On 2009-09-07, Ignoramus11113 wrote:
On 2009-09-06, DoN. Nichols wrote:
On 2009-09-06, Ignoramus11113 wrote:
[ ... ]
I would love to get that info, indeed. Possibly different sets have
different counts. Let me know, this would be great. I will buy 6" 1/2"
thick disks at mcmaster and will drill them.
O.K. From my B&S one, which has somewhat smaller plates
(largest number of holes is on the outside row):
Plate #1 #2 #4
Row == == ==
Outer 20 33 49
19 31 47
18 29 43
17 27 41
16 23 39
Inner 15 21 37
Someone else posted his plate counts on practicalmachinist, and it
seems that it is the same set of numbers. Since I assume that these
numbers have been carefully selected for best usefulness, I will try
to replicate them.
Those are the ones which are most useful for a 40:1 ratio gear.
Four screw holes.
You are saying that your plates have four?
I also found these plates on wttool.com, not sure whether they would
fit.
http://www.wttool.com/product-exec/p...m_source=froog
The first question is the diameter of the hole vs the bearing
boss on the dividing head.
The second is -- how useful will those hole patterns be with a
40:1 ratio head? The ad says that they are for a rotary table, so they
are more likely a good choice for a 90:1 ratio (which is more common in
rotary tables than in dividing heads).
Note that the smallest hole count on this is 26, so you miss all
the patterns on the #1 plate, and the first two on the second plate.
You are totally missing any multiple of 20, 19 is covered by 38,
18 is missing, 17 you have as 24, 16 as 32, and 15 as 30. so you are
missing from the first plate two of the initial circles. And your chart
will be expecting the same number of holes.
Also -- assuming that you did mean that there were four screw
holes in the plate you have, there are the questions of "what size
screw", and "on what circle radius" -- even if the bore is right. (And
it is probably metric on the plates in the ad, and almost certainly some
reasonable fractional inch size in your dividing head.
For that matter -- does '12"' mean the diameter of the plates,
or the diameter of the rotary table that they are to be used with?
[ ... ]
My plates are 5" diameter, and about 1/4" thick. (0.243, 0.245,
and 0.247")
Mine are the same diameter and thickness.
O.K. So other than the problems of the bore and the number of
screw holes, they are similar to my B&S ones.
If your existing plate is 1/2" thick, remove it and check the
other side. It may have a different set of holes there, with the holes
all going only half-way through (except where two line up by accident).
No, 1/2 is the thinnest circles at mcmaster carr. I wish I could get
1/4". I could make them, but it is more work.
Better to go from 1/4" mild steel plate than from discs cutoff
from round stock I think.
This brings up another question, that I would probably ask in a
different thread. The thinnest steel round disks at McMaster are 1/2"
thick.
And what is the thickness of the plate on your index head? If
1/2", check whether they are doing the trick of flipping it over for a
second set of holes. If not -- prepare to turn it thinner -- perhaps
after drilling. Or -- get some 1/4" steel plate, lay out circles on it,
bandsaw to approximate diameter, and turn on a faceplate with a live
center clamping it (and something to spread the force to out near the
edge). Obviously -- put a sacrificial spacer under the workpiece so you
don't turn the faceplate.
I may do this also, it would be a good exercise. I would cut the
approximate bigger circle with a plasma cutter. Also, my handle can
work with bigger plates, easily up to 6" diameter, so I would make
mine 6".
Check how high the crank bearing is from the table -- in all
positions of the head -- to make sure that a larger plate will not hit
the table.
[ ... ]
I may try to use a screw machine length carbide drill, so that I would
not need to worry about lubrication and drill at high speed so as to
waste less time.
Hmm ... you need really good fixturing with the carbide bits.
Any slip to the side and "plink" -- a broken bit.
I think that I can adequately hold them with two vee blocks placed
vertically.
But -- this means that if it *does* wander, it will break
instead of drilling a hole off location.
Do you think that item 27515A11 would be able to
drill steel without center drilling? (in a Bridgeport)
It has a split point, which is important. They are expensive,
but I would suggest that you get several, since it will otherwise bring
your project to a halt if the bit breaks. And carbide *does* break
easily.
What spindle speed can you get?
up to 5k RPM at 100 Hz.
Hmm ... 164 SFM -- a bit slow for carbide in mild steel I think.
How will you hold it? Collet, drill chuck, end mill holder?
If
1/8" collet.
Good -- that should be short enough.
a drill chuck, how much runout at 1/8". (BTW -- you did check the hole
size in what you have I presume. IIRC, my B&S one is closer to 0.100"
holes. But I've never had to drill a new plate, so I could be wrong.)
It was something like 0.1210 or some such, IIRC.
Because the head is already dividing everything by 1/40, there is not
a big need for super tight tolerances on the sizes of the holes.
Agreed. The highest hole count in the standard plates is 49,
which is 7.34 degrees direct, or 0 degrees 11 minutes 1 second error at
the head's spindle. IIRC, the spacing between the holes is pretty close
to the diameter of the holes, so an oversized hole just barely touching
the next oversized hole would be about half that error. Real error
probabilities are even smaller, of course.
But if the holder has a short enough overhang and has minimal
runout, if the spindle speed is high enough, and if the power feed is
slow enough, then it could work quite well.
I thought that I could just feed manually.
That is more likely to break the carbide bit because you are
feeding too fast for the drill to cut and clear the chips. The manual
feed on a Bridgeport is a bit too stiff to get fine control over the
feed of that small a drill bit. No feel. It looks like about 370 SFM
in mild steel, which with a 1/8" bit turns out to be around 11,300 RPM.
If you are *very* careful about your feed, the 5000 RPM may do.
Or feed using the knee crank. :-)
How slow can you set the downfeed? (My Bridgeport is CNC from
birth so it does not have the manual downfeed.) Between setting a slow
feed and an automatic depth stop release (which I seem to remember on
older Bridgeports which I have used) you may be alright. Test it on
some scraps first.
Are you planning to use your perl script to make a list of hole
coordinates?
Definitely.
You might consider using the index head itself to drill the circle
of holes. Which set of holes do you have? Can you use them with
the chart to make the other hole patterns you need? It will give
you good practice using the dividing head and the sector arms. Just
rotate it until the spindle faces up instead of horizontal. You can
divide a temporary plate by hand or using your rotary table, and use
that for drilling the hole circle you need. Note that the 40:1
ratio means that your placement errors in your temporary plate are
reduced by a similar ratio.
This is much more error prone than using a perl script with DRO.
The index head? Assuming you have the right holes to start
with, you can set it up to make it very difficult to make mistakes.
Yes, you have to count full turns of the crank (assuming that you are
making fewer than 40 holes in a circle) but you set the sector arms so
one touches the indexing pin, crank the full turn(s) if needed) and then
continue until you touch the second sector arm and drop the pin into the
hole which is there. Then you move the two sector arms (which are
locked together by a screw) until the first arm touches the pin, and
then you are ready for the next step. If you want to be extra sure, you
set the drill to just barely kiss the plate before the feed stops, and
then go through everything for the required number of holes. Then go
one more, and you should be right over the first "kiss" as a check. Once
you are sure that works out right, you can reset the depth to drill the
holes fully through into a sacrificial material behind the plate.
The chart should give you the number of full turns, then a
choice of one of (typically) several hole circles, and either the number
of holes from first sector arm to the second, or the degrees set between
the sector arms (there should be a scale to tell you that. This way,
you can set by hole count, and cross check it against the degrees, or
vice versa.
Also, I realized that if I make a plate with 540 equally spaced holes
(they would be equally spaced in terms of angle, but not necessarily
on the same radius from the center of the plate, so the holes would
not overlap), then this permits me to set the dividing head to any
minute of angle. I could write a script to give coordinates of these
holes to ensure that they do not overlap.
Hmm ... 540 equally spaced holes -- six circles of holes per
plate, That is 90 holes per circle, while the maximum on the standard
plates is 49 holes in the outermost circle. 90 holes at 0.125"
diameter, touching, will give a circumference of approximately 11.25",
or a minimum diameter of 3.581" And if you have the holes actually
touching, how weak will it be? Remember that you have to have the same
number of holes in each circle. You can get the outer row fairly easily,
but the inner is questionable. Let's allow 0.005" between holes so they
don't actually touch -- That moves the smallest circle out to 3.7242"
You might be able to do this if you can fit larger plates -- and
if you make longer sector arms and a longer indexing arm -- *if* the
larger plates don't interfere with the table.
And -- remember with these plates, you will need to print up a
new set of charts (fire up that perl interpreter and pray to Larry
Walls. :-)
Enjoy,
DoN.
--
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 ---