Is there a minimum practical speed for a shaft to be running for ultrasound to be useful in diagnosing problems with it?

Mike,

The practical low speed is your patience. You want a time duration of several revolutions, so that cage rotation repeats several times. Hand-held contact probes can be both uncomfortable and inconsistent when trying to hold for a long time without movement. A magnet base sensor solves that problem. I have used a notebook/tablet PC with software to plot a strip chart of dB level Vs time to improve the accuracy of measurement over several minutes duration. Low shaft speeds can be a challenge for for vibration measurements as well.

Walt

quote:

Originally posted by Walt Strong:
Mike,

The practical low speed is your patience.

Okay, that's not a problem. I just wanted to make sure that there wasn't a critical surface speed below which the friction isn't 'fast' enough to produce noises in the expected ultrasonic range. Thanks for the reply, Walt.

In my experience I have used ultrasound on slow-speed bearings of as little as 8 rpms. I have had a radar site tell me that they took readings with their ultrasound and magnetic mount on bearings operating at 2-3 rpms for several minutes to an hour and compressed the time to look for defects. Typically the manufacturers of ultrasound may tell you in their manuals that 25 rpms is the lowest. Some identify "slow-speed" bearings as anything under 300 rpm's, for several years the vibration kurus would tell you vibration wasn't useful under 300 rpms so the ultrasound industry apdopted the 300 rpm as the range for slow-speed.
As Walt stated a good magnet and portable laptop/PC and lots of time should yield results.
I have actually taken a PC with a spectral analysis software (SpectraPlus in "RECORDER MODE"), a patch cord, an Ultrasound Instrument tied directly to the input side of my PC, and a magnetis contact probe on the bearing housing and recorded several minutes (30 min.) before on bearings of 10-12 rpms.
This practice is not just limited to slow-speed bearings if you do not have the vibration equipment available or you need a second view on a vibration problem you don't feel your vib. unit is giving you the answer your looking for.
In just a few months I believe the ultrasonic worls will be doing a lot more waveform analysis (time series) than ever before. The problem will be giving everyone the calibration settings and/or a standard to go by to understand what you are looking at.

Hi Mike,
I have used ultrasound on bearings which rotate 120 degrees in about 10 minutes.

Best Regards,
Tom Murphy

quote:

Originally posted by Tom Murphy:
Hi Mike,
I have used ultrasound on bearings which rotate 120 degrees in about 10 minutes.

That is _quite_ slow. If it works on those bearings then it should certainly work for my bearings. It _did_ work on those bearings, yes? Actually, is it easy to tell whether it's 'working' or not? That is to say, can you tell the difference between a poor reading and a good bearing?

Mike

Yes Mike,
it did work and I could find the bearing that had the failing lubrication.
The simplicity with which you can find these problems is often difficult for non-believers to grasp until they have either found a problem for themselves or been on a training course.
I had 24 wide-eyed converts last week on a training course - 24 experienced engineers with considerable vibration and thermography expertise who went away discussing how they were going to implement this simpler ultrasound method for example as a pre-filter to their 18-25,000-point per month vibration programs.

Best Regards,
Tom Murphy

Hi, I have just joined the forum and have scanned a number of postings on this subject. First if you are serious about measuring bearings, especially slow speed and by that I mean as slow as 0.25 rpm then using a high frequency accelerometer is just not up to it. Even 40KHz will not produce a suitable signal to noise ratio. A well astablished product on the market is the Holroyd Acoustic Emission meter. This works at 100KHz and is not sensitive to probe orientation, bearing type or speed. Like true ultrasonics (flaw detection types) you need a couplant (grease) and the sensor can be positioned where the highest reading is found (use earphones). Bearing or gear box Distress(R) is measured. Above a value of 10 needs re-greasing if the levels gow down and stay down then lack of lubrication is the problem, a reading >15 is a cause for concern. The instrument also indicates % to failure. For the slowest speeds the online unit measures 4 parameters: Distress, Peak (for single defects) Intensity (average severity) and Extent (area of damage). The same unit can be used with a sound pressure sensor which is extremely accurate for pressure and vacuum. Messing around with standard FFT analysers and accelerometers willl produce erratic and noisy results that need some expertise. Taking a Distress reading can be done by anyone in a few minutes of instruction

Hi Mike
Like Peter's posting, I use the Holroyd Pro for measuring shaft speeds of 3RPM and have had some good results. The hsoftware isn't as complecated to use as the vib stuff and it is really easy to set up. It will indicate you may have a problem but not what the problem is.