Hello to everyone.
I'd like to say thanks to T. O'Hanlon for helping to keep the board alive and also thank you to Joe Peterson. This board has been a great help for me as a stand alone newbie just by reading all the old posts.
I have a bearing that's showing a multiples of 1x in the Peakvue is this an indication of looseness or possibly 1 impact per revolution?
Viewing the polar plot shows the event happening in the same spot every revolution if I'm using the Polar plot correctly. The standard spectrums not showing any indication of increasing looseness. This is on a belt driven Band saw wheel.
I'll try attaching the data - usually kicks my butt.


Images (1)
Original Post
Duncan - not sure what you mean by "that goes only as high as the inner and outer race characteric frequencies" would you be looking for an Fmax of the 1x of the inner race frequency?

Ron - Mills going good but I'm partially insane. In the process of changing our CMMS over to One World, entering equipment,cat codes and rebuilding outdated pm's and creating ones for equipmnet that was over looked in the origianl set-up. My predictive program is slipping but the days are only 24 hours long. Lots of OT doing both jobs

If you use the AutoCorellation plot of the PeakVue TWF you may clearly see that there is actually some other (higher then 1x) frequency being modulated by 1x. I have, kind of, flipped over the approach as to what is the carrying frequency and what is the modulating one, but why not if that is what you'll see in the plot?

Could you please post the AutoCorellation plot with the cursor on the higher frequency?
Hi Wally,

Post us a normal velocity spectrum with an Fmax of 1K and 800 lines(vertical scale of .1 in/sec)and a waveform (4096 points) .
The Peakvue you are showing here only has a HP Filter of 500 Hz, so whatever Peakvue is showing could be fairly visiable in the normal velocity spectrum and the Peakvue 1X might be sidebands off some carrier or even harmonics carrying from the normal 1x to out past 500 Hz.

Email me the point in a small database if you would like.
David – You mention a cracked inner race, how is it a cracked inner race would show up at a 1x TS in a Peakvue? I've heard it mentioned by the papermill guys but have never seen one or any data from one.

Ralph, David – I'll Post a normal velocity spectrum from work in the morning.

Ralph – How do I create a small database for mailing?
Ralph – How do I create a small database for mailing?


It's pretty "long-winded" to post here.
But I'll try. Smiler

Extraction of Data Using RBMware

1- Open the Database Utility Program
2- OK copy/move
3- On the left half of the screen you should see a database name, click the plus to the left of the Dbase name
4- Click the plus on the left of the word "station list" which should be under the database name YOURS MAY SHOW SOME OTHER NAME LIKE "AREA" All of your names may be different than mine, but that doesn't matter as long as you follow these steps (I HOPE.) ( smile)
5- Click the plus on the left of the station name that contains the data we need showing under the "station list"
6- Click the plus on the left of the "machine list" located under the station name
7- High light the name of the machine that contains the data
8- Click the "file" word at the top of the screen on the left and choose "NEW FILE"
9- You should see a window with 2 boxes, one saying local host if your system is a stand alone and the lower box blank for you to type in a name for the new database. Example: emailtoralph.rbm
10- OK this.
11- You should then see a screen wanting you to choose what RBMview file to associate with the new database, here type in a name example: toralph.mdb and then you choose the "CREATE NEW" option and then confirm
12- You should see a new window with the new database name
13- Go to the top of the screen and click "window" and then "tile horizontal"
14- Click and drag and drop the machine you high lighted in step 7 to the "area list" or station list, whichever is the name shown, of the new database
15- You should see a screen with 4 options, Structure, set, data, and external file options
16- Leave structure options set at whatever comes up on default DO NOT CLICK OK UNTIL ALL 4 TABS HAVE BEEN LOOKED AT AND SET
17- Click the data tab and check all 8 of the same boxes over the fields and set all dates to include data you want to transfer, for say, the last 4 months (if it isn't an awful huge amount, say 100 data sets per point).
18- At the bottom, leave the statistics at ignore
19- Click the "Set" tab and leave as is
20- Click OK. If you get a message that says "no rbmview data can be transfered" click "yes" to continue
21- The transfer should take place now
22- After finished, go back to Plotdata and see if the data is where it should be
23- If not, email me. If it's OK, then send me the toralph.rbm file and the toralph.mdb file
24- If the file is too big, say over 4 meg, you may have to redo and choose a different start date
25- Good Luck.
Well I opened up the bearing and took a look at one of there breaks today everything was as it should be, pulled the data that you guys were asking for and it seems to have pulled it's head back in, physical impacting noise heard thru the headphones is gone, Peakvue isn't showing the repetitive impacting that was there, Hmm where do I go from here, it physically sounded bad thru the headphones, data seemed to be telling me there was the start of looseness, but now it's gone? I'd say it was the amount of imbalance but the 2nd and 3rd spectrums down on the trend had the most definite impacting going on along with the noise to accompany it. Watch and see I guess. What would you do??


Images (1)
Wally, in case the question wasn't answered, an inner race crack will show up as a "non" synchronous peak in most cases. The time waveform will show a series of non-synchronous events spaced by 1x rpm as the crack passes through the load zone. It shows up very nicely in gSE TW or in your case Peakvue. Transient vibration can be a real headache. Did you use your strobe to look at the blade?
- electricpete, I took off the bearing cover to inspect for anything obviously loose like a spanner nut and washer, housing or shaft fit. Bearing sets inside what I call a bearing pot with the cover being an end cap that when removed allows for the bearing to be removed/installed axially on the shaft.
No bearing was not replaced just inspected for anything unusual.

- vibbase, no it wasn't answered, I always assumed a cracked inner race would show as an inner race defect in the spectrum.
Please correct me if I wrong - your saying that the event spaced at 1x TS in the Peakvue could be a cracked inner race and not 1x turning speed?
No I didn't strobe the blade, what should I have looked for? I wasn't concerned by the blade because I took data with different saws being ran on the wheel and they all seemed to be in good shape, not like some saws with rough areas caused by repairs or a bad area around the butt weld where the saw was joined together.

-Ralph, I'll try in the morning

Something I have noticed is that typically PeakVue will pick up damage long before it ever shows in the velocity spectrum (not always, but a lot of the time). The bearing could have possibly had some wear in it but it just didn't look like it when opened up. The PeakVue amplitudes were rather low in the data you posted. Something I do is when PeakVue readings are indicating problems (ie. looseness, bearing fault freqs etc...), I look closely at the amplitude and also see if it shows up in the velocity spectrum. If it doesn't show in velocity and the accel waveform pk/pk is low, I will keep an eye on it. PeakVue is a very powerful tool but you can call things too early if your not careful.
In the waveform, the impacting will be once per revolution which would make one think it would be synchronous. If you zoom, you may be able to count the little piggy back impacts that will equal bpfi. The thing here to remember is the bearing bpfi. It will be the fundemental frequency and normally be surrounded by one time sidebands. It also will be fractional and appear as non synchronous frequencies. The sidebands are due to the modulation of the inner race rotating in and out of the load or impact zone. One frequency altered by another = modulation.

Add Reply

Link copied to your clipboard.