What's new
What's new

Vertical Mill Column Dovetail Clearance

LAmachinist

Plastic
Joined
Jun 19, 2020
Location
NM, USA
I would appreciate any guidance from the more experienced rebuilders here, having searched this question and not found an answer.

Background:
I am rebuilding and scraping a small vertical knee-type milling machine. The column, saddle, and table all use 60 degree dovetail ways with straight gibs. From initial examination, I had thought the condition was OK - just the usual wear and table droop expected for a mill that has seen some use since it was made in 1964. Not claiming this is was the tightest mill out there - but compared to a small shaper I recently re-built, the condition of the ways was in much better shape (wear, but no scoring). I figured I could get away with scraping only, without the need to re-machine much if anything.

Problem:
However, there is one big problem area - the knee-column fit. The column has a large male dovetail - the wide flat surface of the dovetail should not be bearing, the two smaller flats on either side and the angled surfaces of the column dovetail should be the bearing surfaces. Likewise on the back of the knee, the two small flats and the two angled surfaces are bearing surfaces (with the wide flat between the angled surfaces being fairly roughly machined, needing only to provide clearance so the other flats can mate.

The flats of the knee dovetail were unusual looking; they originally looked like someone had been at them with a file (draw filing them). Looks like there was also a casting defect someone filled with babbitt or solder. More concerning was the large (non-bearing) flat in between the angled parts of the dovetail: it had been rubbing against the column dovetail surface, as evidenced by the shiny wear pattern on the originally milled surface. Likewise, there was a bit of wear on the column surface that was clear during rough scraping. Obviously that was not correct, indicating that even before scraping there had been interference in the dovetail geometry.

back of knee.jpgback of knee 2.jpgknee 3.jpg

After scraping the two flats on the back of the knee square, and roughing in the front of the column dovetail (to serve as a reference surface), the two flats of the knee still do not make contact with the two flats of the column (which still need to be scraped but seem within 0.001 inches so it shouldn't take much material removal on these surfaces).

column.jpginterference.jpg

The interference is, at worst, about 0.005 inches (estimated by feeler gauge with the knee resting on the column). So either the top of the column dovetail has to come straight down at least 0.005 inches, or the recessed flat on the back of the knee need to be milled deeper at least 0.005 inches. That would be a bit to scrape but doable with my Biax. Alternatively, I could always take the knee to a larger mill to mill down the flat surface between the ways.

Question:
What is the appropriate clearance for the non-bearing dovetial surface between the knee and the column? I started this project with those two surfaces interfering, scraping the actual bearing surfaces has of course made this inference worse. What clearance should I shoot for as I correct this? I imagine the gap should be big enough that future wear of the dovetail wont lead to these two surfaces contacting each other, but not so big that chips become lodged in it (there is no wiper or chip guard for that surface). I would guess the clearance should be a couple thousands- 0.004? That would mean I need to remove 0.009 since I already have 0.005 inches of interference.

Any guidance would be appreciated!
 
Gee I thought someone on here would have a comment, even if only to pick on my scraping technique! Surely I am not the only one who has run into this issue and needed to re-machine clearance for a dovetail way and wondered what the clearance 'should' be.
 
Most machines I'm familiar with have quite a lot of clearance there. Like 1/16" or more I'm pretty sure. I'm guessing this is to prevent chips getting hung up in there and galling up those surfaces - with the extra clearance the chips can just fall through. So yeah, you probably are about the only one who has run into this, since most milling machines have way more clearance there than they will ever have taken up in normal wear.
 
I don't have any experience with Clausing machines and don't know what their original design intent might have been. In comparison to Cincinnati and Milwaukee machines I have, the Cincinnati has column ways like the saddle ways on your knee. In other words, there's a strip that's a raised part of the casting with a lower section between probably .04 - .06, but I didn't measure it precisely. On the Milwaukee, the entire face of the column is a way surface and the knee contacts only the back of the dovetail.

On your machine I understand you face something of a dilemma. If we assume that there should have been clearance on that surface originally, the column could go onto a large machine and have .030 or so machined off the whole face. That's probably unrealistic. Alternatively, the knee, as the smaller casting, could have clearance machined between the dovetails. Again, in your case probably not realistic.

So you know that there needs to be some finite clearance on that whole face so that the original way surfaces can meet and function as intended. How much is enough? You hazarded that .004 could do it. That might be just enough. And if you get the other surfaces right it might last your lifetime, assuming you're not using it as a production machine. You've scraped that surface and the contact looks good where it contacts, but that's probably irrelevant for your purposes. If you know it needs to be cleared it's a waste of time to try to get good contact. Just lean on the scraper, shovel off the material and develop some clearance. Either estimate about how much you remove on one pass, divide that into .004 and go over it that many times. Or, use a feeler gauge to see when you're getting about where you want at the edges and lay a straightedge various ways to see that you don't have a hump in the middle. A depression in the middle wouldn't matter, you just want some assurance that you don't have "less than" your target clearance.
 
QT: [Gee I thought someone on here would have a comment, even if only to pick on my scraping technique!
I thought you were doing a nice job of it nd enjoyed the photos...
Buck

Guess I would say if you can reach it a shim might sweep through to clean it..at .004 you can't do much of anything..but I'm not the scraper guy..Richar King would know what is best..
 
I'd suggest that a small clearance is all you need.

Chips should not be an issue. Usually there would be a rubber or other material shield over that part of the machine anyway, from column to back of the "saddle", protecting the knee ways.

Maybe there is too much travel for that, and if so, then scrape that surface clear but still pretty flat, and put a wiper across as part of the column ways wiper.
 
I should probably have said for completeness, that even it it was found that Clausing initially had .03 clearance there, expecting him to scrape .03 off that surface is probably unrealistic too.
 
Thanks all! These responses have been helpful. Given that the column face is pretty flat as is (and the column is too big for any mill I have access to), it sounds like it might be easier to mill off 10 or 20 thousandths off the interior of the knee dovetail to re-establish clearance. This would have the added benefit that I could easily mill in a groove at the acute angle of the dovetail to provide clearance for my prismatic straight edge for scraping in the non-gib side of the dovetail. My friend has a very nice Schaublin 13 :drool5: that would probably be convenient to use for this in the horizontal configuration. Alternatively, I could put the knee up on an angle plate and use a bridgeport. With a knee this small, I should have options.

I just wanted a reality check - while I don’t know the history of the mill, it was hard to believe that all of the factory original clearance had been lost to wear. Guess I can blame the previous owner(s).

And yes, I should have detected this before I started this project by doing a proper machine survey. But with a machine where there are so many places I can make adjustments to the geometry, I dove in without too much care. Assuming the knee doesn’t move much after machining, my haste will not end up costing me much time.
 
Just wandered through the shop and had another look at the mill column - I can still clearly see the original machining marks on the column dovetail (original surfaces were milled, not scraped). This machine has no z-power feed, and given the good condition of the column ways, I don’t think there’s any way 0.030 wore away from the back of the knee. That, plus the file-like marks I found on the back of the knee, suggests that someone monkeyed with the ways on the knee. Either that or the machine was built on a Monday.
 
To end this thread, I ended up fly cutting 15 thousandths off the inside of the dovetail, using a radiused tool to cut the relief needed for scraping the angled dovetail surface. Nice to have friends with very accurate milling machines!

This leaves 0.010 clearance, between the knee and the column – I don’t think the knee will ever wear that much, especially on a manual hobby machine. FYI, the dovetails on this Clausing 8520 mill measure 50 degrees, not 60 as I stated earlier.

IMG_7312_b.jpg
 
I hate to bring this up after you've likely reassembled the mill but its generally considered best practice to relieve the middle 30% of any short slide, this being done for a few reasons, it allows that axis to positively lock without rocking even after the mill begins to wear, basically a perfect flat plane will wear at the ends and develop into a convex surface. to some degree this middle relief also allows a thicker film of oil to remain even with the slide locked which is also important as this area is the outlet for your Z axis oilers, probably worth knocking some oil grooves in if you haven't already assembled.

at a guess this mill was fairly worn and was very roughly re-scraped, the previous owner likely hollowed out the middle 30% by draw filing and scraped the 4 pads for bearing with a master rather than match it to the knee and very possibly gave the knee a quick touch up in the areas known to be most worn. I have to admit I've done several similar quick and dirty rescrapes on friends machines when I'm abroad don't have access to my masters/measuring gear simply so their machines don't get any worse and to give them they're full travel back.

The babbit filler seems to be fairly common on the 80's era Asian machines, I actually assumed it was nickel filler the first time I saw it on a Y axis long slide because of the surface of the surrounding area. I can only guess that most western manufacturers would reject any visible casting defects but the lower end of the market will let them slide,
 








 
Back
Top