Question for the No-DRO Crowd

[rmac]

dro_question.png (6.06 KiB) Viewed 5229 times

Suppose you have a rectangular part with a rectangular hole in the middle, as shown by the solid lines. Then suppose you need to enlarge the hole to the shape shown by the dashed lines. Assume you are using an end mill that matches the required radius in the corners of the enlarged hole.

One way would be plunge down at A, then move to each of the corners B, C, D, and then back to A. With a DRO this would be easy, but without a DRO, you wouldn't know where to stop the C -> D cut or the D -> A cut due to backlash.

Another way would be to plunge at A as before and make just the cuts from A -> B and B -> C. Then go back to A, plunge again, and make the cuts from A -> D and D -> C. That approach solves the backlash problem, but requires climb cutting on the cuts from A -> D and D -> C, which may not be a great idea in some materials on some machines.

This has to be something that was done millions of times before DROs came along. But how? Maybe with indicators mounted on the machine to eliminate the uncertainty due to backlash? Or maybe by setting stops to limit the table travel? Any hints? How would you "dials only" folks approach this? For the record, I'm also a "dials only" guy, but more due to circumstance than by choice.

-- Russell Mac

[Bill Shields]

You gotta know the backlash in the machine.

stuff the mill in on position in each corner so that you have a reference. You may want to pick nibble away at it..just go slow.

Mill each side, working to the corner tangent with a smaller than corner radius end mill.

Final light cuts climbing each side.

[Harold_V]

You gotta know the backlash in the machine.

Depends on the tolerance one expects to hold. On worn machines with long cuts that can get you in trouble.

It's easy to do even tight tolerance work when machining windows (or pocket----like the one described). The first two sides, as has been noted, are easy, but the opposite sides, where the screw is reversed, not so much. What I do is cut short of my target, then measure. Once you know what your reversed dial is reading as it relates to the cut, it's easy to hit size. So I don't forget dial settings, on a piece of paper, or even the mill table using a wax pencil, I'll make a
+
then at the end of each arm I write the target dial setting. With this method, a worn screw makes no difference. I always leave a few thou for a finish pass (always climb milling, regardless of material type), so I end up with seamless corners. I find that about three thou is more than adequate, and I try to make my roughing cuts uniform so the end result is predictable. I also always lubricate the cuts with the appropriate fluid. It helps in discharging the chips and improves surface finish immeasurably.

Mill each side, working to the corner tangent with a smaller than corner radius end mill.

I suspect you mean a larger than needed corner radius. If you use one that's smaller and you overshoot the roughing you can end up with a reduced radius when the finished cuts are taken. The larger end mill yields a little more material in the corner, but it's not difficult to remove once the window has been opened.

If I'm roughing and there's a lot of material to remove, I use a larger end mill. Otherwise I tend to rough and finish with the same size, although I may switch to a sharp end mill for finish cuts, especially if surface finish is a concern. Sharp end mills leave less of an edge burr, too, so it's worth the time it takes to make the change.

H

[Bill Shields]

No ..I mean a smaller tool to avoid burying 90 degrees of the cutter and getting a dig

The tengent point of the arc to line intersection, along the length of the line, is the same no matter diameter of the cutter ..

Perhaps I am not explaining correctly -> my apologies

[liveaboard]

I recently made such a cut for the first time; luckily for me, the finished size was unimportant because I had yet to discover of all the issues listed above.
Thanks for the lessons.

[rmac]

You gotta know the backlash in the machine.

Stuff the mill in on position in each corner so that you have a reference. You may want to pick nibble away at it..just go slow. Mill each side, working to the corner tangent with a smaller than corner radius end mill.

This sounds fine for low tolerance work, and sounds like it works even if you don't know how much backlash you have. It occurs to me that you could make it easier to get the tangents just right by bluing the insides of the reference holes before cutting the straight sides.

What I do is cut short of my target, then measure. Once you know what your reversed dial is reading as it relates to the cut, it's easy to hit size.

There you go! Thanks. It may have taken me a while to figure this out.

-- Russell Mac

[Bill Shields]

blue or not...if you don't want your cutter to dig in...stay out of full radius corners...the cutter will pull....

but if it is scribe line tolerance...who cares?

[rmac]

stay out of full radius corners...the cutter will pull....

Been there, done that.

but if it is scribe line tolerance...who cares?

I hear ya. Unfortunately, I'm still in the mode where I need to be careful whether it matters or not, just for the practice. Fortunately, I'm not in any hurry, so it's all good.

-- Russell Mac

[Bill Shields]

better than being careless even when it does matter....

[Harold_V]

blue or not...if you don't want your cutter to dig in...stay out of full radius corners...the cutter will pull....

Not with proper technique. I cut pockets without any issues (assuming I don't miss my mark). Key to success is to not take large cuts when nearing finish size, and to use a sharp end mill so the cut is clean without deflection. The approximately three thou I suggested (previously) works perfectly well with a climb cut and will yield corners without a hint of undercut. I had no choice but to learn to make cuts like that. The aerospace industry is (or used to be) critical of such features.

Hint. When approaching the corner, the adjacent handwheel is backed off so backlash is taken up, which prevents undercut. By the time you're within a few thou of the corner the slide should begin to move, so you get a very acceptable corner radius, free of chatter and underctu, although it may have a thou or two error, but it's not visible. It works for me without issue.

H

[Lew Hartswick]

If the tolerance is critical, I'd edgefind and locate each corner and cut just one side. Solves all your original problems.
...lew...

[rmac]

If the tolerance is critical, I'd edgefind and locate each corner and cut just one side. Solves all your original problems.

Another way to skin the cat. Thank you.

-- Russell Mac