btw, the design proposed is very close from X2 R8 design, except that the X2 use two regular deep groove bearing.
And I have already upgraded my CM3 X2 bearings with taper ones, nothing like P4/P6 precision ones, but definitly an improvement.
Assuming the bottom bearings are preloaded at the factory how do you think the shrink fit will affect the preload. What level of precision do you propose on the shrink fit.
Phil
the later one is a more servicable design but the same principle. the lock ring replaces the press fit. would work very well, as its the same as alot of high speed spindle setups.
again, his design allows easy light preloading of the tail bearing to eliminate any play, making it basically identical to skf's drawing example:
I would be interested to see the general arrangement of a commercial machine tool spindle that uses a double row angular contact bearing at the bottom end?
Phil
But the top end design is much more complex than the OP's and if you look closely you will see that the top bearing is preloaded.
How does his design allow easy light preloading the the top single DG bearing. You did say it was an OK design.
You did also say it doesn't need preloading. So which is it?
Phil
i said it doesnt need preloading because it doesnt. i said you could add it easily if you wanted.
whats not to understand? (besides deliberately not wanting to?)
and who gives a flying f*** what commercial spindles use? if he wanted something exactly like a commercial spindle, hed buy one!
Simplicity, gentlemen, simplicity for pity's sake.
We can all romp in the realms of exotic bearing layouts for whatever purpose armchair engineering leads you, but in the cold hard light of day, when the metal has to be cut, it's the simple approach that gives the best results.....KISS rules.
If'n the requirements were for the ultimate spindle bearing layout that can be possibly envisaged...anything's possible....the combinations of angular contact, tapered roller and needle bearings layout is so vast that the analysis would fill a book, but the making would need a lifetime's experience and the total input of a vast workshop infrastructure set-up.
At this moment in time I can only cite a personal design manufacture that I designed and MADE for a toolpost grinder in 2001.
I worked for a chap, who, since my retirement is now one of my best friends, and at the time he required a toolpost grinder to do what toolpost grinders do....produce round ground surfaces on a lathe in lieu of a regular cylindrical grinding machine....there was also the factor of over a grand in dollars for a commercial model and second hand one at that, which he didn't want to spend.
If'n you can fault the results, you can look to the design or manufacture of the device etc.
Long story short.....I opted to have two angular contact bearings up front with cups facing outwards and labyrinth seal, held to the spindle with a locking collar, and a single sealed semi floating radial ball race at the other end.
It is a common design to have a single angular contact bearing at the other end that is spring loaded for preload, but the spring loading mechanism and lubrication sealing made it more complicated, so the single sealed radial ball race was used......no problem to date.
Anyone who attempts to apply axiall preloading to a radial ballrace will radically reduce it's life expectancy.
The preload assessment was done on the "try it and see" principle......having made it, the grinder was run free for an hour and the bearing housing temp measured (with a finger on the bearing housing)....then a .02mm thick shim was placed between the bearing shells and the test run again.....a temp rise after 5 minutes soon indicated that the original bearing assemble with shells touching front to front or with cups facing outwards, whatever the way you see it, without any extra preloading, was satisfactory, and has run that way for over 10 years now.
At this moment in time I have redesigned the spindle and quill of my Beradi jig borer to have a similar layout with the twin angular contacts and a single radial sealed ballrace etc.......the original design here which I will retain also has the spindle in a quill, and the upper part of the splined spindle moves in an out of a fixed pulley running in two radial ball races.
The pulley with the two ballraces takes all the side loading from the drive belts and insulates any side loading from the spindle which is free to move up and down in the splines......this is common practice in any design where a belt drive head is used.
To ensure longevity and accuracy but more important reliability, I am going to bore the spindle end inside with a parallel cavity, and fit a hardened and ground insert with ISO 30 taper and Locktite it in place.
This way, in the area where most wear and abrasion takes place, making an insert from tool steel that is hardened and accurately ground commercially will ensure the fit and tool mounting capability of the spindle end, and is as good as it can get without making the whole spindle a commercial enterprise with price to boot.
In addition, in future if'n I wanted to change the design to have an R8 collet mounting I can remove the inset and replace it with an R8 one without scrapping the whole spindle....the same would go if'n I wanted to have a Morse taper etc etc.
In milling spindle design, I don't think you can beat the simplicity of having two angular contact bearings in close relationship to one another at the bottom end, in whatever relation suits you, as an initial spindle design requirement, and use whatever upper bearing design you fancy.....better and simpler than twin tapered roller bearings with preload layout requirements that would make your eyes water.
BTW, I'm working on a design that uses 3 sealed radial ballraces at the bottom end and two ball thrust bearings midway, (for axiall location), with a single sealed ball radial to take the side loading if any or to support the spindle at the top.....it may raise a few eyebrows, but has been used in other locations where both radial and axiall accuracy of running is important....I would post a sketch, but at this moment it's only a doodle in Photoshop on my graphics tablet.
It would be relatively easy to make, and have a fairly small housing diam.
Ian.
Has anybody considered we have a Norwegian troll?
Dick Z
DZASTR
I bet Chrisjh is wondering what happened to his thread.LOL
Dick Z
DZASTR
well, hopefully he just keeps working away on his design
i do have a question though. if this is a "DIY on the cheap" spindle would it not be best to use an available shaft, like from the x2, g0704, etc? to me the shaft is the most difficult part for a diy machinist to make properly, and they are cheap enough to buy in r8 tapers.
i have a g0704 spindle shaft i got from weiss for $40. just needs bearings, a housing, locknut, and a pulley and its ready to go. should not cost more than $200 total i think.
Hi, seeing as the spindle part carries all the tool holding capability, and requires to be hardened and ground while at the same time being truly straight, it would certainly be cost effective to buy in this item....not many of us have the capability to heat treat and grind a spindle at all without it ending up like a banana.
Making a housing should be as required for the machine in question, either in alluminium or steel.....the ball races are a buy in item anyway, so are the Nisil seals....why not the spindle too?
The spindle also brings into question the draw bar....it needs to be drilled right through to clear a drawbar of sufficient strength in the end thread without stripping the thread.
You would not get a draw bar more than 9mm diam through a spindle at the drive end of 19mm diam without weakening the spindle.....and drilling a 9mm hole through a spindle really takes some doing.....the steel is not just mild steel but nickel chrome type spindle steel to resist the torque loads.....what compounds the problem is the drive method....if'n the spindle end has a spline running along it this makes the bottom of the spline diam that much smaller in diam again to get the draw bar in.
With this drawbar problem facing myself in my jig borer remake, I'm contemplating not having a drawbar......perhaps an ER40 milling chuck with a 25mm parallel shank can be fitted permanently in the spindle end and loctited in, which will make the spindle so much simpler to make.
Ian.
Hi All,
I had no idea that I could create some much havoc and chaos!!!
I do not profess to be a professional when it comes to spindle design. I am a simple radio technician with an interest in mechanical designs and CNC.
Ihavenofish has a good idea in the possibility to "buy" an spindle shaft and design around it. I knew that the most difficult part of this project was getting the spindle shaft made satisfactorily in a home workshop. I had intended to machine the R8 bore and taper in my CNC lathe using mild steel and getting the internal taper plasma nitrided rather than hardened by other methods. Maybe a bit ambitious?
In fact, a "buy in" spindle shaft was the method I used when I designed and manufactured my CNC lathe. I simply bought a lathe spindle shaft as a spare part for my existing manual lathe and used it. Worked a treat and I now can interchange all my existing D1-4 chucks and tooling from my manual lathe on my CNC lathe. Details here Chrisjh Home Built CNC Lathe
For the CNC lathe I use a 110mm taper roller at business end, and a 100mm double row angular contact bearing at the rear with axial preload applied with a nut. I had problems with temperature increases causing the preload to "tighten" to the point where the spindle would sieze. I solved this issue by adding a 2" 90 durometer ORing between the preload nut and the inner race of the rear bearing. So now when the temperature increases, the ORing compresses and maintains a "sprung" preload. Bit of trial and error to get it right, but I am happy with it now.
This week I have been busy keeping the wolf from the door but will get back onto the design soon.
Regards
Chrisjh
there are special disc springs meant for preloading bearings. most higher end spindles will use them for the exact reason you mention. they should be more consistent and durable than an o ring.
when you use a floating tail bearing and the double row at the nose like your original(and current) design, this wont be an issue.
I agree with ihaven't got no fish...the original design shows great promise for simplicity, but with a few mods as has been suggested etc.
I'm not a great fan of the "must have preload at any price club", the fact that just taking up the slack will give you results that equal the best commercial bearing design layout from the home machinist's point of view.
Preload (unless it has been accurately established) means forcing the bearing to work at the maximum load envisaged by the manufacturer, to offset bearing migration under load, a condition that will never be met by the CNC worker who rarely uses the "plowing" off metal at the highest speed/feed rate and always pecks away with many cuts to allow for the inadequacies of the machine frame, but at the same time having the bearings straining to work with a non existant work loading for all of their working life.
Preload also means reducing the bearing life prematurely for no better reason.
Applying preload by spring pressure in one direction means that the bearing is free to "migrate significently" in the opposite direction should the cut go that way, and CNC is all about multi direction machining constantly, hence the interest in the backlash properties of the ballscrews etc.
Looking at the bearing maufacturers catalogue and the sealed double row A/C bearing package set-up with manufacturer preload set at the factory, I would think that for simplicity and the original concept idea for a design that would be sold for others to make, further complicating the design with added preload at the top end is unnessary.
I fully suscribe to the "fix it at the front end" and allow the back end to float axially brigade.....preloading the back end is a hit and miss manoeuvre at the best of times.
BTW, where does the "lack of adequate preload" in the top end make itself felt.....during a heavier roughing cut?.....not significent......during finishing?....not apparent to justify extra loading, and anyone who can deflect a radial bearing at the top end during machining is doing something out of the ordinary in the drive application.
Ian.