Spindle Bearing Replacement For The Rf-31 Mill/drill

mikey

Platinum
Staff member
Rest In Peace
Joined
Dec 20, 2012
Messages
9,422
Original upper.jpg

During my quest to eliminate the excessive runout on my RF-31 Mill/Drill I determined that the spindle bearings were a major contributor. As you can see, these bearings have seen better days. In addition to being filthy, they have not been maintained and have worn much earlier than they should have. Rather than find a direct replacement I wanted to use sealed/lubricated for life/maintenance-free angular contact bearings because the construction of the head on the RF-31 provides multiple areas for ingress of dust, dirt and chips that will eventually end up on the exposed upper spindle bearing. This is not good for longevity so I wanted to eliminate the possibility.

Surprisingly, this doesn’t appear to be a common mod as information on this subject turned up little. However, one of our very own HM members, Canuck75, did this exact job. I contacted him and he encouraged me to go for it and suggested that I provide a detailed write up for the benefit of the other RF-31 owners on our forum. This post is to honor that request.

The bearings we used are from FAG, a German maker of precision bearings:

Upper: FAG 7206-B-XL-2RS-TVP

Lower: FAG 7207-B-XL-2RS-TVP

· These bearings have the exact same dimensions as the stock bearings but are each 1mm thinner, which is a minor issue I’ll touch on later. Both are rubber-sealed on both sides and lubricated for life.

· They are single row angular contact bearings so they must be used in pairs as we are doing here in order to sustain the bi-directional axial loads a spindle experiences. These bearings have a 40° contact angle, which means they can sustain large axial loads while handling very high radial loads as well.

· The XL designation means the raceways have an improved geometry and are precision-honed to improve their load bearing capacity.

· The TVP suffix means the cage is made from glass fiber reinforced polyamide, not metal.

· They are accuracy class P5 bearings, the equivalent of ABEC 5. They are probably much more accurate than the spindle and quill are.

Cost will vary with your source and luck. I was able to find both brand new for under $100 but retail for them is somewhere around $300.00 or so.

I contacted FAG for instructions on installation procedures, preloading specifications and their recommended break-in procedure for these specific bearings. Here is what they had to say:

Normally, in a spindle, you have a break-in routine to distribute grease because the bearings are usually un-sealed. But these bearings are sealed and lubed for life, and more robust than your average spindle bearing, so there is no need for that. The newer x-life products (-XL suffix) have such finely honed raceways that there is no measurable burnishing of the surface finish. [Therefore] no special run-in required.

There are a dozen variables that could influence the running temps of your bearings, and we know virtually nothing about your system. So it is not possible to predict what temps your bearings might see. But the upper operating limit of your bearings is limited by the cage, seals, and the grease, and should not exceed 100 deg. C. A good target would probably be 75-80 C max. More than anything, you should expect to reach a steady-state temp within about 30 minutes. If you do not reach steady state, or if you rise above these temps in a short time, that might be an indication of excessive preload.

As I said, these bearings must be installed in pairs; they can be installed either back to back or belly to belly, meaning the inscribed bearing designation on one face can face in or out but both must be facing the same way – X/X or O/O but not X/0. Supposedly, X/X is slightly stronger (inscriptions facing out).

FAG emphasizes in their literature that installation forces should be applied only to the race being pressed to avoid brinelling (the permanent indentation of a hard surface) of the precision-honed races. Therefore, pressing adapters that contact only the race involved is wise. They also advise the use of an arbor press, not a hammer. (Sorry, I know us hobby guys love to use our hammers but this is not one of those times.) I used a Dake hydraulic press for this project and it worked fine, albeit with less tactile feel than an arbor press would give.

The press fits are only about 0.0002” so there isn’t a lot of force involved. However, alignment and the application of force to the right area is critical so I made pressing adapters from solid aluminum rod as I had trouble finding pipe of the appropriate size – that’s a lot of boring! The largest adapter is 2.8” OD.

Press adapters.jpg

The adapters on the left and center are used on the lower bearing. The one on the right is used on the top bearing and engages both inner and outer races at the same time; this is fine according to FAG since the two races are the same thickness and pressure applied to both races at once will not stress the internal bearing surfaces. Each adapter is sized to just fit the width of the race it is intended to contact.

Procedure:
Remove the quill from the head. This is a simple job that involves removing the depth stop, quill return knob and quill return spring along with the spring anchoring screw and quill alignment pin. Be sure to lock the quill before you remove the pinion shaft. Once the pinion is removed, catch the quill as you unlock it. The locking cotters can remain in the head.

With the quill removed from the mill, remove the little cap overlying the lower bearing. Mine is aluminum and is a right hand thread that comes off easily.

Now the spindle can be pressed out of the quill from the top. This leaves the lower bearing still on the spindle and both bearing races in the quill. The upper bearing just lifts out and off the upper spindle.

spindle out.jpg

After pressing the lower bearing off the spindle the next step is to remove the races from the quill.

The lower race comes out with some judicious tapping with a 3/8” steel rod and hammer behind the race. There are two access slots in the quill casting to allow a good angle for this. Tap alternately on both sides until the race releases.

Lower race.jpg
Removing the upper race is problematic because you can’t angle a drive rod behind the race. Fortunately, I previously made a race removal tool to do this very job on sport bike steering stem bearings and it came in really handy.

RRT1.jpg RRT2.jpg
The tool is inserted so the fingers face the race. As the tool is pulled up through the quill the fingers snap in behind the race and engage at four points. Then a few solid taps with a hammer on the head of the tool and the race pops out easily. The tool is just a length of thin-gauge stainless pipe with two through-cuts made on the band saw. The fingers are bent outward by hand, and the head is an aluminum slug. It isn’t something you’ll use often but when you need it, you need it.

Once everything is cleaned, the bearings can be installed. The manufacturer makes a big deal about cleanliness when installing their bearings. I wiped each contact surface with alcohol until the paper towel came away clean and then used a microfiber towel before I blew the area out with air and applied a light coat of oil. The bearings were unwrapped just before installation.

Here, the smallest adapter is being used to press the inner race of the lower spindle bearing in place:

LowerB1.jpg Lower on.jpg

Then the lower bearing outer race is pressed into the lower end of the quill:

Lower into quill.jpg

Once the lower bearing is installed, the upper bearing can be pressed onto the upper spindle. The upper bearing is supported underneath by the aluminum adapter that rides on the rim of both the inner and outer races. I applied force to the lower bearing rim with the large adapter and got the upper bearing in but had to also apply a little follow up pressure on the spindle nose to be sure both bearings were bottomed out.

Upper B.jpg

Again, these bearings are installed back to back so that their inscriptions face out.

Inscription.jpg

Recall that I said these bearings are 1mm thinner than the tapered bearings they replace. It turns out that this matters because the preload nut has a limited amount of downward travel. At this limit there is about a 1mm space between the bearing and the tabbed washer so the bearing cannot be preloaded. Therefore, a spacer with the dimensions of 1.4” OD X 1.2” ID X 2mm thick was made from 6061-T6 Aluminum to take up this space. Aluminum is fine for this purpose – it just needs to solidly take up space between the inner bearing race and the tabbed lock washer. You can just see it under the tabbed washer in the pic below.

Finished.jpg

The castellated tabbed washer is placed on top of the spacer (Note: the castellated washer has a downward-pointing tab that must fit in a groove in the spindle) and the preload nut is threaded on – LEFT HAND Thread; the beveled edge of the preload nut faces the bearing. As these bearings do not have a specific preload value I just snugged the first nut very firmly with a spanner and a spindle wrench and engaged one of the washer tabs. The second nut is firmly tightened down on top of the first one and the spindle is complete and ready for installation in the mill. The Luminar 28mm spindle wrench held the spindle solidly while tightening the preload nut and I can recommend it.

I checked runout at the bench with the quill clamped down firmly. Runout was less than but close to 0.0001” TIR. The quill was then installed and the bearings were run at 1800 rpm for ½ hour.

I checked the temps directly at each spindle bearing with an infrared thermometer to be sure that preload wasn’t excessive. Both bearings leveled out at 41.9°C at 25 minutes and then the temperature remained stable.

Run out was checked with a Compac 215GA inside the R8 taper after the bearings were hot; it was 0.002”TIR. While this was disappointing I knew the spindle was running pretty true so the runout I was seeing had to be due to the drive sleeve bearings, which it was. Runout after the drive sleeve bearings were changed dropped to less than 0.0001” TIR. (I’ll discuss the drive sleeve bearings in another thread.)

Overall, this is a very simple maintenance step that will reduce runout and the need for bearing maintenance for the useful life of the bearing. Once the pressing adapters were made it was done in under 5 minutes.

Once again, my thanks to Canuck75 for doing the legwork to find the correct bearings!



Mike
 
Last edited:
Excellent how to Mike. Good job!
One of their (Canadian) outlets is not too far from my shop.

FAG Bearings Limited
Address: 801 Ontario St, Stratford, ON N5A 7Y2
Phone: (519) 271-3231
 

During my quest to eliminate the excessive runout on my RF-31 Mill/Drill I determined that the spindle bearings were a major contributor. As you can see, these bearings have seen better days. In addition to being filthy, they have not been maintained and have worn much earlier than they should have. Rather than find a direct replacement I wanted to use sealed/lubricated for life/maintenance-free angular contact bearings because the construction of the head on the RF-31 provides multiple areas for ingress of dust, dirt and chips that will eventually end up on the exposed upper spindle bearing. This is not good for longevity so I wanted to eliminate the possibility.

Surprisingly, this doesn’t appear to be a common mod as information on this subject turned up little. However, one of our very own HM members, Canuck75, did this exact job. I contacted him and he encouraged me to go for it and suggested that I provide a detailed write up for the benefit of the other RF-31 owners on our forum. This post is to honor that request.

The bearings we used are from FAG, a German maker of precision bearings:

Upper: FAG 7206-B-XL-2RS-TVP

Lower: FAG 7207-B-XL-2RS-TVP

· These bearings have the exact same dimensions as the stock bearings but are each 1mm thinner, which is a minor issue I’ll touch on later. Both are rubber-sealed on both sides and lubricated for life.

· They are single row angular contact bearings so they must be used in pairs as we are doing here in order to sustain the bi-directional axial loads a spindle experiences. These bearings have a 40° contact angle, which means they can sustain large axial loads while handling very high radial loads as well.

· The XL designation means the raceways have an improved geometry and are precision-honed to improve their load bearing capacity.

· The TVP suffix means the cage is made from glass fiber reinforced polyamide, not metal.

· They are accuracy class P5 bearings, the equivalent of ABEC 5. They are probably much more accurate than the spindle and quill are.

Cost will vary with your source and luck. I was able to find both brand new for under $100 but retail for them is somewhere around $300.00 or so.

I contacted FAG for instructions on installation procedures, preloading specifications and their recommended break-in procedure for these specific bearings. Here is what they had to say:

Normally, in a spindle, you have a break-in routine to distribute grease because the bearings are usually un-sealed. But these bearings are sealed and lubed for life, and more robust than your average spindle bearing, so there is no need for that. The newer x-life products (-XL suffix) have such finely honed raceways that there is no measurable burnishing of the surface finish. [Therefore] no special run-in required.

There are a dozen variables that could influence the running temps of your bearings, and we know virtually nothing about your system. So it is not possible to predict what temps your bearings might see. But the upper operating limit of your bearings is limited by the cage, seals, and the grease, and should not exceed 100 deg. C. A good target would probably be 75-80 C max. More than anything, you should expect to reach a steady-state temp within about 30 minutes. If you do not reach steady state, or if you rise above these temps in a short time, that might be an indication of excessive preload.

As I said, these bearings must be installed in pairs; they can be installed either back to back or belly to belly, meaning the inscribed bearing designation on one face can face in or out but both must be facing the same way – X/X or O/O but not X/0. Supposedly, X/X is slightly stronger (inscriptions facing out).

FAG emphasizes in their literature that installation forces should be applied only to the race being pressed to avoid brinelling (the permanent indentation of a hard surface) of the precision-honed races. Therefore, pressing adapters that contact only the race involved is wise. They also advise the use of an arbor press, not a hammer. (Sorry, I know us hobby guys love to use our hammers but this is not one of those times.) I used a Dake hydraulic press for this project and it worked fine, albeit with less tactile feel than an arbor press would give.

The press fits are only about 0.0002” so there isn’t a lot of force involved. However, alignment and the application of force to the right area is critical so I made pressing adapters from solid aluminum rod as I had trouble finding pipe of the appropriate size – that’s a lot of boring! The largest adapter is 2.8” OD.


The adapters on the left and center are used on the lower bearing. The one on the right is used on the top bearing and engages both inner and outer races at the same time; this is fine according to FAG since the two races are the same thickness and pressure applied to both races at once will not stress the internal bearing surfaces. Each adapter is sized to just fit the width of the race it is intended to contact.

Procedure:
Remove the quill from the head. This is a simple job that involves removing the depth stop, quill return knob and quill return spring along with the spring anchoring screw and quill alignment pin. Be sure to lock the quill before you remove the pinion shaft. Once the pinion is removed, catch the quill as you unlock it. The locking cotters can remain in the head.

With the quill removed from the mill, remove the little cap overlying the lower bearing. Mine is aluminum and is a right hand thread that comes off easily.

Now the spindle can be pressed out of the quill from the top. This leaves the lower bearing still on the spindle and both bearing races in the quill. The upper bearing just lifts out and off the upper spindle.


After pressing the lower bearing off the spindle the next step is to remove the races from the quill.

The lower race comes out with some judicious tapping with a 3/8” steel rod and hammer behind the race. There are two access slots in the quill casting to allow a good angle for this. Tap alternately on both sides until the race releases.

Removing the upper race is problematic because you can’t angle a drive rod behind the race. Fortunately, I previously made a race removal tool to do this very job on sport bike steering stem bearings and it came in really handy.

The tool is inserted so the fingers face the race. As the tool is pulled up through the quill the fingers snap in behind the race and engage at four points. Then a few solid taps with a hammer on the head of the tool and the race pops out easily. The tool is just a length of thin-gauge stainless pipe with two through-cuts made on the band saw. The fingers are bent outward by hand, and the head is an aluminum slug. It isn’t something you’ll use often but when you need it, you need it.

Once everything is cleaned, the bearings can be installed. The manufacturer makes a big deal about cleanliness when installing their bearings. I wiped each contact surface with alcohol until the paper towel came away clean and then used a microfiber towel before I blew the area out with air and applied a light coat of oil. The bearings were unwrapped just before installation.

Here, the smallest adapter is being used to press the inner race of the lower spindle bearing in place:

View attachment 132856 View attachment 132854

Then the lower bearing outer race is pressed into the lower end of the quill:

View attachment 132853

Once the lower bearing is installed, the upper bearing can be pressed onto the upper spindle. The upper bearing is supported underneath by the aluminum adapter that rides on the rim of both the inner and outer races. I applied force to the lower bearing rim with the large adapter and got the upper bearing in but had to also apply a little follow up pressure on the spindle nose to be sure both bearings were bottomed out.

View attachment 132859

Again, these bearings are installed back to back so that their inscriptions face out.

View attachment 132851

Recall that I said these bearings are 1mm thinner than the tapered bearings they replace. It turns out that this matters because the preload nut has a limited amount of downward travel. At this limit there is about a 1mm space between the bearing and the tabbed washer so the bearing cannot be preloaded. Therefore, a spacer with the dimensions of 1.4” OD X 1.2” ID X 2mm thick was made from 6061-T6 Aluminum to take up this space. Aluminum is fine for this purpose – it just needs to solidly take up space between the inner bearing race and the tabbed lock washer. You can just see it under the tabbed washer in the pic below.

View attachment 132852

The castellated tabbed washer is placed on top of the spacer (Note: the castellated washer has a downward-pointing tab that must fit in a groove in the spindle) and the preload nut is threaded on – LEFT HAND Thread; the beveled edge of the preload nut faces the bearing. As these bearings do not have a specific preload value I just snugged the first nut very firmly with a spanner and a spindle wrench and engaged one of the washer tabs. The second nut is firmly tightened down on top of the first one and the spindle is complete and ready for installation in the mill. The Luminar 28mm spindle wrench held the spindle solidly while tightening the preload nut and I can recommend it.

I checked runout at the bench with the quill clamped down firmly. Runout was less than but close to 0.0001” TIR. The quill was then installed and the bearings were run at 1800 rpm for ½ hour.

I checked the temps directly at each spindle bearing with an infrared thermometer to be sure that preload wasn’t excessive. Both bearings leveled out at 41.9°C at 25 minutes and then the temperature remained stable.

Run out was checked with a Compac 215GA inside the R8 taper after the bearings were hot; it was 0.002”TIR. While this was disappointing I knew the spindle was running pretty true so the runout I was seeing had to be due to the drive sleeve bearings, which it was. Runout after the drive sleeve bearings were changed dropped to less than 0.0001” TIR. (I’ll discuss the drive sleeve bearings in another thread.)

Overall, this is a very simple maintenance step that will reduce runout and the need for bearing maintenance for the useful life of the bearing. Once the pressing adapters were made it was done in under 5 minutes.

Once again, my thanks to Canuck75 for doing the legwork to find the correct bearings!



Mike




Mike

Looks like you did a great job but if you will forgive me I have a couple of observations. You said that you can install the angular bearings back to back or belly to belly but not opposite. I agree that they cannot be opposite but in this installation they still have a right and wrong way. If you look at the bearings you will see a wide side and a narrow side on each inner race. The same applies to the outer races. The preload in our installation is actually applied to the inner races as the outer races are pressed into the quill top and bottom. In my view that means the wide side of the inner race must be down on the bottom bearing and up at the top bearing. It may well work the way you installed them, but in my humble opinion you are not going to realize the maximum axial or radial strengths the angular aspect of these bearings offer. I will gladly accept any information to prove me wrong but I think I am right.

Secondly, the installation of the top nuts and notched washer is a little opposite to convention. The bottom of these two notched nuts should bear directly on the inner race of the bearing and is tightened as necessary for preload. Then the tabbed washer is put on top followed by the top nut. The tabbed washer has one inner tab and the rest are outer tabs. The inner tab is captured by the spindle spline thus preventing any rotation. Now locate a tab that is aligned exactly with a lower nut notch (after preload) and bend it down thus capturing the bottom nut relative to the spindle and preload. The top nut is just brought up snug to the tabbed washer and bottom nut. Finding another tab that is aligned exactly with a top nut notch that tab is then bent up. I beg your forgiveness but the way you have installed the tabbed washer the tabs will not reach the top nut. They are meant to capture both so nothing can move or back off.

Respectfully
Canck 75
 
Last edited:
Mike

You mentioned that the RF31 spindle has many areas that dust/dirt/chips can ingress and I agree

On my machine there is a large access hole in the bottom of the head casting just behind the quill where dust/dirt/chips can and did get in, and from there to the top bearing as it did on yours. Also going by your pictures I don't see a knockout slot in the actual spindle so therefore a R8 spindle (drawbar), but there is in the quill. These slots also provide an ingress route for crud right into the quill just above the bottom bearing when the quill is extended. It is easy to see why the unsealed roller bearings could get gunked up so easy. By the way my spindle is R8 and the quill does not have a knockout slot. Many of these machines must have had taper vice R8 spindles resulting in some mix and match of parts in manufacture.

Using the sealed bearings is going to solve much of your problem, but further to that, you should make two plastic plugs for the sides of your quill and a plate to cover the bottom opening in the head casting (assuming yours is like mine). I suggest this because it will prevent gunk from getting onto the rack and pinion area on the back of the quill which is equally important.

Canuck 75
 
Last edited:
Canuck,
In my communication with FAG, they made it clear that the orientation I used is what is preferred. I don't know that it matters a great deal because the spindle does not sustain a large amount of axial loading. I will give this a try and see how it goes.

As for the positioning of the castellated washer, I am following the original installation and the order of components in the manual.

ScreenShot016.jpg

Are you sure about your order? I will admit that this is the first RF-31 I have worked on but have done a number of drill presses and all of them have the same arrangement. In a drill press at least, the castellated washer locks the preload nut and the top nut serves as a jam nut. I'm willing to admit I got it wrong but I don't think so.
 
Mike

You mentioned that the RF31 spindle has many areas that dust/dirt/chips can ingress and I agree

On my machine there is a large access hole in the bottom of the head casting just behind the quill where dust/dirt/chips can and did get in, and from there to the top bearing as it did on yours. Also going by your pictures I don't see a knockout slot in the actual spindle so therefore a R8 spindle (drawbar), but there is in the quill. These slots also provide an ingress route for crud right into the quill just above the bottom bearing when the quill is extended. It is easy to see why the unsealed roller bearings could get gunked up so easy. By the way my spindle is R8 and the quill does not have a knockout slot. Many of these machines must have had taper vice R8 spindles resulting in some mix and match of parts in manufacture.

Using the sealed bearings is going to solve much of your problem, but further to that, you should make two plastic plugs for the sides of your quill and a plate to cover the bottom opening in the head casting (assuming yours is like mine). I suggest this because it will prevent gunk from getting onto the rack and pinion area on the back of the quill which is equally important.

Canuck 75

Yeah, my quill has those useless slots. Those Taiwanese copied the drill press design down to the stupid slots intended to knock out a Morse taper, in a machine that does not use a Morse taper. They're nothing if not faithful copiers. When I took the quill apart there was no debris inside, unlike the rest of the machine that had dust, dirt and debris everywhere. My thought was to leave them alone and see if chips made their way in before deciding to block them off. I'll keep your advice in mind, though.

I also have a hole in the head casting behind the quill and have made the template for a block-off plate to stop ingress of chips and debris. I'm trying to clean my shop up so I can move again and I'll get the plate done.
 
Yeah, my quill has those useless slots. Those Taiwanese copied the drill press design down to the stupid slots intended to knock out a Morse taper, in a machine that does not use a Morse taper. They're nothing if not faithful copiers. When I took the quill apart there was no debris inside, unlike the rest of the machine that had dust, dirt and debris everywhere. My thought was to leave them alone and see if chips made their way in before deciding to block them off. I'll keep your advice in mind, though.

I also have a hole in the head casting behind the quill and have made the template for a block-off plate to stop ingress of chips and debris. I'm trying to clean my shop up so I can move again and I'll get the plate done.

On my RF-31 clone I made a thin wall sleeve that I pressed into the quill to block off the slots. I also made a seal carrier that replaces the lower bearing cap and a plate to block off the opening in the head casting. All three mods work great at keeping swarf and coolant out of the lower bearing. If you want to see pictures let me know and I'll post some.

Tom S.
 
Mikey.

Sounds like you have a good handle on the issues at hand. I'm now going to look at the drive pulley bearings to make sure they are okay.

With regard to the angular bearing orientation, this crosssection drawing shows what I am referring to. I don't understand why FAG suggested otherwise. In our particular installation, and the way you have installed them, the preload is driving the bearing races apart. The way I have installed mine, the preload drives the bearing races together - big big difference!

0908%200077%20-%2010000%20w_tcm_12-127750.jpg

Really appreciate trading ideas and information. These are great little mills. Here is a pic of mine hard at work making 1 1/4" wheel spacers for my Ford Ranger.


Cheers
Canuck75

IMG_7367.jpg
 
Back
Top