Teensy Weensy Drilling op help?

Hi all

Having a bit of a nerve wracking time at the moment, hoping someone out there is equipped to say "you idiot, all you need to do is XYZ" :)

I have a 3/16" (4.76mm) OD x 0.9mm wall thickness tube of 316 stainless mounted in my 4th axis and need to drill a pattern of holes in it. Ideally Ø0.3mm. FSWizard tells me a feed rate of 30mm/min and 15000 or so RPM is appropriate with my little HSS PCB drills (stubby with 1/8" shanks).

I slowed it down to half that plunge rate and attempted a G01 F15 Z0 move and, can't say I was surprised, watched as the bit bent a little and snapped.

The drill point is centred over the A/X axis in the Y direction (A running along X) but I can't honestly say I can even see a spot where it tried to start the hole on the surface of the tube.

Any words of wisdom on a better way to do this before I roll the dice on my last spare 0.3mm bit?

Thanks in anticipation...

There is no way 15 ipm will work with a .3mm drill bit.

Max out the rpm and try a 0.0001" to 0.0002" chip load. At 15k rpm that puts you at a feedrate of 3 to 6 ipm. I would stay towards the lower end of the feedrate. Pecking will help clear the flutes out. A 0.015" peck might help.

Second what ERUS said. Spot drill first with a center drill, then peck drill with small depths.

15mm/min is about 6IPM but I'll give it ago at half that, with a 0.4mm (0.015") peck. I don't have a centre drill with a chisel point that small but I could mill a flat land with a 2.0mm slot drill first, so there's no convex curvature to deal with.

Thanks, folks.

I have had similar problems myself drilling 0.3 mm dia holes. I think what is happening is that the tip of the drill (which does NOT come to a single point) is deflecting off the curvature.
You could try starting with a tiny centre drill: either BS0 or custom, and only going in something like 0.1 - 0.2 mm. Cheap brand BS0 may still have an unacceptable 'flat' at the end. I had some carbide centre drills smaller than BS0 made up for me, and they were usefull.

For one job I ended up making a D bit from a broken 1 mm drill bit ground to a very sharp point on a T&C grinder with a diamond wheel. Grinding the flat on the D bit was tricky, but the point itself did provide enough feedback to show where I had got to. In the end I think I went for a (approx) 25 micron overshoot past the actual point.

But that in itself did not solve the problem of the runout in the spindle/chuck. Even if there is only 10 microns runout, that can snap a 0.30 mm drill bit (Sphinx brand, Swiss, very good). In the end, I reversed the entire system. The part went in the lathe chuck and the drills went in an Albrecht chuck mounted on a custom vertical slide on the cross slide (allowing X and Y adjustment).

Step 1: Made a custom jig to hold the (many) blanks without any wobble
Step 2: faced off a single unit so I could see where the centre of rotation was.
Step 3: mounted my precision XY stage with Albrecht chuck on the saddle and aligned the tip of the D bit with the centre of rotation.
Step 4: used the D bt to make a tiny start recess in the job.
Step 5: replaced D bit with 0.3 mm drill bit
Step 6: watching the whole thing with a stereo magnifying headset, used the main drive on the saddle to drill out the hole.
This made many units, each with a 0.3 mm hole.

I have also done something similar with a TV camera, tele lens and medium-sized monitor for a member of my staff once. He was trying to make a miniature air bearing. This may be closer to what you want. With this set-up he could see when the fragile drill bit was starting to bend and back off. That was with a Swiss Aciera drill which had 'very little' runout - and a rather high price. He loved it.

Cheers
Roger

Thanks, Roger.

In the end I flattened off a land with a 2mm endmill, then went 5mm/min 0.2mm peck.

Did a trial run with aluminium tube and it worked beautifully, then brushed the last spare bit with my fingertip and snapped it as I was removing it from the collet.

I’m guessing Ali will do the job :)

I flattened off a land with a 2mm endmill
Oh, the number of times I have done that! I even do it before using a centre drill with some steel. Not cheating, just being smart.

Cheers
Roger

Imagine my frustration (running from the MDI, this didn't seem worth the hassle of CAD/CAM) when I discovered that

G91 G83 A72 Z1.0 R2.2 Q0.2 L5 F5

would crawl to A72, then do 5 peck cycles in the same damned place, rather than [ rapid A+72, do a peck cycle ] * 5.

Ah well, got it done in the end and the thing works well. It's a gas filler stalk for charging 1.25l PET bottles with CO2 from a regulator hanging off a CO2 fire extinguisher, AKA a Mad B#stard's SodaStream. And at $20 for an out-of-test but full D size extinguisher that lasts a year compared to $30 for a SodaStream cylinder that lasts maybe a month around here, I'm happy.

- - - Updated - - -

Imagine my frustration (running from the MDI, this didn't seem worth the hassle of CAD/CAM) when I discovered that

G91 G83 A72 Z1.0 R2.2 Q0.2 L5 F5

would crawl to A72, then do 5 peck cycles in the same damned place, rather than [ rapid A+72, do a peck cycle ] * 5.

Ah well, got it done in the end and the thing works well. It's a gas filler stalk for charging 1.25l PET bottles with CO2 from a regulator hanging off a CO2 fire extinguisher, AKA a Mad B#stard's SodaStream. And at $20 for an out-of-test but full D size extinguisher that lasts a year compared to $30 for a SodaStream cylinder that lasts maybe a month around here, I'm happy.

Mach3Mill_1.84.pdf:
Page 10-25, section 10.7.24, para 3:
Rotational axis words are allowed in canned cycles, but it is better to omit them. If
rotational axis words are used, the numbers must be the same as the current position
numbers so that the rotational axes do not move.

and further down
The L number is optional and represents the number of repeats. L=0 is not allowed. If the
repeat feature is used, it is normally used in incremental distance mode, so that the same
sequence of motions is repeated in several equally spaced places along a straight line. In
absolute distance mode, L > 1 means "do the same cycle in the same place several times,"
Omitting the L word is equivalent to specifying L=1. The L number is not sticky.
When L>1 in incremental mode with the XY-plane selected, the X and Y positions are
determined by adding the given X and Y numbers either to the current X and Y positions
(on the first go-around) or to the X and Y positions at the end of the previous go-around (on
the repetitions). The R and Z positions do not change during the repeats.

In other words, the option of auto-incrementing the A axis with the L parameter is not there.

Cheers
Roger

Smart move making flat landing zones for the drill bit, I will remember that "cheat" if I ever have a similar job! :wave: