[RICHARD ZASTROW]

arizonavideo, there are double row angular contact ball bearings of more than one design. Look in the catalogs to define which one your looking at. The 5200 series are not preloaded, they have axial clearance as you describe.

The split race types can be made with preload (normal) or with clearance (not normal and probably not readily available).

Dick Z

[philbur]

As far as I can see in both SKF and NSK catalogs, double row angular contact bearings with or without a split inner ring come with clearance values only. Nowhere can I find preload as an option. If somebody knows how to order this bearing type with factory set preload could they please post it, because at the moment the manufacturers are doing a very good job of hiding it.

Phil

arizonavideo, there are double row angular contact ball bearings of more than one design. Look in the catalogs to define which one your looking at. The 5200 series are not preloaded, they have axial clearance as you describe.

The split race types can be made with preload (normal) or with clearance (not normal and probably not readily available).

Dick Z

[RICHARD ZASTROW]

The whole discussion/problem can be solved easily and relatively inexpensive by simply using a pair of sealed angular contact bearings of whatever contact angle, quality level and preload you want.

This is a proven and commonly utilized arrangement. Why reinvent the wheel? Unless you just like to argue (or be a troll)?

Dick Z

[handlewanker]

Hi Phil, in the SKF cat, specs for the double row A/C bearing with split inner race state that there is a clearance, but do not mention the term "preload"......this could interpret that with a split inner race and clearance (between the faces) when you fit to a spindle and secure the inner races by a screwed collar you will take up this clearance....applying a loading against the two inner races.......unless you are able to gauge the force applied by the screwed collar you will either get a bearing that runs loose or with a fully tightened up collar one that is under a preload condition.......without being able to measure the loading application, but I think it is hit and miss, unless it is left to the assembler to tighten the collar progressively until the bearing runs without getting hot or rattleing.

However, it would also seem that the split inner race is a design purely to allow dismantling the bearing inner races for assembly seperately to the shaft once it is inserted into the bearing outer shell, ball and cage assembly, a purpose I can't see the reason for, and in the illustration of the split inner race bearing the "clearance is shown to be a distance dimension from the inner race face to the outer race face...puzzling.

I would assume that the double row A/C with single inner race and having clearance, means that the inner race is able to move axially at a certain rate.....possibly not designed with the intention of preload spindle requirements in mind.

Personally, if this is the case in both designs, ie no preload as a design spec, I would definately not be using this type of bearing set for a mill spindle......which is a pity 'cos it seems such a neat package, especially as it also comes with seals too.

As a matter of interest, it would be no great deal to....and here I am suggesting complete heresy....surface grind .02mm off the face of ONE of the inner races which will bring them together and...voila...we have a preload.....monitored by testing the running temperature of the spindle after assembly.....adding a shim if'n you overdo it.

On a regular pair of A/C bearings, I once added a .02mm thick shim to the outer race faces, (cups facing outwards) as they were a pair of odd unmatched bearings I had acquired, and the bearing ran hot after 5 minutes.....even an unmatched pair were a good match in the unshimmed condition, and in the unshimmed condition have been in service for over 10 years now.

So there you have it....a variable double row A/C bearing with split inner races and seals....what more do you want?

I would assume that if'n the need arose for a preloded A/C bearing set up, SKF et al would suggest a matched pair of regular single row A/C bearings in O or X configuration.....however it wouldn't take much to rub the inner race face on one side down on wet and dry paper in the double row bearing set to make them preload as required......more heresy....it would be simpler to use 2 A/C's.
Ian.

[philbur]

There have been an awful lot of unsupported claims made in this thread. If challenging some of those claims makes me a Troll then I am happy with that categorisation. I leave others to categorise those that are unable to support their claims with any credible references.

Phil

[handlewanker]

Hi, If'n the cat doesn't give a bearing type no. to specify when ordering that's a bit offputting......if'n you find out from SKF et al how to order in the preloaded state, post the results, it's probably in the nature of a special order etc.

When a standard bearing arrangement becomes a special order, it probably means the special order is catered for by their standard range.....IE the double row A/C bearing can be ordered with degrees of clearance, (not sure what for), but that is an "off the shelf" scenario and price will reflect accordingly by mass manufacture techniques, but when the same bearing type becomes a "special" it might be dearer than the paired A/C set that it's supposed to replace.

You could always just order the standard split inner race double row A/C set and "adjust" the arrangement to give a preload etc etc, we are after all a bunch of inquisitive experimenters at heart, and the "nothing ventured nothing gained" scenario would apply.

I have no idea of the price difference of the double row A/C bearing with split inner race compared to a regular normal tolerance pair of single row A/C bearings, so the difference might be minimal and the effort to mess around too much.

You know where you are with a tried and trusted bearing design using matched pairs of A/C's etc, so there's no point in reinventing the wheel just to be different......the problem arises when you want to up the speed without adding to the cost for specials etc.

This thread has reinforced my faith in the spindle design with the matched paired A/C bottom end bearing set, but for the top I'm still laying my money on a single row, (possibly double row) sealed radial bearing for the sheer simplicity and trouble free running, but I am also now looking at a roller bearing for the top so that the races can be secured to spindle and housing respectively, and the movement for expansion can take place within the bearing, and I hope the type is also available with integral seals.
Ian.

[packrat]

This thread has reinforced my faith in the spindle design with the matched paired A/C bottom end bearing set, but for the top I'm still laying my money on a single row, (possibly double row) sealed radial bearing for the sheer simplicity and trouble free running, but I am also now looking at a roller bearing for the top so that the races can be secured to spindle and housing respectively, and the movement for expansion can take place within the bearing, and I hope the type is also available with integral seals.
Ian.

It has worked for BridgePort for a long, long time. Go for it

[handlewanker]

I get a warm glow when something I think and want to do is the way it's been done before....sometimes it doesn't pay to stray too far from your box just to see what's round the corner.....if'n Bridgeport found it to be good, who'm I to argue against that logic.

In Cris's case he is designing for a purpose, not to make a Bridgeport lookalike spindle, but to make something that will apply the same principles of spindle support as a Bridgeport, but tailor it for DIY use with materials off the shelf.

It would appear that the jury is still out on the aspect that the double row A/C bearing package has preload or not, which makes it a "no go" for any spindle design that would be suitable for a milling operation and definately needs to be preloaded.

In actual fact, cost aside, it does not involve any more or less machining to fit two A/C bearings in place of the double row A/C set.

There still needs to be a machined housing with bearing cap to retain the outer race/s and a spindle with a retaining nut behind the inner race, and a seal system to suit.

I would not suscribe to a press fit for the bearing outer race/s in the housing, or a press fit for the inner races on the spindle....this would be counterproductive to assembly and maintenance.....so standard pracice for bearing fits in these areas are already established.

That just leaves the cost of a single double row or twin seperate bearing package to sway the argument.

Where the argument gets thin is when the spindle, being for CNC, would like to be used with a high speed aspect as the prime reason for making it.
Ian.

[Chrisjh]

Hi All,
Sorry to be absent so long. I received the dreaded phone call. Someone wanted me to come back to manage some technology transfer between continents. I am supposed to be retired and I don’t know what they will do when I kick the bucket.

I also had a disaster with my CNC lathe. I broke the Z Axis ball screw. Entirely my fault, I had under designed the section where the bearing retaining nut threaded area is undercut at the driven end. Happily I have recovered and the lathe is working again. It will never break again!!!

Recently I haven’t had time to devote to my spindle design. I actually concluded that, for purists, 2 Angular Bearings back to back with a preload nut is the best way forward and have come up with design to this objective. I actually completed the 3D model for this before the phone call. It took me some effort today to revisit and understand what I had done.

I was trying to avoid cutting an external thread for the preload nut on the spindle but relented.

I have a CNC lathe so I can cut threads, internal and external. See video here: [ame=http://www.youtube.com/watch?v=B63tuYJB44E]Using Mach3 to Cut an Internal Thread - YouTube[/ame]

The good news is that I have found a company in Melbourne who specialize in hardening and received some excellent advice from them. I have now decided to use 4140 steel for the spindle and get the tapered section selectively hardened using a plasma nitriding process. The cost is pleasantly low. They gave me a budgetary quote of $5 AUD per kilo for plasma nitriding but have a minimum charge around $40. I thought that this was very good.

As has been suggested, I think that, for hobbyists who can tolerate a small amount of clearance, a single double row angular bearing would be sufficient. Perhaps I can create a 2nd lower cost design using this approach later on.

Regards
Chrisjh

[philbur]

Have you considered the risk of distortion caused by nitriding.

Phil

[Chrisjh]

My understanding is that plasma nitriding presents a low risk of distortion due to the relatively low temperatures involved. Google "Plasma Nitriding" for more info.

Regards

Chrisjh

[philbur]

You have to consider the context. Distortion related to general engineering or distortion related to precision engineering. Are commercial made precision spindles nitrided, if not why not.

Google for more information.

Phil

My understanding is that plasma nitriding presents a low risk of distortion due to the relatively low temperatures involved. Google "Plasma Nitriding" for more info.

Regards

Chrisjh

[RICHARD ZASTROW]

I know back in the day G&L HBM spindles were nitrided. They were suspended vertically in a nitriding chamber that was installed in hole in the floor. I was told it was 75 ft. deep to allow long box way sections to be nitrided vertically as well.

The hardened box ways were installed in scraped cast iron machine frames then bolted in place. That's why they were called "replaceable hardened ways".

Dick Z

[mactec54]

philbur
Have you considered the risk of distortion caused by nitriding.

You don't get distortion when nitriding



Nitriding will work very well for a spindle, as long as the base metal has a good hardness to start with

You don't get any distortion with machining prehard stress relieved material, which are ideal for nitriding


Chrisjh
Make sure that all machined edges have either a chamfer or radius, a radius is best, sharp edges are not good for Nitriding
they must be all smooth, even a chamfer will have sharp edges that need to be made smooth, before Nitriding

[philbur]

Interesting, do you have any links/references.

If it's already hard to start with doesn't he really need to nitride his low cost DIY spindle.

Phil

[ihavenofish]

Interesting, do you have any links/references.

If it's already hard to start with doesn't he really need to nitride his low cost DIY spindle.

Phil

the 4140 htsr should be hrc 32 range. hard enough for everything but the taper in a low power spindle. if he nitrides only the inside taper, it should not pose much warping risk. the taper will need to be ground or hit with a ceramic cutter, and thats about it.

[mactec54]

philbur
Interesting, do you have any links/references.

The nitriding puts a hard skin on the surface, Just like when you Anodize aluminum you get a hard surface, This is not as thick as case hardening but will give a longer life on the taper

As for references, I'm sure there is plenty if you do a Google search, I my self Have had hundreds of parts Nitrided, mostly slides, but some spindle parts

You can see one of my spindle builds here, This is smaller than this project, But I do make them up to a 30 taper
http://www.cnczone.com/forums/cnc_wo...dle_build.html

[Chrisjh]

Hi Mactec,

Thanks for your advice re radii on edges. I appreciate the "voice of experience". I will incorporate a radius on the leading edge of the taper.

The link to your ER Collet design is inspiring. This has given me another idea. I have a lot of ER32 collets. Does your design have a method of locking the spindle for collet changes?

Regards

Chrisjh

[Chrisjh]

Hi Mactec,

Just realized that my question regarding spindle lock was silly. All you need is 2 spanners.

Regards

Chrisjh

[mactec54]

Chrisjh

Yes 2 spanners/wrenches are best for the ER type spindles, the less you have to do to the spindle, the better to keep it in balance, Doing a notch or a hole for a pin is not good, unless everything is symmetrical

The reason I added the extra flats on the ER collet holder, was to keep it symmetrical for better balance

You are on the right track now with your spindle build