Reading chips is one of those skills that separates real deep hole drillers from everyone else. The chips tell you everything — coolant pressure, feed rate, tool condition, material consistency, heat generation. They’re like a conversation the machine is having with you in real time. The problem is, sometimes you don’t want to hear what they’re saying. You’re on a deadline, you’re tired, you’ve got pressure from above to keep the parts coming, and it’s easier to tell yourself the chips look “close enough.”
I’ve seen guys who can look at a pile of chips and tell you within ten percent what the feed rate was. I’ve seen guys who can spot a worn coolant nozzle from the chip color. And I’ve seen guys who ignore every sign the machine gives them because they’re too focused on the production count to stop and think. I’ve been all of those guys at different points in my career.
But the most expensive lesson came from the time I was the guy who knew better and did it anyway.
I learned this lesson the hard way, on a job that still makes me wince when I think about it fifteen years later.
We were running a production order for a heavy equipment manufacturer — track link pins for a D9 bulldozer undercarriage. Forty-one hundred parts, made from 4140 steel hardened to about 30 Rockwell C, needing a 12mm hole drilled 300mm deep. It was a job we’d done before, and the parameters were well-established from previous runs. Feed at forty millimeters per minute, spindle at twenty-eight hundred RPM, coolant at seventy bar.
The machine was a five-year-old BTA drilling center with a hundred-liter coolant tank and a thirty-micron paper band filter. I knew that machine’s sounds and habits the way you know your car’s quirks. It had a tendency to drift on the X-axis if you didn’t warm it up properly in the morning, and the coolant pump had a harmonic vibration at exactly forty-three bar that you learned to ignore. I knew everything about that machine. Or so I thought.
The first fifty parts came out perfect. The chips were textbook — tight cones, about 8mm long, with that slightly silver-blue color that says the cutting temperature is right where it should be. The kind of consistent chip formation that tells you everything is dialed in perfectly. I was watching the chip bin fill up with beautiful little curls of steel and thinking this was going to be the easiest ten-thousand-dollar job we’d had all year. I even mentioned it to Dave on the next shift. “Smooth sailing,” I said.
Around part sixty, I noticed the chips starting to change. They were getting longer and stringier, more like ribbons than cones. Instead of breaking off cleanly at 8mm, they were coming out as continuous strands that wrapped around each other. And they were a different color — a darker, almost blue tint around the edges, with a brownish discoloration that said the heat was building up.
I should have stopped the machine right there. A change in chip shape means a change in cutting conditions. That’s basic deep hole drilling 101. The chips are your direct feedback loop — the most direct feedback you have, more immediate than any gauge reading. If they change, something has changed upstream. It could be the material, the coolant, the tool, or the parameters. Regardless, you stop and check.
But I convinced myself it was nothing. Maybe the material batch had slightly different hardness. Maybe the coolant temperature was a little warmer in the afternoon because the shop was hotter. I told myself the chips looked “close enough” and kept the line running. The production target was in my head — we were supposed to ship five hundred parts by Friday.
I even called the production manager to tell him we were ahead of schedule. “Chips look good,” I said. They didn’t. But I said they did because that’s what he wanted to hear and it was easier than telling him I needed to stop and troubleshoot.
The chips don’t lie. But people do. And I was lying to myself and to my boss because I didn’t want to admit something was wrong.
By part eighty, the chips were coming out as long, continuous ribbons that tangled around the drill shank and clogged the chip box every few cycles. I had to stop every three parts to clear the tangles with a hook. That should have been the second warning, but I was too focused on the count to see it.
By part one hundred, the drill snapped.
Not gradually — instantly. One second it was cutting, the next second the torque monitor tripped and the spindle stopped with a grinding shudder that shook the machine. I hit the emergency stop by reflex, but the damage was done. I pulled the drill out and found the carbide tip was completely gone. The braze joint had failed from overheating — the temperatures had gotten high enough to soften the braze material, and the carbide insert had broken off somewhere inside the hole.
I went back and checked the parts I’d already run. Out of the first hundred, twenty-seven had the same issue — the carbide tip had chipped or broken off inside the hole, leaving a rough, undersized bore. Some of them had broken carbide fragments still embedded in the hole. All of them were scrap. Twenty-seven parts, each with about thirty dollars in material cost and another twenty in labor and machine time. That’s over thirteen hundred dollars in scrap, plus a three-hundred-dollar gun drill, plus the downtime to reset the job.
The root cause? The coolant nozzle in the machine had loosened over time and was spraying at a slight angle instead of directly into the drill’s coolant hole. Just a few degrees off axis. The cutting zone wasn’t getting enough cooling, so the heat built up incrementally, the chips changed shape as the temperature rose, and eventually the carbide failed. The nozzle had probably been loose for weeks, slowly creeping out of alignment with every tool change. The chip change was the first sign — and I ignored it because I didn’t want to stop the line.
Dave, the senior operator who’d been running deep holes since the 1980s, walked past my machine that afternoon and saw the mess of scrapped parts and broken tooling. I told him what happened. He didn’t say anything. He just reached into the chip bin, picked up one of the stringy ribbons from the last cycle before the break, looked at it for a second, and tossed it back.
“You saw this coming,” he said. It wasn’t a question.
“Yeah,” I said.
“Next time stop the machine when the chips change. Even if you think it’s nothing. Even if the production manager is standing behind you tapping his watch. Stop it, check it, prove everything is fine. Then keep running. A ten-minute check is cheaper than a hundred scrap parts and a broken drill. Every time.”
I’ve never forgotten that. Now, if I feel even a hint of doubt about the chip shape, I stop the machine and prove it’s okay before I continue. I wasted years thinking I could “feel” when a job was going wrong. Now I know — the chips will tell you before anything else will. You just have to be honest enough to listen, even when what they’re saying is inconvenient.
Dave retired a few years after that, but he left me with a habit that’s saved me more times than I can count: every time I start a new job or change a parameter, I catch a chip from the first cut and tape it to the setup sheet. I write the date and the parameters next to it. When something starts to feel wrong later in the run, I compare the current chips to that first chip. If they match, I keep running. If they’ve changed, I stop and check.
I showed that trick to the production manager after it saved me from another close call. He liked it so much he made it shop policy. Now every setup sheet in the shop has a section at the top labeled “Reference Chip.” It forces operators to look at their chips from the very first part, which means they can’t ignore changes later.
I’ve taught that trick to every operator who’s come through our shop. It’s simple, it costs nothing, and it catches problems before they become disasters. The physical chip on the setup sheet is better than any memory — you can’t argue with what’s right in front of you.
That hundred-part scrap bin sat next to my machine for a week before it got hauled to the recycling bin. Every time I walked past it, I remembered Dave’s words: the chips are never wrong. Only the operator is. And the good operator is the one who listens when the chips start talking.
I’ve told this story at every safety meeting I’ve ever been asked to speak at. It’s not a feel-good story. It’s a warning. The chips don’t care about your production schedule. They don’t care about your bonus. They don’t care that you’re tired. They just tell you the truth. And the sooner you learn to listen, the fewer parts you’ll scrap.
The reference chip system that Dave taught me is now part of our ISO 9001 procedure. It’s documented, audited, and enforced. Every setup sheet has a taped chip in the corner. If an inspector can’t find the reference chip, they won’t sign off on the first article. It’s become that fundamental to our process.
I called Dave a few years after he retired to thank him for that lesson. He didn’t even remember the incident. “I told a lot of people to stop being stupid over the years,” he said. “Which time was this?” I described the hundred scrapped parts and the broken drill. “Oh, that one,” he said. “Yeah, you looked like you needed to hear it.”
That was Dave. Never one to sugarcoat. But he was right. I needed to hear it. And I’ve been telling his words to new operators ever since. “A ten-minute check is cheaper than a hundred scrap parts.” It’s practically a slogan now.
I still wince when I think about that scrap bin. But I also know that it taught me more than any successful job ever could. The parts that go right don’t teach you anything. The parts that go wrong — those are the lessons that stick.
The coolant nozzle on that machine was loose for weeks before the job. It didn’t cause any obvious problems until the chip load got high enough and the heat built up. That’s the thing about gradual failures — they’re invisible until they’re catastrophic. The only early warning was the chips. And I ignored them.
Now I make it a point to show every operator I train what a good chip looks like versus a bad one. I have a small collection of chips from different jobs, labeled with the date and the problem they indicated. It’s like a chip museum. And it’s more useful than any training manual I’ve ever seen.
The chip museum is my most requested training tool. New operators ask to see it. Senior operators send me chips from their own jobs to add to the collection. It started as a reminder of my biggest mistake. Now it’s the most valuable training resource in the shop.
I keep the first chip from that scrap run in a separate box labeled “Don’t do this.” It’s long and stringy, the kind of chip that says something was wrong from the beginning. I show it to every new operator on their first day. “If your chips look like this,” I tell them, “stop the machine. Then call me.”
That chip museum has grown over the years. It’s got about thirty chips now, each with a story. But the first one — the one from the job that cost me a hundred parts — is still the most important. It’s the one that started everything. And it’s the one I show to every new operator on their first day, because some lessons need to be seen to be believed. I haven’t scrapped a hundred parts from chip negligence since. That’s the power of a good reminder.
The worst part of that job wasn’t the money. It was knowing that I saw the warning and ignored it. The chips were never wrong. Only the operator was. And I was that operator.