Sign up Calendar Latest Topics Donate
 
 
 


Reply
  Author   Comment  
Curran919

Avatar / Picture

Sr. Member / Supporter
Registered:
Posts: 599
Reply with quote  #1 

Related to some previous threads I've started, we are having some trouble negotiating with motor suppliers on allowable vibration levels.

What it comes down to is what are the failure modes of a motor experiencing external vibration. All motor specific standards (IEC and NEMA) do not give vibration limits for an in situ motor. They only give vibration limits for an unloaded, rigidly/compliantly mounted configuration. That's because we generally treat motor vibration as a diagnostic tool to identify a number of motor faults, most of which are not caused by vibration. So when the motor itself is not causing the vibration, you assume there is nothing wrong with the motor.

However, when the motor is mounted less-than-rigidly on other equipment, an otherwise healthy motor will be vibrating at higher amplitudes. The high vibration is not an indicator of a motor fault (or even any fault). If it were a pump, high external vibration is bad, as high casing vibration is strongly correlated to mechanical seal failure, regardless of the source of that vibration. But the motor obviously has no seals. Someone stop me if I am wrong:

  • roller bearings lifetime is much more highly correlated with static force rather than dynamic force. A motor vibrating at 8-20 Hz is not going to affect the bearings greatly. However, coupling the structural system to the rotor system with journal bearings may react poorly to external casing vibration.
  • If the motor vibration is due to resonance that leads to a high deformation at the motor foot, high vibration could lead to mechanical failure of the motor casing/foot. This would have to be really high vibration, and if the motor is vibrating with the equipment it is standing on, this wouldn't matter.
  • The air gap becomes inconsistent which could be a problem (?), but again, this would probably rely on there being lots of relative movement between the rotor/stator, which would probably require a shaft critical speed.

Are those a good reason to blindly follow a useless standard like ISO 10816-3? Regardless, motor vendors (rather, their sales engineers) are not accepting our higher calculated guidelines for motors on vertical pumps, preferring instead a meaningless limit that is lower than the pump limit. This is why overall vibration limits are never even mentioned by NEMA or IEC.

I'm setting up a meeting with ABB on this, but I want to make sure there is not something I am missing.

EDIT: I would be very happy for a reference on this.
OLi

Sr. Member
Registered:
Posts: 1,918
Reply with quote  #2 
I have seen a motor driving a mover/feeder rubber band next to a vibrating crusher at 60+ mm/s for 10+ years (it never failed). Motor was a bit fuzzy but didn't care since the vibrations did not significantly increase bearing load and did not create shorter bearing life as contrary to when a fan is unbalanced and mounted on a motor where the bearing load is directly increased. 
__________________
Good Vibrations since early 1950's, first patented vibrometer 1956 in the US.
http://www.vtab.se
Curran919

Avatar / Picture

Sr. Member / Supporter
Registered:
Posts: 599
Reply with quote  #3 
Quote:
Originally Posted by OLi
I have seen a motor driving a mover/feeder rubber band next to a vibrating crusher at 60+ mm/s for 10+ years (it never failed). Motor was a bit fuzzy but didn't care since the vibrations did not significantly increase bearing load and did not create shorter bearing life as contrary to when a fan is unbalanced and mounted on a motor where the bearing load is directly increased. 


This is my thinking exactly. We had a diesel driver and BB pump on a skid together. The whole skid was resting on antivibration mounts. So naturally the driver shook the skid and everything on it. The pump isn't actually vibrating at those levels. Nothing is wrong with the pump. Stiffening the pump pedestals won't change anything.

Unfortunately, its still a pump, so external vibration is still bad, but you have to also consider where exactly the vibration is coming from.
OLi

Sr. Member
Registered:
Posts: 1,918
Reply with quote  #4 
I know there is a connection btw. mech seal life and pump vibration, some years ago a papermill made an effort to reduce vibration of the site machines including pumps and the most noticeable was that consumption of mech seals reduced by near 50% in that year of the project compared to the previous. If that is a similar problem, that the vibration force must pass the seal area like with the bearings or if it is any vibration as when you shake the complete assembly, I don't know, have not done any test of that as far as I know. I am afraid that it is more like any vibration is bad for mech seals but I don't know.
__________________
Good Vibrations since early 1950's, first patented vibrometer 1956 in the US.
http://www.vtab.se
electricpete

Sr. Member
Registered:
Posts: 651
Reply with quote  #5 

Quote:
what are the failure modes of a motor experiencing external vibration

I agree with your thought process. Nevertheless here's my attempt to brainstorm effects of vibration, roughly in the order of the importance  that I perceive them.

You mentioned sliding bearings.  I agree those can be problematic. For one thing, relative motion of shaft/bearing can disrupt the oil film. I've had several sleeve bearing motors that damaged their bearings due to vibration. I think this is the basis for establishing relative displacement limits on the order of half the diametral clearance (above this the likelihood of oil film disruption increases).    

Vibration may be a symptom of misalignment.  Misalignment can load the bearings.  (Yes, you verified the alignment, but that was at shutdown conditions, can you prove it's good during operation?)

Bearing insulation or electrically insulated mechanical parts can wear, loosen, and move under the influence of vibration, leading to a mechanical failure.   I have two case histories of this, although I wouldn't attribute them specifically to external vibrations rather than internal vibrations.  Age was also a factor (those electrical insulating parts were never replaced in 20+ years in both cases). 

Vibration may facilitate increased bearing movement within their housings, which over time can cause accelerating looseness, especially in presence of a rotating load.  Debris from the fretting can get into grease or oil.  I tend to think it is less a concern in vertical machine greased bearings where any debris from bearing housing wear would fall below the bearing and new grease is added from above the bearing.

Vibration can cause movement of windings in the slots and endwindings. Although in general these areas are well supported against movement from electromagnetic forces.  VPI winding pretty much eliminates any concern about slot vibration.

Bolted joints can loosen under the influence of vibration.  (not if properly tightened and locked, but who knows).

Quote:
roller bearings lifetime is much more highly correlated with static force rather than dynamic force. A motor vibrating at 8-20 Hz is not going to affect the bearings greatly.

I agree. But my intuition suggests that skidding might be more likely in presence of vibration.  In the portion of the race where vibration is perpendicular to ball/race interface it may lower the effective load. In the portion of the race where vibration is parallele to ball/race interface, the rolling speed of two rolling elements 180 opposite is influenced in opposite directions.  I've never seen it, just a thought. 

Curran919

Avatar / Picture

Sr. Member / Supporter
Registered:
Posts: 599
Reply with quote  #6 
Oli, indeed ANY vibration is detrimental to the mechanical seal, so my example wasn't very applicable.

Pete, thanks for your insights. I think most people just consider vibration as a catchall of dynamic forces that can be treated monolithically. And ISO 10816-3 sure doesn't help. I found one paper that discusses the failure modes, but I guess it is such a broad topic, that it can't specifically talk about the characteristics of the vibration that causes failures. Tables I-V just attribute possible failures to vibration or 'periodic dynamics'. Obviously, they don't describe associated magnitudes either. Obviously anything is going to fail if the external vibration is strong enough.

What I extracted from the article was that shaft vibrations (derived from unbalance) is a particularly common root of failures. Besides that, external vibration is bad for bearings when not in motion (false brinelling or fretting), but that doesn't apply to my case. I am fairly convinced that external vibration without relative vibration between rotor and stator is not as harmful to a motor's health as it is treated.
ivibr8

Sr. Member
Registered:
Posts: 163
Reply with quote  #7 
Curran
Am I misreading your comment about ISO 10816 being useless?
It clearly allows for higher vibration levels for FLEXIBLE mounting as compared to RIGID mounting for all the reasons discussed above.

Jim P
Curran919

Avatar / Picture

Sr. Member / Supporter
Registered:
Posts: 599
Reply with quote  #8 
Quote:
Originally Posted by ivibr8
Curran
Am I misreading your comment about ISO 10816 being useless?
It clearly allows for higher vibration levels for FLEXIBLE mounting as compared to RIGID mounting for all the reasons discussed above.

Jim P


Chapter 3 specifically.

It was supposed to be a catchall as ISO slowly developed the subsequent chapters (last version of chapter 3 expunged pumps from the list because chapter 7 had recently been published), because it is better than nothing (arguably). However, when they came to making a motor-specific chapter, they couldn’t get the necessary involvement/support from the motor vendors, who didn’t think a 10816 motor standard was necessary. From the vendors' points of view, NEMA, IEC and API 541 already were covering the important aspects of motor vibration: that being the internal excitation sources and primary fault detection. 10816-3 is overly broad, it applies equally to everything from everything from 8000rpm 50MW steam turbines to 120rpm 20kW DC motors. Unfortunately, its also authoritative. That means, in situations where it makes no sense, we are still forced by customers to use it. Its not good for the vendor or the customer.

For new machines, we have 2.3 mm/s for rigid and 3.5 mm/s for flexible. IEC also gives 2.3 mm/s for a rigid limit, but this is for uncoupled motors. NEMA only gives compliant configuration, but it is comparable to IEC's compliant mounting limit. Clearly, a coupled motor will have higher vibration (generally much higher), which means 10816-3 is much more restrictive than the motor specific standards. Even at the 3.5 mm/s flexible limit, this is always lower than the limit for VS pumps from any pump standard, including 10816-7. Why would the motor have a lower limit than the pump it is mounted on top of? What would the point be of having a VS pump limit then if the motor will always have higher vibration and the limit is lower than the pump limit?

If it weren't unnecessarily restrictive for motors, is giving higher flexible limits enough? Perhaps. Though I would argue that having two motors of equal speed and 'flexibility' (defined then by the first natural frequency), would react differently. Mostly, is the flexibility contributing to the low natural frequency (and therefore the flexible designation) driven by flexibility in the mounting (e.g. flimsy pedestal) or in the motor casing. In other words, what is the ratio of the motor reed frequency to the assembly reed frequency. If it is low, most deformation during vibration will occur in the motor casing, which is bad for reliability. If it is high, most deformation during vibration will occur in the mount/pedestal, which is better for reliability (though probably equal for rotor vibration).

Therefore, flexibility should be defined (at least for motors) as the ratio between the motor reed frequency to the assembly reed frequency. Not as a function of separation margin like it is in ISO 10816-3. And a lot of people DO use this as a criteria for flexibility. I got this from an unpublished paper by ABB, "HOW TO ENSURE LOW VIBRATION OF 2-POLE MOTORS ON FLEXIBLE FOUNDATION":

"Further, the vertical motors are excluded because the origin of harmful vibrations of these motors is usually the reed frequency [rather than foundation flexibility]. Even if the reed frequency problems are strongly related to the flexibility of the foundation, the phenomena of vertical motors are better suited to be considered separately."

A little vague, but I think we are getting to the same point.
ivibr8

Sr. Member
Registered:
Posts: 163
Reply with quote  #9 

Quote: That means, in situations where it makes no sense, we are still forced by customers to use it.

OK....I think I understand better based on the above quote.

When I started our Site's PdM program in mid 80's, I was also charged with writing a NAVSEA document to describe how the program would run. One of the chapters specified how we were going to evaluate the Severity.
I wrote it up in a manner I thought consistent with the available documents I could find. There was not a lot of ANSI/NAVSEA/ISO/etc. specs at that time and no internet to speak of (I didn't invent it yet [biggrin]).

The governing agency (those who would approve the final document) wanted me to specify a SINGLE level of vibration to determine when a machine was "Good" or "No Good - It needs to be Repaired" condition. l spent a lot of time, frustration, heartaches and headaches arguing that there was no such god level. The engine room's HPAC's and the noise-quieted hull-connected machinery were totally different.
After MUCH effort, we finally came to wording that allowed me to evaluate severity based on trending while referenced severity guides were an "initial" jumping off point <---that's why we are in the business of condition monitoring. 

But customers still need some quantitative assurance that a new machine will be delivered, designed as specified, installed with a high degree of quality and function properly for many years. Now that I am retired, many of my jobs are simple "does it meet spec" criteria from initial construction firms. I don't seem to be asked to perform condition monitoring anymore....so I am not seeing the frustration you currently feel. 

I have been to many VI conferences and recall a number of well-respected members of the governing committees discuss how the various specifications are administered, evaluated, voted upon and revised. I had the impression (and I guess would expect) there were many differing opinions depending on what was approved. No politics here, right?

Anyway, I've certainly rambled enough. Thanks for explaining your main concern.

Regards
Jim 


Curran919

Avatar / Picture

Sr. Member / Supporter
Registered:
Posts: 599
Reply with quote  #10 

Thanks for the insight.

I think you can see how this would frustrate me, at an OEM more than a reliability guy. Ultimately, most of y'all can use your discretion to give recommendations, but the limits are always there as a backup. I am handcuffed by them everyday and it is becoming increasingly frustrating. I've written specs on how to interpret the vibe standards for use in our factories, but the guys on the floor say it doesn't matter, because the customer is still specifying vibration limits from 10816-1:1995.

I know that I would be incapable of creating absolute limits that are more meaningful than what already exists. If there was just one authoritative source, I would probably not even notice if the levels were too low. However, its when one source disagrees with the rest that it becomes very obvious that the one source is probably not reliable, especially since it has the widest scope of any standard giving vibration limits.

OLi

Sr. Member
Registered:
Posts: 1,918
Reply with quote  #11 

Well I find you use 10816-1 when you buy, 10816-3 or/and others when you sell so if nothing is stated at the purchase that it rarely are, discussions do follow. There are problems both ways, number controlled purchasers that strictly require 1-2mm/s on a small VFD driven fans where background level is twice that to OEM wanting to have 14-15mm/s on a gbx right on the plant floor. It should be agreed upon to reasonable levels that are also cost effective compromises.
10816-1 was consumer driven 10816-3 are more supplier driven and later ones even more so as ISO participation can only be paid for and motivated by 3-letter OEM's these days so it will be even more so.
We have a document here from paper industry that we wrote the first of many years ago based on ISO at that time, it is now in it's 6th revision and I find requirements for 2mm/s on papermachines, I rarely see that in machines I monitor and some OEM declared it would be 10's of thousand engineering hours to get that. So it's a struggle.

 


__________________
Good Vibrations since early 1950's, first patented vibrometer 1956 in the US.
http://www.vtab.se
Curran919

Avatar / Picture

Sr. Member / Supporter
Registered:
Posts: 599
Reply with quote  #12 
When you say the 10816-1 limits were user driven, do you mean that VDI 2056:1964 was user driven?

I can definitely see how 60+ years ago and back to rathbone, most of the decisions were made independently by the users and the OEMs didn't get involved much.
OLi

Sr. Member
Registered:
Posts: 1,918
Reply with quote  #13 
Well, my father was there at that time in the ISO and they had meetings just the same as now but he was on the OEM side for most parts but maybe not the ISO 2372 that was based on VDI 2056. Remember he had the first vibrometer patent 1956 in the US. We had ISO2372 stickers on the backside of the vibrometers in the 1970/1980's and was conforming to that spec. I still have them in the museum shelf and some still are in use but more the 1980-1990 vintage that one or two come back for calibration yearly but I saw one 1970 vintage at a gasturbine plant last year, still working fine, better than the digital monitoring anyway. So I taught a long time that VDI 2056 was based on a scientific research etc. I have later had a different information that it was set during a long and hardworking dinner session but basically, compared to the standards and levels I see popping up in the new standards today I do see 10816-1 as consumer driven but maybe in comparison to the later documents, maybe not from it's birth. You may be correct if I adjust that to "machine driven"? Gearbox acceptance level 14+mm/s? What is that? Never in my foot can that be good for the machine....... Maybe just grumpy difference btw 10816-1 and 10816-3 2.8mm/s to 4.5 mm/s may not be so much to argue about so I can live with that. I am not Buscarello who claimed that "you can always save money by lowering vibration" I think it depend on what level you start also.
__________________
Good Vibrations since early 1950's, first patented vibrometer 1956 in the US.
http://www.vtab.se
RustyCas

Avatar / Picture

Admin
Registered:
Posts: 1,810
Reply with quote  #14 
Pete does a good job of laying out the concerns. I’ve never seen an issue on other than rolling-element-bearings (REB). For REB motors, external vibration on standby machines (installed or not) is a killer - death by a million pin pricks. The other major concern is loosening of motor hold-down bolts when exposed to high-cycle vibration. I don’t like split-ring lockwashers (the only type I see), but I use them when doing alignment on machines that see high-cycle external vibration from an adjacent or driven machine. But that’s a system failure issue, as opposed to a motor failure issue.

Other than those 2 situations, in 30+ years I’ve never seen a motor failure that I would attribute to external vibration.

__________________
"The trend is your friend"
Previous Topic | Next Topic
Print
Reply

Quick Navigation:

Easily create a Forum Website with Website Toolbox.