Problem Details
Job: 6000-series aluminum, two-diameter hole. A Ø6 mm hole 100 mm deep, followed by a Ø3 mm hole 40 mm deep. Tooling is standard HSS jobber drills ground for aluminum. Machine is a vertical CNC mill with flood/external coolant only — no through-spindle coolant.
Symptom: the Ø3 mm drills break. The Ø6 mm operation runs. A calculated starting point of S2600 / F250 was rejected by the operator as too aggressive.
Root Cause of the 3 mm Breakage
At 40 mm depth a Ø3 drill is running at 13.3×D. Three failure mechanisms stack up:
- Chip packing. Aluminum produces long, gummy, work-hardening-free chips that weld into the flutes. Without through-coolant the flutes are the only evacuation path, and at 13×D there is no way for external coolant to reach the cutting lip.
- Chip re-entry on peck return. The spindle is vertical and the hole opens upward-down into gravity. Every full retract dumps some chips back into the bore. On the next plunge the drill point meets that debris, the point jams, torsional load spikes, and a 3 mm HSS shank twists off. This is the dominant killer in blind vertical deep holes without internal coolant.
- Insufficient surface speed / rubbing. Too low an RPM at a very light feed makes the drill rub instead of cut, generating heat at the margin and built-up edge on the lips. Aluminum then welds to the drill.
Cutting Data That Works
Two data sets are in play. The first is field experience for 3 mm HSS in aluminum; the second is a verified production run on this class of job.
| Parameter | Experience-based range | Verified run (0 broken drills / 90 holes) |
|---|---|---|
| Spindle speed | 4000–5000 rpm | 2500 rpm |
| Feed rate | F150–F200 mm/min | F100 mm/min |
| Peck increment | Required if no through-coolant | 10 mm (Q10) |
| Coolant | Through-spindle strongly preferred | External + full peck retract |
| Result | — | 90 holes, zero drill failures |
Derived values for a 2-flute drill, Vc = π × D × n / 1000 and fn = F / n:
| D (mm) | n (rpm) | Vc (m/min) | F (mm/min) | fn (mm/rev) | fz (mm/tooth) |
|---|---|---|---|---|---|
| 3 | 2500 | 23.6 | 100 | 0.040 | 0.020 |
| 3 | 4000 | 37.7 | 150 | 0.038 | 0.019 |
| 3 | 5000 | 47.1 | 200 | 0.040 | 0.020 |
| 3 | 2600 | 24.5 | 250 | 0.096 | 0.048 |
| 6 | 2500 | 47.1 | — | — | — |
The pattern is clear: every surviving combination lands near fn ≈ 0.04 mm/rev (0.02 mm per lip). The rejected S2600 / F250 setting is 0.096 mm/rev — roughly 2.4× the chip load of the proven data, at only 24.5 m/min cutting speed. That is the arithmetic reason it broke drills: heavy chip load at low speed means high torque and a thick chip that cannot clear a 13×D flute.
fn constant — torque per revolution stays the same, chips get thinner and the cycle finishes sooner. What breaks 3 mm drills is chip load and packing, not spindle speed. Limit RPM only by spindle runout, tool balance and how well the drill runs true in the holder.Peck Strategy and G-Code
Without through-coolant, a full-retract deep-hole cycle is mandatory. Use G83, not G73. G73 (chip-break) retracts only a small amount and never clears the flutes over 40 mm — it will pack the drill.
(3 mm HSS DRILL - 6000 SERIES ALUMINUM - 40 mm DEPTH)
T03 M06 (3.0 mm HSS, short as possible in holder)
G90 G54 G17 G21
G00 X0. Y0.
S2500 M03 (Vc = 23.6 m/min; scale to S4500 with F180 if rigid)
M08 (flood - aim nozzles at hole mouth)
G43 H03 Z10.
G83 X0. Y0. Z-40. Q10. R2. F100. (4 pecks of 10 mm, full retract to R)
G80
G00 Z50. M09
M05Cycle mechanics with these values: 4 pecks, 40 mm of cutting at 100 mm/min = 24 s of feed time plus rapid retract/return per peck. If the Ø3 hole is a continuation of a 100 mm Ø6 bore, set R just above the Ø3 entry point (e.g. R-98.) so the drill does not rapid the full 100 mm of empty bore on every peck — but verify the bore is clear of chips first, because a G83 retract to a deep R-plane will not evacuate the upper bore.
Peck sizing rule
| Depth reached | Peck increment (Q) | Reason |
|---|---|---|
| 0–3×D (0–9 mm) | Full 10 mm acceptable | Flutes still open to atmosphere |
| 3–8×D (9–24 mm) | 10 mm | Proven on the verified run |
| 8–13×D (24–40 mm) | Reduce to 5–7 mm if torque rises | Evacuation path is longest; chip re-entry risk peaks |
Practical Fixes Beyond the Numbers
- Get coolant to the lip. A through-coolant carbide or HSS-Co drill in 3 mm changes this job from marginal to routine: it can often be run in one plunge with no pecking. If capital allows, this is the correct engineering answer to a 13×D blind hole.
- Spot and pilot. Spot with a stub or 90°/120° spot drill, then start the Ø3 with a stub-length drill to 3×D, then swap to the long drill. This removes wander-induced side load, a common cause of small-drill snapping at entry.
- Minimize stick-out. Clamp the drill so only what is needed protrudes. A 3 mm drill's torsional stiffness scales with D&sup4; every extra millimeter of free length lowers the twist-off margin.
- Air blast between pecks. With the vertical spindle, program a dwell at the R-plane with an air nozzle aimed into the hole mouth, or add an M-code air blast, to push chips out rather than let them fall back in.
-
Retract fully to a clearing plane periodically. Every 2–3 pecks, retract to
Z10.and blast. Costs seconds; saves tools. - Watch for built-up edge. Inspect the drill lips every 10–15 holes. Aluminum welded to the margin is the warning sign before a break. Replace or re-point at first sign of adhesion.
- Do not dry-run this hole. HSS in aluminum without lubricant at 13×D will gall and seize.
Verification
- Run one hole at
S2500 F100 Q10and stop. Withdraw the drill and inspect flutes: clean, curled chips = correct. Packed, ribbon-welded chips = reduce Q and increase coolant/air. - Check spindle load percentage on the control during the last two pecks. A load trace that climbs peck-over-peck indicates chip accumulation, not tool wear.
- Measure hole depth and check for a bell-mouthed entry (indicates wander/side load — add a pilot).
- Once stable, scale up: raise to
S4000 F150keepingfn = 0.038 mm/rev, and re-run the load-trace check. If load is flat and chips are clean, continue toS4500–5000 / F180–200. - Log tool life. The benchmark to beat is 90 holes with zero breakage at the conservative settings.
FAQ
What RPM and feed for a 3 mm HSS drill in 6000 series aluminum?
A verified production setting is 2500 rpm at F100 mm/min with 10 mm pecks (0.040 mm/rev). Experience-based data supports 4000–5000 rpm at F150–200 mm/min, which holds the same 0.038–0.040 mm/rev chip load at higher surface speed.
Why do 3 mm drills break at 40 mm depth in aluminum?
Chips packing in the flutes and falling back into the vertical blind hole during peck retract. On the next plunge the point jams on that debris, torque spikes, and the small shank twists off. Excessive feed per revolution — for example 0.096 mm/rev at 2600 rpm — makes it worse.
Should I use G83 or G73 for a 13xD hole without through-coolant?
Use G83. Its full retract to the R-plane clears the flutes completely each peck. G73 only breaks the chip with a short retract and will pack a 3 mm drill long before 40 mm.
Can I drill 40 mm deep in one plunge with a 3 mm drill?
Only with through-spindle coolant delivering lubricant and pressure to the cutting lip. With external coolant only, peck drilling with full retracts plus air blast is required.
How do I calculate cutting speed for a 3 mm drill?
Vc = π × D × n / 1000. At D = 3 mm: 2500 rpm gives 23.6 m/min, 4000 rpm gives 37.7 m/min, 5000 rpm gives 47.1 m/min. Feed per revolution is fn = F / n; divide by 2 for per-flute load on a standard twist drill.
