When you (someone who has an active ultrasound program) are using airborne ultrasound "contact" probe for bearing analysis, how are you determing failure? Is is by comparing (Comparative Method) the bearing in question to another similar bearing, Trending (Historical Method) the bearing over time and/or are you listening to the bearing and saying it is bad (SnapShot Method). When do you determine the bearing is to be REMOVED?
What is your experience?
Please NO generic ultrasound manufacturers answers. I would like to know what you the technician are doing...

This message has been edited. Last edited by: Jim,

Jim,

Are you using an ultrasound microphone that is some distance from bearing to measure airborne emissions, or are you using a contact sensor to measure structureborne emissions? A microphone (in meter or remote) can be used, but it is more susceptible to background noise.

I use a contact sensor and the following procedure:
1) Meter value (typical level and peak/maximum level)
2) Sound characteristic (friction or impacting)
3) Dynamic level (type and amount of level variation)
4) Severity Chart (developed by experience for level and sound)
5) Plot Comparisons (points on same machine or same points on similar/identical machines)
6) Plot Trend (if several surveys are available)

There are no industry standards for ultrasound, so interpretation must be based on instrument, measurement method and experience. I typically use both vibration and ultrasound measurements for bearings and rely on one or both together (depending on the situation) for making recommendations for maintenance actions.

Walt

Thanks Walt, I rephrased the question. Although there is no industry standard, I know that there are techs using numbers that manufacturers have suggested. The majority of airborne ultrasound users do not use vibration with ultrasound.
Thanks
Jim

Jim,
I listed the method that I use. It is based on a SDT 150 meter with a 097/095 contact probe. I have no problem relying solely on ultrasound to find faulty bearings, but I find combining with vibration to be more effective. It is my experience that instrument vendors offer only very general guidelines or numerical criteria. These can be trouble if followed exactly. I have tried several ultrasound meters over the last 10-years and none worked better than what I use. Perhaps a new product will come along that equals or exceeds the performance and is easier to use.

Walt

Jim,
Based on the "Views" for this topic, it appears that not many are interested in using ultrasound for bearings. Are there any folks using ultrasound effectively? Have any folks tried ultrasound for bearings and been dissatisfied with the results? In either case, it is important to know what instruments and measurement methods were used. I know from my experience in using several meters and sensors that only a few are very useful. I know that Electricpete has degraded the ultrasound method on the Engineering Tips Forum. If anyone has not done bearings correctly, then you don't know what you are missing!
As stated earlier, I don't rely on the Trend-Method (described in your Uptime Magazine article). I use the Snap-Shot-Method based on numerical level, level variation, and audio characteristics. This is supplemented by my own Ultrasound Severity Chart and the Comparative-Method when I typically use MS Excel to plot measurement points by machine or by multiple machines.

Walt

Walt,

I am interested in using ultrasound for bearings. Unfortunately I don't have any information to share here. Gosh knows I would love another method of determining a bad bearing. It is like rusty said on a topic in the vib section, If I could taste a bad bearing I would do it.

Aubrey

Aubrey,

An associate and I gave a presention on "Ultrasound for Bearing Condition Monitoring" in October 1996 at the CSI Reliability Week conference in Nashville, TN. CSI published the procedings. This came well before CSI's ultrasound meter was designed or sold. We updated the PowerPoint presentation in 1999 for conducting training programs for marine engineers for shipboard machinery applications. We have been using the same meter, sensors and procedures for over 10-years. You can find many manufacturers and sales persons that depict ultrasound as deceptively easy for bearings, yet there are few that can apply ultrasound consistently well on a wide range of machine types, bearing types, speeds and loads. There is really a lot to learn.

Walt
w_f_strong@msn.com

Hi Walt and Jim,
since Jim posted "Please NO generic ultrasound manufacturers answers" in his original post, I have sat back and kept quiet.
Like you Walt, I have had many successes using SDT ultrasound both on its own and also in conjuction with vibration. I have also presented many papers on this subject and will be doing so again at PdM2006 where I will be comparing data taken under both methods on the same bearings.
In answer to your original question Jim, all three methods are valid and have their uses as they do in all branches of predictive maintenance.

Best Regards,
Tom Murphy

Aubrey,
You said you have an interest, but do you own any ultrasound instruments that may be suitable for bearing fault detection? If you do, then you need only to focus on good meaurement procedures. If you don't, then you need to "try before you buy" several meter that are advertized for the task.

Walt

Have you got any example of good / bad low speed conveyor bearing and conditions of measurement (frequency of sampling and sensitivity)?

Or some example of rotating dyer supports which rotate more or less 10 RPM.

We have a lot of difficulties to isolate ultrasound noise from bearing from other ultrasound sources because lot of competing ultrasounds.

Do you consider some sensitivity display limit for bearing testing in noising environment? UE users talk about sensivity limit (right narrow blinking), but in some environment extra important noise is heard in this sens limit. Comparing with similar bearing is ok, but sometimes it is difficult to find the right bearing.

If you could find some file to my email address, please indicate with files measurement characteristics.

Compare please this file from one steam dryer with the file in next post, which in similar condition from the other steam dryer. Both files with sensitivity 55 and frequency 34 Khz with Ultraprobe 10000

Regards,

Allen

FSV210_S55_f34_DRY11.wav (545 Kb, 38 downloads)

Other file from other dryer.

Similar measurement characteristics.

Regards,

Allen

FSV310_S55_f34_DRY11.wav (647 Kb, 33 downloads)