Go
New
Find
Notify
Tools
Reply
  
-star Rating Rate It!  Login/Join 
Posted
Does anyone have experience measuring acceleration online with an accelerometer with 4-20mA output?

I have a ‘FORMER CUSTOMER’ who has had several catastrophic bearing failures on one of his critical blowers. The fan is belt driven operating at approximately 2500 RPM with rolling element bearings. He’s asked what type of remote monitoring system would be the most effective for alerting operations of an impending bearing failure. He uses vibration switches with 4-20mA outputs on a few machines now, but he doesn’t have much faith in them (nor do I). I have several customers who measure velocity with 4-20mA accelerometers mounted on or near the bearing housings. These systems are generally used with sleeve bearings and primarily alert them of lower frequency vibration such as unbalance. In this case I fear velocity may not give them enough early warning to react to a rolling element bearing issue.

Any thoughts on using acceleration as opposed to velocity? Or are there better options?

I used the term ‘FORMER CUSTOMER’ above as I only do reactive diagnostic work and balancing for him these days as we where bumped out because the very large company that owns this plant has a corporate agreement with a very large reliability company. The very large reliability company sends a guy from very far away to visit the plant once a month and then emails a report. Then when a machine crashes and burns, typically on a Friday evening, I get the call to come out in the middle of the night to make sure everything is good when they start up!!! Sorry…this post wasn’t meant to be a rant! It’s just that the Corporate folks don’t see the added value from the local companies who know their customers, know their customers equipment, AND PROVIDE SERVICE!!!

I guess my last paragraph should’ve been saved for a later post. Anyway… any thoughts or recommendations that might help me sleep in on Saturday mornings would be much appreciated.

Thanks…


Michel
 
Posts: 53 | Location: Illinois | Registered: 04 March 2005Reply With QuoteEdit or Delete MessageReport This Post
Posted Hide Post
Sorry I don’t have direct experience with the main question you asked. In principle, velocity collected from the casing/housing of rolling element bearings can be used with good results for early fault detection. However, I would assume there might be a few general checks to consider to come up with a better evaluation.

1. Ensure the system correctly measures the required signal
a. Test the reading of the system and compare it against an independent vibration system (e.g. portable vibration analyzer).
b. The location of the sensor must be carefully selected. Signal may be significantly attenuated if the “resistance & damping” between the two locations (source of vibration and point of measurement) are large.

2. Ensure the system correctly sends the required signal.
a. Set point is correctly adjusted to reflect the genuine vibration level to activate the alarm.
b. Simulate the vibration signal using a shaker table to find out if the remaining loop functions well or not. This also will show if the trigger is linear with same amplitude at different frequencies or not.

I dealt with a similar system but the signal was integrated to mils. It did not activate the switch at the same amplitudes of mils at different frequencies.


For the other part of your post, I’d like to inform you that we do not sleep on the mornings of the Saturdays!! Saturdays are working days here. Big Grin


Regards- Ali M. Al-Shurafa
 
Posts: 129 | Location: To the east of Saudi Arabia | Registered: 07 June 2005Reply With QuoteEdit or Delete MessageReport This Post
Posted Hide Post
Michel,

Azima might have what you are looking for.

We do remote/online monitoring of accels, prox probes, pressure, current, etc.

The customer can login to the Advantage Reliability Portal and view their data (spectrum, timewaves, trends). Plus the system is setup with alarms to contact the plant when an alarm is triggered. Data is sent over the plants network or can be sent using cellular service. Azima analysts will also be viewing the data and reporting on issue that appear.

You can view a demo at www.azimainc.com

You can contact me if interested in learning more, jschlottman@azimainc.com. I am a field analyst for Azima. I can put you in touch with sales.

Jacob
 
Posts: 44 | Location: Evansville, Indiana | Registered: 21 December 2005Reply With QuoteEdit or Delete MessageReport This Post
OLI
Posted Hide Post
Vendor,
We have a version of our VA43T that from a single velocity or accel probe give 2x4-20mA one for 10-1000Hz vibration and one bearing condition, we have also made a package so 3 converters give 6x4-20mA for vibration, bearing and temp from 2 transducers accel/temp combo probes on 2 bearings for the protective minded. Specially made for fluff transport fan bearings in diper plants where the wooof factor is hi when bearings overheat. Normally the vibration alone will shut off before it´s too late but the margin may be slim, some more signal processing give some more margin to plan service and repair. Setting of alarm, warning and trend limits is as always most important. Olov


olov dot li at vtab dot se
www.vtab.se
 
Posts: 594 | Location: Linköping | Registered: 03 October 2004Reply With QuoteEdit or Delete MessageReport This Post
Posted Hide Post
If it is a critical bearing, use acceleration. I suggest 'peak' as opposed to RMS because it will show impacting that may be lost in RMS readings. Usually, well before lower frequency defects show up in velocity, you will see a steady increase in acceleration, and often there will be a dramatic increase just prior to failure. Acceleration gives the earliest warning and won't cause a panic over a simple imbalance or looseness (hold down bolts) issue.

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


Regards,

Rusty
 
Posts: 1254 | Location: Arkansas | Registered: 20 February 2005Reply With QuoteEdit or Delete MessageReport This Post
Posted Hide Post
Non Vendor

I suggest Rockwell XM122. I have installed several of these at local Aluminium plant and work very well!
Measures Velocity and acceleration (gSE) multiplexing, and allows either integration to control system (DCS), or direct configuration and analysis (spectrum) via USB. Also 4-20 mA output. Attached is outline of product for your interest.

Wayne

Word Doc122.doc (614 Kb, 31 downloads) XM122
 
Posts: 49 | Location: Dubai | Registered: 16 August 2006Reply With QuoteEdit or Delete MessageReport This Post
Posted Hide Post
Michel,
I have experience with both accelerometers and transmitter. I prefer the latter. They will give early bearing fault indication. As the element failure mode increases up to about what is considered Stage 2 to Stage 3 they will produce spikes in the output signal.

The ones we use now are METRIX Vibration transmitters. Model # ST6917.

If you want earlier detection IMI has a 640B series that also works great.


Barry Crawford
 
Posts: 56 | Location: Tennessee, USA | Registered: 23 February 2005Reply With QuoteEdit or Delete MessageReport This Post
Posted Hide Post
Hey Mitchel

You might check out a wireless vibration monitoring system from goulds pumps called ProSmart. Its low cost and collects a FFT every 5 seconds. You can set filter alarms that will alert you by cell phone, Email or Pager.

Regards,
Jack
 
Posts: 4 | Location: Houston, Texas | Registered: 22 December 2005Reply With QuoteEdit or Delete MessageReport This Post
Posted Hide Post
Michel, I agree with the others, acceleration for monitoring the bearings. There are many choices offered here and they are on the right path.
CTC offers a dual output type accel, that is powered by the "Loop" system and can be monitored in acceleration, and also provide a dynamic output for use with a portable data collector. The key to all of these solutions is in specifying the correct frequency spans and full scale vibration levels, and then education of the equipment owners to set the ALERT/ALARM levels correctly. My biggest challenge doing this is, that we often are trying to set these units up with a Process Control expert, with very little understanding of the vibration data.

I can offer assistance, if you would like, to help educate the owners to the advantages of offering some sort of reasonable cost sensor, that can be incorporated into their existing system. The key to explaining this, is to use the "DOLLARS" card.

I can help explain this to the Responsible parties, if you feel the need. (I often hear that the owner doesn't hear from their own back yard very well, but amazingly can hear quite clearly, across the country!!)
I can be reached daily at 1-800-999-5290, ext 831. Gary Kaiser


With mistakes, comes wisdom ;>Wink
 
Posts: 12 | Location: Rochester, New York | Registered: 26 November 2005Reply With QuoteEdit or Delete MessageReport This Post
Posted Hide Post
Our transmitters are offered in Velocity, Acceleration, Enveloped Acceleration and Displacement with dynamic output, loop powered or 24 Vdc. They are available with our label, or SKF, CSI, Goulds, Howden, Dresser, Dynalco, Woodward and others. Wireless solutions, internet, intranet, cellphone, HMI, etc. available. Happy to discuss offline.
 
Posts: 377 | Location: Gulf Coast - Texas | Registered: 14 July 2004Reply With QuoteEdit or Delete MessageReport This Post
Posted Hide Post
Gary,

I've had some email exchanges with a Sales Manager within your company. I've asked him to have someone put together a quote for everything we'll need. My experience is with portable data collectors and analysis, so this stuff is somewhat foreign to me. I shouldn't have any trouble identifying the alarm settings once the system is setup. My problem is knowing what I need in order to go after a quote for the customer.

I don't have any desire to sell something I cannot fully support, however if I could help out an old customer to whom I've remained loyal despite our current business relationship, it would be worth the time invested.

I'm sure others at CTC will get back with me, but if you could send me an email (chris@elect-mech.com) detailing what I need and what I can do, I would be grateful.

Thanks,


Michel
 
Posts: 53 | Location: Illinois | Registered: 04 March 2005Reply With QuoteEdit or Delete MessageReport This Post
 Previous Topic | Next Topic powered by eve community  
 


Copyright © 2004-2008 NetexpressUSA Inc. All rights reserved.