CNCToolCat's Cat-House - Page 9
Close
Login to Your Account
Page 9 of 11 FirstFirst ... 7891011 LastLast
Results 161 to 180 of 209
  1. #161
    Join Date
    Feb 2009
    Country
    UNITED STATES
    State/Province
    Texas
    Posts
    940
    Post Thanks / Like
    Likes (Given)
    150
    Likes (Received)
    159

    Default

    How’s the repair coming?

  2. #162
    Join Date
    Sep 2006
    Location
    Abingdon, VA
    Posts
    3,281
    Post Thanks / Like
    Likes (Given)
    4295
    Likes (Received)
    2956

    Default CNCToolCat's Cat-House

    I knew I had to machine the outer and inner front bearing spacers thinner. And that was going to be a challenge — as the rings need to be flat and equal-thickness within a couple of microns....let’s round up a little, and I work in inches, so we’ll call it a tenth.

    Even though the ring thickness’ should be match ground, I will attempt the work by turning, using one of my beloved QT15’s.

    I had successfully done this prior, for my first QT28 spindle rebuild back in 2011. And the QT15 has a fairly new MMK 10” chuck — tight and accurate as a drum.

    Why do the rings have to be machined thinner you ask?

    Well, it’s necessary to accommodate the industry-standard 7024 precision thrust bearings that I use, rather than the 120BT..., which are a slightly thinner 7024.

    I’m too cheap to source the special bearings, so as with most pricey things I buy, I go to EBay first.

    So here’s the print showing the big NN3024 roller in front, the spacer rings, and the 120BT
    thrust bearings.



    I used the other ring itself to pre-load the cast iron pie jaws, using parallels to load it away from the shoulder. A 4mm grooving tool made several spring passes over the z-shoulder that would locate the face of the ring.



    I had about 0.070” to take off, so I took .005” facing passes on the ring. I labeled the rings with 6 points, and on the bench would mic each point after a facing pass.

    I could tell all 3 jaws were about a tenth high around the bolt holes, so I used a flat diamond hone to work the jaws, taking another pass after honing, measuring each time.

    I knew had to get the flatness I needed before getting to the final pass, and was able to just do it!

    I ended up with the outer ring flat to less than a tenth, and overall thickness to my target dimension + .0001”.

    I can live with that!



    Now on to the inside ring.

    After using a face grooving tool to cut a slot in the face of the jaws, I clamped on a scrap part to machine the jaws for the ring.



    Shown mounted on the jaws, I was able to get this ring flat to about a tenth, while matching the outer ring’s thickness as close as I could measure with new Mitutoyo digital mics.



    On the bench, I used a light touch of Arkansas stone to tune the rings even closer, getting both well under a tenth in parallelism and thickness.

    So, the bearing spacer rings would be good to go...., as far as this “shade-tree machine tool mechanic” is concerned!

    On to re-assembly.
    Last edited by cnctoolcat; 02-26-2019 at 08:34 AM.

  3. #163
    Join Date
    Sep 2006
    Location
    Abingdon, VA
    Posts
    3,281
    Post Thanks / Like
    Likes (Given)
    4295
    Likes (Received)
    2956

    Default CNCToolCat's Cat-House

    First step to installing spindle bearings is to clean all the “sticky, gooey, anti-rust crap” from the bearings.

    I used mineral spirits and a clean toothbrush in a clean plastic pan, with lots of scrubbing:



    My eBay bearings had been in the box unopened for a couple decades, and the oily coating had congealed:



    Once the bearings were thoroughly dry, it was time to add the Kluber grease. Based on a recommendation from “machtool” here on PM, I used Staburags NBU8 EP instead of the popular Isoflex NBU15. The NBU8 is a more extreme-pressure grease, and although it has a lower max rpm rating, I wanted the heavier duty grease for this big lathe application.



    Notice the aluminum foil, which I wrapped around each bearing for pre-heating in the shop’s little toaster oven. (Remember, I’m a ‘git-er-done’ kinda guy...no hot plate available, haha).



    The pre-heated (maybe 200 degrees F) bearings dropped right on, and made assembly fairly simple:



    The large roller bearing install is a bitch, if done by the book. You have to grind the locating shoulder to precisely locate the tapered-bore bearing axially, to set the proper radial clearance.

    All that sounds like too much work for me! So, I take a gamble and simply install the bearings, and torque the dog shit out of the retainer nut — which sets the preload for the big roller bearing by pushing it further up the taper of the spindle, expanding the inner race to create the running fit with the outer race.

    By using both hands to rotate the outer race back-and-forth, you can tell how tight the big roller bearing is getting — as you’re tighten the big nut.

    But remember, the ground shoulder stops how far the roller bearing can be pushed up the taper! So, the “gamble” is counting on the new bearing to be made with about the same tolerances/clearances as the original. (And we’re talking millionths of an inch here...)

    The two thrust bearings are pre-loaded by the “capture” when clamped inside the spindle housing.

    Since I machined the bearing spacer rings to create the exact bearing stack dimension of the original, the bearing’s OD stack should lock up inside the spindle with the same clamping pressure as original. (In hillbilly theory at least...??)

    For three prior spindle rebuilds, this hillbilly method has worked for me. Although it is a risk, as you won’t really know if the spindle bearings are preloaded and locked in correctly until after installing and testing.

    So, with fingers crossed, it’s time to get the spindle back into the machine...
    Last edited by cnctoolcat; 03-02-2019 at 05:04 PM.

  4. Likes david n, wheelieking71, CBlair, SND, Mstcnc and 2 others liked this post
  5. #164
    Join Date
    Sep 2006
    Location
    Abingdon, VA
    Posts
    3,281
    Post Thanks / Like
    Likes (Given)
    4295
    Likes (Received)
    2956

    Default CNCToolCat's Cat-House

    Next up is the install of the spindle.

    An idea from “cwtoyota”’s thread about rebuilding his QT10N, I used M8 threaded rod to slowly wrench the spindle in.

    It worked great, spindle went in like butter.

    It beats the way I did my first QT28 spindle rebuild in 2011, where I ended up using a sledge hammer and 4x4 to pound the spindle in! To this day I’m still amazed I didn’t damage the bearings when installing that one.










    For the rear roller bearing, I installed the outer race before inserting the spindle, but didn’t install the bearing itself until the spindle was in final position.




    Next up: finish up the re-assembly, and perform run-in testing....
    Last edited by cnctoolcat; 03-06-2019 at 01:35 PM.

  6. Likes david n, Ox, wheelieking71, Mstcnc, SND and 5 others liked this post
  7. #165
    Join Date
    Sep 2006
    Location
    Abingdon, VA
    Posts
    3,281
    Post Thanks / Like
    Likes (Given)
    4295
    Likes (Received)
    2956

    Default CNCToolCat's Cat-House

    After getting the spindle back together, I first dialed in the runout on the actuator. You vary the torque on the actuator mounting bolts to adjust runout, it’s neat how it works.







    And after a little run-in time, I put the tenths indicator on the tapered nose of the A2-8.

    Now it could change after putting it to work — (it is hillbilly after allCNCToolCat's Cat-House), but runout is currently about 30 millionths of an inch.

    I will take it!

    The new bearings sound really nice and quiet, and after a little warming (expected) on initial run-in, the headstock stays nice and cool, after hours of continuous running.

    So far so good!

    And a couple pictures of the original bearing races, no wonder she was so loud:





    Next up is mounting the new S-26 Hardinge collet chuck, stay tuned.
    Last edited by cnctoolcat; 03-10-2019 at 09:07 AM.

  8. Likes Ox, david n, dkmc, Mstcnc, chip_maker and 4 others liked this post
  9. #166
    Join Date
    Jan 2002
    Location
    Upstate NY -In the Flats next to the corn fields
    Posts
    8,900
    Post Thanks / Like
    Likes (Given)
    1414
    Likes (Received)
    2334

    Default

    Nice job!
    What RPM's and times for the run-in procedure?

  10. #167
    Join Date
    Jul 2009
    Location
    Peoria, IL
    Posts
    11,615
    Post Thanks / Like
    Likes (Given)
    49
    Likes (Received)
    8847

    Default

    Thanks for giving us the money shot of the bearing races! Always nice to see that your hunch was right!

  11. Likes wheelieking71 liked this post
  12. #168
    Join Date
    Sep 2006
    Location
    Abingdon, VA
    Posts
    3,281
    Post Thanks / Like
    Likes (Given)
    4295
    Likes (Received)
    2956

    Default CNCToolCat's Cat-House

    For spindle run-in, I started at 150 rpm’s, then increased by 100 up to 1500, with each step running 15 to 30 minutes.

    I try to baby my CNC machines, so I de-rate the maximum rpm’s (via parameter) of the spindles...especially the ones I’ve rebuilt.

    For the QT28’s, which have factory max rpm’s of 3,000, I limit mine to 2,000 max. For the work we do, that’s more than enough for these A2-8 machines.

    The initial running was without the chuck and the actuator, as I wanted to be as easy as I could on the new bearings.

    I ran in the forward direction the first day, then reverse the next day. I like to run left-handed insert holders for roughing and finishing...makes it easier to inspect and index the inserts, and it hedges your bets for machine rigidity.

    The Hardinge factory collet-chuck draw-nut was the perfect length for the draw tube, but I did have to enlarge the female thread from M78x2.0, to M80x2.0.

    After boring out, you can still see a little evidence of the original M78 thread:



    I used some dial caliper attachment thingys to get a pitch diameter dimension from the original Kitagawa draw-tube nut, and simply replicated that dimension on the new nut.

    It’s always a risk when ID threading without a good gage, but for this job I happened to find laying around the shop another draw-tube nut exactly the same as the new Hardinge one!

    So I had a “setup” part so to speak, which eases your mind some when you go to unclamp the part...knowing if it’s not right you most likely won’t be able to re-clamp it properly to thread again.

    But, not only did it fit the draw tube real nice, I got lucky and was able to get the new thread directly over the slight remnants of the original M78 thread...for both nuts, now that’s what I call luck!



    And then it was time to mount up the new S-26 collet chuck, which was sourced via EBay for $800 (about 80% off new price) along with several sets of collet pads at $35 per set.

    Yes, I’m cheap.

    But the reality for me is that I can usually find what I need on eBay with just a few mouse clicks (inserts, tooling, spare parts, supplies, etc.), and save while doing it.

    There’s a “treasure hunt” aspect to Ebay as well, and that to me makes it fun buying the stuff I have to buy anyway.



    Next up is test cutting, which unfortunately reveals some surprises, although not related to the spindle rebuild.

    To be continued....CNCToolCat's Cat-House
    Last edited by cnctoolcat; 03-10-2019 at 04:10 PM.

  13. Likes Ox, CBlair, SND, wheelieking71, Panza liked this post
  14. #169
    Join Date
    Jan 2003
    Location
    Canada
    Posts
    11,074
    Post Thanks / Like
    Likes (Given)
    5045
    Likes (Received)
    3190

    Default

    Thanks for posting this rebuild process, pretty neat to see and looks worth attempting.
    Are the bearings any special accuracy spec?

  15. #170
    Join Date
    Sep 2006
    Location
    Abingdon, VA
    Posts
    3,281
    Post Thanks / Like
    Likes (Given)
    4295
    Likes (Received)
    2956

    Default CNCToolCat's Cat-House

    Quote Originally Posted by SND View Post
    Thanks for posting this rebuild process, pretty neat to see and looks worth attempting.
    Are the bearings any special accuracy spec?
    Most machine tool spindle bearings are relatively “standard” items, albeit very expensive due to the precision involved.

    Most are either P4 or P5 precision, with P4 being more precise.

    P4 is equivalent to class 7, I believe. There is actually a class 9 bearing, which is “ultra precise”. For grinders and such I reckon....?

    Doing a spindle by the book, the back-to-back thrust bearing set should be factory matched.

    Actually, matched-sets of thrust bearings are becoming less common, as the manufacturers have the variances so small, they don’t offer as many options for matched sets. You just buy two of the same P/N thrust bearings and get ‘er done...like I did!

    ToolCat Greg

  16. Likes SND, wheelieking71, Antarctica liked this post
  17. #171
    Join Date
    Sep 2006
    Location
    Abingdon, VA
    Posts
    3,281
    Post Thanks / Like
    Likes (Given)
    4295
    Likes (Received)
    2956

    Default CNCToolCat's Cat-House

    So, after giving the new spindle bearings a good run-in in both directions, it was time for some test cuts.

    Well, how did the test cuts go you ask?

    Here’s a clue:



    As it turns out, the spindle bearings were not the only thing totally “fu-barred” on this machine.

    A few tests with a crowbar showed the turret had
    about .005” vertical play!

    Good grief, no wonder the machine sounds like a crapped-out Chinese bench-top lathe trying to take .250” per side in inconel!

    As it turns out, brass chips had collected and packed under the X-axis way covers, and as the turret could come down in X, these chips managed to damage the end cover on the bottom-left linear rail truck.

    Well, when you consider those very plastic end covers are what return the balls inside the truck, those end covers are pretty dang important to the function of the linear guide system.

    No good end cover = no linear truck balls!

    No linear truck balls = machine tool no worky!





    In what HAS to be the most convoluted way-cover system ever devised by man, all it took was a little brass intrusion to cause the ToolCat a big headache!

    The X-axis covers consist of 2 stainless strips of sheet metal, that are fixed into position, and they pass above the linear guide trucks via spacers to the rear of the carriage:



    These flexible strips lay flat in front of the turret. But if they get damaged, chips can find their way under them.

    Now, Mazak did use fabric bellows under these sheet metal strips, as additional protection for the linear rails.

    But no matter, if enough chips get under these strips, they can pack in place, and damage the fragile plastic end covers on the linear trucks:



    Mazak did engineer slots at the bottom of the X-axis, so chips that get in there can get out by gravity.

    But....

    Brass chips tend to pack, especially when mixed with a gooey concoction of lube oil and coolant.

    So...

    With much chagrin, I started to dismantle the turret. It ALL has to come out to replace the linear rails and trucks.

    What the hell, I’ve already rebuilt the spindle, right?

    This is a 1995 model machine. And the irony is that Mazak uses this same X-axis way cover system on most of their “higher-end” machines, like the Integrex, to this day!

    And this very way-cover system will create all kinds of hijinks as I go through the X-axis rebuild process.

    Mazak really outdid theirselves on this design — which was intended to never be reverse-hillbilly engineered.

    But this hillbilly engineer-turned-machinist knows the way of the Samurai, and will not be bested by a convoluted design.



    Stay tuned...CNCToolCat's Cat-House

  18. #172
    Join Date
    Jan 2006
    Country
    UNITED STATES
    State/Province
    Connecticut
    Posts
    1,563
    Post Thanks / Like
    Likes (Given)
    647
    Likes (Received)
    751

    Default

    You'd be surprised where those little brass chips can get into, Well, maybe not you personally CNCToolCat's Cat-House.

    Sent from my SM-G930V using Tapatalk

  19. Likes wheelieking71 liked this post
  20. #173
    Join Date
    Jan 2013
    Location
    Gilbert, AZ
    Posts
    5,620
    Post Thanks / Like
    Likes (Given)
    6887
    Likes (Received)
    7040

    Default

    I find 7075 to be almost as bad as brass (when turning). But, yea, brass is bad! (plastic is the worst!)

  21. #174
    Join Date
    Aug 2002
    Location
    West Unity, Ohio
    Posts
    24,722
    Post Thanks / Like
    Likes (Given)
    4850
    Likes (Received)
    7549

    Default

    Am I seeing this right?
    Are the ball runs in the rails undercut?

    That should greatly help fretting/skidding, but I haven't seen that design before.

    Or maybe in this case - gives gunk a place to pack in?



    -----------------

    Think Snow Eh!
    Ox

  22. #175
    Join Date
    Jan 2003
    Location
    Canada
    Posts
    11,074
    Post Thanks / Like
    Likes (Given)
    5045
    Likes (Received)
    3190

    Default

    Can you replace that with new roller rails that don't need X covers instead of that old ball rail dirt catcher design?
    Maybe their height/size don't match though but would be a real good upgrade if they did.

  23. #176
    Join Date
    Sep 2006
    Location
    Abingdon, VA
    Posts
    3,281
    Post Thanks / Like
    Likes (Given)
    4295
    Likes (Received)
    2956

    Default

    Am I seeing this right?
    Are the ball runs in the rails undercut?

    That should greatly help fretting/skidding, but I haven't seen that design before.

    Or maybe in this case - gives gunk a place to pack in?

    The original rails are NSK "LY45" series, which have the double-row grooves ground on each side. The grooves are actually a gothic-arch shape (not round), so the balls only touch at one point on each side of the grooves. The relief you see along the bottom of the grooves helps facilitate grinding this "gothic-arch" profile....I would imagine?

    With properly functioning way-covers, this part of the linear rail should never see any contamination.

    Makes we wonder how many of the machine tool builders get by with exposed X-axis rails these days?

    Can you replace that with new roller rails that don't need X covers instead of that old ball rail dirt catcher design?
    Maybe their height/size don't match though but would be a real good upgrade if they did.
    Those were my exact thoughts at first! But, as it turns out, these NSK 45mm linear guides are the "lower" version, which are only 60mm tall from the bottom of the rail to the top of the truck.

    Standard 45mm guideways are 70mm tall....

    (Just one of the MANY Mazak design choices that were primarily done to force you to buy all the parts from them!)

    Upon researching linear-roller designs, it appears they are all the 70mm tall series.

    So, I was limited to using the 60mm tall series ball-style...no roller-style on this Yamazaki...

    But, I was able to source some THK rails/trucks that have significantly better load and stiffness ratings than the original NSK's...and that had me giddy with excitement!

    To be continued....
    Last edited by cnctoolcat; 03-21-2019 at 01:07 PM.

  24. Likes Ox, wheelieking71, SND, chip_maker, yardbird and 1 others liked this post
  25. #177
    Join Date
    Jan 2013
    Location
    Gilbert, AZ
    Posts
    5,620
    Post Thanks / Like
    Likes (Given)
    6887
    Likes (Received)
    7040

    Default

    I just love watching your projects Greg.

  26. Likes cnctoolcat, alan speyrer liked this post
  27. #178
    Join Date
    Jul 2009
    Location
    Peoria, IL
    Posts
    11,615
    Post Thanks / Like
    Likes (Given)
    49
    Likes (Received)
    8847

    Default

    I find that brass/bronze chips are the worst for slivers and cuts while working on a machine. Stainless is also very painful...

  28. #179
    Join Date
    Aug 2002
    Location
    West Unity, Ohio
    Posts
    24,722
    Post Thanks / Like
    Likes (Given)
    4850
    Likes (Received)
    7549

    Default

    Brass chips are easy to find a few days or so later.
    It starts festering up and you know right where to cut with the Exacto.




    ---------------------

    Think Snow Eh!
    Ox

  29. #180
    Join Date
    Jan 2013
    Location
    Gilbert, AZ
    Posts
    5,620
    Post Thanks / Like
    Likes (Given)
    6887
    Likes (Received)
    7040

    Default

    Quote Originally Posted by Ox View Post
    Brass chips are easy to find a few days or so later.
    It starts festering up and you know right where to cut with the Exacto.




    ---------------------

    Think Snow Eh!
    Ox
    Yea, way easier to deal with than stainless


Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •