Sign up Calendar Latest Topics Donate
 
 
 


Reply
  Author   Comment   Page 2 of 2      Prev   1   2
Noknroll

Avatar / Picture

Sr. Member / Supporter
Registered:
Posts: 843
Reply with quote  #16 
CSI has a function that removes those dominant frequencies you mention which brings the less dominant frequencies to the fore. Must admit I've never used log seriously, only dabbled.
RustyCas

Avatar / Picture

Admin
Registered:
Posts: 1,810
Reply with quote  #17 
Quote:
In VA Tab, right click just outside the X or Y axis and select Log


VA Tab.... what is that? [wink]

Quote:
very useful in some situations particularly crushers, reciprocating engines and vibrating screens 


That is an excellent point, and really not something I've ever thought of, perhaps because I tend to rely a great deal on "maximum peak acceleration" trends which are mostly immune to "running speed" events which are typical of crushers and vibrating screens/conveyors.

__________________
"The trend is your friend"
Ralph Stewart

Avatar / Picture

Sr. Member
Registered:
Posts: 235
Reply with quote  #18 
Quote:
VA Tab.... what is that? [wink]


You are probable like I am. Still using Plotdata more than the new program "VA" ("Vibration Analysis").[smile]


The tabe should be in the upper left corner of the screen.

__________________
Ralph Stewart
AlertAnalytical.com
RustyCas

Avatar / Picture

Admin
Registered:
Posts: 1,810
Reply with quote  #19 
Oh, I know where the VA tab is... I just don't go there unless absolutely necessary.  Kind of like the proctologist.
__________________
"The trend is your friend"
JasonTranter

Member
Registered:
Posts: 29
Reply with quote  #20 

Seeing that John made me the mad professor of log scales [wink] I should explain.

So, first, I learned log scales at DLI, I did not teach them log scales. I was a linear boy before DLI.

Second, it is the log amplitude scale I am referring to; the log frequency scale is just weird [wink]

Third, at DLI they used VdB - you had to get used to the numbers, but one thing for sure, you never had to adjust the amplitude scale - everything from the highest amplitude peaks to the lowest amplitude peaks showed on one scale. I used a fixed scale and at a glance I could learn an awful lot - once my brain was calibrated.

Fourth, the fact is that the log amplitude scale will show you patterns that you can miss in the linear scale - but only if there is a peak in the linear scale that is high enough so that the detail in the "noise floor" could not be seen - but you probably aren't analyzing super-low amplitude data.

So, what could I see in the log scale that could be missed on the linear scale?

1. I often saw pole-pass sidebands that were very hard to see in linear.


2. You could see harmonics and sidebands way before they showed in the linear spectrum.


3. You could clearly see bearing faults much earlier. They would be invisible in linear. Now, if you don't want an earlier warning then you would not bother with log (or PeakVue, shock pulse, enveloping, etc.)


4. You could easily see where the resonances were lurking. Whether you were concerned about the damage resonance can do or not, it certainly helped to see where the amplitude of 'normal' peaks were inflated due to resonance, not due to the underlying fault condition.

5. You could also see other issues that caused the noise floor to lift - impacting (e.g. due to looseness), wear, rubs, lubrication issues, cavitation, turbulence, etc. But you need to compare to older data.

You don't have to use it permanently, but if you want to see the future, toggle to log.  But don't do it too often because you may like it and there is no going back [wink]


John from PA

Sr. Member
Registered:
Posts: 963
Reply with quote  #21 
Quote:
Originally Posted by JasonTranter

Seeing that John made me the mad professor of log scales [wink] I should explain.



Ah, but I did not place blame upon you, and in fact never knew of your association with DLI.  What I said is
Quote:
Azima, formerly DLI, professes that log is better.


and I do go further to state that in some circumstances it is better for catching early distress.  This thread is also a spin off from http://www.machineryanalysis.org/post/vibration-fail-pattern-unknown-anybody-know-about-this-9768889?highlight=pattern&trail=15 where someone sought help in identifying a "pattern".  We all had difficulty trying to identify what the OP meant by "pattern".

JasonTranter

Member
Registered:
Posts: 29
Reply with quote  #22 
Sorry John, I was referring to a different John; the person who made the original post. And I know him well enough to have a joke with him.
Vibe-Rater

Avatar / Picture

Sr. Member / Moderator / Supporter
Registered:
Posts: 1,398
Reply with quote  #23 
Yes definitely Jason.  Sorry to dob you in.  As you know I don't delve into the tiny peaks much.  Once you start looking in there I think I would end up in an asylum and a paper machine would take me 3 months to analyse.  And clients would be doing an enormous amount of maintenance.  The handy thing with paper machines is that the felts only last so long. 6 - 8 weeks which dictates the stoppages during which lots of other work is done.  Very different to say a refinery where there are lots of standby's so they can be swapped and repaired without disrupting production.  So in paper there is much more opportunity to "run it to the next shut".

So I stand corrected, I always thought that you were the reason DLI pushes log amplitude not the other way around.

Anyway, I was just interested to see how may might fall into each camp.

I guess I came up with the question because of recently meeting, wait for it, a double Phd in condition monitoring and ISO Cat 4 (you will have met him Jason).  That boy is obviously extremely smart. Unfortunately he is not that practical. Had to explain why i called a changeout on a pump bearing (which he misdiagnosed previous survey). Man did he go technical whilst the deteriorating waterfall plot was right their in front of our eyes  inner, outer race and rolling elements clearly there.  Maintenance didn't carry out the work.  Then all went flatline.  Waiting on the old bearing if it ever shows up. Luckily there is a sister pump / standby. we'll see how it goes.  Will be at that site again Mon and Tue. rgds


Previous Topic | Next Topic
Print
Reply

Quick Navigation:

Easily create a Forum Website with Website Toolbox.