Up Time Magazine- April/May 2008 - (Page 61) 2. ‘Spectrum only’ analysts - I believe too many analysts rely too heavily on spectra alone. The spectrum only reveals part of what the machine has to say about the fault condition. The time waveform and phase relationships (and other condition monitoring and performance data) help to complete the picture. Too few people collect time waveforms (and many of those who do are not collecting it correctly). And to top it all off, the settings used to collect the spectrum may not be ideal anyway. When you explore the limitations of the spectrum, and when you better understand the mechanical forces involved with all of the fault conditions, you will see that time waveforms and phase readings (which are now so easy to collect) reveal essential information that will allow you to correctly diagnose the nature and severity of the fault condition. 3. ‘Bearing centric’ analysts - I believe that too many people focus too heavily on detecting rolling element bearing faults. There is so much more to vibration analysis, and reducing downtime, than just detecting bearing damage. If the correct tests are performed, faults such as resonance, unbalance, misalignment, and many others can be detected and corrected at an early stage. With these conditions corrected, the bearings will have a longer life. What’s more, if the bearing detection techniques (measurement practices and analyzer settings) were performed correctly, the fault would be detected earlier, and remedial action could be taken to extend the life of the bearing. This could open a long discussion about bearings analysis, root cause failure analysis, and the relationship between vibration analysis and reliability – but we’ll save that for another time! 4. ‘Default setting’ analysts - I believe that too many people do not have their analyzers set up correctly when taking measurements (and the measurement axis and mounting method may be inadequate). People are using analyzer settings because “that’s what the manual says”, or “that’s what they said to use in the training course”, or “that’s the way we have always set it up”. And even worse, some measurement settings have been selected because someone found that by adjusting the averaging, resolution and settling-time options the measurement completed more quickly. Speed is great, but repeatability and measurement quality are far more important. When you understand what averaging, windowing, resolution, dynamic range and other settings and specifications mean, and how www.uptimemagazine.com they relate to the fault conditions you are trying to detect, you will make the best possible selections. There is another important point worth making. The capabilities of data collectors (analyzers) have changed greatly over the years. Not that long ago the measurement settings were chosen as a compromise because of limited resolution, memory, speed, and dynamic range. Modern data collectors do not have those limitations; therefore your settings should also evolve. What Is The Solution? The solution, in my opinion, is to make sure that analysts understand everything related to their field. They should understand the analyzer, and what all of the settings mean – averaging, overlap, resolution, triggering, etc. They should understand the signals – where do harmonics and sidebands come from, what happens when there are impacts and transients, why does the noise floor lift up, and so on. And they should understand the machine – resonance, unbalance forces, why looseness occurs, what happens as the bearing begins to fail, etc. I don’t mean that you should just recognize the patterns in the spectrum. I mean that you should understand what is happening inside the machine, what forces are being generated, and why the phase relationships and vibration patterns (in the spectrum and waveform) change the way they do. Now, some may look at this list and think that it is not realistic. You may believe that you can get by without understanding all of these issues. You may even question whether you could ever truly understand these concepts, especially if it has been a long time since you attended school. Well, I am here to tell you that you should understand them, you can understand them, and you will benefit enormously if you do understand them. And there are now thousands of analysts out there who would agree. Figure 5 - A 3D animation demonstrating looseness The solution was to put the computer to work. Everyone is familiar with the 3D graphics used in computer games and on television. And everyone has seen software simulators that mimic real life electronic and mechanical systems. So why shouldn’t they be put to work in vibration training? Using hand-drawn sketches to explain the FFT, or bearing vibration, or gearbox time waveforms never did cut it, and it sure does not cut it today. 3D animations make the machine come alive; especially for mechanical people. All of the sketches and arm waving in the world cannot duplicate the power of an animation like the one shown in Figure 5 (and you are only seeing it in its 2D, static format). Simulation software makes it so much easier to understand vibration analysis principles. Sixty minutes of sketches, arm-waving, and long discussion may not be enough to explain windowing, the FFT, modulation, resonance, misalignment forces, or bearing vibration – and the list goes on. But when you can click a mouse, adjust some settings, and make it all visual, the students will truly understand it (maybe for the first time) and it will take much less than 60 minutes! There is nothing sweeter than hearing “Ah, now I understand!” Apart from anything else, 3D graphics and How Do You Achieve These Goals? Thus far I have discussed all of the great reasons for receiving effective training, but I have also listed all of the reasons why this used to be very difficult to achieve in practice. So something had to change. Figure 6 - Bearing vibration simulation 61 http://www.uptimemagazine.com
For optimal viewing of this digital publication, please enable JavaScript and then refresh the page. If you would like to try to load the digital publication without using Flash Player detection, please click here.