Choosing Spindle Bearings

[2_many_hobbies]

I have a dream... and now I are confused.

I have a ER16 C20x150mm collet chuck and six bearings (62042RS) And a chunk of hot rolled steel tube with an 1/8th wall.

My idea was to face and boar the tube and press the chuck in with a bearing on each end and I'll have a nice sturdy spindle but as I kept reading I got worried about lateral motion causing runout. I then learned about angular contact bearings.

But what about thrust bearings I said? I originally saw an issue with screwing a nut onto the end of the chuck to keep it held in place but thought that a standard bearings inner part is not rated for any up/down forces. I figured I would need to put a thrust bearing that spreads the nuts downward force to the outer part of the spindle bearing so that the inner part is left free to keep the collet concentric. But then...

I was poking around and found a tear down of a MaxNC spindle and a Taig ER spindle and I see it looks like theirs an angular contact bearing on both ends with what looks like two regular bearings in the middle.

So now I believe angular contact bearings where designed to keep the best of both worlds, they make contact on a chosen angle (say 45deg) and will allow forces in both the x & y planes (though not as much as the thrust/regular bearing combo) and then you have two regular bearings in the middle to keep everything aligned. Am I correct in this whole line of thought? Ive gone through so many ideas that I have confused the heck out of myself and don't even want to touch this project until I'm sure.

Please forgive the MS Paint layout I usually do it in solidworks but its to late to bother. But is my layout the correct method of making the most sturdy and accurate spindle?

Thanks!

[RICHARD ZASTROW]

Depending on the speed, axial load etc. your design might work.

If there is any chance for heat build-up, you would be better served to have both the angular contact bearings secured near the chuck end. By having the "regular" bearing on the opposite end and allowed to "float", axial length growth due to heat expansion wouldn't effect the pre-load on the angular contact bearings.

I don't think you'll need 45 deg. bearings. The larger the angle, the greater the thrust capacity but the lower the radial capacity. Conversely, the smaller the angle, the lower the thrust capacity and higher radial capacity.

Typical A/C angles are 15-25 deg. with light, medium or heavy preload. For a milling spindle taking a lot of abuse you might want a 25 deg. with heavy preload. For a high speed router with small high speed carbide cutters, 15 deg. and light preload may be more appropriate.

Those 45-60 deg. A/C bearings are typically used in axial thrust only applications like fixed end ballscrew bearings.

Hope that helps

Dick Z

[2_many_hobbies]

Ok so I understand A/C bearings right then, and the reason you use two is to allow for loading in more than on direction. I never would have thought of heat build up but it makes sence to put both on the bottom thanks.

I still wonder about a holding nut on the collet chuck putting to much force on the inner part of the upper spindle bearing though, I see on the LMS MT to R8 change plans that their spindle simply has a single nut on top to keep everything tight even though its a press fit. I guess you don't crank it down and its just snug to make sure the spindle doesn't slowly slip down.

New design will be as follows:

ER16 C20-150mm Collet Chuck
2 1/8 hot rolled pipe 1/8 wall
Two A/C Bearings 20/47/14
Two Regular Bearings
Double hex nut or castle nut with cotter pin

Spindle is driven by a 8000rpm 110V DC motor on 2:1 Reduction via timing belts and sprockets. I can add air cooling later on if it gets to that, ive seen an idea on putting a fin around the spindle and it draws/expels its on cooling air. I thought it was a neat idea.

Do I have the A/C bearings the right directions?

[jalessi]

Angular contact thrust bearings typically have a contact angle of between 25 and 40 degrees. They can be paired together in three arrangements, namely tandem, face-to-face or back-to-back. The tandem arrangement increases the thrust capacity in one direction but has no provision for thrust in the opposite direction. However, back-to-back or face-to face pairing of two angular contact bearings provides bi-directional thrust capability. The back-to-back arrangement is the most common configuration. Refer to the manufacturer’s drawings for the proper bearing orientation prior to installation.

[RICHARD ZASTROW]

As mentioned above, bearing manufacturers have the data and info you need. It's all published in their catalogs and spec.'s. All is available on the net and down loadable.

For a spindle I'd be hesitant to use angles greater than 25 degrees.

Ref: your drawing, the back to back angular contact bearings should be on the collet end and the floating "regular bearing(s)" on the opposite end.

The angular contact pre-loaded pair are installed against shoulders to prevent moving from either axial force or heat growth. This is also pictured in the bearing manufacturers catalogs.

Keep on, you'll get it.

Dick Z

[Herbertkabi]

Spindle needs spindle bearings, are precision and super precision bearings, angular contact, mostly 15 degrees.
This steel tube you planned to use like housing ... Im not sure you will like the result.
Hereby my ER11 based spindle 33.000 rpm. High Speed Spindle for SIEG X2 Mill - getting technical
I have used ER20 as well.
Cheers,
Herbert

[2_many_hobbies]

Richard: You say manufacturers have all the data and even example uses available for download but I have no idea where to get this. I have only ever bought bearings from VXB.com and they have nothing at all for download. Can you post a link with the materials you use?

Herbertkabi: WOW.. I just spent 6 hours reading threads you have posted in and I'm amazed by your work, makes my pathetic attempts look like a kids first play-dough creation lol Im an electrician by trade and through posts and Wikipedia I now understand BLDC motors 10x more and even want to make my own. I have two questions for you though: Do use use direct connected motors just to make compact spindles or is direct drive always better as I would imagine a 30,000RPM BLDC on 3:1 timing belt would give you insane torque and 10,000rpm which would be enough to eat through steel. And the second is you find scrap rotors and modify them to your needs is this because of the machining difficulties of silicon steel (brittle, hard to source?) I don't even know if you can buy silicon steel wafers and machine your own lamination but I have seen many people use powdered iron and resin in their designs.

Now back to my spindle. The original spindle idea came from this site. Ill show todays train of thought so we can all follow. I read this article on spindles to learn more about the design but the "click here for pictures" doesn't work so I have no idea what their talking about when they say "the inner races are relieved. When the inner races are clamped together, the relief clearance is eliminated"

My original spindle design was to use the LMS R8 spindle conversion and build my own housing and when I decided on the ER16 collet chuck instead I read the LMS Spindle Swap Document And they only use two regular bearings so I figured my spindle will be handling the same material so my bearings should work fine. Lower bearing is 6007 actually and mine are hybrid ceramic so mine should be better even...

Herbertkabi, you said 'This steel tube you planned to use like housing ... Im not sure you will like the result' I chose steel for rigidity mostly but im thinking that because its not as conductive to heat as the collet chuck or the bearings that they will take all the heat? Should I go down to 6060 T6 aluminum as that would take the heat out of the bearings quite quickly.

I guess i should mention the type of machine I am running all this on, Its by no means an X3 quality machine as its all built by myself, table travels on C20 pillow blocks and has a travel of 110x160mm run by 112ozin steppers. The tables primary use is engraving but I wanted to have the ability to machine T6 and sence the table cant be used to hog the material through a 2000rpm spindle I need a much faster one to take the load off the table.

My engraving spindle is a 1" x 5" piece of T6 as the housing with a 3/8" T6 spindle that uses the collet of a dremel tool to hold the bits, im quite happy with it as its runout is only ~0.0003" and I run it a t 8000rpm via a cordless drill motor and pulley.

Im going to use Herbertkabi's picture and describe what I think things are for as I learn better from pictures and examples.

Ok, The front taper, dark green, orange, and green pieces in the front are seals and such to keep contaminants out of the bearings so lets not worry about those... Two A/C bearings put back to back with their outer races touching. Im assuming they are in aligned like this:

|#\O|#####|O/#|
|#/O|#####|O\#|

Hopefully you get what that's going for, I assume this setup because downward force on the collet would put the first A/C bearings against its race which makes sence and you pre-load the 2nd bearings so that when then downward force takes place the 2nd bearing is still well seated on its race. (is the inner race of the first A/C bearing sitting on a collar or anything or is it just a press fit?)

Pre-loading on the 2nd A/C bearing is achieved with the long (yellow in on pic, green in the cutaway) cylinder that is fit between the inner races of that A/C and the top bearing. The rear bearing carrier and ripply spring is there to allow the spindle to deform from heat and now cause the bearings to go out of alignment or pre-load right? It also appears that some amount of pre-load adjustment can be had by the amount of pressure on the top bearings inner race, in this case the motor connector could be pressed on further for a heaver pre-load.

Provided that I am understanding all of this correctly I should be able to draw up a good solidworks idea using the proper techniques.

Thanks!

[Herbertkabi]

Earlier I built direct drive spindles wheres motor connected via flex coupling, motor was vibration dampered. It is because adequate balancing was too hard for me. Now I do all on one shaft.
When you like to use Aluminium for housing then ... by my opinion ... if you choiced super precision bearings then only when spindle will be water cooled, because different t expansion of Aluminium/Steel. I use Stainless for housing.

I like direct drive spindles. When brushless motor (even rc) and rc-controller, then you only need to add some feedback electronics will keep any ordered RPM as stabil as sustained with different cutting forces (load).
Most of my spindles are 24V...36V and >1kW range - typical no-load amperage only 0.5...2 A , 2A...5A when normall milling, 10...20A when some hard works - when Constant Speed is switched out, when switched in then you can see how current plays 2A...20A ... even..30A and more... but dont worry when 60A constantly permissed and 80A peak (for example).

I did not want to be so censorious and too critical - I was?
Then please excuse me - was most of all because my English is very "DIY".
I have built very robust and somple spindles myself as well - all devolves on the money you like to spend and purpose.

But anyaway - new super precision spindle bearings you can find from ebay and cheap, often these are just wrongly described, it could take several hours to find out these "hidden" items and you must to reapeat searching every week. My last found was 2 x matched pair of 102 Barden Super Precision Bearings, new and sealed - was described like just bearings and very nebulous photo about ... cost was $1, none bidded instead me, bought for $20 shipped. Today I have really nice collection of super precision spindle bearings of most sizes I need - ebay. Only twice I bought (from De.factory) for EUR 180 each.

Hereby scetsh(!) of my latest spindle, ATC for 1/8 bits(only),
2 x SNFA VEB15(UM15) and 2 x VEB8(UM15)
60.000 RPM, Brushless ...
Now Im busy with second one - first I gave to guy who was able to make
two sets of shaft and tool holder for this spindle.

Cheers,
Herbert