What's new
What's new

Peck Drilling

Metalcutter

Titanium
Joined
Sep 14, 2005
Location
San Diego
The Drilling Cycle

Two important factors in the drilling cycle contribute to premature drill wear.
Absence of dwell at the bottom of a hole or peck and rapid traverse down to the
exact bottom of a hole for succeeding pecks.

The reaction time of CNC equipment is all but instantaneous. When a drill,
feeds its way down into a hole and reaches the bottom of the hole; or the end of
any peck it changes direction to 'Z' up. It also changes into rapid traverse.
In other words the drill is literally snatched out of the hole.

Twist drill geometry incorporates right hand spiral flutes. The flutes
intersect the cutting edge to form a positive back rake angle. Under feed a
chip flows up the flute and therefore lays over the cutting edge.

This overlay condition resists the retraction of the drill. Through this, resistance forces
develop which work against the cutting edge trying to flake it away from the
drill body.

Drill geometry is designed to support the drill point under
tremendous cutting pressure while spiralling down into the work. Yanking the
working drill from the hole causes these forces to change direction on the
cutting edge at an angle down and away from the drill body creating an effective
high negative back rake.

On retraction the drill shears the chip off using the
fragile cutting edge. The result is either abnormal abrasion, chipping or
breaking off the cutting edges.

Also after leaving the hole a crested bump is
left at the hole bottom because the drill point did not finish making the chip
before retracting. If the returning drill is going in for another peck it could
rapid to the bottom catching this bump and experience excessive cutting forces
again chipping the cutting edges.

A fix for this problem is to program a dwell long enough to hold the 'Z' axis drill
position for one to two spindle revolutions letting the chip thin out to zero.

Retraction at this point would not damage the tool. This length of time at 3000
RPM would be approximately 0.02 seconds for one revolution. To figure this time
increment divide the RPM into 60 seconds. Three hundred RPM would equal 0.2
seconds.

Shallow Holes

Most machining centers have several canned drill cycles which are used to drill
holes. A G81 spotting or shallow hole drill cycle has no programmable dwell,
although popular and okay for through holes it is not a good choice to use in
blind holes.

A G82 drill cycle has dwell that can be programmed. This G82 cycle is a good
replacement for a G81 when used for drilling holes through or three drill
diameters deep. In fact a G81 can be completely eliminated.

Deep Holes

Peck drilling is generally required when drilling holes deeper than three and
one half drill diameters. Most canned pecking cycles allow for starting the
feed rate just above the last peck to prevent the drill from hitting the hole
bottom at rapid traverse rate.

The difficulty is that no dwell is available at any peck bottom except the last. So no effective method exists to prevent the chip breaking action from causing severe tool damage. Some people think this tool damage comes from heavy feed rate. Light feed rates are then programmed to solve a problem which should not exist in the first place. The net result is lower productivity.

A New Subroutine

A temporary work-around is to build a special pecking cycle. The pecking cycle
would incorporate a dwell at the end of each peck for a time length sufficient to
allow one or two drill revolutions.

Another approach would be to change the drill feed to up, or Z plus, for a few thousandths. This would allow the chip to thin out at the end of the peck just before retraction.
This approach would not be a dwell but a feed reversal and would thin the chip and also position the drill away from the hole bottom to prevent work hardening the material.

The permanent fix is to talk the machine tool builders into creating drill pecking cycles in
their controls which could provide programmable dwell or feed reversal between
every peck. This enhancement would allow drilling to proceed at the best feed
rate and production could go up across the board.

An exercise:

Take a one inch travel indicator and mount it to the machine head and point it down, and "touch" something.

Then program a drill cycle very slowly. Make the rapid slow enough to see as well.
Then run it and watch how the drill cycle acts in your machine.

Notice too, the amount of clearance between pecks before cutting takes place again.

That clearance between pecks is important. If the clearance is .030 and you are drilling a .031 hole you'll be waisting time, especially if you are pecking every 1/2 diameter.

The Haas I used once had the ability to change the clearance between pecks. I use about 3-5% of the drill diameter for clearance.

The Special Peck Cycle

When drilling the tough alloys SST, Inconel, and like materials, drills last longer if you can feed reverse before retracting. In these alloys it's probably worth the trouble to figure one out.

Regards,

Stan Dornfeld
 

metlhed

Stainless
Joined
Jan 17, 2007
Location
Ohio
FYI, some peck cycles allow for a dwell call at the bottom of the peck. Most do not, but there are 'bout 1000 different peck cycles available 'tween controls and builders.
 

HuFlungDung

Diamond
Joined
Jan 19, 2005
Location
Canada
Would or could turning on 'Exact stop check' be of any benefit in a peck drilling operation? Might it slow the machine motions just enough to do the same job as a special cycle filled with dwells ?
 

Metalcutter

Titanium
Joined
Sep 14, 2005
Location
San Diego
I honestly don't know THAT much about G-Code. I only had a Fadal for a few months.

It is a drill cycle and I thought exact stop was for milling.

Please try it and if you have time get back to me on what you found.

Also thank you for the rating... who ever you are. *s

Regards,

Stan-
 
Last edited:
Joined
Nov 15, 2005
Location
Illinois
Metalcutter,
Terrific post. I agree with everything you said.
I have noticed in many situations that I get more tool life from my drills when I peck a little less. compared to pecking more often.

One way to help the shock given to drills is to slow down the acceleration of the z-axis in the parameters. It is just one or two parameters.
 

Metalcutter

Titanium
Joined
Sep 14, 2005
Location
San Diego
Metalcutter,
Terrific post. I agree with everything you said.
I have noticed in many situations that I get more tool life from my drills when I peck a little less. compared to pecking more often.

One way to help the shock given to drills is to slow down the acceleration of the z-axis in the parameters. It is just one or two parameters.

Thank you Mari..

Regards!
 

CarbideBob

Diamond
Joined
Jan 14, 2007
Location
Flushing/Flint, Michigan
Well this certainly is a different metal cutting theory.
I wonder what happens if you let the tool dwell so the chip formation transfers from shear at full chip load to burnishing at zero chip load in a nice piece of stainless?
I see more tools worn out by too light of a chip load than by too heavy.
This is one of the reasons milling tools won't last as long as turning tools. As the tool exits and the chip thins you don't maintain enough chip thickness to establish a clear shear plane which results in increased heat in the tool. Rubbing (zero chip load) is never a good thing for cutting tools.
Write yourself a macro and give this a try. :eek:
Bob
 

g-coder05

Titanium
Joined
Mar 5, 2006
Location
Subic Bay
how about
G20G17G40G80G90
G0E1X5.0Y5.0
T01 M06
M3 S3000
G0Z.1 H1
G8 (hi precision stop on)
M88
G83Z-1.0R.1Q.1F16.1
G80 Z.1 M89
G9 (hi precision stop off)
M5

Ive used this on a fadal on accident. i forgot to turn the exact posistioning of after a finnish pocket pass. then i noticed that there was a slight hesitation between moves coming out of and going back in the hole.

just curious if this would work on any other machine.
 

metlhed

Stainless
Joined
Jan 17, 2007
Location
Ohio
g-coder05,

With those exact codes, no. Fadal uses some codes that you don't see very often on other contols. For instance, exact stop on a Fanuc 21i is M70, and lots of control options don't include exact stop. In theory, yes, but it depends on control, machine and purchased options. Fanuc also uses a P (with no decimal: P500 is 1/2 sec dwell) for a dwell at the bottom of drilling cycles, and I believe Okuma uses an E for dwell at the bottom of the peck (with decimal: E.5 is 1/2 sec dwell), which Fadal uses as it's positioning (like G54, 55, etc on Fanuc). Also, the M70 on the Fanuc is not modal and does not need turn off. Nice example, though. I'm gonna pass that off to the Fadal guys here for an example, thanks.
 

Boris

Titanium
Joined
Oct 4, 2005
Location
England
Having drilled horrid tool steels with 1mm drills Its time to stick my 3 euros in

What I observed while doing that drilling was that if you put a dwell in the hole, no matter how short, you would always end up rubbing the bottom of the hole, does'nt matter that its 'only' 1 or 2 revs, its still a rub/work hardening situation,
Now in those tool steels the hole would work harden anyway on peck, but the trick to get over that was to keep the feed high .08mm/rev for example, so that the drill point would punch through the work hardened bit and start drilling, if you fed in slow, then the drill would initially struggle to get through the work hardening, then snatch as the pressure increased and the drill went through the work hardening, thus increasing the chances of the drill breaking.

The usual guideline peck we used for sub 2mm drills was a peck of no more than 1D .

Changing our drilling cycles to the above rather than the mid speed/slow feed system we used previously extended the drill life by about 100% which ment 2 broken drills to edm out of a die instead of 4.

Boris
 
In theory, this is true. In reality, only a percentage of it is true. Given that even the fastest of machines has a momentary "dwell" when doing canned cycles. The reason is the control having to "switch" from Feed to Rapid and Rapid to Feed. There is no real "instant" but some machines can react quicker in the switch than others.

Some of you pointed out concerns over work hardening materials or even simple stainless and your concerns are justified. A dwell at the bottom can cause more grief from hardening as well as heating up the drill tip causing edge fracturing and soon, a broken drill anyway. Generally, the best solutions for this type of drilling is a G81 to full depth or thru. Get in, get out. But if I had to peck in tough stuff, no dwelling here. It only takes a fraction of an "instant" to burn up the end and work harden in many cases. In softer materials???? Forget it, the cheap cost of a drill over the added cycle time of a dwell just isn't worth it for me.

On a similar note regarding rapid retract, .... stew on this. I said a bit ago that the machine reaction to having to switch to rapid mode causes a momentary delay anyway even with the fastest of machines. Most of you should be able to try this and get similar results. I have a lot of parts with gobs and gobs of drilling. As a test, I took one of them that has 14,000 holes to tweak the drilling cycle. One of the drills runs for 21 hours on its own. So it rewrote the drilling cycle to do a high feed out instead of the usual canned cycle. Everything else was the exact same... Same depth, same peck, same speed/feed, same reference plane, same retract, etc. The only thing different was that the macro feeds the drill on retract instead of rapid. By doing this, the cycle time dropped by over 7 hours. Why? Because the control doesn't have to "think" about the necessity of 'Rapid'.

You can test this yourselves. You don't need a 14,000 hole part either. Just take a plate and program a pattern of 2 or 300 holes. Just simply test by using a spotter in G81. Time it. Now, write the code in long hand with high feed retracts.. something like 500IPM or faster. Run that program and time it. You should see a measurable time difference. Additionally, if you Feed in between the holes instead of rapid positioning, you should see even more of a difference. So, do you quit using canned cycles? Hell no. The situation needs to warrant it. Even with a hundred holes, it may not be worth it unless you're deep drilling all of them with a lot of pecking. Some more food for thought....
 

metlhed

Stainless
Joined
Jan 17, 2007
Location
Ohio
Psychomill,

Excellent point...I've noticed some of the same with fast controlled feeds over rapids, but I ask you this...if you have 50 parts total 2 on the table and a 6 hole bolt circle with the std +/-.01 on the circle, but +-/.002 per hole. Ya got 50 parts and on #11&12 you find the hole runs big by .003. Do you scrap #11,12,13,14 'cause the cycle runs too fast to make a tool change or check your taper for chips? That kinda stuff has it solid place, but not a panacea for all runs. Great point, though.
 
Of course not Metlhed, what's your point? In your case I would use the Laser Tool Eye in my machines to verify the tool length and diameter at a pre-determined interval, have a macro tool change to a back up tool already pre-set in my 330 tool hive if it detects wear, then a macro would trigger my FMS robot loader and notify it which part it had to change a tool on, segregate it at the load station for a manual check and verify.

Doesn't everybody have this stuff????? ;)


:D:D
 

Metalcutter

Titanium
Joined
Sep 14, 2005
Location
San Diego
In theory, this is true. In reality, only a percentage of it is true. Given that even the fastest of machines has a momentary "dwell" when doing canned cycles. The reason is the control having to "switch" from Feed to Rapid and Rapid to Feed. There is no real "instant" but some machines can react quicker in the switch than others. ....

Thank you for your contibution psychomill, I'd like to add my experience with a FADAL.

I was spot facing aluminum with a .250" end mill at 9000 rpm and 15 ipm G81. I noticed a witness line across the diameter of the spotface. I couldn't feel it, but it was there.

I couldn't believe it could happen at 9000 rpm. I did however change to the G82 code with a .02 sec dwell. It went away. That's why I say the change is instantanious.

Your thoughts?


Best regards,

Stan-
 

ARB

Titanium
Joined
Dec 7, 2002
Location
Granville,NY,USA
Hey Pyschomill.

Can I come and shadow you for a week?:)

That must be some kind of outfit you work for!

On the rapid thing. I have found this to be a time saver myself on certain jobs where seconds count.
 
It's only money right??? Can never have enough... always wishing for more.

Thank you for your contibution psychomill, I'd like to add my experience with a FADAL.

I was spot facing aluminum with a .250" end mill at 9000 rpm and 15 ipm G81. I noticed a witness line across the diameter of the spotface. I couldn't feel it, but it was there.

I couldn't believe it could happen at 9000 rpm. I did however change to the G82 code with a .02 sec dwell. It went away. That's why I say the change is instantanious.

Your thoughts?

Given the machine you're using, most likely the reason you saw the mark was because the machine over compensating the next move (or under compensating, depends on how you look at it). Even at 15ipm, the machine was trying to react to the rapid retract and was "short" cutting the hole. This is actually quite common. The dwell therefore made it disappear because you gave it a command that forced the spindle to remain at its target depth long enough before retracting.

Since most machines/controls don't allow you to use LookAhead during canned cycles (or isn't actually active even when commanded) and Exact Stop or Position Check (G60, G61) doesn't fuction for Z in a canned cycle, you chose the wise option (for this case) to add a dwell.

So, there's two scenarios. Surprisingly, several "cheaper" machines do react quicker (for this type of switching) than a high end one. However, target numbers are usually poor, the switch can be more violent and have an inconsistent reaction. The accel/decel is more linear than curved. This is also the tendency for lower powered controls and slower processors.
 








 
Top