News

I had a job a few years back. Stainless steel shafts, about 300mm long, with a threaded end and a cross-hole. Nothing exotic. We machined them, they looked great, off they went to the customer.

Two weeks later, the phone rang. The shafts were bending in service. Not a lot—maybe 0.1mm—but enough to make the assembly bind up. The customer wanted to know what we did wrong.

We didn't do anything wrong in machining. The problem was we machined them from annealed stock, and the customer put them into a application that needed more strength. What they really needed was heat treated material. They just didn't know it.

That conversation happens more often than it should. People treat heat treatment like a black box—something metallurgists worry about, not something machinists need to understand. But if you're making parts that need to actually work, you need to know what heat treatment does, when to do it, and how it affects your machining.

I watched a batch of forty parts get scrapped last month because of a burr you could barely feel.

The parts were aluminum valve bodies. Machined beautifully. Dimensions perfect. Surface finish right where it needed to be. They went to assembly, and the technician slipped an O-ring over one of the ports. The burr—tiny, almost invisible—caught the rubber and shaved off a sliver. Not enough to see, but enough to compromise the seal. The assembly passed pressure test? No. It leaked.

Forty parts. Two shifts of work. All because someone decided that deburring could wait and inspection would catch it.

This happens every day in shops around the world. Deburring gets treated like an afterthought. Something you hand to the new guy with a file and a scraper. But in reality, that "afterthought" is often the difference between a part that works and a part that fails.

I had a customer call me last week with a question I hear all the time. He had a prototype part in hand—a small aluminum housing, nothing complicated—and he was running his fingernail across the surface.

"It looks good," he said. "But what's the Ra? Can you get it smoother?"

I asked him what the part needed to do. He wasn't sure. He just knew that smoother was better, or at least that's what he'd heard.

This comes up constantly in this business. Engineers put surface finish callouts on drawings because they think they should. Purchasing agents ask for "a good finish" without defining it. And machinists are left guessing what "good" actually means.

So let's talk about what surface finish CNC machining can realistically deliver—not what the brochures say, but what you can expect when you put a part on a machine and run it.

I watched a part get scrapped last week over something you could not see with the naked eye.

The inspector pulled out the profilometer, ran the stylus across a sealing surface, and shook his head. The print called for Ra 0.8. The part measured Ra 1.2. Off to the scrap bin it went. The machinist was frustrated—the part looked fine, the dimensions were good, everything fit. But the engineer had specified that roughness for a reason, and reasons cost money when ignored.

Surface roughness is one of those things that seems abstract until it bites you. A seal leaks. A bearing overheats. A shaft cracks after a thousand cycles. And when you trace it back, the root cause is often those microscopic peaks and valleys that everyone ignored during machining.

So let's talk about Ra—what it actually means, why anyone should care, and how you keep it under control without driving cycle times through the roof.

I have lost count of how many arguments I have seen start over a tolerance block.

A part comes off the machine. The machinist checks it. It is 25.02mm on a feature that calls out 25.00mm ±0.05. He stamps it good and sends it to inspection. The inspector looks at it, sees 25.02, and writes it up as non-conforming because the print says ±0.05 and the part is 0.02 over.

The machinist says, "It's within tolerance." The inspector says, "It's over by two hundredths." The engineer gets called in. And then someone notices that the drawing actually says 25.00mm ±0.05 and the machinist read it as 25.00mm ±0.05. Everyone was looking at the same numbers and seeing different things.

This happens every day in shops around the world. Not because anyone is wrong, but because tolerances are a language. And like any language, you have to know the grammar.

I have been in this industry long enough to remember when 0.1mm was considered good enough. You bored a hole, it fit the shaft, and everyone went home happy. Then came aerospace. Then medical devices. Then the whole world started asking for tighter numbers without really understanding what they were asking for.

Now every drawing that lands on my desk has ±0.01mm somewhere. Customers write it in the notes like it is nothing. Like asking for 0.01mm is the same as asking for a chamfer.

It is not nothing. So let us talk about what that number actually means when you are standing in front of a machine.

I was talking to a shop owner last month. He had a job come in—stainless steel, 12mm thick, about fifty parts, some holes and cutouts. Nothing crazy. But he sat there staring at his equipment list, trying to figure out which machine to put it on. Laser was busy. Plasma would do it, but the edges would need cleaning. Waterjet could handle it, but it would take all day.

He looked at me and said, "I have three cutting machines and I still don't know which one to use."

That is the reality of running a fab shop today. You have options. And options are great until you have to make a decision. So let's break this down. Not with marketing brochures, but with the kind of talk you would hear on the shop floor.

Walk through any toolroom that has been around for a while, and you will notice something. The wire EDM machine is usually sitting off in its own corner. Maybe it is behind a plastic curtain to keep the mist contained. And if you ask the foreman about it, he will probably tell you the same thing: "That machine only runs when nothing else can handle the job."

That is the truth about wire EDM. It is not a production machine for cranking out simple parts. It is the last resort—and sometimes the only option—for work that is too precise, too hard, or too delicate for conventional machining.

So what ends up on that wire machine? Let's talk about the parts that justify the slow speeds and the setup time.

If you have spent any time on a shop floor, you know the drill. A job comes in, and the part travels. It goes from the lathe to the mill, maybe to another machine for a secondary operation, and if you are unlucky, it sits in a queue waiting for the next setup. Every time you move a workpiece from one machine to another, you are gambling with tolerance. And you are definitely burning hours.

For shops looking to stay competitive—especially in aerospace, medical, or automotive supply—the conversation has shifted toward doing more with less. That is where mill-turn machining comes in. It is not a new concept, but it is one that continues to prove its value: load a billet once, walk away with a finished part.