Automatic Centering Under Mill Spindle

[alsinaj]

Another challenge: How to mill a slot on the centerline of a set of parts that have slightly different widths? The parts are rectangular: 1.25" high; about 4" long; and about .5" wide, plus or minus a random amount. The slot goes into the ~1/2 x 4" face, on the long (2") centerline.

I have about 300 of these things to do. Is there a quick way to get them centered under the spindle without measuring, touching off and offsetting each time, or should I just cough up $200 for a self-centering milling vise?

[Deus Machina]

How accurate do they need to be?
If you get a little leeway, I've used the trick of marking the top of my chuck jaws with layout die, scribing a line in the center, and then align the spindle to that. Repeat the dykem and scribe line on the part.
For the other axis, I used one of these.

[jlevie]

While a self centering vice would be faster, it isn't necessary.

Measure each of the parts and write the width on each part. The touch off the first part and move in Y to the centerline. On each succeeding part calculate the difference by taking half of the difference between the current and the preceding part and move in Y. A DRO, or a Dial Indicator will make this a fast operation.

[Jaakko Fagerlund]

Isn't there any other features on the parts? And please tell us what the tolerances are, "random amount" and "slightly different" are not units of measure.

[alsinaj]

Deus, The slots need to be within a few thou of center -- the closer the better, as any deviation has to be compensated for elsewhere later. The type of center finder in your photo unfortunately won't work because these parts are rectangular, not round, and the edges are rounded off randomly from wear, so the reference point has to be the sides, not the edges.

[Davidhcnc]

Break open a set of feeler gauges 002" to 030"

line up one part in the vice with a 015" feeler at the fixed jaw

lock the Y axis

Cut the slot

measure how much thicker or thinner the next part is and insert the appropriate feeler/shim.

cut the slot

(refinements and dimensions to suit the job and resources)

edit...always refer to the first part or you will get a cumulative error

[RussZHC]

"cough, cough"...other ways will drive you nuts but then I am a bit paranoid [as david points out, a wrong measurement or a cumulative error from a previous part(s) is a real possibility here] and like you said making corrections later elsewhere maybe a worse option (plus, IMO it adds another op, potentially)

Edit: IMO, even if you group them it still means a series of "extra" measurements...supposed that 300 you had 20 batches of 15 parts that each were within whatever tolerance when you got to machining them you could do batches with minimal resetting but you first have to measure them all accurately enough to group them...

Edit II: you could also mount a self centering 4 jaw chuck on the mill table...if you have one is would save some cash

[bobw53]

I see 2 simple solutions...

Bring them all to the same width, which they probably should have been from the beginning...

Or group them... If you have a power feed, you can measure up 20 or so and get started, then measure the rest while the machine
is doing the work for you....

Just put them in labeled piles... Where I used to work, I had all the benches and cart tops painted with a thick white oil based paint, I could just write on the
top with a Sharpie and then wipe with acetone when done.. re paint ever couple years.

The other thing... Are they the same thickness throughout the whole part, or are they tapered slightly, you may have no choice but to bring them to a similar thickness.

Measuring 300 parts in one dimension is no big deal, even if measuring in 3 places, probably take less than an hour to sort them all, thats a lot cheaper than a self centering vise,
which is probably not as accurate as you want it to be anyways.

[loose nut]

Use 1/2" x 1/8" stock, cut 4" long.

Make up a jig to hold the parts, they will all be the same size, drop each part into the jig, clamp and mill the slot. They will all turn out the same way and you only have to do one setup.

[Jaakko Fagerlund]

I join in the group of measure the parts. Don't know your dimensions, so I just say digital calipers. Zero out the calipers on the first part. Next parts: measure the part and jot down the deviation measurement on it with a permanent marker. Then on the mill, put stops on the table so that it locates the part from the feature you have measured. Put down the first part, zero out your DRO on the place you want to be in and machine it. Put the next one in, move your Y or X axis half the deviation and machine it. Rinse and repeat for the rest of the parts.

And no, it doesn't matter what the calipers actual reading is, you are doing comparative measurements and the digital calipers are precise enough in this thing.

[Jaakko Fagerlund]

Use 1/2" x 1/8" stock, cut 4" long.

Make up a jig to hold the parts, they will all be the same size, drop each part into the jig, clamp and mill the slot. They will all turn out the same way and you only have to do one setup.

Can't do a jig when the width varies between parts.

[macona]

If all the parts are centered, yes, get the vise. You wont regret it not for messing around with 300 of the things.

Otherwise a cnc mill with a touch probe.

[jep24601]

The self-centering four jaw chuck certainly sounds like an easy approach but I was wondering if something could be accomplished with a more commonly available self-centering three jaw chuck. The part could be centered between two jaws with the third jaw aligned with the centerline of the part. If the part is wider than a jaw then it could be centered in the wedge action of two bars clamped to the centerline jaw and captured in the other jaws clamping the width of the part with an appropriate spacer for the angle of the bars. The bars and spacers would be the same for all widths of the parts as the part would slide in until clamped in the wedge of the angled bars.

[Rich Carlstedt]

Make two identical metal triangles One long leg, and one short leg that is shorter than the minimum width of the parts
So if the parts are 2" to 2 1/4" wide the triangle short leg should be about 1 3/4" .

Clamp the part, with some part of the jaws exposed , and overlay the triangles- long legs against the jaws, and they will form an intersection
that is centered on the part. Place a washer that has a hole the size of the endmill on the part . Enter the hole with the endmill and move the washer towards the triangle intersection and touch both parts.
If you are heavy on one side , the short legs (sides) will not be coincidental and seperate .
Move y axis and when they are equal you are centered.
Slip the washer out and the triangles and go for it.
It really is a fast and easy way without any measuring.

Rich

[darryl]

Mount a regular vise. Find the widest part and clamp it in the vise. Find center and lock that axis. Mill the slot. Measure and mark all other parts with the width. Mark the widest part also. Then calculate the difference between the widest part and the next one in line to be slotted. Cut that figure in half and build up shims to that dimension and place against the fixed jaw before clamping the next piece. A kit of shims can be made up from many things- paper being one. Onion skin is pretty thin, ordinary paper is thicker, etc. Mark the thickness on each shim, then you can just select them to make up the calculated figure for each part.

I suppose it would be just as easy, if not easier, to just move the axis by that figure, as long as you always approach the final position from the same direction.

Tedious, all that measuring and adjusting. A self-centering vise could be handy

I'm mainly concerned that if there's a self-centering mechanism, the grip on the part might not be all that strong. In contrast, a fixed vise jaw is going to be a steady reference position, particularly if you lock the table axis.

How much do the parts vary in width?