Drill Drift

>
>
> <putting on my teacher's hat and firing up the verbositizer...>
>
> Drills drift with depth when the axial thrust is large enough to:
>
> A. Cause the drill to flex
> B. Defy correction by the balance between the opposed radial forces
> from the cutting edges.
> C. Both of the above
>
> Try this for a test. Put a small jobbers length drill (1/8 diameter or
> less) in a drill press. One of your Costco specials will work just fine.
> Start it cutting on something sorta tough, probably not aluminum. After it
> gets started, gradually feed it harder and harder, and watch it carefully.
> As the feed pressure increases, the drill will flex, taking on the shape
> of a vertical jump rope. If you could freeze it in that state and examine
> it carefully, you'd see that the jump rope curve extends all the way to
> the point of the drill. This means that the point is no longer pointing
> along the same centerline as your drill press spindle. It's now pointed
> into the work at an angle, which is determined by the severity of the jump
> rope effect.
>
> For a while, a drill in this state will scrape and claw and drill an
> oversize, bell-mouthed hole. After a few revs, though, being that it's a
> drill, it'll figure out how to start drilling like it (sorta) should -
> exept that it's still pointed into the work at an angle. So, it'll drill
> at an angle. Often, it'll drill at an angle that continues to increase as
> it goes deeper into the work. That's because, once it finds a way to
> start drilling, the jump-rope effect gets compounded by the new angle of
> attack. Things just get worse and worse. I've seen small drills cut
> sideways by as much as a whole diameter, when drilling only a few
> diameters deep, even though they were held in decent spindles, and didn't
> have any serious runout problems till they were flexed by excessive infeed
> pressure.
>
> Normally, a good grind on the drill will keep the radial forces
> balanced, and will try to correct for most of the above. But the possible
> correction is limited, and is easily overwhelmed by infeed pressure and
> drill flexibility.
>
> There are two possible solutions. One, use a drill with a properly
> split point. This will have cutting edges that go almost all the way to
> center, and will have very little chissel point. It's the chisel that
> requires a lot of axial thrust, since it's just squishing and squeezing
> metal out of its way, and can't really cut like a real cutting edge.
> Cutting edges themselves need torque; but very little infeed force, unless
> they're really dull. Get rid of the chisel, and you get rid of the thrust
> problem. You can feed faster with a split point, and still not flex the
> drill. Which means you'll have an easier time drilling straight holes
> instead of crooked or curved ones.
>
> Second solution should be obvious. Use carbide. It's much more rigid,
> and won't flex like HSS. Simple.
>
> Apply both of the above solutions at once, and you should get straight
> holes easy.
>
> With regard to stacking plates, there can be some problems. When you
> start drilling a hole, you've got the point of the drill, center only,
> pushing on the work. It's the center that does the most pushing, even on
> split point drills. As the drill goes deeper, you're using both the point
> and the cutting edges, creating maximum push. Whatever force it takes to
> drill well will try to deflect the workpiece, unless the work itself is
> very rigid or very well supported.
>
> Suppose, just for the sake of argument, that the aluminum plates you're
> drilling are able to flex a little bit - maybe just .005" when the drill
> is pushing to the max. So you start drilling, with a nicely chosen
> feedrate of, say, .010 per rev. And the work deflects a bit, and then
> resists, and drill goes fine - right up until the point breaks through the
> other side of the plate. Then, all of a sudden, the biggest part of all
> the pushing force is gone, and all that's left in the plate are the
> drill's cutting edges, which don't push all that hard. In fact, they can
> actually pull at the work instead of pushing it, just because of the way
> they're shaped. So, at the instant the point breaks through, your plate
> goes instantly from being deflected .005 away from the drill to being back
> at flat and normal, or maybe even being pulled toward the drill a little
> bit. What started out as a nice neat .010 per rev feedrate suddenly
> becomes .015 per rev, or more. Bad news. The drill bites too hard,
> probably pulls the plate even farther toward the spindle, which increases
> the feedrate even more, and so on, until you either break the drill, or
> decide to turn down your feedrates so they're way less than optimal during
> most of the cut, just to be tolerable when the drill breaks through.
>
> Now, take all of that and stack the plates 4 or 6 deep, so the total
> possible deflection is much bigger, and the drill has to do 4 or 6
> breakthroughs during every cycle. It gets real messy in a big hurry.
>
> If you want to test this, just go back to the drill press you used
> earlier, put a decent size drill in the chuck, and drill through a test
> plate. When you get to where the drill is just about to break through,
> you'll probably instinctively back off the feed pressure and grab tighter
> with the hand that's holding the work. That's because you know that the
> work is going to jump, and buck, and try to screw itself right up the
> drill bit, if you're not real careful. Same thing happens in a CNC
> machine, even if the clamping and smooth infeed help reduce it a bit.
>
> So, you'll need to support your plates underneath. Don't bridge them
> across blocks or parallels, but support them everywhere with a sacrificial
> plate underneath. That prevents downward deflection. Then, find a good
> way to clamp everywhere over the tops of the plates, so they can't bow
> upward when the drill's edges tug at them. A sacrificial plate on top,
> with pre-drilled holes, works best, as long as it's substantial and rigid
> enough to truly hold the plates down and keep them flat.
>
> If you support the plates properly, and pin them down adequately, then
> drilling a stack of plates shouldn't be any different than drilling
> through a thick piece of solid. If you don't contain them well, and use
> the right kind of drill, you're gonna need a lot of drills, and probably
> some good drugs or alchohol to get you through the day.
>
> <verbositizer to idle, teacher's hat back on the hatrack...>