Evaluate CNC machining machine import options

• 1
  What a CNC machining machine import really changes
• 2
  Matching axis count and travel to your part mix
• 3
  Reading accuracy class and repeatability claims
• 4
  Spindle, torque and duty cycle on imported machines
• 5
  Service access, spares and the real cost of ownership
• 6
  When an imported machine is the wrong answer
• 7
  CNC machining machine import options compared

An imported CNC machine does not make parts better by itself. It changes three variables: how many setups a part needs, how much heat the structure absorbs during a long cut, and how repeatable the tool change is over thousands of cycles. Those three drive your scrap rate and your cycle time. Everything else on the spec sheet is a cost or a convenience.

Setup count is the easiest one to test. If a part needs four faces machined, a 3-axis machine needs three or four fixtures plus operator re-clamping. A simultaneous 5-axis center can reach those faces in one setup, so the locating error from re-clamping disappears. For a part with tight true position between bores on different faces, that single change often decides whether the job is feasible at all.

Thermal behavior is slower and harder to see. A machine that holds ±0.005 mm for the first hour may drift to ±0.02 mm by hour six as the spindle and ballscrews warm up. Cast iron and polymer concrete beds absorb and redistribute that heat differently. This is why a 20-minute test cut proves very little about a machine that will run a 9-hour shift.

Tool change repeatability is the third variable. A worn tool changer or a weak spindle taper shows up as runout that varies from pocket to pocket. On a machine cutting 10,000 parts a year, a 5 μm variation in tool seating becomes a scrap statistic. Ask for a tool-to-tool repeatability test, not just a spindle runout number.

Start with the largest part you will actually run, not the largest part the machine can hold. A machine rated at 4,000 mm travel spends most of its life cutting small parts if that is your order book, and you pay for the envelope in floor space, power draw and foundation cost. List your top ten parts by annual quantity and write down their bounding box. The machine should cover the biggest one with the fixture clamped, because fixtures and vises eat 100–200 mm of travel.

Then count the faces that need machining. Prismatic parts with features on three or fewer faces usually run faster on a 3-axis machine with good fixtures than on a 5-axis machine with a slower rotary table. Five-axis earns its cost when the part has compound angles, deep pockets with undercuts, or tolerances that stack up across setups. For impeller-like or medical implant geometry, 5-axis is not a luxury; it is the only practical route.

Also check the rotary table diameter against your part. A Ø400 mm table cannot swing a 500 mm part without interference on the far side of the cut. Draw the part on the table in CAD before you commit.

Travel figures in brochures rarely state whether they include the tool change position or the table at full stroke. Ask for a machine drawing with the actual usable envelope marked. On some imported machines the usable envelope is 10–15 percent smaller than the headline number, which matters when your part is close to the limit.

Positioning accuracy and repeatability are different numbers and only one of them predicts your part tolerance. Positioning accuracy is how close the machine gets to a commanded point. Repeatability is how close it returns to the same point over many cycles. For production, repeatability matters more, because your offsets absorb a constant error but cannot absorb scatter.

A machine quoting ±0.005 mm positioning should also quote repeatability, usually as a smaller figure like ±0.002 mm. If the supplier quotes only one number, ask which one it is. Vague answers mean the spec sheet was copied from a catalog.

Thermal compensation changes how those numbers hold up. Many imported machines now ship with scale feedback on X, Y and Z, and some add spindle growth compensation. Scale feedback removes ballscrew pitch error from the loop, which matters on machines with long travel where screws expand several micrometres per degree of temperature rise.

Do not stop at the machine spec. The tolerance you can hold on a real part also depends on the fixture, the tool holder, the cutter and the material. A machine with ±0.002 mm repeatability running a cheap ER collet and a worn end mill will still produce ±0.05 mm parts. Budget for shrink-fit or hydraulic holders if your feature tolerances are tight.

Spindle power tells you what the machine can rough. A 15 kW spindle with a good torque curve will take a heavier cut in 4140 steel than a 25 kW spindle tuned for aluminium at 20,000 rpm. Look at the torque curve, not the peak kW. Ask for continuous torque at the rpm you plan to run.

Duty cycle is the number most buyers skip. A machine rated for continuous duty can run 24 hours at full load. A machine rated for 40 percent duty needs cooling breaks or it derates. If your shop runs two shifts, buy continuous duty or you will replace spindles early. Ask for the duty rating in writing.

Tool interface matters too. BT30 and HSK-E40 suit small tools and high rpm. BT50 and HSK-A63 handle heavy cuts and large diameters but spin slower. A machine with the wrong interface for your work will force you into light passes and long cycle times.

Coolant strategy belongs in the same conversation. Through-spindle coolant at 70 bar clears chips from deep holes and lets you run higher feeds in stainless and titanium. Without it, deep-hole drilling becomes a pecking operation and your cycle time doubles.

The purchase price is roughly half of what the machine will cost you over ten years. Spares, service visits, training, power and tooling make up the rest. Before you sign, ask how many spare parts are held in your region and what the lead time is on a spindle, a ballscrew and a servo drive. If the answer is eight weeks by sea freight, price that downtime into the decision.

Ask who installs and commissions the machine. A machine that arrives with no local commissioning support can sit for weeks while you find someone who knows the control. Ask for the name of the commissioning engineer and confirm they are certified on that control platform.

Training is often bundled but rarely sufficient. Two days on a new control is enough to jog the machine, not enough to program a 5-axis job safely. Budget for a week of on-site training or send one programmer to the builder's facility before delivery.

Finally, put the acceptance test in the contract. Define the test part, the material, the tolerances and the measurement method. Run the test at the builder's plant before shipment if possible. A machine that passes a written acceptance test is far easier to argue about later than one that arrived with a verbal promise.

Importing is not always cheaper. If you need one machine for low-volume prototype work, a used domestic machine with local service may cost less per good part than a new import with slow spares. Run the numbers over five years, not over the invoice.

If your parts change every month and your fixtures are simple, a 3-axis machine with a good CAM post may beat a 5-axis machine that you cannot keep fed with work. Capability you do not use is still a cost.

If your team has no experience with the control platform, add the learning curve to your lead time. A new control can take a competent programmer two to four weeks to reach full speed, longer for 5-axis with RTCP setup.

And if your tolerance requirement is loose, say ±0.05 mm, do not pay for a high-accuracy import. Spend the difference on better fixtures and tooling instead. That is usually where the accuracy is actually lost.