On 2010-11-10, Michael Koblic wrote:
The purpose of center drilling is to start the hole exactly where intended
without the drill point wandering all over the place, yes?
Actually -- that is the purpose of spot drilling -- with a spot
drill, not a center drill.
The purpose of a center drill is to make a conical depression
with a precise 60 degree angle and a bit of clearance to allow room for
the point of a live or dead center to stablize the end of a workpiece in
a lathe.
Yes, people do use center drills for the purpose of spotting
drills -- but that is not what they were made for.
This is then
normally followed by a twist drill of the desired size etc. From this
concept I would assume that the axes of the drills are concentric, or in
other words the hole drilled by the twist drill is exactly concentric with
the hole started by the center drill.
This does not seem happening in my case and I am wondering why.
Example: Using my mini-mill, I start the hole with a No.1 center drill and
then change to a twist drill (say 7/64"). X and Y are locked. As I bring the
drill down it is clearly off centre - today I measured it and it is quite
consistent: The drill point moves 0.010" "east" and 0.005" "south" to enter
the starter hole. If the full hole is then drilled it is slanted ever so
slightly - perhaps 0.001" over 0.25" length. This happened with two
different 7/64" twist drills.
Hmm ... sounds like a tramming error -- the axis of the spindle
is not perpendicular to the bed. And likely the column axis as well.
Your drill is longer than the center drill, so you have to move the head
up the column. If it is not perfectly vertical, this will drift the
center line of the axis to the side -- and front to back as well. The
column on a lot of the small mills tilts, so it provides adjustment for
the side-to-side part. However, the fore and aft takes a bit more
tricky work to get straight -- using shim stock is the usual approach.
Get a machinist's square (a blade with a heavy right-angle
piece) and resting that on the bed, check for a slight angle between the
blade and a piece of precision ground rod in the chuck. (or better, in
the collet, as chucks can introduce errors too.) (Actually -- get *two*
machinist's squares, and check them against each other as well. They can
come out of square to start with.)
Is yours one of those whose column has a pivot at the bottom, so
you can drill holes at an angle? If so, have you checked that the
column is truly vertical?
Or is yours one with a head which pivots on the carriage which
moves up the column? In that case, that might not be set quite right.
I tried a different No. 1, I tried both ends, same result. Looking at the
slowly rotating point with a magnifying glass it describes a small circle
which is not obvious when I bring it down on the metal.
That is either runout in the chuck, or a bent arbor for the
chuck. I forget -- does your machine use R8 collets? If so, how are
you holding your drill chuck? An R8 arbor, or a straight shank arbor
held in a collet? In either case, the arbor might be slightly bent.
However, there is
perceptible vibration of the mill which is absent if I drill with the twist
drill. I interpret this that the mill head is doing the circles while the
point is embedded. If I had a more rigid set-up the circle would perhaps be
apparent.
If you had a more rigid setup, the center drill would simply
make the hole a bit oversized as it swept around. And depending on the
orientation of the flutes to the eccentricity, it would either rub hard,
or cut oversized like a fly cutter.
I tried the same experiment with a No.2 and No.3 - same result.
Did you try a runout indicator on the shank of the drill? How
about with a longer rod (best would be a drill blank -- or an unusually
long dowel pin, but you might get away with some drill rod if it has not
been bent.
I thought I'd better find out which is the true center: The "center drill"
or the "twist drill" one. This was even more complicated than I expected. I
used two centere finders on small punch marks. They both showed center
differently! The centre found by the barrel-type coincided with the center
drill point, the wiggler type was quite significantly off (I use 10x
magnifying glass to get the best accuracy with both).
I presume that you started with the wiggler a bit off center,
and pressed a finger against the side until it stopped wiggling? *That*
is what sets it on center.
So the questions at this stage we
1) Is this a normal behavior? I thought unlikely...
2) Is this because of cheap Chinese center drills?
3) Is this a function of the mill chuck?
4) Is there some other reason?
Possibly the chuck or its arbor -- possibly the tram of the
column and head. More likely, a combination of the two.
The chuck or its arbor will produce the circles.
The tram of the head will produce the shift of the center point
as the head goes up and down.
I was wondering about the way the drills are clamped in the chuck and I
tried different degrees of tightening.
The proper way is to tighten equally using all three holes, one
after the other.
The last effort involving only light
tightening of the chuck both for the centre drill and the twist drill I
managed to hit the centre-found punch mark with both the centre and twist
drill.
Hmm -- the chuck is not a firm fit on the arbor? The arbor is
not a firm fit in the spindle?
Is it possible that over-tightening the chuck throws things out of kilter? I
hope to repeat this with the bigger center drills tomorrow but I would
appreciate any insight.
First get the runout indicator and check for runout on a ground
hardened rod (e.g. a drill blank or a dowel pin) both just where it
exits the chuck and down closer to the end. If it is better where it
exits the chuck, then the arbor may be bent. If it is equally off down
the length of the rod, it is likely a bur on one of the jaws in the
chuck.
Lots of things to check.
Good Luck,
DoN.
--
Remove oil spill source from e-mail
Email: | Voice (all times): (703) 938-4564
(too) near Washington D.C. |
http://www.d-and-d.com/dnichols/DoN.html
--- Black Holes are where God is dividing by zero ---