sleeve bearing scraping and blue-check / dry-roll check: common misconception

Many references recommend to do a blue-check or dry-roll check of new sleeve bearings upon installation into an electric motor which determines the need for scraping (to be done only by experienced qualified people). That used to be a little mysterious to me, but I think I have figured it out.

The purpose of scraping new bearings upon installation into a motor is NOT to make them rounder or smoother or to the correct radius (as some people seem to think) .... the bearing manufacturer is perfectly capable of doing that as long as correct specs are used.

The purpose of scraping bearings upon bearing installation into electric motors is to compensate for any misalignment between as-installed position of the shaft/bearing induced by misalignment between the two bearing housings or perhaps gravity sag of shaft. (sleeve bearings are not particularly tolerant of misalignment, especially as ratio of length to clearance increases).

If you have misalignment in the horizontal plane between bearings of horizontal motor, then you may see diagonal pattern. (the angle shows up much more dramatically than you’d think... if you have 3” diameter shaft with for simplicity 3 mils clearance and a horizontal angle misalignement of 3 mils per three inches, then angle you see is NOT 1/1000 which would be invisible.... it is closer to 1:1 since the pattern goes from one side of bearing to the other over the length of the bearing). [I added a new message below with example of this].

If you have misalignment in the vertical plane between bearings of horizontal motor, then the blue-check/dry-roll check pattern will be only on one end of the bearing, and not the other. That is the idea of the 80% contact spec for blue pattern/roll pattern.... it should be at least 80% of the LENGTH of the bearing (NOT the width!)

There seems to be a common mis-conception that 80% means you want 80% contact over the entire bearing (including width). That is just plain silly because the journal has smaller radius than the bearing and digging out the bottom of the bearing to form similar radius of curvature for full width would destroy the ability to form a convergent wedge.

How common is this misconception? See attached excerpt from article “Sleeve Bearings: A Modern Use for an Old Technology” by Richard Nailen (normally very accurate imo), published in Electrical Apparatus magazine, February 2008. On the 2nd page of the pdf, he says

quote:

[Nailen (2nd page of pdf):]
In assembling the bearing, then, the essential step is what’s sometimes called “blueing and scraping.” The make sure the babbit contacts the journal as uniformly as possible throughout the contact angle of Figure 2 calls for a fitting procedure like that given in the accompanying box [Figure 15].


Figure 2 is on the first page of the pdf and shows contact angle extending for 120 degrees centered on bottom of bearing.

The accompanying box [Figure 15] is dead wrong. The “before” of Figure 15a looks like pretty good contact to me. The after Figure 15b is full contact (ouch!). The caption for figure 15b says:
quote:

[Figure 15b caption, 2nd page of pdf]
Final fitting accounts for the difference in radius between the two curved surfaces, leading to the more uniformly distributed contact at (b).”
.

All I can say is ouch.. dead wrong. The difference in radius is a design feature and we don't want to change that.

He goes on to talk about various repair specifications with imprecise terms [I agree], but he never tells us that Figure 15 is dead wrong and in fact the entire context of how he refers to figure 15 and the caption of Figure 15 suggest this is supposed to tell us the correct pattern in Figure b.

I think I have an idea why there misperception that you want 80% or 90% over the entire bearing (rather than just a narrow strip down the center whose length is 80 – 90% for vertical alignment as it should be). When machinists work with FLAT surfaces they use percentage of contact of the ENTIRE AREA as determined by blue check to describe how flat and smooth the surfaces are. Sleeve bearing 80% contact should never be confused with that type of check... it is completely different. The after shown in Figure 15b would kill a bearing.
 
 
 
View Printer Friendly Format
By the way, not really related, but here is a photo of blue check from horizontal sleeve bearing motor which I believe represents horizontal misalignment (diagonal pattern of blue removal and slight wear from lower left to upper right of the photo). After about 30 minutes working on it with a gray scotch-bright and WD-40, put it back and tried again and the patern was right down the middle.

I don't know what the WD-40 was for. That's what the mechanic used. If anyone wants to explain it to me, please do.
 
 
 
Photos (1)
IMG_0113
 
Like Like (0 likes)
PermalinkView Printer Friendly Format
Thanks Pete... always learn from you.

During bearing installation, I always hear our millwrights discuss "twist and tilt"... and adjustments they make to the bearing to compensate for these misalignment characteristics by either shimming or making minor adjustments to the lateral axis to tweak the bearing position right in. I believe they use either plastigage or dial indicator swing-checks to make these adjustments. If these type of adjustments are possible, I assume it would be preferable to align the bearing in this fashion rather than scraping a bearing. If my assumption is true... do some installations have limitations to the millwright's ability to adjust a bearing's alignment position such that scraping is necessary?

Thanks for sharing your insight Pete. I always get smarter when I take time to read your posts.
 
Like Like (0 likes)
PermalinkView Printer Friendly Format
WD40 to scotch brite pad is like adding water to a piece of emery paper Pete. It allows you to get a much smoother finish than dry wiping. there will be no scratches when you get done, no matter how miniscule the depth of them may have been had you not used it. I've used and seen used about any lubricant or penetrating oil to help the scotch brite pad.

Thanks for the information on the installation. I used to get in some arguments with a knot headed supervisor I had in a strip mining job when we would pour babbitt bearings, and he wanted the full 80% contact.

D
 
Like Like (0 likes)
PermalinkView Printer Friendly Format
thank you epete for sharing...
 
Like Like (0 likes)
PermalinkView Printer Friendly Format
My experience with ScotchBrite is it breaks down physically and creates debris. I assume the debris includes some abrasive material.

I'm more than a little scared about using abrasives around bearings made of stuff that has embedability, like most any grade of babbitt.

http://www.clevite.com/techbul.../bearings/TB2076.pdf

Many Bronzes and aluminums can be successfully finished by honing even though the honing stones break down in use and produce lots of grit.


Finishing babbitt (and aluminum and bronze) is best done using a cutting tool even though the finishes the produce lack honing's benefits of crosshatch and plateaued finishes.

Most of our industrial bearings are pretty lightly loaded compared to some engine bearings, so we can "get away with" a lot.
 
Like Like (0 likes)
PermalinkView Printer Friendly Format
The concern raised by Dan is both valid and critical. I have seen too many inexperienced millwrights using emery cloth and/or crocus cloth in the absence of the proper materials. Industrial grade emery consists of the mineral corundum (aluminium oxide) often mixed with other compounds such as magnesia, mullite, and silica (sand). These particles can get embedded in a bearing surface and do tremendous damage to the journal when the machine is placed in service.
 
Like Like (0 likes)
PermalinkView Printer Friendly Format
Thanks for all the comments and interest.

MachineryWatch – Thanks for your comments. I interpret that most of your discussion applied to clearance at shaft-to-bearing and fit at bearing to housing. Plastigage checks these clearances. Swing check for vertical machines and lift check for vertical machines are also indirect ways of checking the combined clearance shaft-to-bearing plus beairng to housing. Shims can play various roles: If crush/interfernce exists at bearing to housing fit, then shims can be used between splitline for plastigage check.... shims might possibly be used on top of bearing to compensate for excess clearance between bearing and housing although we have to be careful if bearing is insulated and also have to ensure shim does not come loose to interfere with oil ring (I'm not positive whether this is good practice but I've seen it done). I have not heard of using shim for adjusting bearing alignment on a horizontal sleeve bearing motor....it is not mentioned anywhere in EASA or EPRI documents. If shimming is an acceptable method to improve contact pattern for horizontal motors (I’m not saying it is), then we would have to add shim at the rib(s) of the housing beneath the bearing and it would only provide an ability to compensate for vertical plane misalignment, not horizontal and probably not sag.


Dan and John – I definitely appreciate your input and respect your opinion.
At a minimum it highlights the need for very careful inspection of the babbit surface after using Scotchbrite. Whether or not it is acceptable to use Scotchbrite to begin with is something I’d like to discuss more.

Here are two references to consider:

Electric Power Research Institute (EPRI) Report GS-7352 “Manual of Bearing Failures and Repair in Power Plant Rotating Equipment” states:

quote:
EPRI
6.1.1.3 Journal Bearings. Minor jounral bearing repair, too can be accomplished in the field using hand tools. ...


Full, circular jounral bearings can be repaired in the field using a wide blade hand scraper, 0.0012 inch (30 micron) Mylar film, and Bear Cloth or Scotchbrite. Care must be used not to remove too much material thus changing the dynamics of the bearing. The contact pattern of the bearing to the shaft can be obtained by applying Prussion blue to the shaft, restin gthe bearing on the shaft, and visually inspecting the contact areas. Scrape the areas that show contact with the shaft until an even contact pattern is achieved....

Elliptical, three-lobe or lemon bore jounral bearings are more difficult to refinish using this procedure. Extreme care must be taken not to change the special configurations o fthe bearing bore. Normally these repairs are best done by the original manufacturer.

Pivoted-shoe journal bearings are the most difficult to repair in the field. Normally the radial thickness of the shoe is held to a 0.0005 inch (13 micron) tolerance. For this reason, only a minimal amount of material can be removed. Minor scratches should be treated with nothing more abrasive than Bear Cloth or Scotchbrite. Field patching of pivoted shoe jounral shoes is not recommended, as it is difficult to blend the patched areas back to the original contour without a proper mandrel.


Electrical Apparature Service Association (EASA) “Mechanical Repair Fundamentals” states
quote:
EASA:
Fitting. Fitting a new sleeve bearing is an important part of the assembly process. Install the bottom half of each bearing and then spin the rotor with the bearing journal dry (or with a small amount of oil wiped onto the journal) to establish a wear pattern quickly. Thrust the shaft axially several times while it rotates. Scraping is generally done using a babbitt knife or bearing scraper, followed by polishing with a Scotchbrite pad.

The bearing should be thoroughly cleaned after each fitting before being rolled back in for further fitting. The objective is a minimum of 60% contact centered in the bottom half, with no contact at the corners or top.


Do you think both EASA and EPRI completely missed the boat in their recommendation to use scotchbrite on babbit? Maybe it is a new concern identified by recent operating experience they were not aware of? Or maybe the concern is different because they are suggesting to use it in a slighly different way that I described (they suggest it as a finishing stage rather than the sole means to re-shape the babbit) ?

I did notice most of the references talk about using a scraping tool. Personally I am a little bit more nervous about a scraping tool than scotchbrite since I picture the scraping tool as a quick / rough change in geometry and the scotchbrite as a slower finer finish. (I realize it also depends alot on who holds the scraping tool). For the same reason, I would be a little bit nervous about using a cutting tool, although I guess it’s not such a big deal if you have a spare bearing ready to go in case there’s a “whoops” moment. If we decide not to use scotchbrite, then what type of cutting tool or other tool would be used? And is the tool incapable of creating rogue embedable particles?

Also I know there are different colors of scotchbrite. I assume these give different roughness and perhaps different hazards for embedable particles.. Any comments on whether grey would be a good grade to use if we use scotchbrite?
 
Like Like (0 likes)
PermalinkView Printer Friendly Format
electripete, at my former employer (the gear manufacturer) we used a material that resembled a piece of screening, not unlike the mesh of window screening. We called it "screen cloth" although I'm not certain if that was a trade name and Google doesn't come up with anything by that name. One of the objections I would have with Scotch-Bright pads is that they are made with ultra fine grade silicon carbide (see http://solutions.3m.com/wps/po...WS904glLP8NVZTQJBbl). This presents a scenario exactly like, in my experience you want to avoid; the embedding of abrasive material into the babbitt surface. This is supported by what Dan Timberlake states..."My experience with ScotchBrite is it breaks down physically and creates debris. I assume the debris includes some abrasive material. I'm more than a little scared about using abrasives around bearings made of stuff that has embedability, like most any grade of babbitt." Dan's link to information concerning a Tri-metal bearing is also interesting in that we made a line of gearboxes that used these bearings at rotor speeds from about 14000 RPM and up. They were selected because of very high babbitt fatique resistance, which as you know declines rapidly with thicker babbitt. As the aritcle points out, the babbitt was 1 mil thick, so virtually any abrasive product would remove the top layer (babbitt) and leave the next layer of copper.

By the way, a related interesting thread exists at http://maintenanceforums.com/e...451/m/2151038503/p/1
Scotch Brite.
 
Like Like (0 likes)
PermalinkView Printer Friendly Format
JohnPa,

I do know that your (and my) former employer from which you recently retired has used 'Scotlchbrite' to clean bearing surfaces. I would hesitate to call this scrapping.

If you limit the material removal to less than 0.0001 inch you may be ok. 'Scotchbrite' can clean the surface without changing the geometry. Actual scraping can significantly alter the bearing geometry. This is not the best practice for critical machinery.

My advisor for mechanical engineering worked at a machinery company. He related a story about him picking up a scaping tool (around the late 50's early 60's - times have changed), and a senior technican informing him that only certain people were allowed to scrape. This was because of the companies experience with getting machines to run. In all likelyhood, the mechanic was adding a similar benificial profile to his bearings.

There was also a story about bearing crush, which adds a preload. Not everyone could do this well. Today, these items are taken care of by the design and manufacturing of the bearing, not by trial and error scrapping.
 
Like Like (0 likes)
PermalinkView Printer Friendly Format
I haven’t heard of tri-metal or bearings with babbit that is 1 mil thick. That is different equipment than I work on.

My personal interest is sleeve bearings in electric motors.

If I understand what you and Dan are saying, it is that the concern is based on your own intuition, not an actual experience where using abrasives on babbit caused a particle to become embedded in the babbit and/or resulted in shaft damage, right?

It’s still good input for consideration either way... I just wanted to clarify the basis.

Let me add some more references:
EPRI TR 100897 – “Repair and Reconditioning Specification for AC Squirrel-Cage Motors with Voltage Ratings of 2.3 to 13.2 kV” [December 2000]
quote:
EPRI:

9.4.2 Sleeve and Tilting Pad Bearings
9.4.2.1 Minor babbitt scratches shall be repaired by polishing the damaged surfaces with
a polishing pad such as Scotch-brite cloth.


The exact same requirement as above appears again in the December 2008 rev of the EPRI motor repair specification document (re-numbered as EPRI 1016679). The EPRI document GS-7352 referenced earlier dates back to 1991.

The quote above from EASA Mechanical Repair Fundamentals also appears in the EASA “Technical Manual” as well as EASA Technical Note 38 – “Repairing Sleeve Bearings”

EASA and EPRI are pretty reputable resources in my opinion. EPRI supports the US power industry which probably has more industrial sleeve bearing machines than any industry other in the US. EASA sets the standard for motor repair practices (for customers from all industries) across the US.

These documents are the closest thing that I know of to “standards” for this particular type of activity. I assume it is the same for many other people in the power and motor repair industry. I assume therefore many people follow the practices recommended in these standard-like documents (EASA repair shops are supposed to follow EASA practices, all repair shops are supposed to follow customer specs).

Despite what I would presume to be widespread usage of these practices over a period of time based on documents published 1991 thru 2008, I have not heard of any problems attributed to the practice of using scotchbrite on babbit.

That’s not proof of anything. But it’s something to think about and it weighs pretty heavily in my own personal thoughts about the subject. Everyone is free to draw their own conclusions.

The more information, analysis and opinion that is offered on any side, the better all of us can do.
 
Like Like (0 likes)
PermalinkView Printer Friendly Format
A dry waller's screen is basically what John is referring to or so I think.

I have scraped bearings via the ole time various styles and shapes of scrapping tools that use to be available. Most or all of these jobs were due to the hand pouring technique.

Some large ball mills were poured on-site and hand scraped on-site.

I have hand scraped turbine bearings as well. This is not common practice and generally done to get things running vs waiting on a new bearing.
 
Like Like (0 likes)
PermalinkView Printer Friendly Format
Hi Pete,

My comment was just personal opinion.
I hold API, EPRI, AGMA and other groups composed of practicing, practical folks in real high regard.
If they are comfortable with applying Scotchbrite to produce a finish a motor bearing's babbited surface, I guess I would be too.

I'll still be concerned about machines with projected area bearing loading over 200 psi.

The Scotchbrite I've used are the commonly available green and dark red grades.
I'd like some of 3M's rotor-peen flappers for shaft straightening.
http://multimedia.3m.com/mws/m...n=61-5001-1739-7.PDF
 
 
 
Photos (1)
 
Like Like (0 likes)
PermalinkView Printer Friendly Format
Thanks Dan. That is something to think about. If Dan/John or anyone wants to add any final thoughts about abrasives or scotch-brite on babbit, please do. I’ll let you have the last word on that (I promise I won’t post any more references or use the “we’ve always done it that way” argument again.)

quote:
Today, these items are taken care of by the design and manufacturing of the bearing, not by trial and error scrapping.


I think you are looking at the purpose of scraping differently than me. My viewpoint as stated in my original post was:

  • ”The purpose of scraping new bearings upon installation into a motor is NOT to make them rounder or smoother or to the correct radius (as some people seem to think) .... the bearing manufacturer is perfectly capable of doing that as long as correct specs are used.

    The purpose of scraping bearings upon bearing installation into electric motors is to compensate for any misalignment between as-installed position of the shaft/bearing induced by misalignment between the two bearing housings or perhaps gravity sag of shaft.”

As-installed misalignment is obviously something the bearing manufacturer cannot address.
 
Like Like (0 likes)
PermalinkView Printer Friendly Format
We poured many a babbitt bearing on draglines, scraped them, and smoothed them with various means, scotchbrite being one of them. Our application was very low speed . We never poured them for the MG sets. When we poured the last ones I can remember (Marion 7800), we poured 1000# at a time with a crane. We always cleaned the shells with various liquids from varsol to diesel fuel and rags. If the rags would not slide over the babbitt surface without snagging, then it needed more polishing. Never used a microscope to look at them for embedded particles because again, it was slow speed and huge. I could see where the concern is though, as scotchbrite does break down as you use it, even when you use a liquid to fine it down.
I would probably be more concerned with a turbine/generator, but unless I used a lot of force during the scotchbriting (new word) and forced the particles way down in the material, I would count on the oil flow taking small loose ones away.

D
 
Like Like (0 likes)
PermalinkView Printer Friendly Format
quote:
EPRI supports the US power industry which probably has more industrial sleeve bearing machines than any industry other in the US. EASA sets the standard for motor repair practices (for customers from all industries) across the US.


In an oil field one may have many recips, engines and compressors. There can be many sleeve bearings in these applications (tri-metal may be used). I assume industrial means other than vehicles, because someone like the post office would have many bearings.
 
Like Like (0 likes)
PermalinkView Printer Friendly Format
OK Bill. I withdraw the suggestion that EPRI is largest user of sleeve bearings.

In view of my post dated 19 December 2011 03:05 PM, would you care to withdraw or explain your comment "Today, these items are taken care of by the design and manufacturing of the bearing, not by trial and error scrapping." ?
 
Like Like (0 likes)
PermalinkView Printer Friendly Format
Technique and methodology holds tolerances to 0.000n" (even if not an OEM shop or related); while,

Design locates the bearing on concentric center-lines with other machine components on concentric center-lines.

Remove setup distortion and you should be where you want.

It's been a long time since I've seen anyone scrape motor bearings. The large one's in ball mills for example are basically 3' radius and hand poured on-site, then fitted.

Per adventure and hopefully I haven't over-stepped here with this answer. But, this is what and how I see it.

But; still some drop a wrench on the babbit - and scrape the fault or similar incident. Ouch!
 
Like Like (0 likes)
PermalinkView Printer Friendly Format
quote:
Design locates the bearing on concentric center-lines with other machine components on concentric center-lines.

As you know, machines don’t always meet design. And we have frame distortion from foot problems, foundation settling, heavy teminal boxes. And these things can make the real world different than the paper world.

That’s why we do the blue check, to find out what we've got.

If blue check is fine, you’re right, concentricity is close enough that we don’t expect any effect on the bearing.

If blue check fails, the condition of our machine doesn’t meet industry standards, we have to decide what to do about it. Scraping (possibly Scotchbrite if you’re comfortable with it) seems like a reasonable thing to do about it if you have someone available that is qualified and experienced in that task. As stated in the EPRI quote above, applies to plain circular journal bearings, not tilting pad or lemon bore etc.

Just a side comment – The mechanical foreman at one of our local motor repair shops described for me their practice that if the two sleeve bearings on a given motor differ from each other by more than 5C during uncoupled run, they stop the run and scrape the bearing on the hot side until the difference becomes less than 5C ....if time allows (usually means if there’s nothing waiting behind it to use the test stand... they just leave the motor on the stand, do the scraping and re-try the run). He said it’s not uncommon for motors not to meet this on initial run and he’s never had one that he couldn’t bring to within 5C. (I guess if the two bearings are equally stressed from the misalignment, the comparison doesn't work) There may have been a little salesmanship involved in that story... take it with a grain of salt.
 
Like Like (0 likes)
PermalinkView Printer Friendly Format
There are other means to remove distortion and induced stresses and those should be employed.

A motor's foot should never be enlarged is an example. IMHO; neither should the bearing's geometry changed to correct an underlying fault. Address the fault.

The motor in question; is it of a poor design? Can it be improved? Can engineering correct the problem? Can engineering get out of the way and let the skilled craftsmen do their job - sometimes that is THE solution;-D
 
Like Like (0 likes)
PermalinkView Printer Friendly Format
quote:
The motor in question; is it of a poor design? Can it be improved? Can engineering correct the problem?

I hadn’t given it a lot of thought until now.

Regarding the motor whose bearing photo was posted above and also attached: 3” shaft with specified 5-8 mils clearance. In my judgement “typical” mounting of motor to bedplate and foundation. No cracks in foundation. Motor frame includes a lot of thin surface pieces which creates an impression of flimsiness but I’ve never had a chance to carefully inspect the structural skeleton inside (photo of identical spare motor in slide 1)

Terminal box that you can see is not large, in fact smaller compared to most for this size motor (600hp 3600rpm 4kv horiziontal sleeve bearing). Most of our motors have term box mounted on the center of the side of the motor. This family of motors also has term box on the side, but it’s to the outboard end of the side of the motor. From a qualitative standpoint, we can imagine that term box mounted in the center of the side moves both inboard/outboard ends of the frame equally and creates no misalignment, whereas termbox mounted on the end like this pulls the outboard bearing horizontally to the right (simplistic view of a complicated 3-d distortion, and quantitatively I have no idea how significant the magnitude would be). The photos in slides 2 and 3 are the blue check pattern of the outboard bearing that we have been discussing. I’m not sure offhand which end of this bearing points toward the winding, but it actually doesn’t matter for our purposes... if you rotate slide 2 top view by 180 degrees the contact/wear pattern would still be upper right to lower left of the slide. If we envision that the distortion from weight of the term box pulls outboard bearing horizontally to the right and leaves inboard bearing alone, then the direction of the wear pattern we’re seeing is consistent with this distortion. That doesn’t prove much... there was a 50/50 chance this would be the orientation of angular mislignment in the horizontal plane even if not caused by the term box.

My gut feel, as small as this particular term box is it wouldn't make a difference. I guess next time we find outselves in similar situation, we could try jacking on the term box to see if it affects the blue check. (I probably would do it now that my curiousity is up). But I can imagine that turning into a little bit of research project, and even after you spend all that time, there’ no guarantee you’ll find the cause/solution and you may end up scraping the bearing anyway.

But I’d like to circle back to your comments which if I read between the line suggest there must be something dramatically wrong with this particular motor to fail a blue test and we should never scrape a bearing but instead should find/fix the cause of the failed blue test. I’m not disagreeing with the value of digging to find the cause of things that deviate from specification in some cases, but certainly it is not always practical. I would submit that the quasi-standard documents cited above do not support the view that the reaction to failed blue test should be searching for a cause, instead they suggest the response to failed blue test is to perform scraping.

It leads me to a question: Of those who perform blue test during bearing replacement, what fraction do you find fails the check?
 
 
 
Like Like (0 likes)
PermalinkView Printer Friendly Format
Let me try the attachment again.
Can you guys read it?
 
Like Like (0 likes)
PermalinkView Printer Friendly Format
Pete neither one opens for me.
 
Like Like (0 likes)
PermalinkView Printer Friendly Format
quote:
It leads me to a question: Of those who perform blue test during bearing replacement, what fraction do you find fails the check?


Excellent question:

Q1; Did all pass muster - proceed to GO - collect $200.

Q2; Needs corrective action.

S1; Scrape the bearing per advised best method?
S2; Further investigate for - soft foot and initial alignment setup.
S3; in conjunction w/S2 further checks for induced stress.
S4; 1, 2 & 3 passed....... further checks for design flaw.

Conclusion: scrape bearing as a last resort??????????

One would think incorporating a valid checklist should be the way to go.

Some flaws are excited from poor setup that otherwise would go unnoticed had the setup been good. Poor setup also induces stresses as well as external factors; i.e., piping, etc...

Finding statistical data would be a bear. In my experience I would not enter a 'guess' as to how may encounter problems. Even in the same plant, one department may be experiencing problems that go unnoticed elsewhere within the organization. Or, I could be wrong.
 
Like Like (0 likes)
PermalinkView Printer Friendly Format
Thanks Barry.
Here is the attachment for my post 21 December 2011 11:26 AM
(it finally worked)
 
 
 
Like Like (0 likes)
PermalinkView Printer Friendly Format
I have worked in a large motor rebuild shop and repaired a lot of sleeve bearing motors,in the 1000s. I will state that why the scrapping is needed is that on any machine there are tolerances. Co-plane tolerances for the feet, center tolerances, tolerances for end brackets; and so it goes. These tolerances do add up. At some point, the sum of the tolerances is totaled. That is the bearing location,position,performance, what ever you want to call it. Scrapping is the final correction for this.

The use of emery on babbit should be banned! I have some old GE books on babbit bearings and even back then it was frounded upon. Scotch Bright is some times used today, not to remove metal, but to dull the surface so you can see the new high points. You can also use a cloth rag and oil to dull the surface. Scrape the areas needed, dull the surface, reinstall for the next trail turn. Remove the bearing and repeat process until you have a good surface pattern. Emery, including the screened emery should never be used. The particals are too hard and can be embedded in the babbit and damage the shaft.

Bluing is a difficult process for most people to do. The amount of bluing on the shaft makes a huge difference in the result. Too much will act as a lub and you get poor results due to lift. This proceedure is best left to the experts. The dry method is easier for most.

Proper scraping will remove minimal metal and should only enhance performance. Tempatures can be lowered by proper scraping. As well, you can ruin a good bearing by not understanding what you are doing. I've seen this way too many times. Cross-hatching the surface can also improve performance, but must be done correctly.

Hope this helps in some small way.
 
Like Like (0 likes)
PermalinkView Printer Friendly Format
Very good, lake man.

Obviously the hands-on skilled craftsman that's actually doing. I have not been in your environment that cast me into a role of repetitive correct work. Mine is limited with time passing between.

I still do fool with engines - have boats - will work on engines; grown to have somewhat of a hate relationship with 350CID. But like Jag, MB and Ford. Again, limit work on my own stuff.

Often the skilled craftsman needs to take the lead and make the call at certain functioning levels in the mechanical maintenance world - possible electrical too.

Sometimes plants run in spite of management.
 
Like Like (0 likes)
PermalinkView Printer Friendly Format
Good info, Lakeman. Thanks
 
Like Like (0 likes)
PermalinkView Printer Friendly Format
From a maintenance and process view:

Say one scrapes a bearing with measurable (significant) material removal and this results in the best vibration ever seen or not seen, 0. Did you measure the geometry of the bearing so that this could be repeated, on purpose that is? No 3-d measurements? The boss tells you to do it again with motor B.

On the other hand the vibration is awful. No measurements again? How do you avoid this situation.

In either case the bearings 'had' (were) taken out and 'corrected.' This is re-work without fixing or investigating the root cause.

If it is good enough for a motor, is it good enough for your main generator? Main turbine? The logic is the same, and for all I know about an individual facility this is practiced.

The bearing doesn't fit. Why? Poor manufacturing? Machine issue? Installation proceedure/methods? Find root cause and fix it once.

'Cross-hatching ' was mentioned. Interesting? Patterns on the surface can icrease load carrying capacity (or destroy it). I saw something on this years ago. I don't see bearings designed like this in my applications. To do this by hand could be risky.
 
Like Like (0 likes)
PermalinkView Printer Friendly Format
I wonder how much a fat motor rotor bows it's shaft, and could create edge loading in beautifully colinear/concentrically machined bearings. I think including features to make inserts self aligning are smart, not a cover up for "poor" machining.

High output car and motorcycle engine bearings can be subjected to loads our <200 psi machines never will. Distortion of the connecting rod big end bore from loads like one side of this.......
http://t0.gstatic.com/images?q...CJCrEh7Fq8K5LOXmcCPM
have meant the con rods evolved from this, barely out of the steam era with low capacity babbitt linings...
http://www.mtfca.com/discus/messages/50893/74890.jpg
to these, with big ends with husky proportions in an effort to maintain housing bore roundness, since bores that "pinch" tighter in one direction at 8000 rpm will cut thru or otherwise destroy the lovely hydrodynamic oil film
http://www.turbonsx.com/pist_rod2.jpg

Seem likely the crankshaft's bending would induce a certain amount of edge loading that does not exist under 4000 rpm, and reverses direction to boot.

Add to that the fact loading is high enough, and oil films thin enough to make even the direction of polishing important on some crankshaft materials.
http://www.stealth316.com/misc...ank-grind-polish.pdf

The folks at navsea apparently never had to concern themselves with this, simply since their bearing loadings just are not that high.
That's a nice place to be, if you can.
 
Like Like (0 likes)
PermalinkView Printer Friendly Format
Read with interest this post..I have machined, fit and checked quite a few sleeve bearings (large motor, fan and pump)over the past many years. The orginal post alluded to scraping for misalaignement issues, I believe that is a very narrow viewpoint. Scraping is one of the final checks in equipment assembly along with tilt and twist. In self aligning bearings (some type of spherical seating) scraping is rarely needed. But getting 80% contact is also difficult where oil rings have damaged the journal area. The second type of sleeve bearing has a flat seating area, I am most familiar with these on motors and high pressure feed pumps. It is a rare occasion where these do not need scraped. The scraping is more a function of correcting machining error problems than alignment problems. In a bearing with a clearance of .005" a missed setup in the lathe of .001" is a lot. Also if there are alignment issues this would be picked upon and corrected on the tilt-twist check.
As for removing so much babbitt that correct clearance is sacrificed, I have only seen that once and believe me the mechanic did not understand how to fit bearings. Relize scraping removes .0000nth' of material on each scraping (depending on the experience of the mechanic). As noted opening up a bearing by scraping for cooling is not an unusual practice and is readily done.
Addressing the scrothbrite, I was trained to use steel wool, as a final polish, pre scrothbrite days and there were many issues using steel wool. Scothbrite is a wonderful final finish to a sleeve bearing. Again the mechanic must understand this is a sanding medium and not just grind away, it is a polish and clean, then polish and clean over and over until desired finish is obtained...note all the cleaning steps.
For a last paragraph...when I started in a shop of 30 or so mechanics over 20 mechanics understood the methods of fitting sleeve bearings. Today I see three plants totaling over 100 mechanics and there are less than 3 that understand the "art" of bearing fits. The emphasis over the last 10+ years and attaboys are for speed and condemnation for quality, coupled with retirement of some fine mechanics have left sleeve bearing fits a lost art IMO...and that is a shame.
 
Like Like (0 likes)
PermalinkView Printer Friendly Format
Hi J Heber,

You said " there were many issues using steel wool."

Could you describe that a little more?

thanks
 
Like Like (0 likes)
PermalinkView Printer Friendly Format
Steel wool tends to facture and embed quite easily it also rusts and can leave a high spot at the rust point and leads to a hot spot, could be seen a a small teardrop shape on disassemble. Many woodworkers stopped using steel wool and went to scothbrite beacause of the same issues. I had a woodworking issue where a piece of steel wool embeded and rusted under the finish.
Since I am back will comment on cross hatching. A lot of older millwrights/ mechanics did not use blueing. They cross hatched with a small stainless steel wire brush, when the shaft was rolled over the cross hatching any high spots would polish for scraping. On a fun note when I trained I was encourged to develope my own cross hatching design, always done radially nvere axially. For many years I did this with -0- ill effects. In fact it was always enjoyable to tear down a machine and see the cross hatching still there and by that signature you knew who did the work, and they were proud of their work.
As for comparing a motor bearing to a generator bearing, kinda apples and oranges...worked on a lot of 100Mw to 550Mw machines and all had spherical seating can't remember ever working on a motor with spherical seats, probably did and don't remember it.
Enjoyed the thread...thanks all
 
Like Like (0 likes)
PermalinkView Printer Friendly Format
If you really want to test your skills, try split bearings with a tapered OD. The bearing is drawn in to adjust the clearance. These were used on large grinder motors in the steel mills and old vertical motors with pivot shoes.

Also to add to the line, pressure fed vs oil rings. Do you work them different?
 
Like Like (0 likes)
PermalinkView Printer Friendly Format
A lot of good points and discussion all around.

quote:
Bill wrote:Say one scrapes a bearing wth measurable (significant) material removal and this results in the best vibration ever seen or not seen, 0. Did you measure the geometry of the bearing so that this could be repeated, on purpose that is? No 3-d measurements? The boss tells you to do it again with motor B.

The purpose of scraping is to address the as-installed condition on a given machine which is affected by very small unknown factors such that you don't know what you've got until you actually check it, by doing the very-sensitive dry-roll/blue check. So, trying to record measurements from one machine and apply them to another machine sounds pretty silly to me (it makes more sense to do the blue check on the 2nd machine rather than extrapolate from the first, I hope anyone would agree). Additional notes: the intended result of scraping is to reduce temperature (rather than vibration). Also, we have a pretty good idea about the quality of the scraping in accomplishing it’s objective by repeating the dry-roll / blue check after scraping and before returning the machine to service (it helps the craftsman decide when to stop scraping.)

quote:
J. Heber wrote:Scraping is one of the final checks in equipment assembly along with tilt and twist

I'd like to hear a little more about tilt and twist checks. Are these common checks?
I have heard of 4-corners check (similar clearance between shaft/bearing by feeler gage on all four corners at splitline). i Imagine you could put dial indicator base on shaft and try to read a face of the bearing. Are these the checks you have in mind? What specs would you use?
quote:
J. Heber wroteself aligning bearings (some type of spherical seating) scraping is rarely needed.


....As for comparing a motor bearing to a generator bearing, kinda apples and oranges...worked on a lot of 100Mw to 550Mw machines and all had spherical seating can't remember ever working on a motor with spherical seats, probably did and don't remember it

Good points. It makes good sense this would not be required for spherical-seat (self-aligning) bearings.

quote:
Dan wrote:The folks at navsea apparently never had to concern themselves with this..

I dont’ get the NAVSEA reference. (seriously, maybe I’m just slow). Can you explain it?
 
Like Like (0 likes)
PermalinkView Printer Friendly Format
Surely, you at least take closing clearances (across the bearing) as a measurement. I mentioned significant material removal, a geometry change. If you find a new geometry that works, why not duplicate it. We do that by design with an analytic approach. If one geometry works for one motor it should work for an identical motor.

OK, what is the mechanical difference regarding bearings for a 25 MW motor and a 25 MW generator (running at the same speed - At this size they are likely to both be synchronous machines.).

I did work on a reversing hydroturbine/generator of the 450 + MW. It was a motor and a generator, no difference. One does see motor/generators in some refineries and some LNG plants, no difference.

Some motors do have spherical seats. I have not been convinced that they always self-align with heavy rotors.
 
Like Like (0 likes)
PermalinkView Printer Friendly Format
EPete quote:
Dan wrote:The folks at navsea apparently never had to concern themselves with this..

I dont’ get the NAVSEA reference. (seriously, maybe I’m just slow). Can you explain it?[/quote]

=============

The plain bearings on this motor rotor were "superfinished" by Northropp Grumman using close fitting shoes with some kind of abrasive.
http://defensetech.org/2007/03...ture-of-navy-motors/

The results they were looking for was improved finish, still defined by lower RMS value I think I recall.
When I inquired about the direction of polishing there were blank stares, even from the highly experienced engineers, some who dabbled in motorsports.

I'm guessing the projected unit loading was probably < 200 psi.

High reliability is somewhat easier with conservative designs. Low stress designs can "get away" with more.
Sometimes the rules about what's important or acceptable change with application.

When I worked in a specialty machine shop we'd have to report the sad news of failed Mag-particle or dimensional inspections and would recommend extra corrective machine work or part replacement. The added expense sometimes met with resistance.
Some of the customers with race engines would say " but it only has to run two more races. "
Some of the customer's with street rods or restoration projects would say " but it's not a race engine."
The 5 national champs, and folks more familiar with mechanicals said " do it. When will it be finished?"
 
Like Like (0 likes)
PermalinkView Printer Friendly Format
Lake Man, split tapered OD.. no thanks, I get my butt kicked enough as is. As for pressure feed verses oil bath (sump). No difference for babbitt to shaft interface, standard tilt-twist and blue check, reliefs around where the oil enters the shaft. The pads underneath on the bearing to bearing housing require a blue fit up but that is a little off this thread.

Electric Pete, tilt-twist: The twist part is what you described as a 4 corner check. The tilt is to verify the shaft is parallel to the bearing (and bearing housing). On tighter clearances plastigage is used an larger clearances fuse wire or indum wire. With the lower half of the bearing in place a long piece of plastigage or wire is laid on top of the shaft on then the top half of the bearing is assembled and torqued done. Top half removed and clearance is checked from end to end on the plastigage or if wire was used the wire is mic'd.

Bill, In my narrow little spot in the universe, 10 years ago opening and closing clearances were documented and recorded on all babbitted machines. Today only turbine closing clearances are documneted (some contractors still record openning clrs.). Smaller machinery go undocumented. I'm with you, self-aligning bearings do not self-align. Today I get involved with machinery that was rebuilt and failed. One major cause I believe is some people don't understand 'self-aligning bearings do not self-align' wear patterns have taught me that. Still can't beat a good initial setup.
 
Like Like (0 likes)
PermalinkView Printer Friendly Format
Thanks everyone.

Dan - good info. I will have to go back and read what you're written to make sure I understand. No doubt there are a lot tougher applications than electric motors out there and probably the motor guys can learn a lot from the experience of people that have worked on tougher applications.

Thanks, J Heber, that makes good sense (4-corners check for twist = in horizontal plane and axial comparison of plastigage results for tilt = vertical plane). All of our bearings have one or two oil rings cutouts in top half of bearing, but we could still compare the remaining areas of the top-half where there is no oil ring cutout. .

Now an interesting thought comes to mind about relationship between tilt and twist inferred from by platigage/4-corners check to the tilt and twist inferred form dry-roll/blue check.

If bearing and shaft are perfectly cylindrical, then tilt or twist inferred from plastigage or 4-corners check should be correlated to the tilt or twist inferred from blue-check/dry-roll.

But it doesn’t hold when we scrape the bottom half.

Let’s say we observe tilt by blue-check/dry-roll and then scrape the bottom half and improve the blue-check/dry-roll-check results and presumably temperature, the tilt pattern inferred by plastigage laid axially on top of shaft would not change.

And let’s say we observe twist and scrape the bottom to compensate and get good roll-check, the 4-corners check is not going to get better either because we’re not going to scrape all the way up to the splitline or axial ends of the distribution groove (in fact I can imagine that the 4-corners check may appear slightly worse ...let’s say you see contact at the NW and SE corners of the bottom and scrape those... afterwards the shaft is going to settle more toward the NW and SE corners and 4-corners check will look worse).

Also, I tend to think the dry-roll / blue check is more sensitive to detecting small deviations than those other checks.

Regarding spherical seats....

At our plant, up to and including 2500hp horizontal sliding bearing motors, we have no spherical seats on the bearings. We have one family larger than that and they do have spherical seats. These are 7000hp 2-pole 13.2kv horizontal sliding bearing motors with tilting-pad radial bearings.... bearing assembly mounted in self-aligning spherical bearing seat. The outboard bearing is insulated (slide 1), the inboard is not (slide 2). After removing/replacing bearings, I saw the repair shop field service craftsman used a deadblow hammer to help the bearing self-align itself. He has a lot more trouble on the outboard than the inboard, I think maybe (?) that was due to the white electrical insulation which I presume was added by the motor manufacturer (not part of bearing manufacturer's design), looks kind of “cheap” in the way it’s applied, and has a rubbery texture which seems to me would resist sliding... making it harder for that insulated bearing to re-align itself.

Whether is re-aligns correctly or not, you have far less motivation to scrape self-aligning bearings than fixed bearings. If you believe they re-align, then there is less need to scrape as was mentioned since you are not combatting potential misalignment. If you don’t believe they re-align and presumably think the position is somewhat non-repeatable, then I can see scraping would be much trickier because you have to ensure that the position of the shaft when you do your blue-check/dry-roll to determine where to scrape has to match the position during operation. And of course we have to remove those bearings after the blue-check dry roll to do the scraping. I think I’d be a lot more nervous about scraping self-aligning bearings for that reason. But it’s not a decision I’ll face because the only motors we have with spherical seats are those 7000hp motors which also have tilting pad bearings... which already tells me they should not be scraped.

I have no idea what gets done on our large turbines or generators (or even what kind of seats they have) – that equipment is outside my responsibility.

I can understand why there would be some hestitation about scraping because of concern for significantly altering the geometry of the bearing. My understanding is the amount of babbit removed is not supposed to be signficant in comparison to bearing clearance (a fraction of a mil removed) and is not supposed to “significantly” alter the geometry (clearance, "effective" radius of curvature, things that might affect dynamic performance). I would presume a final plastigage check of clearance is in order once we decide we’re done scraping (based on good as-left roll pattern), but even that is a pretty gross check of a complex 3-dimensional geometry. In the end I think the nature of this particular task is that it requires trust in the skill of the craft doing the job. In some industries, relying on “skill of the craft” rather than quantifiable measures has a bad connotation. It makes me slightly uncomfortable as well, but ignoring industry standards also makes me uncomfortable. Perhaps there is some room for judgement depending on the situation (who’s doing the work, is a spare bearing available, have the bearings historically run hot, etc).
 
 
 
Like Like (0 likes)
PermalinkView Printer Friendly Format
quote:
I wonder how much a fat motor rotor bows it's shaft, and could create edge loading in beautifully colinear/concentrically machined bearings. I think including features to make inserts self aligning are smart, not a cover up for "poor" machining.

I know you, Bill, me and others know how to do that calc. But I’m not sure I have ever done it before. So out of curiosity, I tried it out with an example rotor that I have detailed info on: 2500hp, 1800rpm, 13.2kv horizontal sleeve bearing motor (rotors weighs 4,000 pounds.)

Results including rotor description are attached.
Slide 1 = rotor outline.
Slide 2 = Cross section of rotor spider.
(moment of inertia determined in http://www.eng-tips.com/viewthread.cfm?qid=274310)
Slides 3-4 = rotor photos
Slide 5 = The simplified model used for calcs
Slide 6 = Includes pdf icon – double-click to shows detailed calculations in Matlab’s Mupad symbolic program. Slide 6 also includes results:
Results:
* Slope at bearing due to weight of rotor is 0.00015
* For bearing length of 4.5”, this is approx 0.7 mil change in elevaation from one end of bearing to the other.
* Since nominal bearing clearance is 6 mils, this is approx 0.7/6=11% of bearing clearance.

That's more than I expected. I don't know for sure how motor OEM's deal with that, but I suspect they neglect it and machine everything concentric from end to end. I also suspect this much misalignment doesn't have a huge effect on bearing operation.

========
** UPDATED 5pm 1/1/12 to correct L2 to 25", which decreases bearing slope to -0.00015 and results in 0.7mil change in elevation instead of 1.0mil
 
 
 
Like Like (0 likes)
PermalinkView Printer Friendly Format
 
Post Reply