Achieving reproducibility

I think I have a number of small issues; all adding up to give me the errors reported above.

I agree. You also have a lot of cumulative errors and also at the limit of of your caliper/equipment resolution. You can often measure a part with a caliper 3X and each measurement will be different. Just switch from a drill to a boring head and changing the height of the head or knee cen change position, just lots of sources of errors.

You are really at the limit of what you can get out of the mill as setup, and are doing pretty decent at that. You can get a DRO for as little as $250 or less with 5 micron scales (which have their own deviations), this takes some guesswork out of the dials and table movement locking, which can easily account for the error you are seeing. It would also give you more precise center to center hole dimensions. As others mentioned, just locking down the gibs can introduce significant error, although there are a few tricks to minimize it. Mechanical edge finders and also the basic electronic touch ones, have not been very accurate in my experience, certainly not at the 0.001" level. I use a mechanical Haimer which is good for both edge and center to ~0.0004" and also have an electronic probe center finder which is about the same accuracy or a bit worse. You also have eccentricity of the collet, angular changes in the head, etc. Part of the process is reproducibility of procedure, but you still have movement of the machine, and also the process of drilling or boring holes. Aluminum can deform or twist on clamping. Drills tend to walk very easily, the longer drills can deflect quite a bit. When I need hole location precision I center drill the hole with a Keo countersink first and than follow with a stub drill. On precise boring dimensions I use a reamer. My goal is to hit 0.001" or better tolerance when it is required. Without a DRO, decent edge finder, etc. I was lucky if I was at 0.003" on a good day.

If you are making multiple parts, then I agree with others in setting up a jig with alignment pins that reference off of known holes.
 
I’m going to simplify your process to keep from being wordy. If I understand correctly you have some type of fixture and prep the pieces and bore one end of the beams. Then you step off the center distance and verify, lock everything down and proceed to bore the opposite end. In essence all three beams have been removed and reinserted into the fixture. If this is the case a .003 variation is pretty good, in the shops I worked in would not use that procedure unless we had at least a +/- .01” tolerance. The center drill and drilling process should have nothing to do with the error as the boring should take care of any discrepancies.

You have my procedure correct.
 
If positional accuracy of holes is your ultimate goal there are a lot of jigbore machines that no one uses today floating around used, tooling for them may be a bit more then you are willing to spend however, an old Moore would be a good choice.
Good Luck
 
I thought I could amuse the group here with how the original topic turned out:

cell-stuff02.JPG

Here we see one of the beams after it has had its back shaved and one end lightened so that the CoG is coincident with the center of the axle. It is also sitting on its holding bracket. It is balanced to within about 0.1 grains and has very low friction, at the ends of the beam are ball transfers. 3 beams and 6 ball transfers support the back side of a 13" telescope mirror 1.1" thick and 10.3 pounds in weight (or just under 2 pounds per ball).

Originally the idea was to have the bearings pressed onto an axle and into their races. But after screwing around with that for a few hours, I could never get the bearings to rotate smoothly and at low friction. So, I used the micrometer and measured the ground shaft at 0.125,1". I sanded them down on the lathe with 1000 grit and WD-40 to 0.124,9 and they would fit into the 1/4-1/8-3/32 bearings. It was this set of shafts that would not roll smoothly. So I took the shafts down to 0.124,6 and the beam would spin 20+ times when kicked with the finger (snap). However, this (also) meant the beam could slide on the axle and was not laterally located.

cell-stuff03.JPG

a couple of brass spacers locate the axles well enough (about 0.003). Originally the 4-40 screws were to allow lateral placement of th beam, but now this will be done with the brass spacers.

So now we have 3 sets of beams pivoting in bearings, and we need a means to hold the beams in exactly the right spots and build the rest of the framing around them.

cell-stuff12.JPG

I carefully drilled 6 holes in a flat piece of round (scrap) plywood (baltic birch) to locate the ball transfers. I also machined some bushings that just fit the ball transfer bores at the ends of the beams, and slotted the back sides of the brackets to fit the y-frame holding these bracketed beams. Two of the brackets are simple slots at 30º, the third bracket has a linear slot and two 60º slots.

The original idea was to have the slots grab onto the y-frame. Due to drill drift in the beam bearings and in the brackets, I was forced to relieve the slots so the y-frame can giggle--thus inducing no stress on the axle--thus placing the ball transfers at exactly the right position. Later, after the upper frame has been attached to the lower frame and the things sanded and painted, these brackets will be epoxyed onto the y-frame while in the support plate.

cell-stuff01.JPG

You might notice that the beams and brackets have lettering to denote which bracket goes with which beam and in which orientation. The balance is close enough that the ball transfers are also indexed to a particular bore in order to remain balanced.

As it sits, the y-frame can jiggle in the brackets and pop in and out effortlessly--inducing no stress on the axles. I did assemble this flipped it over, put in the ball transfers and used another flat wooden plate to check friction--it is very low (lazy susan low).

I also checked deflection from the axle-shaft-4-40 screw and found around 0.0060" with 8 pounds of force on a single beam which is "acceptable". Should this prove problematic, I can support the axle with little brass triangles between the floor of the bracket and soldered to the spacing tube.
 
  • Like
Reactions: Z2V
Back
Top