CNC Mill Basics Quick Start: How a Milling Cut Removes Metal

• 1
  CNC Mill Basics Quick Start: What the Cut Removes
• 2
  Rotation, Flutes and Chip Load
• 3
  3-Axis, 4-Axis and 5-Axis: When Each Wins
• 4
  How the Part Is Held Decides the Result
• 5
  What Tolerance and Finish You Can Realistically Hold
• 6
  Design Moves That Keep Milling Cheap
• 7
  What Happens Between Quote and Shipped Part
• 8
  Choosing the Right Milling Setup

A CNC mill is a spinning cutter that moves along controlled axes while the workpiece stays clamped. The tool has multiple flutes, and each one takes a small chip per revolution. Metal is removed, never formed, so the finished part is whatever the cutter left behind.

The controller reads G-code, a list of coordinates and feed commands. Every line tells the machine where to go and how fast. That file decides surface finish, tool load, and cycle time far more than the machine model does.

Milling is subtractive, so the first question is always what the cutter can physically reach. A deep pocket, a sharp internal corner, or a tall thin wall sets limits before any tolerance does.

This page is a CNC mill basics quick start for engineers who need to reason about those limits without standing at the control.

The spindle turns the tool, typically between a few thousand and well over 10,000 rpm depending on diameter and material. Surface speed is the speed of the cutting edge as it sweeps past the metal. Small tools need high rpm to keep that surface speed up.

Feed per tooth, also called chip load, is how far the tool advances for each cutting edge. A 10 mm four-flute carbide end mill in 6061 aluminum might run around 0.05 mm per tooth at 8,000 rpm. Push the chip load too low and the edge rubs instead of cutting, which work-hardens stainless and burns the tool.

Too high and the flute breaks or the part moves. The window is narrower in titanium and Inconel because they conduct heat poorly and keep it at the edge.

Climb milling, where the cutter travels in the same direction as the tooth at the entry point, throws the chip behind the cut and generally gives a better finish on modern machines with low backlash.

A 3-axis mill moves X, Y and Z. The tool always approaches from one direction, so every feature must be reachable from that side. Simple plates, brackets, and housings with open pockets are cheap and fast on three axes.

A 4-axis machine adds rotation about one axis, usually A, so the part turns while the tool cuts. This suits cylindrical parts with flats, slots, or holes on multiple faces, and it cuts setup count.

A 5-axis machine adds two rotary axes, letting the tool tilt. That matters for undercuts, deep cavities with angled walls, and organic surfaces. It also lets a short, stiff tool reach a deep feature by tilting instead of hanging out long.

Tilting is not free. Simultaneous 5-axis motion is slower to program and to run, so use it where the geometry demands it, not as a default.

Clamping is the quiet variable. A vise holds a block rigidly but only reaches the outside. Soft jaws machined to the part profile hold thin or irregular shapes without crushing them.

Vacuum plates suit flat, thin plates where clamps would block the cut or distort the part. Fixture plates with dowel pins locate a part repeatably across thousands of cycles.

The setup also sets the datum. If the drawing dimensions from one corner but the operator zeros from the center of a bore, every feature shifts by the difference. Agree on datums in the DFM stage and the first article usually passes.

For thin walls, light finishing passes with a sharp tool and generous support beat heavy roughing. The part deflects under cutting force long before the machine runs out of accuracy.

General milling holds around ±0.05 mm on a well-set-up 3-axis machine. Tightening to ±0.005 mm is possible but only on features with good access, stable material, and a controlled thermal environment.

Tolerance is not uniform across a part. A bore can be held tighter than a long thin rib, because the rib moves under cutting load. Specify the tight number only where the function needs it.

Surface finish follows the same logic. Ra 3.2 μm is a normal as-machined result. Ra 0.8–1.6 μm needs a finishing pass with a sharp tool and light chip load. Below that, expect a second operation such as polishing.

Harder materials and deep cavities push both tolerance and finish wider. A 200 mm deep pocket in Inconel is a different job than a 10 mm slot in aluminum.

Corner radius is the first thing to check. A cutter leaves a radius equal to its own, so an internal sharp corner is impossible. Set internal radii to at least one third of the pocket depth.

Keep the part reachable. If a feature sits behind a wall taller than the tool can clear, the job needs a second setup or a different machine, and both add cost.

Standardize hole sizes to common drill and end mill diameters. A 9.8 mm hole costs more than a 10 mm hole with no functional gain.

Avoid unnecessary tight tolerances on non-critical faces. Every tight callout adds inspection time, and inspection is a real line item on any quote.

A quote is a promise about a process, not just a price. The shop reviews the drawing, checks reach and tolerance, then decides the machine, the workholding, and the number of setups.

Programming follows, then a first article. The first part is measured against the drawing before the run continues, because a wrong datum found after 500 parts is expensive for everyone.

In-process checks catch drift from tool wear or thermal growth. Final inspection confirms the last part and, on request, the report ships with the parts.

The whole chain matters more than any single step. A perfect toolpath on a loose fixture still produces a loose part.