Deckel spindle bearings

[donie]

Iam thinking of getting a Deckel. I was looking at a view of the spindle bearing set up. This is a very simular set up to one that I work on often.
The rollers at the nose can be delt with using a lap piloted by the ball bearings at the rear. With care this can be done to extreem acuracy with little effort.
An adjustable internal lap would have to be made with a length to guide it from the rear bearings.
Looking in a earlier post, Ross said that the bearing are availible in oversize.
Are these bearings available from Deckel in oversize? If not what is the diameter and length of this roller.

Don

 

[TNB]

As far as I know, needle bearing cages are supplied with under-sized needles compared to the nominal size.

There are different classes depending on the under-sizement of the needles.

When the radial play of the setup increases, you're supposed to be able to fix it by ordering replacement cages of the next dimensional class.


The needle cages used in Deckel mills are standard ones, and you should be able to get them from any good bearing supplier assuming :
- He's a metric guy
- He accepts to fuss around and allows you to choose the precise dimensional class you want to order...

I must admit that the first bearing suppliers I asked the question didn't even understand what I was talkin'about...

Try this link : http://medias.ina.de/medias/en!hp.info/K*UK*4;agWLJNYdu_2a

 

[donie]

On this type of bearing, when going to oversize rollers. It would always be nessesary to lap the outer race, as the wear could and would cause failure or at least bad performance.
Also the bearing outer race would likely wear in an egg shape in the direction of normal milling.
One of the worst case problems on this is if the outer race needs to be replaced. It would be impossible to replace the outer race without distorting it. So the outer race would have to be undersize for the lapping tool to true it.
This type of bearing is found in alot of antique and modern machinery, this the first I have seen in a mill.
Another problem is the bearing cage, but in most all applications the cage can be reused. Its hard but not impossible to make a cage.
The last and maybe major problem would be the condition of the spindle itself.
Iam going to post a picture of the type of lapping tool nessesary to do this on another web site for a different application. If there is interest, maybe I can link to it.

Don

 

[AlfaGTA]

Donie:
The needles are avaliable from Deckel DMG. I am not sure what size incriment they come in. There are other sources possibly the link that Thien supplied.
As to size and length, i am not sure that all the needles are the same, depending on the specific model of Deckel and the year of manfacture. For example, the needle rollers on the horizontal spindle of my early FP2 were made with an inner race for the nose end of the spindle. The top needles rode on the spindle directly, same as my 83' FP4NC. The interesting thing about the FP2 horizontal setup was that the inner race had a tapered inner bore that fit on the tapered OD of the spindle. There was a ring nut to push the bearing down the taper. By tightening the nut you could effectively take up small changes in the bearing clearance.
If you need it i have the specs for the spindle clearance as supplied by the current Deckel service geru Volker Spitz.

Another source for the rollers might be FPS in Germany. I contacted them some time ago about getting some rollers, adn they replied that they could supply, but they were more interested in doing the spindle rebuild themselves.
There is also the matter of grease. The factory reccomends only Kluber Isoflex Super Tel
If you are going to rebuild your spindle contac me off line i might be able to help.
Cheers Ross

 

[AlfaGTA]

Donie:
" Also the bearing outer race would likely wear in an egg shape in the direction of normal milling."

You are correct about the uneven wear. Problem with the Deckel is that thre is no outer race! The Quill is the outer race. There are no removeable parts here. Also the later machines (unlike the FP2 i described in the post above) don't have any inner race that is replaceable either. The rollers run directly on the spindle, again ro replaceable parts.\
Cheers Ross

 

[donie]

Wow,this is interesting, but it looks like the spindle is rebuildable.
I dont have one of these yet.
Ross, I was looking at a Job that you posted, you were Wohlhauptering a casting in the horizontal postion.
It looks to me that the machine would be ideal for the stuff that comes in, that I dont have a fixture for.
Don

 

[AlfaGTA]

Donie:
The horizontal settup is excellent. Better than the vertical for boring. I have a boring head setup using the Sandvik "Capto" system (i think the best tooling system in the world!). I have bored automotive cylinders up to 101m/m in diameter. On the FP4NC i have done bores that are well over 10" deep. I can hold .0002" on diameter. The tool will repeat within .0003" when removed from the spindle and replaced. Pretty incredible stuff. A note here is worthwile. The Deckel spindle because it uses needle roller bearings will have some clearance between the rollers and the spindle. the result is sone slight play in the spindle...it cannot have any pre-load unlike an angular contac ball bearing. Holding boring tollerences that your Moore jig boring machine holds is a challenge for any Deckel until you get to the very late machines with ball bearing spindles. The Deckel spindle was designed with the heavy loads of milling as the prime design peramiter even in the CNC machines.
For precise boring the accessory "precision boring head" is better as its spindle has ball beearings.
Now all that said, for light finishing cuts with a boring head in the horizontal plane the weight of the spindle and boring head tend to cancel the bearing clearance, and as i said before i have had good results with mine. This is a matter of degees. As i am sure you have noted from some of my Monarch posts the accuracy i am demanding here might not be to your standards. For my work, holding .0005" for most stuff is just fine, and the Deckel does that pretty well. (FP4NC)
The best part about the FP4NC setup is that for repair work on engines the universal table can be used to get teh block or case in tram with the machine...I just move the table to get everything in proper alignment.
If i were looking for a machine today i would consider another one just like the FP4NC i already own. It is the best mix of CNC capibility along with ease of operator use. (Don feel free to jump on me here!) For eveyday use it runs so much like a manual machine that i somtimes forget about the control or programs. The best combination (IMHO) is The FP4NC with the Dialog4 control. Fitted with the universal tool makers table it is a very flexible tool. the down side is that the contrrol is dated and getting old. Parts are only avaliable through Deckel (DMG). Also by modern standards the machines are slow. Max spindle speed is 3150 RPM, unless you get one of the machines that has the optional 2 speed range on the vertical spindle. That setup gives 6300 RPM.

As for the spindle rebuilding,i think the drill is to hone, lapp or precision grind the quill ID and re-fit oversize rollers if your spindle requires. There are many Deckel users here that have been running their machines with the original spindle and rolles with no problems. Now again it is a matter of requirements. Making a hole or holding milling to a thousandth or two is pretty easy even with a less than perfect spindle. Holding a tenth or two is a different matter.
As far as i know my FP4NC (built in 83') has never had either spindle rebuilt. I have overhauled both by cleaning, and re-grease and re-assemble. The bearing clearance is still in spec, so i judt keep running the machine. I do notice the whe boring a deep hole that is a long distande from the spindle nose using the vertical spindle that i can't get a perfect (that's a relative term) round hole. I suspect that the Quill is worn slightly and the wear translates into an oval hole. Perhaps you would share your ideas on a lapp, as someday i guess it would be worth my effort to correct the problem on the Verticak spindle.
Cheers Ross

Ps. i know we are some distance apart, but you are welcome to come by anytime you are down in the Bay Area. I would be glad to give you a demo on any of my Deckels.....

 

[donie]

Great info Ross, The Jig Borer is not at all a good mill.
I need a good mill. Thanks for the invite will come by somday.

Don

 

[rklopp]

Regarding needle spindle bearings having clearance: Note that, when the spindle is turning, the oil film is hydrodynamically wedged into the roller gap, taking up the clearance. Like a plain bearing, the oil film can have incredible radial stiffness.

It would not work well to build a needle bearing spindle with preload, as the overall assembly is very very stiff in the radial direction. Even a very tiny amount of interference fit results in huge forces, which in turn would lead to early fatigue failue. The reason one can preload an angular contact ball bearing is that it is comparably springly.

 

[donie]

Absolutly right, no preload allowed on straight roller bearings. Also interesting is this spindle is grease packed.
Ross had made the point about nessesary clearance in this bearing not causing much problem because of the weight of spindle and boring head will tend to cancel it out.

Don

 

[Bjorn]

How much is the nominal radial play for a Deckel spindle? How much play is allowed?

 

[AlfaGTA]

Bjorn:
The following information was supplied to me through a factaory service tech that installed my FP4NC in 1993. The machine came used from Germany through the US distributer. When the machine was powered up it had a noisy spindle, so we overhauled it and cleaned and re-packed the bearings. Now remember this is for a spindle in a CNC machine that has a top spindle speed of 3150 RPM, so the specs might be somewhat diffeent for the ealier manual machines......But on my FP2 the spindle fell into this range.

The prociedure is as follows. Get everything spotless...Wash bearings and housing ane dry with lint free cloth...no dust dirt or even fingerprints!
Assemble the bearings (use gloves and keep everything clean) to include the thrust bearings.
Place an indicator (accurate) with the plunger touching the spindle nose close to the Quill, and at right angles to the spindle axis.
By hand push the spindle from side to side.
The movement should be a total of 10 microns. (10 micro millimeters)
Allowable tollerence= 10-15 microns. (.010-.015M/M) (.00039-.00059") DRY

The thrust adjustment is made on final assembley,and is done with grease .
End play measured at the nose of the spindle should be : .005 M/M (.0002") with heavy hand pressure.
Part of the secret is the grease, only Kluber Isoflex Super Tel.
Also it is imperative that you not use very much grease when repacking the bearings. A slight smeaar on each roller and a dab between on the cage is plenty. Over greasing will cause the bearings to run hot and overheat.
Cheers Ross

 

[J. Elliott]

Keep it up, Ross and Donie, this is a VERY interesting thread. I've learned a good bit already.

I've had an occasional email with Donie concerning the bum spindle on my 40 taper milling head. I've got 8 thou of runout at the spindle nose, which causes the whole machine to shake at certain resonating speeds. I'm using it as-is for the time being, if for no other reason than to see what can be done with a bad spindle. I'm still a part-timer, so have the luxury of not losing my mind over it.

Without the need to say so, the subject of bearings and their operation has become a topic of interest to me....

Off the wall here and totally un-related to the thread, does anyone know of a good basic primer on typical spindle/bearing setups you'll come across in normal machine tools?

 

[AlfaGTA]

J.
Wow .008" runout on a spindle, that is pretty bad i would think! I would venture to guess that the spindle is bent! What kind of machine is this on?
Would guess that this can't be doing the bearings any good either. I think i would investigate further to find the cause and possibly correct the problem. I know that the sliding head Cincinnati vertical mills used tapered bearings in the spindle and they are adjustable while the machine is assembled. On thoes machines if the adjustment is out they will have a very loose spindle.
Cheers Ross

 

[AlfaGTA]

J.
Forgot to mention, there is a great site on precision bearings at the Barden Bearing site. Lots of information on setup, and care of precision bearings as used in machine spindles.
Cheers Ross

http://www.bardenbearings.com/

 

[donie]

The high class spindle bearings are the most accurate and sensative items that will be ecountered.
Barden angular contact bearings are all Grade 9.
These bearings have an error of about 12millionths. A finger print is about 50millionths thick, a speck of dust about 100 millionths thick.
So its important not to touch these bearings. It is also impossible to have a completly dust free room.
When these bearings are removed from the package, the exposure to any clean room has to be minimized.
You really dont want to gage them. Instead the actual size OD and ID is provided for you.
You need to have calibrated gage blocks and master gage rings to set your gages for your shafts and bores. The press fit ranges from line to line to a tenth or two in most cases. A press fit of .001 would distort a bearing with 12 millionths runout.
The acceptable fits are provide by the manufacturer.
The error or high spots for the inner and outer rings are marked on the rings. This minor error can be used to advantage to offset errors on the shaft and housing bores.
So its best to clock the shafts and bores and log the measurments so as to determin the placment of the bearings high spots in relation to the shaft and housing errors.
This can only be another compramize as nothing is perfect except maybe the spindle shaft out of a Moore Jig Borer.
The last spindle I did was for my ebay Hardinge Chucker. By clocking the spindles bearing surfaces and collet bore, I was able to place the bearings high spots to get the best compramize and end up with a spindle runout exceeding curent factory specs.
The actual installation only took minutes, but preparation took many hours.
The shafts and housing surfaces must be burr free under 30X magnification. Removing the old bearings will cause scratches and ridges that will have to be smoothed off.
Its simple to do these and hard also because of the preparation.
It will be noticed that very few bearing house personel know anything about these and really dont want to sell them to the general public because they dont want to hear any crying if it dont work out. If it dont work, its your fault not the bearing maker.
If the bearings are grease packed that is another consideration. There are several types of grease for differant applications. Exxon Andock ultrafilterd grease cost about $90 for 90cc.

Don

 

[AlfaGTA]

Donie:
All good information! As a side note when replacing spindle bearings one should also check the spacers between the bearings. Many angular contac bearings have both inner and outer spacers that locate the bearings and in some cases set the bearing preload. It is important that you know when buying spindle bearings if the bearings were ground with the preload in the bearing (inner race slightly narrower than the outer race), or if the mearing was ground with the spacers intended to provide the preload. It is also important that the spacers be very parallel. Any slight out of flat here will cause distortion in the bearing and cuse running problems and inaccuracy.
Fornaturly the Deckel arrangement has no spacers or preload due to the needle roller setup as described above.
Not as accurate or easily replaced should they be dammaged , but by most accounte they continue to work very well without problems or falure.
Just clean and grease them from time to time and for most applications they will be fine!
Cheers Ross

[ 11-23-2004, 04:41 PM: Message edited by: AlfaGTA ]

 

[donie]

If the bearing spacers are a differant length, thats when things can get scary. Thats means a custom preload.
The machine tool manufacturer may decide that the aplication needs a modified preload. A custom ordered preload from the bearing company would be rediculously expensive, so the preload offset is done with the bearing spacers.
Here is where it gets dangerous. if the preload is high, say twice the normal preload it may be too risky to swap out the bearings without actualy testing the bearings deflection under the actual load.
I had run into a situation like this. The bearing manufacturer said dont use those spacers with that much offset!
The machine manufacturer said you must send the spindle to us for a base price of $4000 + whatever.
I did somthing sneeky. I called back the machine company, told the girl that I had a bunch of spindles being done, she connected me to the spindle builder, who must be the lonlyest person in the company and he told me how they do it.
And I will share with you all.
Fabricate a hydraulic cylinder with 1" piston area with a flexible tube to an accurate pressure gage, atatch this to the ram of an arbor press. This is a hydraulic weight scale.
Fabricate a holder to hold the bearing outer ring. Fabricate a shouldered plug for the bearing inner ring. Rig up a tenth indicator to measure the inner ring deflection in relation to the outer ring as you add the load with the arbor press. Do this to both bearings, add it up and you have a reasonably accurate spacer offset!
One thing, this means more bearing handling and more chance of contamination. Has to be done fast and every has to be clean.
Here is another tip on these kind of bearings.
Write up a list of assembly steps, and follow it.

Don

 

[Don UK]

What is the procedure for adjusting Deckel precision boring head bearings? There seem to be 2 types:-

Early Fvb2 with 4 cup springs & later 2021 with 1 spring spacer.

I have a late 2021 which appeared unused & still covered in wax.

I stripped the spindle & found some of the balls pitted by corrosion. The tracks were, fortunately, in better condition.

I obtained suitable replacements from a specialist supplier,cleaned,lubed,adjusted & backed off 1/4 turn as I seem to remember reading somewhere.
After running in, the spindle float under sideways hand pressure seems a bit more than I would expect (almost .001").

What should this be for a boring head?

Don

 

[AlfaGTA]

Don:
Sorry i don't have any specific information on the precision boring spindle bearings. I would guess that .001" of side movement at the spindle nose is not correct. With angulat contac ball bearings that the spindle has, some preload would be required to keep everything stable. This head is designed to locate the poring head precisely, and with freeplay at the node of the spindle i don't think you would get teh desired accuracy.
Checked the boting head i have (never been apart in my use) Serial number 2021-1385 that came from an LK . The play on the nose of the spindle is .0000-.0001" depending on which direction you make the measurement. Done with a tenth reading indicator and hand pressure with no tool in the spindle....
Cheeers Ross