No announcement yet.

My Mill Slipped Tram Due To Knee Sag

  • Filter
  • Time
  • Show
Clear All
new posts

  • My Mill Slipped Tram Due To Knee Sag

    I haven't trammed my BP in a few years as there was no reason to do so until yesterday when I had to mill some .500 wide slots .375 deep in some 1 1/4" CF sq. bar.
    I didn't square the bar up as this job isn't that critical and this piece of bar was pretty square. Sometimes that's can be the cause, along with the work kicking up in the vise.

    A part had to fit all the way to the bottom of the slot. What I saw happening is the slot was cut slightly wider at the top on both sides.
    The bottom of the slot was .491 and the top was cutting at 496.
    To verify what I thought might be the problem, I put a gage pin in the collett and brought it up to the fixed jaw of the vise. What I saw was the pin was touching the upper part of the jaw and there was a very slight gap towards the bottom of the pin, couldn't quite get a .001 shim between the pin and the jaw but I could see light.
    I guess it's time to re-tram. I take it this is common wear. It was never that way the last time I trammed it.
    Looks like the knee fell forward a bit.

    Side note: A couple months ago I did tighten up the knee gib ever so slightly and it's caused some sticking at points out of the normal place. But if anything that should have the opposite of what I'm seeing know. Tightening the knee gib would make the bottom of the gage pin touch the vise jaw first.

    Last edited by JoeLee; 10-17-2018, 01:13 PM.

  • #2
    Originally posted by JoeLee View Post
    I take it this is common wear. It was never that way the last time I trammed it.
    Looks like the knee fell forward a bit.

    Side note: A couple months ago I did tighten up the knee gib ever so slightly and it's caused some sticking at points out of the normal place. But if anything that should have the opposite of what I'm seeing know. Tightening the knee gib would make the bottom of the gage pin touch the vise jaw first.

    It could be normal wear. Or it could very well be that the head of you mill moved. You've got a lot of bolted connections to hold the head in the right orientation and those can squeak around on you a bit, especially if you've been doing a lot of hogging, experienced chatter, crashed!, etc. One of the big reasons I migrated to fixed head milling machines was metal removal rate and accuracy. On a turret mill - even a Lagun that was quite a bit stouter than the average Bridgeport - it's hard to push cutting tools to their max without causing the alignment of the mill to drift a bit (and when I say "a bit", I mean <0.002" unless I crashed or did something very wrong... which has never* happened). On fussy things, I would rough it all out, then tram the mill, then finish mill it. That intermediate tram step often revealed an out-of-tram on the order of 0.001" or 0.002" over the 8" swath I was testing.

    Regardless, I was taught to tram a turret mill before each project / start of the shift / etc., including sweeping the vise for square if using the vise. It only takes a few minutes and gives you confidence that things are as expected. You may not always need to adjust the tram, but you at least sweep it. Now that I have my own BP, I probably won't tram it as often since I know other ham-fisted monkeys aren't beating on it, but I still check it regularly.

    *and by "never", I mean: "I'm not going to tell you about those times "

    Edit: Also, reading your numbers, you have a 0.0025" difference in 3/8" (since you're getting a trapezoid shape, I'm assuming you did this in two passes for a total of 0.005"). But you're measuring less than 0.001" in what, 1" (the height of the fixed jaw you were gauging against)? The trapezoid shape you're measuring in your slot is likely due to end-mill flex more than out-of-tram condition.
    Last edited by Fasttrack; 10-17-2018, 01:35 PM.


    • #3
      1) Cutting the slot along the x-axis I presume. Did you cut the slot to full depth in 1 pass or multiple? If multiple did you change z with quill or knee? If I'm thinking straight a sagging knee will still cut a parallel slot, just not square to the work, assuming the head is trammed in Y (nod). If the head is out of tram in Y it will also cut a parallel slot if the slot was cut to depth in one pass, but will cut tapered as you described if cut in steps. One side will be flat but tilted, the other will be "straight" but stepped, resulting in an overall taper larger at the top.

      2) Your test does not prove a sagging knee unless the head nod is perfect. This is not an assumption I'd make given your current position. Actually, with a machine with adjustable nod, I don't think I'd ever assume this. Your test also assumes something about the vice. A better test of knee sag would use a precision square on the table and an indicator mounted on the head/column somewhere. It'll give you much more resolution than using the height of a vise jaw too. Run the knee up and down and see what the indicator says. If you don't have a precision square, use whatever square you do have and do the test twice with the square facing 180 degrees and average the results. You'll learn something about your knee and square. This test assumes the table is of even thickness, which may or may not be a safe assumption. You can check that with an indicator on the table while moving it in y. Even if the knee is sagging, this will show if the table is wedge shaped.
      Last edited by JCByrd24; 10-17-2018, 01:57 PM.


      • #4
        To answer your questions, I did cut the slot using the X axis and I did rough out the slot before the finish cuts. Finish cuts along the sides were done in increments of about .005 per side with a new carbide end mill 3/8" dia. so there was no tool deflection.
        Then I flipped the part over and did the repeat on that side, always making sure that I kept the same side against the fixed jaw.
        The slots are not parallel, the calipers tell me that pretty clearly. The slot narrows towards the bottom on all sides.

        I just checked the surface of the vise as shown in the picture. Using the chunk or round stock that I just ground both ends.
        I can move that round anywhere on the vise surface and there is no dial deflection. Locking the knee makes about .0001 difference in dial reading.

        Sweeping the dial is another issue. I can't make a full 180 deg. rotation with the current set up but with the rotation that I could get between the jaws I saw about .0002 front to back. I'm guessing that if I could swing a full 180 I would see more.

        During the summer I had to take the vise off for another job and decided to check it on the surface plate. The surface of the vise was no more than .0002 out and teh fixed jaw was dead square to it. So the head must have moved some. I have run into some chatter on previous jobs, it happens now and then and perhaps that's what caused it.

        I really need to get a ring about 8" or larger ground dead flat on both sides for tramming the head off the table.

        A couple years ago I bought a 12" diameter brake rotor specifically for this purpose. I need to part the hub off of it and have it ground dead flat on a surface grinder but I can't find anyone with a grinder big enough to do it.

        I'll have to take the vise off tomorrow and check the tram.

        The two "H" blocks are what I machined yesterday. When I first noticed the issue I thought that the slot had opened up from removing the material, but I guess not.

        The thing that confuses me here is that if the head drops the slot should cut wider at the bottom and narrower at the top, I'm seeing the opposite here.
        The head needs to come down....... when heads go out of tram they don't rise do they??? I would guess it's always a drop. The force of gravity is always working to pull it down.

        Last edited by JoeLee; 10-17-2018, 03:27 PM.


        • #5
          Well what you are describing is very odd indeed. Final slotting passes of 0.375" axial DOC x 0.005" radial DOC with no change in quill or knee position should result in parallel sides to the accuracy of the end mill, regardless of knee sag or head nod. I'd measure your end mill for diameter (hourglass shaped) or being bent. Even bad spindle bearings would seem to show up in other ways before allowing that much angular deflection.

          Given the shape I wouldn’t rule out part deflection either. Try the same slot in only one side of a beefier part.
          Last edited by JCByrd24; 10-17-2018, 04:37 PM.


          • #6
            I checked both end mills, there is no taper in the sides of either one.

            I also indicated the top of the work and tightened up the vise just to see if there was any jaw lift..... nothing 0.

            I put another gage pin in the collett and brought it up to the vise jaw and it's dead parallel. The light you see in the picture is from the flash, the tram is fine.

            So, I'm at a loss as to why this happened. Tomorrow I'll try slotting another piece and see what happens.


            • #7
              If your knee is sagging try this. put a flat bar, parallel, in the vise (or on the table) so that its length is parallel to the Y travel. Tram the spindle Y axis, nod, using a square off of the column ways. Now zero your indicator to the top of the parallel in the vise and then move your table in the Y direction. If the knee is sagging you will measure a rise as the table moves towards the column.

              In edit, Another thought, with a good .25" drill bit drill a hole in a piece of stock with the quill at near full extension. Replace the drill bit with a .25" rod or dowel pin and confirm that it will slide into the hole at full extension. Now raise the knee and then lower the quill and reinsert the pin in the hole. If all is good the dowel will fit good at both ends. If not then your nod is off or the table is not rising vertically with respect to the spindle.

              no neat sig line
              Last edited by larry_g; 10-17-2018, 11:09 PM.


              • #8
                It's very difficult to try and check squareness of the knee against the column visually, and even more so trying to extend the vise surface out far enough with a parallel to set a square on it, too much room for error.

                The other method of drilling a hole and replacing the drill with a gage pin to see if it slips through the hole is no better since drills may cut slightly smaller or larger.

                There is about .001 climb measured in the table when it's moved all the way up against the column, but I knew that last time I trammed the head. It should still cut a slot with parallel sides.

                I did climb cut the sides which I failed to mention in my previous posts but with a new sharp end mill and .005 - .007 DOC on the sides it should not matter. If I was taking heavy cuts I could see where it may cause the end mill to deflect out at the bottom.



                • #9
                  It's got to be deflection of the work piece due to material removed right? Off all operations in a mill that can have error, this is just not one of them, even on the most worn out mill. The symptoms suggest only a couple of causes, assuming it's not just a measurement error:

                  1) Bent or hourglassed endmill, or tool deflection
                  2) Head nod changing between cuts
                  3) Table angle changing (rocking) between cuts (with only 1/8" movement in Y)

                  These are all incredibly unlikely. A slot cut that leaves more material both below it and to either end of the work I would bet would be parallel.