Sign up Calendar Latest Topics Donate
 
 
 


Reply
  Author   Comment  
ram

Member
Registered:
Posts: 27
Reply with quote  #1 
For some equipment I observed bearing fault frequencies are prominent in axial direction rather than in radial direction. That's also 2 to 3 bpfi or bpfo. Any one experienced this.
Vibe-Rater

Avatar / Picture

Sr. Member / Moderator / Supporter
Registered:
Posts: 1,397
Reply with quote  #2 
SRB's (Spherical rolling element bearings) particularly but also tapered rollers, axial is best.  Based on 25 years of doing this stuff - full time. rgds
Big Al

Sr. Member
Registered:
Posts: 102
Reply with quote  #3 
On what sort of equipment have you experienced this?

It's quite common to collect data on paper machine bearings in the axial direction as many people believe the defects are more prominent axially.

I'd also expect any machine using tapered roller bearings (or thrust bearings of any description) to show defects axially. Gearboxes are generally well worth axial readings for this reason.
OLi

Sr. Member
Registered:
Posts: 1,915
Reply with quote  #4 
I always thought that on papermachine cylinders it was due to the strange things used to cope with thermal expansion axially that made the bearings loaded axial and giving the good readout. So having CARB bearings in the machine it would be different?
__________________
Good Vibrations since early 1950's, first patented vibrometer 1956 in the US.
http://www.vtab.se
ram

Member
Registered:
Posts: 27
Reply with quote  #5 
I have experience with induction motors driving centrifugal pumps.my doubt is why it shows only 2 _3 harmonics of fault frequencies. Is it the case with ball bearing also.
Vibe-Rater

Avatar / Picture

Sr. Member / Moderator / Supporter
Registered:
Posts: 1,397
Reply with quote  #6 
Oli,  Carbs are the same.  Axial is best for them too.  I believe it has to do with contact angle which transfers signal to axial direction. I was on a paper mill once who asked for advice because they had regular dryer cylinder bearing failures.  They were taking vertical measurements at the tender side and at the drive side had permanent sensors located vertical behind the exposed gears.  I asked why and the answer was - recommendations by previous "consultant".  I told them to go axial as soon as practical.  Obviously time consuming to re locate the permanents, and virtually no more failures after work complete.

A famous Finnish paper machine company also places their sensors vertically arguing that imbalance is detected that way. Oli, I ask you. when have you seen an out of balance dryer cylinder? Compared to bearing faults they are virtually none in comparison in my experience. (zero in my experience but you might have seen them).

What I have seen is when a dryer cylinder fills with condensate (problem) they generate a noticeably high 1x - but that is also easily seen in the axial direction.

Again, this is all my experience over 25 years full time on paper / tissue machines.  Others may have different experience. tifwiw. rgds
OLi

Sr. Member
Registered:
Posts: 1,915
Reply with quote  #7 

CARB don't have a angle it is rollers on a tube for inner ring so that's why I am curious, they do fail eventually, I have seen data but don't remember how it was collected.

Well you know, they have some funny ideas, we once was asked to define the resonances on every bearing in a machine so they could "block them in the system" we did

argue it was crazy and didn't do it. We have in connection with machine speedup balanced cylinders in situ with a hydraulic drive to drive one or two far apart since they influence each other badly so you can not run a set of 3 or so as they may be grouped when operating. You also need a tiny guy to enter the weights and it takes some machine downtime. You should also balance as close to proposed running speed as they behave like spaghetti. Finnish cylinders are also balanced from factory with a plane in the middle that is good. I also once seen a machine run with a slight over fill of condensate that did compact a number of concrete foundation pillars. We could after the event confirm cracks by looking at transfer function of vibration each side of the suspected cracks during operation and it was in turn confirmed by core drilling..... So some more concrete required. However I guess that event would be found even axially or in the coffee room..... So no I see no need to monitor vertically either part from the CARB case where I am still suspicious. Installations I have seen are not really setup the way I would like either but it is a different story. Mostly better looking data compared to other brand installations where the dominating was LF hum from ground loops anyway.....


__________________
Good Vibrations since early 1950's, first patented vibrometer 1956 in the US.
http://www.vtab.se
OLi

Sr. Member
Registered:
Posts: 1,915
Reply with quote  #8 
In my view, if you plot 10x calculated brg. freq's and minimum 3 match with real data then you have a match and find that fault when taking it apart. If you match at the higher end then the damage is newer and if you match lower down it is older as the damage flat out when running giving lower freq's. If you match the primary calculated freq. and even worse, from several fault types, you need to swap bearing pretty soon, thus only my 1 SEK. 
__________________
Good Vibrations since early 1950's, first patented vibrometer 1956 in the US.
http://www.vtab.se
Previous Topic | Next Topic
Print
Reply

Quick Navigation:

Easily create a Forum Website with Website Toolbox.