic systems.

In Chapter 3, some fundamental concepts of sampled signals are presented. Starting with the sampling theorem and continuing with digital filter theory, this chapter presents some important applications of digital filters for fractional octave analysis and for integrating and differentiating measured signals.

Chapter 4 introduces some applied statistics and random process theory from a practical perspective. It includes an introduction to hypothesis testing as this tool is sometimes used for testing stationarity of data. This chapter also gives an introduction to the application of statistics for data quality assessment, which is becoming more important with the large amounts of data collected in many applications of noise and vibration analysis.

Chapters 5 and 6 provide an introduction to the theory of mechanical vibrations. I anticipate that the contents of these two chapters will already be known to many readers, but I have found it important to include them because my presentation focuses on the experimental implications of the theory, unlike the presentation in most mechanical vibration textbooks, and because some later chapters in the book need a foundation with a common nomenclature.

In Chapter 7, the most important transducers used for measurements of noise and vibration signals are presented, specifically the accelerometer, the force sensor, and the microphone. Because piezoelectric sensors with built‐in signal conditioning (the so‐called IEPE sensors) are widely used today, this technology is presented in some depth. In this chapter, I also present some personal ideas on how to become a good experimentalist.

The analysis techniques mostly used in this field are based on the Discrete Fourier Transform (DFT), computed by the FFT. Spectrum analysis is therefore an important part of this book and Chapters 8–10 are spent on this topic. Chapter 8 introduces basic frequency analysis theory by presenting the different signal classes, and the different spectra used to describe the frequency content of these signals.

In Chapter 9, the DFT and some other techniques used to experimentally determine the frequency content of signals are presented. The properties of the DFT, which are very important to understand when interpreting experimental frequency spectra, are presented relatively comprehensively.

Chapter 10 includes a comprehensive presentation of how spectra from periodic, random, and transient signals, and mixes of these signal classes should be estimated in practice. Chapter 10 also includes a comprehensive explanation of Welch's method for PSD estimation, including overlap processing, as this is the method used in virtually all commercial software. The treatment of practical spectral analysis in this chapter should also be of use to engineers outside the field of acoustics and vibrations who want to calculate and/or interpret spectra by using the FFT.

In Chapter 11, the design of modern data acquisition and measurement systems is described from a user perspective. In this chapter, both hardware and software issues are penetrated.

Chapter 12 addresses order tracking, which is a common technique for analysis of rotating machinery equipment. The chapter describes the most common techniques used to measure such signals both with fixed sampling frequency and with synchronous sampling.

Frequency response functions are important measurement functions in experimental...