Even before the maintenance crisis crashed onto the stage, Plant Services believed two truths about our favorite discipline. First, we hold that no maintenance professional can get too much education. Second, we say that cross-training in technical subjects holds no downside risk. In fact, those beliefs have been among the motivations that drive this monthly column into topics that have some rational relationship to how your maintenance department operates.
It’s well accepted that predictive maintenance is one of those best practices that can help keep your company’s market offering from morphing into an import that wreaks havoc on our international balance of trade. The two predictive technologies featured here are infrared thermography and vibration analysis. As you explore the Web citations, we hope you learn something new, and then see if it works in your current reality on the plant floor. So, join me in search of practical, zero-cost, noncommercial, registration-free resources that will help you decode the shakes and see the invisible. Remember, we search the Web so you don’t have to.
Vibratory big picture
As is the case with any predictive technology, the idea is to evaluate some measurements and, if necessary, engineer a fix. What happens between those end points is the essence of maintenance excellence. A roadmap might be useful for achieving some continuity and best practices within your operation. So, for the case for vibration, you could do worse than to shake a mouse at www.machinedyn.com/revised/training/tutorial.pdf to read “A brief tutorial on machine vibration.” It’s an article in which Victor Wowk, P.E., founder and president of Machine Dynamics Inc., Albuquerque, N.M., argues that not all vibration is bad. Afterward, his 10-page article includes a decision tree that can take you from performing vibration problem diagnostics to identifying possible remedial measures that can stop the shakes. He provides a tabulation of suggested vibration limits for electric motors, generators, fans, pumps, compressors, gearboxes, internal combustion engines and turbines. Finally, you’ll find hints for analyzing the data, tips for evaluating the conclusions, ideas for root-cause analysis and a list of some corrective measures that might work.
Scribd is the name of a company founded in 2006 by Adler, Friedman and Bernstam in San Francisco. It claims more than 20 million visitors each month and more than 17 billion words in its library, which purports to be five times the size of Wikipedia. The idea behind Scribd is that it’s a vehicle by which people can publish their own material for the benefit of the rest of us who don’t. How altruistic. An example of the relevant content you can find there is a 72-page article titled “Vibration Analysis.” The target audience is anyone concerned about vibration problems with industrial equipment. Wiggle over to www.scribd.com and enter the phrase “vibration analysis” in the search box. Look for the entry with the correct title and page count. The meat of the piece begins on page six, where the lesson starts with springs and simple harmonic motion. Some 50 pages later is where you finally get into machine vibration-monitoring and tips that can help you make the most effective use of your vibration monitor. You can print the document or read it online. But, there appears to be no way to download it anonymously.
Become an expert
The great advantage of collecting vibration waveforms digitally instead of through analog methods is that you can use digital signal processing in the form of the Fast Fourier Transform (FFT) to convert the time-domain signal you collected so easily to the frequency domain signal that you can analyze so easily. Don’t view that conversion process as some sort of black magic lying beyond the comprehension of mere mortals on the plant floor. Instead, get a copy of “The Scientist and Engineer’s Guide to Digital Signal Processing” by Steven W. Smith, Ph.D., president and founder of Spectrum San Diego. This 640-page, 34-chapter book has been available for years and has undergone several revisions. The author claims that digital signal processing is one of the most powerful technologies that will shape science and engineering in the 21st century. He explains the eight reasons you should master the technology. He even offers the software that performs the signal processing. The best part of this amazing discovery is that everything mentioned here and much more are available as free downloads at www.dspguide.com. All I can say is that you’ll be doing yourself a disservice if you don’t investigate this opportunity.
Some people have a natural talent for visualizing the effect that changing the variables has on the outcome of a particular analysis. They think in numbers. Others need something more tangible. They might need a picture to see the effect of tweaking the unknowns. That fact seems to be the impetus behind the Wolfram Demonstrations Project, an open-code resource that uses dynamic computation to illuminate mathematical concepts. Developed by Stephen Wolfram at Wolfram Research Inc., Champaign, Ill., the project features more than 3,400 demonstrations that can help you get a better grasp on the technology. For example, you might want to explore the FFT. Shake that mouse toward http://demonstrations.wolfram.com and enter “fast fourier” in the search box. Then, look for the demo titled “Frequency Spectrum of a Noisy Signal.” At this point, you’ll only be able to preview a canned demo. If you want to explore on your own, you’ll probably need to download the free Mathematica Player first. Just follow the download instructions. Don’t stop there, though. Go back to the home page and find additional relevant demos that might be useful for coaching and educating your maintenance technicians.