CNC turning vs CNC milling: how geometry decides the process

The short answer

CNC turning rotates the workpiece and moves a tool into it. CNC milling holds the workpiece still and rotates the tool. That single difference decides most of the process choice before volume, tolerance or cost are discussed.

If the part is a body of revolution, meaning a pin, a bush, a sleeve, a threaded insert or a connector barrel, turning is the natural route. If the part is a block, a plate, a bracket or a housing with pockets and hole patterns, milling is the natural route. The difficult cases are the ones in between: a shaft with a cross hole, a round flange with an off-centre slot, a turned body with two milled flats.

Those mixed parts are not automatically milled from solid. Often the cheapest route is to turn the round form and add the off-axis features in a second operation, either on a mill or with live tooling on the lathe.

What turning does well

A lathe holds the workpiece in a chuck or a collet and turns it at a set surface speed. A single-point tool travels along the axis or across the face, so the machine can produce outside diameters, bores, faces, grooves, chamfers, radii and threads in one continuous programme.

The strength of the process is that the part stays on one centreline for the whole cycle. Diameters, shoulders and thread callouts share a single rotational datum, so concentricity is a property of the machine rather than the operator. On a bar-fed lathe, the machine also cuts the part off and feeds the next length without stopping, which is why turned parts hold their cost at volume.

Turning is weaker on anything that is not round. A flat needs the tool to interrupt the cut, a cross hole needs a different axis, and a pocket cannot be produced by a single-point tool at all. Those features are where a second operation starts.

What milling does well

A milling centre holds the part in a vice, a fixture or a chuck on a trunnion, and a rotating multi-tooth cutter removes material along three or more axes. End mills cut slots and profiles, face mills flatten a surface, drills and taps make hole patterns, and a ball nose cutter produces a contoured surface.

Milling handles asymmetry, which is exactly what turning cannot. A rectangular housing with a pocket, a row of holes on a pitch, a mounting face at an angle and a set of tapped bosses are all normal mill work. Five-axis machines add the ability to reach undercuts and angled faces without re-fixturing, at a higher hourly rate.

The cost driver is setup and fixturing, not just cycle time. A part that needs soft jaws, a custom fixture and a probe routine carries that cost in the first article and again at every changeover. For low volumes, that setup can be a larger share of the price than the cutting.

The comparison that matters

The table below is a starting point for a first conversation, not a rule. A shop with the right machine can move a feature across the boundary, but the economics usually follow the geometry.

The features that force a decision

A few specific features change the answer more than the overall shape does. Read the drawing for these before choosing a route.

• Cross holes and radial holes: these need a second axis. They can be drilled on a lathe with live tooling, or in a second milling setup. Either way the position tolerance is no longer a pure turning capability.
• Flats and slots on a round body: a pair of flats is often cheaper to mill in a second operation than to broach or turn, unless the volume justifies a custom tool.
• Thin walls: a turned thin wall deflects under chuck pressure. A milled thin wall deflects under cutter force. Both need light passes and a support strategy.
• Deep pockets and narrow slots: these need a cutter with enough length to reach the floor. Cutter deflection sets the achievable depth-to-width ratio, not the machine.
• Threads off the centreline: a turned thread is generated by the spindle. An off-axis thread has to be milled or tapped, and the callout needs a different inspection method.
• Interrupted cuts: a flat on a turned diameter puts an impact load on the insert. That shortens tool life and can chip a finishing insert.

When the part needs both processes

Mixed parts are common in electrical hardware. A copper terminal might be turned from bar for the barrel and then milled for a palm with two bolt holes. A brass insert might be turned, knurled and then cross-drilled. The question is not which process wins, but how many setups the route needs and where the datums land.

A mill-turn machine removes a setup by combining both operations in one programme with one datum. That improves position tolerance between the turned bore and the milled hole pattern, and it removes the handling risk on a small part. The trade-off is hourly rate and availability, so mill-turn usually earns its place on parts with tight feature-to-feature position or on parts that are difficult to re-fixture.

Where mill-turn is not available, plan the sequence so the most critical features are made in the same setup. If the bore and the mounting face must be perpendicular, cut them together. If a cross hole only needs clearance, put it in the cheaper second operation.

Tolerances, finish and what they add

Both processes can hold tight limits, but they get there differently. A turned diameter is controlled by tool wear, thermal growth and machine rigidity, and it can be measured directly with a micrometer at the machine. A milled position is controlled by the machine's positioning accuracy, the fixture and the probe, and the error compounds across the part.

As a working guide, a turned diameter at ±0.02 mm is routine and ±0.005 mm needs a finish pass, temperature control and a capable machine. A milled feature at ±0.05 mm is routine, and ±0.01 mm needs a probe routine and a stable fixture. These are illustrative starting points. The real limit is the supplier's measured capability on that material and that machine.

Surface finish follows the same logic. Turning produces a fine helical lay that suits a sealing surface or a bearing journal. Milling leaves a scalloped path that depends on the cutter geometry and the stepover. If the drawing calls Ra 0.8 µm on a milled face, expect a finishing pass and a longer cycle.

What to send for a process recommendation

Send the drawing with the functional datums marked, the material and temper, the annual volume and the release size. Say which features are critical and which are clearance. A supplier who knows that the bore is a bearing seat and the cross hole is a cable exit will spend the tolerance where it belongs.

Add the blank form if it is fixed, for example whether the part must come from bar or from plate. Add the finish and any masking. If the part is still in design, send two versions and ask for the cost difference between a turned route and a milled route. That comparison is usually more useful than a single price.