What Can I Make With My CNC Machine?

• 1
  Subtractive manufacturing sets the rules
• 2
  Rule 1: Can the cutter reach every surface?
• 3
  Rule 2: Can the part be held while it is cut?
• 4
  Rule 3: Does the material cut cleanly at your tolerance?
• 5
  Rule 4: Does batch size change the answer?
• 6
  Rule 5: What does the part have to survive?
• 7
  Which process fits which part

A CNC machine does not build a part. It removes material from a solid block with a spinning cutter that follows a toolpath. That single fact answers most questions about what you can make with my cnc machine. If the cutter can reach the surface, and the block can be clamped without moving, the part can be cut. If the cutter cannot reach, the part cannot be cut on that machine in one setup.

The cutter has a diameter, a length and a flute geometry. Those three things define a minimum internal radius, a maximum depth-to-diameter ratio, and a floor finish. A 6 mm end mill cannot cut a 2 mm internal corner. A tool that is 50 mm long will deflect under load. Designers who respect tool geometry get parts that cut cleanly the first time.

Additive processes build up; subtractive processes cut away. That difference matters at the edges. CNC leaves sharp corners, tapped threads and tight bores that printing cannot hold. It also wastes more material, because chips leave the shop as scrap. For one-off functional parts, that trade is usually worth it. For hollow lattice shapes, it is not.

The rest of this page breaks the decision into five rules: geometry access, workholding, material behavior, tolerance demand and batch size. Each rule has a clear pass or fail. Run your part through all five and you will know whether it belongs on a mill, a lathe, a 5-axis center, or somewhere else.

A 3-axis mill cuts from one direction, along the Z axis. Any feature on the side walls, under a lip, or on a backward-facing face is either cut in a second setup or not cut at all. This is the first limit most designers hit. Pocket floors and top faces are easy. Undercuts and side holes usually are not.

A 4-axis mill adds rotation about one axis, normally the X axis. The part turns while the tool stays fixed in Z. That lets you cut a series of faces around a shaft or a cylinder without re-clamping. Flats, slots and hole patterns around a diameter all become single-setup work. The tool still cannot lean into the cut.

A 5-axis center adds two rotary axes, so the tool can tilt and the table can turn. The cutter can now approach a surface from almost any angle. Deep pockets, steep walls and contoured faces cut with a short, stiff tool instead of a long, flexible one. This is where complex aerospace and medical geometry becomes practical.

The test is simple. Walk the toolpath in your head. If the shank or the holder touches the part before the tip reaches the surface, the feature needs another axis or another setup. For most parts under 4,000 mm, GreatLight can reach the geometry with 3-axis, 4-axis or simultaneous 5-axis work across 127 machines.

• 1
  3-axisOpen faces, pockets, through holes, one-direction work.
• 2
  4-axisFeatures wrapped around a shaft or cylinder.
• 3
  5-axisUndercuts, steep walls, contoured surfaces, short tools.

Workholding is the quiet constraint. A part that cannot be gripped rigidly will chatter, spring or move, and the tolerance will fail no matter how good the machine is. Thin walls, long slender shafts and unsupported plates are the usual offenders. The cutting force has to go somewhere, and the fixture has to absorb it.

Milling usually holds the part in a vise, on a fixture plate, or in soft jaws cut to the profile. Turning holds it in a chuck, a collet, or between centers. A Ø400 mm rotary table handles round parts that need angular indexing. For production runs, a dedicated fixture pays back quickly because it repeats the same location every cycle.

Thin-wall parts need support, not force. A common approach is to leave sacrificial material, cut the walls in light passes, then remove the support at the end. Another is to fill a hollow with low-melt wax or a fixturing compound. Both add cost. A designer who keeps walls above roughly 1 mm in aluminium and 1.5 mm in steel avoids most of that cost.

Ask one question before you finalize a drawing: where does the clamp touch? If the answer is a finished cosmetic surface, the process will mark it. If the answer is a thin web, the part will deflect. Good DFM feedback catches this before the first chip is cut.

Material choice changes what is possible. Aluminium 6061 and 7075 cut fast and hold tight tolerances, which is why they dominate prototype housings and brackets. Stainless 303 and 304 machine well but work-harden if the feed is too light. Titanium Ti-6Al-4V and Inconel cut slowly and wear tools, so the same feature costs more.

Plastics behave differently again. POM and PEEK hold dimensions well when the tool is sharp and the coolant is controlled. ABS and PC can gum up if the spindle runs too slow. Carbon fibre machines like a dust hazard, not a chip, and needs extraction and edge support to stop delamination at the cut line.

Tolerance is a cost curve, not a switch. General machining sits around Ra 1.6–3.2 μm and holds typical shop tolerances. A fine finish at Ra 0.8–1.6 μm is normal for sealing faces and bearing bores. Pushing to Ra 0.2–0.8 μm or ±0.005 mm takes slower passes, more inspection and often a second operation.

The practical rule: pick the loosest tolerance the function allows. A bracket that only has to fit does not need a bearing-grade bore. A hydraulic manifold does. Matching tolerance to function is the fastest way to control cost without losing performance.

One prototype and ten thousand parts are different problems. At low volume, setup time dominates. A 5-axis center that cuts the part in one setup often beats a cheaper 3-axis machine that needs four fixtures, even though the hourly rate is higher. Fewer setups also mean fewer chances for a location error.

At higher volume, the balance flips. Dedicated fixtures, pallets and mill-turn centers cut cycle time. Features that were expensive at quantity one become cheap when the tool never leaves the cut. Die casting and vacuum casting start to compete once the geometry repeats, because the mold cost spreads across the run.

The crossover is not fixed. It depends on feature count, tolerance and material. A simple aluminium bracket may never justify tooling. A complex housing with twenty tapped holes usually does. GreatLight runs from a single prototype to 10,000+ part runs with no minimum order quantity, so the same shop can carry a part from first article to production.

If you expect volume later, say so early. A small change to a corner radius or a draft angle at the prototype stage can save a mold rework six months down the line. That conversation costs nothing at the quote stage.

Function decides finish and material more than the machine does. A part that sees salt spray needs anodizing or plating. A part that slides against another needs a low-friction surface and a controlled Ra. A part that carries load needs the right alloy and a fillet at the stress riser, not just a tight tolerance.

Surfaces are added after machining, and each one has a rule. Anodizing builds up a few micrometres and can round a sharp edge. Hardcoat anodizing builds more and can crack at tight radii. Electroless nickel holds dimensions well. Laser marking needs a minimum character height of 1.5 mm to stay legible.

Heat treatment changes size. A part machined to final dimension before hardening may move out of tolerance after quench or stress relief. The safe sequence is to machine oversize, treat, then finish grind or re-cut the critical faces. That adds a step, so only apply it where the hardness is actually needed.

The last check is inspection. A part that must be verified to ±0.005 mm needs a measurement plan, not just a good cut. Reports can be supplied on request, and every shipment passes raw material check, in-process monitoring and final inspection before it leaves the shop.