I have always been very intrigued by the phenomena and science behind vibration condition monitoring. It is a branch of engineering study that intricately combines material science, motion & dynamics, instrumentations and advanced mathematics. When I finally understood how a plot of amplitudes versus frequencies can provide different insight than the usual amplitudes versus time, it was a major “Wow, Fourier was a genius” epiphany moment of my youth. That was Year 2 Engineering School (July 1985) by the way.
I am sure all of us sensible, rational people can easily see that Joseph Fourier is therefore way more cool, sexy, significant and attention-deserving than Louis Vuitton or Christian Dior.
Let me state it right from the start – unlike LV or CD, Joseph Fourier is not a French Fashion house.
Yes, Joseph Fourier was French – A good looking French mathematician and physicist (1768 to 1830) who did mathematical research leading to what we termed as Fourier Transform today. Much of our vibration analytics today are still based on Fourier Transform which analyzes complex vibration by superpositioning (i.e. geek way of saying combining) a variety of sines and cosines waves of different but related frequencies.
When my colleague Steven Png told me a few years back that measuring and analyzing a different kind of waveform from the usual rotation and related frequency spectrum via Fourier Transform, I remember feeling a little skeptical and also protective about my hero Joseph Fourier.
“More useful than our usual rotation vibration spectrum?”
“Well – not more useful but useful in a different way. …the waveform we are trying to capture is NOT the rotating vibration, but the impact stress waveform detected when metal surfaces knock on each other. If there are micro cracks or abrasive wears on those colliding metallic parts, stress waves can be detected for a very short period. We called this the PeakVue waveform.”
“Wow – so potentially if we can capture those PeakVue
waveform… we can detect early micro cracks on bearings? I mean anti-friction bearings would have metal parts colliding all the time, am I right?”
“Yes – and also early fault detection of gear teeth.”
I am impressed – I mean just think about the kind of early fault prediction you can achieve. However, Steven went on to explain that while the stress waveform phenomenon is potentially useful, the data analytics is not straightforward – a combination of spectral analysis (Wow – Fourier again!), autocorrelation, separation of periodic and non-periodic data, etc.
“Sounds like we need a system of sensor plus PC/Analyzer to be connected to every bearing?”
“Not if you have Emerson’s condition monitoring solutions with PeakVue Plus embedded. Even our standalone Wireless Vibration Monitor has PeakVue Plus capabilities embedded. I can tell you more during the Asia-Pacific Emerson Exchange…”
Well, there you have it – sign up for the Emerson Exchange Asia Pacific to find more from Steven Png who will be presenting on the topic of Leveraging Machine Protection Systems for Predictive Maintenance on 26th Oct 2021. I don’t think he will be wearing any Louis Vuitton or Christian Dior. Remember – he is pure geek.