wonky thrust bearings?

Advert

New Replies

Report Bug

Help/FAQs

wonky thrust bearings?

This topic has 19 replies, 8 voices, and was last updated 3 July 2020 at 12:44 by Oily Rag.

Viewing 20 posts - 1 through 20 (of 20 total)

Author

Posts
27 June 2020 at 17:03 #482551
Mark Davison 1
Participant
@markdavison1
I'm converting my Harrison vertical mill to CNC, triggered by the original lead nut having about 0.050 of back lash and needing yet another VFD if I wanted to use the power feed. I've replaced the X axis lead screw and gear box with a 2005 ball screw and NEMA 34 stepper. I used standard fBK15 and BF15 fixed and floating bearing blocks (that I made myself as no-one had the size I needed) and a pre-machined ball screw. All was well. I'm now in the process of doing the Y axis. I can't use the standard bearing blocks as I'm very restricted for bearing height (shaft centre to top surface of knee), so rather than use angular contact bearings I've used a pair 51102 thrust bearings, mounted into seats in a 35mm thick block of steel that mounts to the front of the knee. This is a very similar setup to the original design but the larger block allows me to mount a plate/box that will house the timing belt/pulleys and serve as a mount for the stepper. To take the radial load from the timing belt I've used a single 6002 ball bearing on the same 15mm journal that the trust bearing are on. The problem I have is when I bolt the 10mm thick aluminium plate to the front of the bearing block (this plate has the seat for the 6002 bearing) the spindle/bearing assembly acquires a considerable tight spot. When I remove the 10mm plate from the front and reassemble everything with a dti, the 6002 bearing is out of concentricity by almost 0.25mm. all 3 bearings are on the same 15mm journal and it is spot on size (15.00mm). the only plausible explaination I can think of is that the thrust bearings are duff, or one of them is. They are supposed to be 9mm thick, and one is spot on 8.99, but the other is 8.89 which is quite some error for a bearing surely? The other explaination, would be my bearing seats in the 35mm block aren't concentric, but surely this would just serve to give me a spindle that isn't perpendicular and bearings that aren't loaded evenily, it wouldn't cause 0.25mm of concentric error? I'm pretty sure the bearing seats are within 0.02mm or less anyway, so I can't see how that could cause the massive error I'm seeing.

Thoughts anyone?

I'm going to order a pair of FAG of SKF thrust bearing (these were Dunlop) and see if that fixes things.

The spindle in the first photo is the original lead screw, the other photos are my 1605 ball screw with an entension sleaved and bonded onto it to give the additional diameter i need to mount the thrust bearings. I've spent today remaking the spindle/extension thinking that was the problem, which it wasn't (and couldn't have been, 4hrs wasted).

This video shows the issue. Here the dti is on the 12mm part of the shaft that carries the taper bush for the timing pulley, but the same thing happens with the bearing that is on the 15mm journal.

https://www.youtube.com/watch?v=HC2h8gP6kkQ

Edited By Mark Davison 1 on 27/06/2020 17:16:11

Edited By Mark Davison 1 on 27/06/2020 17:18:29
Advert

27 June 2020 at 17:03 #15283
Mark Davison 1
Participant
@markdavison1
27 June 2020 at 17:13 #482555
old mart
Participant
@oldmart
You could make sure the thrust bearings have plenty of room radially on the shaft and in the housing, or replace them with the needle type which have flat races.
27 June 2020 at 17:24 #482556
Ex contributor
Participant
@mgnbuk
I've used a pair 51102 thrust bearings

Have you fitted the races the right way round ? The two "washers" in a ball thrust race are not the same – one is a "housing washer" with a precision OD & clearance on the ID, the other is the "shaft washer" with precision ID and undersize OD. Putting the "shaft washer" into the housing causes problems – found this out the hard way on someone else's horizontal borer a long time ago (built wrong at the Russian factory).

Nigel B.
27 June 2020 at 17:40 #482560
Mark Davison 1
Participant
@markdavison1
Yep, they are in the right way round, the one that is the snug fit on the spndle went on first and the one with the clearance on the spindle went in the housing, visa versa the other side.

I'll look and see if they do needle thrust bearings in this size and I'll try those as well if they do.
27 June 2020 at 17:47 #482562
Bazyle
Participant
@bazyle
For a thrust bearing it doesn't matter how thick it is, just that it is a consistent thickness all round, and remains so when rotated. Have you tried putting it on a surface plate with the DTI on top and rotated it? You would have to ensure the pressur of the DTI didn't tip it, perhaps with a 1in gauge block on top.
As OM above implied the track may not be dead concentric with the inner / outer diameter but that whould be given clearance anyway. The one thing that can be a serious fault with a ball thrust race is the track itself not being a true circle, perhaps due to distortion in heat treatment, stress relief etc.
27 June 2020 at 17:55 #482564
Mark Davison 1
Participant
@markdavison1
I haven't left room for the tracks to float, neither on the spindle or in the housing, should I have?
27 June 2020 at 18:10 #482566
HOWARDT
Participant
@howardt
Why not replace whole set with 7002 ac bearings. A reasonable accuracy set will be way better although do not over tension the drive belt.
27 June 2020 at 18:18 #482568
Mark Davison 1
Participant
@markdavison1
unfortunately i don't have the height above the spindle centre line,it would break through the top surface of the mounting block, hence using the thrust bearings (28mm OD as opposed to 32mm). There is a plate that slides over the top of the block, it prevents swarf gettting into the leadscrew/nut. That's what the recess is for in the top of the block. The 6002 bearing sits further forward and I have the height there.

Edited By Mark Davison 1 on 27/06/2020 18:21:53
27 June 2020 at 19:21 #482579
HOWARDT
Participant
@howardt
Have you room between the thrust bearing to install a needle bearing with inner race. This will help position the shaft and not rely on the thrust bearings which give no support. I assume from what you have said before the housing washers are locating on there outer diameters correctly just that they are not concentric.
27 June 2020 at 19:28 #482580
Mark Davison 1
Participant
@markdavison1
I was just considering putting an online bush in there actually. There is about 13mm length between them so possibly. The only limit to the OD is the need to support the washers.

Yes the washers are located on their OD in the steel block, I suspect the tracks aren't concentric with the OD, or with the ID on the spindle washers. They look OK to the naked eye but something is clearly wrong!
27 June 2020 at 19:32 #482581
Ady1
Participant
@ady1
They have the same sort of "thrust bearing" on the Drummond M and although you can't describe me as mr fussy I was never very impressed with it, very prone to tight spots

The needle roller bearing conversion at the front on the other hand was a revelation

Edited By Ady1 on 27/06/2020 19:33:33
27 June 2020 at 19:47 #482582
old mart
Participant
@oldmart
Simply Bearings do a 15 x 28 x 4 needle type thrust bearing, the loss of length could be made up buy using the race plates from the other bearings.
28 June 2020 at 07:22 #482622
Mark Davison 1
Participant
@markdavison1
I found cylindrical roller thrust bearings the same size as the ball but they are between £15 and £25 each. They seem to be designed for almost complete overlap between the opposing surfaces and I have dont have that, in fact I have slightly less on the spindle than that required for the ball thrust bearing. I've ordered some FAG 51102 and will keep my fingers crossed that the are better than the (cheaper) first pair.

Edited By Mark Davison 1 on 28/06/2020 07:23:53
28 June 2020 at 10:26 #482657
Ex contributor
Participant
@mgnbuk
I've ordered some FAG 51102 and will keep my fingers crossed that the are better than the (cheaper) first pair.

Did you sepcify a precision class for these , or just go with general purpose bearings ?

As with other bearings, ball thrusts come in various precision classes & for a ballscrew support application where minimal axial variability is required I would expect to specify P5 or P6 grade.

I once spent a day with a plie of ball thrust bearings sat at a surface table with micron dial gauge checking to find the best of the lot for a spindle bearing installation on a Jones & Lamson lathe – non of our usual suppliers could source the correct precision class from stock, so BSL sent their entire stock of standard grade bearings for us to find the one with least axial runout.

In the end, one was only a few microns out from precsion spec. The customer didn't complain about the rebuilt machine performance, so I guess it was OK. These were much bigger than you are specifying, so easier to check – I sat an assembled bearing on the surface table, put a weight on the top race & checked the axial runout of the top race while rotating the assembly under the DTI. The variations in runout were suprising, with examples at both ends of the tolerance range from the samples supplied – the one chosen was only just outside precison grade, while others were barely within general purpose grade.

Nigel B
28 June 2020 at 14:37 #482714
Mark Davison 1
Participant
@markdavison1
They will just be general purpose, i didn't order anything special. I can accept microns of runout, it 100s of microns of radial runout that I didn't expect! I'll see what these are like and reconsider the cylindrical roller bearing if they aren't considerably better.
2 July 2020 at 19:14 #483569
Mark Davison 1
Participant
@markdavison1
Issue resolved. Cheap bearings replaced with FAG bearings and everything is now fine. The surface finish of the track on the cheap bearing is very poor in comparison to the FAG bearing. They might as well have been ground with a rock. I need to find who I bought them from and ask for a refund. They are the last 'Dunlop branded bearings that I'll waste my money on.
3 July 2020 at 11:33 #483655
Oily Rag
Participant
@oilyrag
Another case of Far Eastern brand appropriation in the belief that it will give a product a lustre to which it is not entitled. I would want not only the price refunded but also some compensation for a product which is 'not fit for purpose intended', after all you spent valuable time in discovering the root cause of the problems these bearings created.

Glad to read you got to sort the problem out eventually.
3 July 2020 at 11:54 #483658
Hopper
Participant
@hopper
There have been complaints about Dunlop bearings here before. LINK

They seem to be part of the Dunlop Tyre concern, using the same logo and a UK address. Standards have slipped i guess. Sad. At least FAG and SKF are still good. Japanese Nachi and also Koyo are also good.
3 July 2020 at 12:44 #483672
Oily Rag
Participant
@oilyrag
Following Hopper's link is interesting!

The statement at the bottom of the website tells you all you need to know – The Dunlop name and flying D trademark are used by Dunlop BTL under licence of Dunlop.

Dunlop Tyres is now, I believe, owned by Bridgestone with the exception of Dunlop Aerospace tyres (for aircraft) which are still manufactured at Fort Dunlop in Birmingham. My money is firmly on the fact that Dunlop BTL are a front company for Far Eastern produced bearings. I can never remember Dunlop of old being in the rolling element bearing business.
Author

Posts