Every cut you make prints a report and throws it on the floor. The chip carries the shape of what happened in the cut and the heat that came out of it, and reading it takes about thirty seconds and no instruments. Most shops sweep it up without looking. Here is how to look, in the order that actually tells you something.
Read the shape first. Shape works in every material.
Color only means something in steel. Shape means something everywhere, which is why it goes first.
What you want off most jobs is a chip that breaks on its own into short pieces—tight C shapes, commas, or short helical curls that clear the cut and fall. That chip has been bent far enough to snap, it leaves the cutting zone immediately, and it carries heat away with it.
The two failure directions look nothing alike:
| What you are sweeping up | What the cut is doing | First thing to change |
|---|---|---|
| Long stringy ribbons, bird's nests | The chip is never getting thick enough to break. It wraps the tool and the part. | More feed, not less |
| Dust, powder, glitter | The edge is rubbing instead of cutting. Heat goes into the tool, not the chip. | More feed, and check the edge is not already worn |
| Thick, violent, fractured chunks | Too much load per tooth. This is where edges chip. | Less feed, or less depth |
| Good shape, then stringy ten minutes later | Nothing changed in the program. The edge wore. | Index or change the tool |
Then read the color—and know that the shop is split on it.
Ask five machinists what a blue chip means and you will get two answers, delivered with equal confidence.
One camp says a blue chip is good news: the heat generated in the cut is leaving with the chip instead of soaking into your tool and your part, which is exactly where you want it. The other camp says a blue chip means the cutting zone is too hot and you are shortening the edge. Both camps have run plenty of parts.
They are both right, and the thing that decides which one applies to you is not the color. It is what else is hot.
Color is not a verdict. It is a thermometer that tells you where the heat went. So use it with the other two readings you already have for free:
- The chip is colored, the part is cool to the touch, the edge still looks like an edge. The heat left in the chip. That is the cut working as intended, and it is common and normal running carbide hard and dry.
- The chip is colored and the part is hot. The heat is not leaving. Your coolant is not reaching the cutting zone, or there is not enough chip to carry it. This is the one to act on, and the part—not the chip—is what told you.
- The chip is colored and the edge is breaking down fast. Too much speed for the grade. Cutting speed drives heat harder than feed does, so take the surface speed down before you touch anything else.
- Chips that come off silver when you are running carbide in steel at a sensible speed. Worth a look. It can mean the cut is being flooded so well it never gets warm, which is fine, or that it is slow enough to be rubbing, which is not.
The chip you must not cut twice
A chip that falls clear is a chip that is done with you. A chip that stays in the pocket gets cut again, and recutting is one of the fastest ways to lose an edge that was otherwise fine. The chip has already work-hardened; it is loose, it is abrasive, and it arrives at the edge from an angle nothing was designed for.
This is where a long stringy chip stops being a cosmetic problem. It wraps the tool, packs the flutes, and feeds itself back into the cut. In a deep pocket or a slot, evacuation is as much a part of the job as the cutting is—and it is worth one more look at the metal removal rate you think you are getting, because a recut chip is removal you are paying for twice.
Flute count is a lever here that people forget they have. Fewer flutes means more room between them for the chip to get out. That is the whole argument for running a 2-flute in aluminum and in deep slots: not that it cuts better, but that the chip leaves.
The thirty-second routine
Do this once at the start of a run and once when anything sounds different.
- Pick a chip off the floor. Not out of the pan—off the floor near the machine, where it landed while the cut you care about was happening.
- Look at the shape. Short and broken is the target. String and dust are both telling you the feed is too low.
- Touch the part. Not the chip. The part is the honest witness for where the heat went.
- Look at the edge. Thirty seconds with a loupe on the tool beats any amount of guessing from the chip.
- Change one thing. Feed first, then speed, then depth. Change two and you have learned nothing.
Questions we get
Are blue chips bad?
Not by themselves. A blue chip means heat left the cut with the chip rather than staying in the tool and the part, which is often exactly what you want, especially running carbide dry. What makes it bad news is the company it keeps: if the part is also hot, the heat is not leaving and your coolant is not reaching the cut. If the edge is breaking down quickly, the cutting speed is too high for the grade. Judge by the part and the edge, not by the chip's color alone.
My chips are long and stringy. Should I slow down?
Usually the opposite. A chip breaks when it is bent past what it can take, and it only gets thick enough to do that if the feed per tooth is high enough. Backing the feed off thins the chip further and pushes the edge toward rubbing. Increase the feed first. If something has to come down, bring the cutting speed down rather than the feed.
What does it mean when the chips are more like dust or powder?
The edge is rubbing instead of cutting. Either the feed per tooth is far too low, or the edge is already worn enough that it cannot engage the material. Both put the heat into the tool instead of the chip, which is the worst place for it. Raise the feed, and look at the edge before you blame the numbers.
Why do my aluminum chips never change color?
Because color in machining comes from steel oxidizing as it heats. Aluminum, brass and most non-ferrous material do not do that, so there is no color to read no matter what the cut is doing. In those materials you are reading shape, sound, and whether material is welding to the edge.
The chips looked right for twenty minutes and then went stringy. What changed?
Nothing in the program—which is the point. A chip that degrades during a run with no change in feed, speed or depth is telling you the edge wore. The tool stopped forming the chip the way it was forming it. Index the insert or change the tool rather than chasing the numbers.
How much does recutting chips really cost me?
More than most people expect. A chip that stays in the cut has already work-hardened, and it arrives back at the edge loose and abrasive from an angle nothing was designed for. It is a common cause of an edge failing early on a job where the numbers were right. In slots and deep pockets, getting the chip out is as much of the job as cutting it, which is why fewer flutes—and the room between them—is often the better tool there.
Two companion pieces: Feed Rate in Turning covers the number that makes the chip in the first place, and Runout explains why some of your flutes may be making chips while the others watch.
Chip reading is a diagnostic habit, not a measurement. The guidance here describes how experienced machinists use chip shape and color together with part temperature and edge condition, and it is offered as a starting point for your own setup, not as a specification. Tell us the material, the tool and what the chips look like, and we will work through it with you.