Suggestions on bearing assembly sequence?
This a '42 round dial backgear, and I'm looking for suggestions on how to assembly things with new bearings
1) back gear shaft. In my limited experience, most bearings are press on one and tight slip on the other. I also am aware to only apply force the interference fit race. On this shaft, there are 6203 bearings at each end that are a press fit to both the shaft and bore. i'm guessing the assembly will have to be a combination of press and heat, ie thermal expansion, but I like to get the take of some guys more experience at bearing installationand suggestsion for a sequence. The parts are laid out in order below - also the front casting is pinned to the body casting so is figety to get in place. It needs some tapping etc and if had to thermally expand it for the assembly process, it might make the pins/holes no line up.
2) output shaft bearings has a pair of angular contact bearings. There is one spacer between the inner races. The outboard bearing has retaining ring groove which I assume is to set is press fit to the right depth. These are a press both on the shaft and bore....how do I accomplish this without applying force on the wrong race? press on the shaft then heat the casting to take the assembly? Challenge with that is that with the shaft mounted, the bearings would have to go in from behind - ie the end with the retaining ring groove first making it dang near impossible to get there depth right?? Also, there is no way to preload, unless it's just via a press on the shaft?
thanks for any and all help
It's been a few years since I went through my similar backgear. Had to review my photos to remember some of it. Mine (1952) did not have any bearing fits with press fits on both inner and outer races, for instance the angular contact bearings on the output shaft were a push fit into their bore, not a press fit. Do diameter measurements verify press fits on both races of the bearings?
Bearings can take axial force without damage within their structural limits, it's not always necessary to have one race be a slip fit. A real press fit takes more force than most bearings can take though so lets be sure we're talking about real press fits.
The sealed 6203 bearings shown are not original, someone has been in there before and all bets are off as to how original the assembly is.
I didn't replace the angular contact bearings in mine so didn't take photos or make any notes. The snap ring is what positions the bearing pair and thus the output shaft. Was it missing?
i carefully measured everything, in all cases the bores in the cast iron are about a 10 micron/4 tenths interference fit where as the bearing ID to shaft doesn't seem to have any measurable difference. Measurements done with 1 micron Mit bore gauge. I'd hold a mic in a stand, mic the shaft or bearing, lock the mic then compare bore gauge readings between bearing IDor bore and mic - point being measuring this small holds its challenges, but doing so via comparison i think should have accurate results. For the 6203's shaft and bearing seem exactly the same....yet it takes quite a bit of force get them apart/together. I was convinced there were interference fits all the way around. thanks for the push
With the angular contact bearings, the two are a press into the casting with a spacer ring only between the inner races. I can't see any way to press them into the casting without the bearings taking the load because with the spacer the force has to go through the bearing. Maybe what I should do is warm the casting so the required force is minimized? Once they're in, pressing the shaft in is easily done by supporting the stack via the inner races
thanks again for the help
PS, I have the snap ring and those bearings are just ones I was playing around with, seeing how i could make it go together
There should be both an inner and an outer matched pair of spacer rings between the output bearings, forgot to mention that. There's a 1942 manual available online which shows the parts laid out, the two spacer rings are clearly shown.
Hi, if the 203's are a tight fit, look for burrs at the edge of the casting and the end of the shaft. It is not uncommon for burrs to be raised during previous disassembly. A careless hammer hit can swedge the bearing seat on the shaft, making the bearing tight fit. I would mount the gear in another lathe and use a fine India stone to carefully clean up the bearing seat. If there are burrs or a swedged area, it will bite the stone while cleaning the seat up, and leave a shiny area on the shaft. Making a nice radius on the end for the front cover. Assemble the gear into the gearbox and guide the shaft into the bearing in the front cover. The bores need a careful inspection and cleaning with 400-600 grit paper. It is easier to stone the shaft down to a tap fit in the bearing, than to clean up the bores for a looser fit on the bearing.
For the output shaft, they look like standard ball bearings, not angular contact. It was common to turn them into angular contact for low precision applications. There needs to be a spacer for the outside races. Clamp the bearing inner races together and try forcing a feeler gage between the outside races. You may find that you can get .002" to .005" feelers in between. With a little work you can figure how much thicker the spacer for the outside race needs to be to give you a light drag preload. Then make a new spacer, if preload is too loose, grinding a few microns off the inner will increase the preload, and off the outer to decrease it. I would press the bearings into the housing first, and then the shaft into the bearings, with the inner race of the outside bearing supported. Insure that the inner spacer is lined up as the shaft reaches the spacer while pressing .
I would use a hand operated arbor press, NOT a hydraulic on all bearing presses.
mystery solved...here's what I found out in detail (to save the next guy from going down the wrong path)
that manual, if we're looking at the same one I am, is very confusing. The one I have only shows one spacer laid out between bearings....but the confusion comes in that the parts list just does not match the diagram! #40 is clearly a bearing not a retaining ring for example...and I can't see where 37 or 38 are visible anywhere in the diagram. I'd always thought it was a bad scan, pages from two different manuals or something....but I have a '43 hard copy and it is identically confusing. Anyway, there was only only one spacer when it came apart and I think I was the first inside it.
Originally Posted by DaveE907
Then the ball dropped....
the don't do they? I'd sort of thought the same but wrote it off as my ignorance....but the arrangement confused the heck out of me...how could a preload be set up without a way to clamp from the other side? They're 3209 which a quick google suggested was a standard angular contact number.
For the output shaft, they look like standard ball bearings, not angular contact.
Well I figured it out, they're 3209 NEW DEPARTURE. Found a substitute guide, Tinkem and looked them up....they're the same as Timken 209 Doh!
which is a standard radial ball bearing, 45x85mm
it now makes complete sense why there is one ring.....and why would angular contact bearings be needed here?
Bill, thanks for all the advise on fitting the 6203's.
Do you think pressing in the 209's with the single spacer is ok so long as the force is across inner and outer races will will work? I suppose it has to. all press work done with an arbor press
Last edited by Mcgyver; 07-04-2012 at 11:51 AM.
This had me second guessing enough to investigate my backgear. I assembled mine about three years ago, definitely recalled an outer spacer but didn't photograph it. So... I just disassembled mine to have a look and sure enough there is an outer spacer between those bearings. See the photos below.
It was worth the work and a set of new gaskets to be certain it was correctly assembled.
There is preload on the outer races: there is no side-to-side backlash and the ring can't be moved with moderate force. Additionally there is perceptable bearing drag when rotating the gear.
For another sanity check see Monarch drawing 204506 which shows a later version of the backgear. The output bearing arrangement is the same. Two spacer rings are shown between the bearings. I've never seen a drawing of this earlier type gearbox we have.
My output bearings are a press fit onto the shaft and push into their bore, location is via the snap ring and its cover. Same fit situation for the smaller ball bearings on the countershaft: press onto shaft, push into bores.
Last edited by DaveE907; 07-05-2012 at 07:22 AM.
The preloading of bearings on the output shaft the way I described creates a duplex back arrangement out of standard ball bearings, and you set your own preload.
With the spacer missing on the outer races, there will be more wobble of the pulley due to the internal clearances in the bearings.
Once the 4 bolts are installed that clamp the pulley on, the bearing breload will now be determined by the spacers since the snap ring is captive to prevent in/out motion of the output shaft. The outer races are not clamped in this application. There will be very little "slop or wobble" with the bearings preloaded.
If you want to use only one spacer, press the bearing with the snap ring in first (with the snap ring on). Then press the other bearing in the housing after inserting the inner race spacer. Support the bearing by the outer race while pressing into the housing. Last step is to press the shaft into the bearings. Support the bearing with the snap ring by the inner race while pressing the shaft thru.
The proper drawing is Electrical Drive Unit EE3 sheet 101 showing 2 spacers and bearings (Conrad) like Mcgyver has. The 1941(flat belt) I have worked on has two L13 angular contact in a duplex back setup. My Wiad has 2 SKF 7209 BEA in a duplex back. The shafts are splined and there are 2 spacers between bearings.
I appreciate both your efforts - Dave disassembling is above and beyond the call of duty!
It's frustrating when there is so much seemingly conflicting information. Based on the info you guys have I'd almost think i'd lost my mind and somehow didn't notice that outer ring....but there's an old thread by DaveC that shows the set up I have. In post 9 of 10EE back gear bearing removal - how?
there is a pic (now just a red x box) top of post 9 showing a freshly extracted assembly with only a inner ring. The pic was there there other day....now the link is bad, but its still in googles cache. Hopefully this calls it up
Mine is a '41, flatbelt, woodruf keyed shaft. It came with the MG gone and the cut off gearbox loose inside. I would be surprised if the gearbox wasn't the orginal, but I can't be sure either - ie there is a remote possibility its a later one like the one in DaveC's thread. It seems surprising there would be so much variation (angular contact vs radial ball, one spacer/two spacers) with lathes made right at the same time.
I don't think my gearbox was opened up or messed with...and DaveC's pic of a similair setup suggests to me I'm not missing any parts.. My conclusion on 45x85mm radial bearings is also what DaveC found...vs angular contacts.
Should I make an outer ring? ITs only going to do something if its slightly different than the inner - Dave is yours still apart; can you measure the thicknesses?
what about this idea.....Any preload force is limited to the force created by the press fit - ie even with angular contacts, the snap ring fixes things from one side but there is nothing to create force from the other...so the force preloading the bearings can't be greater than the press fit. Bill, the approach you described, in pressing the second bearing in, still loads the balls up by the press fit force - ie you're pressing on the outer ring (or both) and its stopped by the inner race hitting the ring. Therefore we should be safe in saying the radial ball bearings are up to withstanding this force (assuming I'm not wrong about the original assembly being just one ring on my lathe)
IF we assume mine isn't missing parts, and if the above paragraph is correct, pressing the second bearing in and having it stopped by the inner ring would preload the two bearings to some extent. To preload it less is where you'd need the two rings.
I'm as much asking questions as making statements.... I've not a lot of roller element bearing experience and just a wee bit of theory.
PS, also odd the descripancy on which element is the press fit. I carefully measured but will do so again
thanks for all the help!
Last edited by Mcgyver; 07-05-2012 at 03:40 PM.
Fafnir and others used standard ball bearings and ground the inner races to make angular contact duplexes out of standard Conrad type bearings. Bridgeport used them for spindle bearings, and it appears that Monarch uses them here. The war effort probably forced many "fit together whatever we can get" machine builds.
Since the inner races are clamped when assembled, and the snap ring keeps things from shifting, there is a lot of leeway that can be done and things will run fine for years.
"Bill, the approach you described, in pressing the second bearing in, still loads the balls up by the press fit force - ie you're pressing on the outer ring (or both) and its stopped by the inner race hitting the ring. Therefore we should be safe in saying the radial ball bearings are up to withstanding this force (assuming I'm not wrong about the original assembly being just one ring on my lathe)"
Yes that is a safe assumption on all points. The spacer for the outer ring (missing) just prevents the loss of preload as the outside race shifts towards the other bearing from the belt load. Even if this happens, things will still work. Bdding the second spacer you can set the preload (even with well worn bearings) and it will stay set till the bearings wear much more.
Bill makes good points.
If you want to make an outer spacer and be entirely safe from putting preload on the new bearings (not ground for preload in a matched pair) simply make the new ring the same thickness as the inner ring. It would prevent the inner bearing outer race from moving beyond what it could if the bearings were clamped together. If you measure the axial movement of inner versus outer rings from co-planar towards it's mated bearing and make an outer ring with width equal to the sum of the inner ring width plus the measured movement of both bearings you'll get an assembly with no preload and no axial play under no-load conditions. That would be better situation than no outer ring. Additional outer ring width would add bearing preload.
My bearings are pressed onto the shaft and would require a bearing separator to remove thus likely damaging them. The best I can do is provide the measurement across both outer races and the spacer: 1.7525. Of course I can't measure the inner or outer spacer widths.
thanks both of you for all the help...without the discussion and guidance I would have likley made a mess of things!
I'm going to recheck and my fit measurements to make sure I've got it right on what presses to what and then get this done.