There Are Bullet Points in the Tool: What They Mean and What to Do

• 1
  Key takeaways
• 2
  What bullet points in the tool actually are
• 3
  Why heat, not sharpness, drives the weld
• 4
  Chip load and engagement: the two numbers that matter most
• 5
  How to remove a bullet point without scrapping the tool
• 6
  What bullet points cost you when they are ignored
• 7
  Step by step: diagnose and correct a bullet point problem
• 8
  Which materials produce bullet points, and why

A bullet point is a small, rounded lump of workpiece material welded onto the cutting edge. It looks like a droplet sitting on the rake face or the nose radius. Operators often mistake it for a chip that never cleared. It is not. The lump has bonded to the carbide at the atomic level, and the bond is strong enough that prying it off removes substrate too.

The mechanism is pressure welding. Cutting generates heat at the deformation zone. If the edge runs hot enough, and the contact pressure between chip and rake face is high enough, the two clean metal surfaces weld. Once a lump forms, the effective rake angle changes. Cutting forces rise, the lump grows, and eventually a fragment breaks away and tears a crater out of the insert.

That sequence matters for diagnosis. A chipped edge at the end of a long run is often the last stage of a problem that started as a bullet point twenty minutes earlier. If you only look at the failed edge, you chase the wrong variable.

• 1
  Signs in the cutRising spindle load, squeal that was not there before, and a burr that appears on one side of the part only.
• 2
  Signs on the partRa drifts from Ra 0.8–1.6 μm toward Ra 1.6–3.2 μm, then smears or tear-out appear.
• 3
  Signs on the chipChips change colour and curl tighter; aluminium chips start breaking instead of flowing.

Every cut puts roughly 80 percent of its heat into the chip. The rest goes into the workpiece and the tool. That split is not fixed. It depends on the thermal conductivity of the material and the geometry of the contact zone. When the split shifts toward the tool, the edge temperature climbs, and the weld threshold gets closer.

Tool wear follows the same curve. Flank wear and temperature move together, and edge strength rises with temperature, so higher hot hardness delays the weld until the temperature climbs further. The practical target is a cutting temperature that is high enough for smooth plastic flow and low enough to stay under the weld point.

Coolant helps, but it helps most where it reaches. Flood coolant often never touches the contact zone because the chip blocks the path. Through-tool coolant delivers pressure directly under the chip, lifts it, and drops the rake-face temperature at the same time. That is why a coolant change sometimes beats a tool change.

Feed per tooth decides chip thickness. Chip thickness decides where the heat goes. When the chip is too thin, the edge rubs instead of cutting, the material work-hardens, and the next pass welds to the tool. A minimum chip thickness of about 0.05 mm per tooth keeps the cut in the shearing regime for most steels. Aluminium tolerates less, but running at 0.02 mm per tooth invites rubbing.

Radial engagement has a similar effect. A narrow radial width concentrates heat in a small arc of the edge and raises the local temperature. Widening radial engagement spreads the same material removal over more of the flute, which lowers peak temperature. This is the reason trochoidal toolpaths with small radial engagement work well for hard materials only when they are paired with high feed rates.

Axial depth controls how long each tooth stays in contact. Deep axial cuts with light radial passes shorten the time per revolution that any point on the edge is cutting, so heat has time to dissipate between engagements. On a 16 mm end mill in 6061, 12 mm axial depth with 30 percent radial width is a workable starting point. Program a test cut and watch spindle load before committing the run.

• 1
  Aluminium starting point0.10–0.25 mm per tooth, 8,000–15,000 rpm, high-pressure coolant.
• 2
  Stainless 304 starting point0.08–0.15 mm per tooth, 120–180 m/min surface speed, never dwell.
• 3
  Titanium starting point0.06–0.12 mm per tooth, 40–60 m/min, through-tool coolant at 70 bar or more.

Do not pick at it with a scribe or a diamond file. The weld is stronger than the surrounding cobalt binder, so the lump usually leaves with a piece of the edge attached. If the insert is indexable, rotate to a fresh edge and treat the damaged one as scrap. That is the cheapest decision in most shops.

If the tool is a solid carbide end mill with a brazed or ground geometry you need to keep, an aluminium etch bath removes the welded layer. Sodium hydroxide solution dissolves aluminium without attacking tungsten carbide. Warm the bath to about 50 °C, immerse for 10 to 20 minutes, and rinse thoroughly. Do not use the same bath on steel or titanium welds; it will not work and it will damage the binder.

For steel and titanium, submerge the tool in a hot caustic or specialized pickling solution and use ultrasonic agitation. Check the edge under 10× magnification afterwards. Any remaining crater or micro-chip means the tool goes to scrap, because a damaged edge will weld again within a few minutes of cutting.

Prevention is cheaper than removal in every case. Log the edge condition at fixed intervals: every 30 minutes for aluminium, every 15 minutes for titanium. Photograph the edge with a phone through a loupe. A pattern emerges after three or four tools, and it points at the parameter that needs to move.

• 1
  On indexable insertsIndex to a new corner and record the failure mode in the tool log.
• 2
  On solid carbideCaustic etch for aluminium only; ultrasonic pickle for steel and titanium.
• 3
  On coated toolsCoating loss at the lump site usually means the tool is done regardless of geometry.

A welded edge does not fail instantly. It degrades slowly, and that slow degradation is expensive. Surface finish drifts out of tolerance, so the part needs a second operation or a rework pass. Dimensional scatter widens, and on a ±0.005 mm feature the difference between a clean edge and a welded one is often the difference between pass and scrap.

Tool cost is the visible part. A carbide end mill that welds after 40 minutes instead of 4 hours costs six times more per part in tooling alone. The hidden part is spindle time. A machine running at 70 percent of its safe feed rate because the operator is afraid of welding is a machine producing 30 percent less than it could.

There is a quality-system angle too. Under ISO 9001:2015 and IATF 16949:2016, process monitoring has to be demonstrable. Edge-condition logs and the parameter changes they drive are the evidence an auditor asks for. Shops that track tool life by part count alone cannot show why a lot drifted. Shops that track time-to-weld can.

For medical work under ISO 13485:2016, the same logs support traceability. A welded edge that sheds a fragment into a machined feature is a foreign-object risk. Documented edge inspection and controlled tool change intervals are the practical answer.