How would you handle this work positioning problem?

My situation: manual mill (Deckel FP1/FPS 300M) with DRO, have a long narrow work piece to which I need to drill a series of randomly spaced holes accurately to 0.001" position along x-axis (y-axis fixed). The issue is that the work is longer than the reach of the mill table. Therefore I need to drill the first set of holes, then move the work piece along x-axis, re-register it and continue drilling the remaining holes. My thought for handling the re-registration was to put a pin into one of the previously drilled holes and key off that. Any other ideas for most convenient way of handling this?

That sounds fine, or if the holes are large enough you could use an indicator to locate the last hole you drilled and go from there. Good luck!

That sounds like a perfectly good approach. I've used a variation with parts that needed to be repositioned. IIRC they were held in the vise, but a stud mounted to the table allowed moving the part, registering on the fixed pin and doing the next operation. If yours aren't through holes it would work a little differently.

drill the last hole,leave the drill in place,unclamp the piece, move over the table, clamp the piece,continue drilling?

.001 is tight, figure out the thermal expansion, and then ask how the customer is gonna inspect this.

randomly spaced to within .001? Your idea sounds fine.

Nerdlinger said:

That sounds fine, or if the holes are large enough you could use an indicator to locate the last hole you drilled and go from there. Good luck!

Click to expand...

Holes too small for this approach.

TGTool said:

That sounds like a perfectly good approach. I've used a variation with parts that needed to be repositioned. IIRC they were held in the vise, but a stud mounted to the table allowed moving the part, registering on the fixed pin and doing the next operation. If yours aren't through holes it would work a little differently.

Click to expand...

That's actually quite an interesting idea, would allow for re-registration against a known spot on the table....easier to do repeat parts.

janvanruth said:

drill the last hole,leave the drill in place,unclamp the piece, move over the table, clamp the piece,continue drilling?

Click to expand...

I fiddled with this approach, but its kinda awkward

Could you use a split collar clamp that fit the part well? I guess first step would be to offset the collar from a known reference using anything with a known length and flat ends. A vice jaw or the side of the vice, if ground, could work well. After the first op, you slide the collar up to the original reference surface and start the next series of drilled holes using the length reference to adjust your DRO. Of course all the surfaces of the collar, reference surface and the known length reference would have to be pretty flat/parallel to meet the tolerance spec. Good luck and let us know how it goes!

You aren't going to drill to .001 anyway. How are you going to finish the hole? If larger, pick it up with an indicator. If smaller, make your bored hole fit a gage pin, and pick up the pin.

drcoelho said:

I fiddled with this approach, but its kinda awkward

Click to expand...

Longer than table traverse ain't a lot, actually.

How long is the item? How far apart are the two most extreme holes?

What sort of shape, mass, hole-size, and material are you working with?

Do you need to do this more than ONCE?

Drill first set of holes and move the job over enough to locate from the last 3 holes, then take the best fit position from all 3 holes. Hopefully mitigating any error from a single hole position.

Or better yet, ream/bore the few holes you will know be used to locate for a extra bit of confidence.

Or make a fitted pin for those few holes and indicate round the outside of the pin if the holes are too small for easy internal indication.

drcoelho said:

My situation: manual mill (Deckel FP1/FPS 300M) with DRO, have a long narrow work piece to which I need to drill a series of randomly spaced holes accurately to 0.001" position along x-axis (y-axis fixed). The issue is that the work is longer than the reach of the mill table. Therefore I need to drill the first set of holes, then move the work piece along x-axis, re-register it and continue drilling the remaining holes. My thought for handling the re-registration was to put a pin into one of the previously drilled holes and key off that. Any other ideas for most convenient way of handling this?

Click to expand...

That ain't gunna happen - not as long as you have your fundamental orifices.

Limy Sami said:

That ain't gunna happen - not as long as you have your fundamental orifices.

Click to expand...

And another thing! Are you using a drill chuck? If so, you're already not accurate. Have you calibrated your DRO. DROs are inherently off by a few thou in full table travel. Oh bullshit! I've got an Anilam and it's perfect. I did too...and it wasn't. It's easy enough to calibrate it, just a pain in the ass.

Limy Sami said:

That ain't gunna happen - not as long as you have your fundamental orifices.

Click to expand...

Agreed.

Did a couple of 14 ft and a couple of 12 ft 2" x 4" x 1/8" square tube parts on the Bridgeport about 3 years back. Around 20 - 25 holes, 1/2" diameter in both.

Tried a number of ways to get a feel for accuracy even though that particular job wasn't precision. Bridgeport has a decent DRO which made life easier.

Registering by leaving the drill in the last hole and moving the table under it gave around 0.050" positional error if I just mashed the drill through in one. Good enough for that job. To my surprise, and delight, cumulative errors pretty much got lost in the wash.

Maybe 0.030" if I did the centre drill, pilot drill and open out thing. Close enough to notice cumulative issues.

Registering with pin in chuck rather than drill was maybe 5 thou + better.

Fixed pin on table got to maybe 10 to 20 thou error range depending on exactly how inspector meticulous I felt like being.

I imagine that using a fixed pin for rough registration and picking up the hole with an indicator could get under the 5 thou error range hole to hole. Cumulative might be all over the shop tho'. But its a drilled hole so accuracy is limited even if you go four facet and use guides. Cleaning up by boring dead to size with tight fitting register pin in bottom and straddle held pin in top with indicator reference might get close to 1 thou error band hole to hole. Cumulative will be worse.

With a really long job like mine support becomes important. The effect of it not sitting naturally flat relative to the table is noticeable.

Shifting the Bridgeport round a bit let me poke my job out through the secondary shop door and over the veggie plot. Fine tuning the machine position so the job didn't get caught up in the runner beans isn't a common machine rigging requirement.

Clive

Really you could move it any amount, travel to the hole and bump into it with an edge finder and you'd know your new position. I doubt this would be good to .001" but I would expect it to be within a few thou.

If you can't get the inside of the hole, then slip in a pin and find the outside edge.

I used to work in a FAA heavy maintenance facility. We made seat tracks for the aircraft from extrusions 20-40 feet long. The length was supported to eliminate bowing and the holes were put in with a center cutting end mill. The spacing was regular and at the end of the travel we released the vises (3) and with the end mill in the last hole moved in X to the other end of the table and tightened the vises and began again. But we didn't pretend to hold .001".

Yeah .001" ain't happening. Unrealistic expectations abound.

All good feedback, thanks to all. My workpiece is roughly 36" in length so not nearly as daunting as some of the examples mentioned. And ya, 0.001" is probably unrealistic, but I like to shoot for the moon sometimes. In this project, a little slop in hole diameter is fine, just don't want cumulative x-axis errors on hole positioning.