Some Spindle Design Questions

[Curtis_F]

Over the past few days I’ve been reading through the custom build posts on the board looking for info on how people have been designing & building their spindle. I pretty much got information overload.

So, I still have some questions, which mainly come from my lack of knowledge of how industrial machine tools are built.

Any help on these will be appreciated. Thanks.

----------

I’ve read that the most recommended bearing setup in spindles is using two angular contact (AC) bearings (mounted either face-to-face or back-to-back) at the chucking end of the spindle. Then one or two high-radial load ball bearings at the pulley end of the spindle. The bearings are positively held a distance apart from each other by use of spacers or steps in the spindle housing that they are pressed against.

I’ve also seen some posts where all the bearing in the spindle housing are AC bearings.

Is there an advantage to using all AC bearings as opposed to a mix of AC and standard ball bearings?

----------

In one post it was mentioned that the high-radial load bearings on the pulley end should be a slip-fit on the spindle to allow for heat expansion & contraction in the spindle itself.

For all you guys that have worked on machine spindles, is that an actual practice??

----------

In the drawings of spindles posted the chucking end of the spindle has a face that is pressed against the inner race of the leading angular contact bearings.

This means that any force applied into the spindle (lathe cutting towards the chuck or drilling / plunge cutting in a mill) will be transferred directly into the inner race of the AC bearings. However, only the press fit of the spindle in the bores of the bearing handles the pull forces generated in a lathe when tool is cutting away from the spindle or in a milling operation with an end mill where the helical teeth of the end mill apply a downward force on the spindle.

Does there need to be something on the spindle shaft (like a threaded on collar) that positively locates the spindle to both sides of the AC bearings or is the one face against one bearing and the press fit of the two AC bearings enough?

----------

Are there generally seals between the Spindle and Spindle Housing (Head Stock) to prevent debris from getting into the bearings, or is something around a .010” gap enough to let the spindle turn and keep stuff out?

----------

A question specifically about lathe spindles; Is a VFD capable of being used to single point cut threads or does it require using a Servo for a spindle motor?



Thank you all for your time,

Curtis F.

Similar Threads:

• Design Questions
• Newbie- design questions
• Spindle Design and Bearings questions
• Newbie- Design questions
• Design Questions

[RICHARD ZASTROW]

Curtis, I'm not trying avoid typing long dissertations. OK, I lied.

If you "google" the precision bearing manufacturers like FAG/BARDEN, TORRINGTON, SKF etc. and study their tech sections, most of your questions will be answered.

Also, you might do a search on the "Zone" for NC Cams and his contributions on bearing questions. Our departed friend was a bearing engineer that knew and shared much valuable information. A bit irascible at times but a real help.

If you still have questions after perusing those tech sections, we'll be happy to help.

Dick Z

[RICHARD ZASTROW]

Curtis, I have a source of info for you.

Barden Bearings, Machine Tool catalog

www.bardenbearings.com/pdf%20CD/MACHINET.PDF

Dick Z

note: I keep trying to get the URL to work but I guess you'll have to do it the hard way.

[Curtis_F]

Dick,

Thank you for the replies! I had searched Torrington and SKF's websites for information and found their organization greatly lacking. The Barden/FAG catalog you pointed to has some good info and more food for thought.

The note to search for posts from "NC Cams" gave me a whole new slew of posts to read as well...this is going to take awhile.

The end goal is to see if it's viable to make a CNC mini-turret lathe (4" x 10") for machinging special screws, studs, nuts and pins. To turn a brass screw with a #1-72 thread I'm looking at a requirment of around 30,000 rpm, so I need to be serious about the bearing selection and spindle design.

Heat expansion will be a problem, along with oil lubrication, which arn't addressed much on these forums (that I've seen) so I've been having a hard time really pinning down any kind of "common practice" for this stuff.

I have pages of notes for the bed design, the crosslide, tool turret, tail stock, electronics, etc. But high speed bearings and spindles...not so much.


So, really any help is appreciated.


Thanks again,

Curtis F.

[RICHARD ZASTROW]

Curtis; I'm not the electronics guy, my background is mechanical. I doubt you'll be able to thread @ 30,000 rpm. Seems a bit much for accurate control of the tool path. Just me and my lack of expertise in things electronic.lol

Mechanically a 30,000 rpm spindle is possible, but for a lathe spindle extremely unlikely. Consider the balance problems; the centrifugal forces generated; the super precision bearings required; noise and vibration as well as heat/cooling problems.

An old rule of thumb on the bearing max rpm listing is to max the spindle rpm at 10% of the spec. ex.: bearing listed @ 10,000 rpm max., machine normally operated @ 1,000 rpm.

Did you find the tech section in the Barden Machine Tool Bearing catalog? Lotsa info there!!!!

Dick Z

[Curtis_F]

I doubt you'll be able to thread @ 30,000 rpm.

No, NO! Just turning the major diameter of the threads at that speed. The threading operation is much slower.

Mechanically a 30,000 rpm spindle is possible, but for a lathe spindle extremely unlikely. Consider the balance problems; the centrifugal forces generated; the super precision bearings required; noise and vibration as well as heat/cooling problems.

That is precisely why I'm doing this research. And even if I can't reach 30k, I'd like to get as close as I can.

The design I have in my head is based entirely around the 5C collet. The spindle shaft, as sketched up, has a max OD of 2.3" and is less than 7" long so there's very little rotational mass. I may end up with a longer spindle shaft, but I'm trying to avoid going larger in diameter to keep the mass as close as possible to the axis of rotation.

An old rule of thumb on the bearing max rpm listing is to max the spindle rpm at 10% of the spec. ex.: bearing listed @ 10,000 rpm max., machine normally operated @ 1,000 rpm.

According to SKF their super precision hybrid angular contact bearings are designed to run at 27,000 to 43,000 rpm depending on bearing materials. Of course they are expen$ive, but they'll be one of the only parts I wouldn't make myself (or have a friend make).

At present it looks like there will be three bearings in the spindle. A matched pair of SKF angular contact bearings on the head end, and a Barden/FAG floating bearing (45k RPM rated) on the drive end.

Did you find the tech section in the Barden Machine Tool Bearing catalog?

Yes! Yes, I did.


Thank you,

Curtis F.

[veteq]

Hey Curtis,

We really mis NCcams over here, he would have made some nice comments on this projects.

I have never seen a 30k lathe spindle, do you really need that kind of rpm?
What is your target for runout etc.
Why not buy a commercial spindle?

regards,

Roy B.

[veteq]

What is diameter off the piece that has to be turned?
If it is say: 0,5 inch and you put your stock material in there
(that isn`t ballanced) it would be hard for the bearings on 30krpm

[Curtis_F]

Roy, Veteg,

I arrived at the goal of 30K based on the following:

The Machinery's Handbook 26th edition says; when cutting Brass with a Carbide cutter the optimal Surface speed is in the 440 to 1170 Feet Per Minute range.

The major diameter of a #0 thread is: .060"
The major diameter of a #2 thread is: .086"
The major diameter of a 1/2" thread is: .500"

Therefore:
Optimal RPM for turning the major diameter of a #0 screw is: 28,000 to 75,000 RPM
Optimal RPM for turning the major diameter of a #2 screw is: 20,000 to 52,000 RPM
Just for comparison;
Optimal RPM for turning the major diameter of a 1/2" screw is: 3,400 to 9,000 RPM

After researching bearings a little I found that the off-the-shelf ones of the size needed to fit around a 5C collet spindle (45mm I.D.) topped out at a little over 30,000 RPM if oil lubrication is used.

Really, the machine would spend most of it's life under 10,000 RPM, but I want to shoot for the head room to be able to hit optimum speeds when I need it.

On the occasions when it would get into the 25,000+ RPM range the material will most likely be less than .19" diameter/hex stock and only sticking an inch or so out of the collet.

And if 30K isn't possible, well that would just make my life easier.


TIR: I could live with a TIR of .001", but with the kinds of speeds I'm looking at I'm going to really have to get it down as much as possible. .00005" would be nice. There will be honing and centerless grinding involved on the spindle parts that's for sure.


Roy,

As to commercially available spindles, I have looked, and Dunham Tool makes some nice ones for the price of a new car and only go to 7,000 rpm.

FAEMAT is another good spindle company, and they go up to almost 90,000 rpm, but you start getting into pressurized oil and air bearings, etc. To pay for one and the supporting equipment I'd have to sell a kidney, half my liver, a lung....


Thanks for the questions guys, it keep me thinking!


Curtis F.

[Curtis_F]

Just wanted to add that the more of NC Cams posts I read, the more saddened I am by his passing. He left some brilliant posts!

...and I'm starting to see that I may be on a fools errand for anything more than 10k rpm. I still have more to a lot more reading to do though!


Thank you all again,

Curtis F.

[veteq]

Hey Curtis,

The specs off your application are more clear to me now, if you have the possibility to use all that hightech grinding equipment i would say go for it.
It would be a hard task but if i had the opportunity i would do the same for sure!!! Make sure you make a nice project log on the forum here.

About that kidney, we should talk, maybe i will buy one in the future!!!

regards,

Roy

[RICHARD ZASTROW]

Curtis, My apologies for this late response to post #6, I had a lightening strike and was knocked out of commission while typing the response.P'd me off.

While reading your post I noticed this "max rpm 47k spec" in SKF precision bearing description.

I have a copy of the "SKF HIGH-PRECISION BEARINGS". In their tech section under "Speed Ratings" to paraphrase;

Valid provided the bearings are lightly preloaded,by means of springs [lathe spindles are normally heavily loaded by means of positive clamping]

Transport heat away from bearings is good. [Difficult without external air or liquid cooling]

When single bearings are adjusted against each other or matched sets are to be used, the speed values must be reduced. [pair of preloaded angular contact bearings]

The higher the contact angle, the greater the axial capacity & lower the speed. The lower the contact angle the greater the radial capacity & speed.

In short, the general bearing requirements for a lathe spindle tend to reduce the max. speed capacity of the bearings. The opposite is true for a high speed engraving spindles.

In reference to your bearing arrangement I'd recommend a pair of back to back medium/heavy preloaded, 25 degree contact angle bearings at the chuck end positively secured to both the spindle shaft and the housing. The rear bearing, a radial bearing secured to the spindle shaft and allowed to "float" axially to compensate for spindle length growth due to heat expansion.

If you intend to operate over 10k rpm, you'll likely need an air/oil mist lube system to avoid overheating. Even a recirculating oil system can get hot at elevated rev's.

Is the 2.3" dia. you mentioned the ID of your bearings? 55 mm? 60 mm?

Dick Z

[Curtis_F]

Dick,

Thank you for looking up the SKF specs. Proof that I need to get more bearing catalogs. I have catalogs on needle and roller bearigns, but I'm lacking them for ball bearings, espically these high-precision spindle types.

In the Barden catalog I had noticed that they note the use of springs for the preload too, I'm still learning what exactly that means. I believe that your assesment is right. I may need 100k RPM rated bearings to run at 30k with all the speed restricting requirments!

Is the 2.3" dia. you mentioned the ID of your bearings? 55 mm? 60 mm?

The 2.3" was the largest OD of the spindle itself. Though with all the reading I've done lately I don't think it would be feesable as it leaves no room for mechanical or labyrenth seals. (see attached PDF) My inital spindle design is also lacking the clearance for a collet draw tube.

The bearings I was considering were 45mm ID x 75mm OD x 16mm wide. I believe they're refered to as ABEC-7 #7009 size bearings.

In looking at the need for labyrenth seals on the head end and a nut on the drive end to apply preload on the bearings I'm probably going to have to go with a larger diameter spindle than I had planned, which also means larger bearings. Probably 50mm I.D. on the bearings.

In reference to your bearing arrangement I'd recommend a pair of back to back medium/heavy preloaded, 25 degree contact angle bearings at the chuck end positively secured to both the spindle shaft and the housing. The rear bearing, a radial bearing secured to the spindle shaft and allowed to "float" axially to compensate for spindle length growth due to heat expansion.

I like that arrangment. Having the matched pair bearings right next to each other was my initial thought, but based on what I've been reading it would be better to put spacers (ground to the same length) between them to increase the spindles stability and help decrease the TIR.

I could also use a second matched pair of bearings and have four (4) on the head end...but it'll be cheaper to make the spacers. Though I'll make the spacers the width of two more bearings just for the option of putting in more bearings at a later date.

If you intend to operate over 10k rpm, you'll likely need an air/oil mist lube system to avoid overheating. Even a recirculating oil system can get hot at elevated rev's.

Yet more research to do!


One thing my research has convinced me of is that once I work out what appears to be a good base design I will be going to a bearing manufaturer or two for their reccomendations. No matter how much reading I do, it won't equal years of real experience working with bearings.


Thanks again,

Curtis F.

[RICHARD ZASTROW]

Curtis; Yes, simple sounding things can be very complicated!!!!!!!!!

Ex., I printed your obsoleted print and noticed 45 +/- 0.35 mm. Definitely wrong for a bearing seat. 35 microns maybe and that in my mind would be excessive.

3.5 microns is recommended by Barden/FAG for 30-50 mm spindle bearings. That's .000138".

The higher the speed the higher bearing precision required. The higher the bearing precision the more precision required in the spindle shaft and housing.

The housing bore tolerance for a 75 mm OD is fixed bearing +5/-3 microns (+.000197"/-.000118"). Floating bearing is +11/+3 microns (+.000433"/+.000118").

Keep on diggin', this stuff is fun or I wouldn't still be doing it after retiring 7 years ago. lol

Dick Z

[RICHARD ZASTROW]

Curtis, FYI just about all the bearing mfg.'s catalogs are available by "googling" on the internet as pdf's. Copy em' for yourself. I have all mine copied in my "documents" area.

Dick Z

[Curtis_F]

Dick Z.

Thank you again for all the moral support and information. It's much appreciated.

Unfortunately, my goal is unattainable by me. It can be done, but going down that road isn't where I think I'll get the most out of the endeavor.

The Achilles Heal of the design is the 5C collet around which I want to design the spindle. There are times I can foresee myself needing to stick material through the spindle (at lower speeds), so I'd be using a draw tube to secure the collets, which bumps my spindle I.D. out to about 1.375 (34.9mm) minimum.

So to use the 5C requires bearings of a minimum 50mm I.D. At that diameter I'd have to go to ABEC-9 bearings and use a force feed oil lubrication system to achieve 30,000 rpm reliably.

If I were to forgo the 5C collet, or at least forgo the need of having a hollow spindle shaft I could use smaller ABEC-7 bearings and get away with it. A milling spindle would be relatively easy to hit 30k rpm.

But my #1 criteria is the 5C, not the RPM.

Thus, having decided on: No force feed oil lubrication. No ABEC-9 bearings. No spring Pre-Load on the bearings.

That brings me to about 15,000 RPM on the bearings MAX. If I use hybred (silicone nitride balls) 25deg AC bearings that speed is achievable with good grease lubrication (or maybe a mist oil system) and a positive air system in conjunction with a labyrinth seal to keep contaminates out.


Well, back to reading bearing catalogs.


Glad I'm not in a hurry!

Curtis F.

[RICHARD ZASTROW]

Curtis, You might check out sealed, grease lubricated, angular contact spindle bearings.

They're available from Barden, FAG, SKF and others.

In fact I have an SKF catalog 5332 E dedicated to sealed high precision AC BEARINGS.
They start at 30mm ID, 47mm OD, 9mm wide, max 30,000 rpm with grease, 45,000 with oil-air. S71906 CD 15 DEG.
and 26,000 rpm with grease and 40,000 rpm with oil-air S71906 ACD 25 DEG.

Also, there are smaller collet systems. Check out Hardinge and Royal on-line catalogs for something else that might solve your problems.

Dick Z

[RICHARD ZASTROW]

Curtis, I checked out the Hardinge catalog. 1c collets have 1/4" capacity, 2c 11/32" capacity and 3c 1/2" capacity.

1c drawbar thread .322"x40 rh, 2c drawbar thread .422"x30 rh and 3c .640"x26 rh .

30mm (1.181") should allow you to build a spindle that will accept one of these collet systems.

Dick Z

[RICHARD ZASTROW]

Curtis, Are you recovering from vacation?????????/

Dick Z

[Curtis_F]

Curtis, Are you recovering from vacation?????????

Between work & vacation I haven't had much time for the lathe spindle design.

I did find the time to download the Hardinge catalogs and looked into other collets than the 5C, but the versitility and shear variety just isn't there in the smaller collets.

If this was going to be a second cnc lathe I'd seriously consider going with the 3C collets. Since this is going to be my primary cnc lathe I need to stick with the 5C collet system.

Curtis F.