CNC Processing Historical Travel: How Machine Motion Changed the Parts We Can Make

• 1
  Why cnc processing historical travel began with a positioning problem
• 2
  From three linear axes to simultaneous 5-axis travel
• 3
  What the travel envelope actually decides
• 4
  Accuracy, finish and the real limits of a travel envelope
• 5
  Setup count is the hidden cost behind historical travel limits
• 6
  Matching geometry to travel and axis count

In the 1940s the limits of a machined part were set by the reach of a human hand. A machinist turned a leadscrew, watched a dial, and cut. The machine could physically reach a large envelope, but nobody could hold position across a long contour while cranking. Accuracy fell away as the cut got longer. Travel existed. Controlled travel did not.

The work of John T. Parsons with the U.S. Air Force on helicopter rotor blades and aircraft skins forced the issue. Templates and manual tracing could not hold the curve. The answer was to let numbers drive the tool, and the first numerical control systems fed positioning data from drilled paper tape into a modified milling machine. Tape stretched. Tape tore. But the machine followed the numbers.

That is the pivot point. Before NC, travel was a physical property of the machine. After NC, travel became a programmable quantity. A shop could decide, in code, where the tool went inside the envelope. Every step since then, from punched tape to microprocessors to simultaneous 5-axis motion, has been an expansion of that idea.

So when an engineer asks whether a part fits, the real question is not how big the machine is. It is how much of the envelope the control can coordinate at once, and how accurately.

• 1
  Manual eraReach was large, repeatability was poor on long contours.
• 2
  NC eraPositioning became programmable inside a fixed envelope.
• 3
  CNC eraMultiple axes coordinated in one setup.
• 4
  Current eraSimultaneous 5-axis plus Ø400 mm rotary travel.

For decades the standard machine moved on three linear axes, X, Y and Z. That covers a lot of work: plates, brackets, housings, pockets and holes. Where it struggles is a compound angle, an undercut, or a contoured surface that wraps around the part. Those features need the part repositioned by hand, sometimes three or four times.

Each reposition adds error. The operator re-clamps, re-zeroes, and the stack-up of two setups is worse than one. Cycle time grows because the spindle is idle while the part is being moved. This is the practical cost of three-axis work, and it is why axis count matters more than nominal envelope size on complex geometry.

Adding a rotary axis changes the math. A fourth axis indexes the part to a new face without releasing the clamp. A fifth axis tilts the tool or the table so the cutter can reach a surface at a natural angle instead of a grazing one. After that, the tool reaches surfaces in one continuous pass.

GreatLight runs 16 simultaneous 5-axis machining centers, 12 four-axis mills, 27 three-axis machines, and 16 mill-turn centers. A Ø400 mm rotary table handles the round work. The point of that mix is not machine count. It is matching axis count to geometry so a shop does not burn setup time on a part that only ever needed three axes.

The engineering rule is simple: if a surface normal points in more than three directions, or the feature sits behind another feature, plan for a fourth or fifth axis. If every face is orthogonal and open, three axes plus a vise is faster and cheaper.

• 1
  3-axisOrthogonal faces, open pockets, through holes.
• 2
  4-axisIndexed work around a rotary table or tombstone.
• 3
  5-axisCompound angles, undercuts, contoured blades.
• 4
  Mill-turnTurning and milling in one program on one machine.

Travel is the distance an axis can move, but the usable envelope is smaller than the spec sheet. Subtract the vise, the fixture plate, the tool holder and the clearance for a safe retract. On a 750 × 1,150 × 550 mm machine, a part that measures 700 × 1,100 × 500 mm may already be tight once clamping is added.

The numbers to plan against are the ones below. GreatLight's largest envelope is 4,000 × 400 × 150 mm, which suits long extrusions, rails, structural profiles and long plates. Medium envelopes of 750 × 1,150 × 550 mm and 600 × 600 × 600 mm cover most housings and enclosures. Compact envelopes of 500 × 500 × 450 mm and 500 × 310 × 200 mm cover small brackets, manifolds and connector bodies.

Height is usually the hidden constraint. A 4,000 mm machine with 150 mm of Z travel will machine a long rail, but it will not drill deep bores into a tall block. Designers who check only the X dimension get surprised at quote. Check all three, then check the diagonal on a tilted part.

When a part exceeds every envelope, the answer is rarely one big machine. It is usually splitting the part, changing the datum scheme, or choosing a different process such as casting plus finish machining.

• 1
  Long parts4,000 × 400 × 150 mm suits rails and extrusions.
• 2
  Boxy parts600 × 600 × 600 mm and 750 × 1,150 × 550 mm.
• 3
  Small parts500 × 500 × 450 mm and 500 × 310 × 200 mm.
• 4
  Round partsØ400 mm rotary table for indexed diameters.

A large envelope does not automatically mean tight tolerance at the far corner. Thermal growth along a 4,000 mm bed, leadscrew pitch error, and the rigidity of a long overhang all show up as deviation. This is why ±0.005 mm is realistic on well-supported work, not a blanket promise across the whole travel range.

Surface finish follows the same logic. As-machined surfaces land at Ra 1.6–3.2 μm. A controlled finishing pass reaches Ra 0.8–1.6 μm, and the finest work in aluminium or brass can reach Ra 0.2–0.8 μm. Deep pockets with long tools flex, so the finish at the bottom of a 10:1 depth-to-diameter pocket will not match the finish on an open face.

Material choice moves the boundary too. Aluminium 6061-T6, 7075 and 6082 cut cleanly and hold thin walls well. Stainless 316L and 17-4PH work harden, so light radial cuts and constant engagement matter. Titanium TC4 and Inconel push tool wear up and slow the feed, which stretches cycle time on long travel passes.

For engineers, the practical takeaway is to state tolerance and finish per feature, not for the whole drawing. A datum face at ±0.005 mm and a cosmetic surface at Ra 1.6 μm is a normal, sensible split. Putting the tightest callout on every dimension raises cost without adding function.

• 1
  Tight zonesDatum faces, bearing bores, sealing surfaces.
• 2
  Loose zonesClearance holes, cosmetic faces, covers.
• 3
  Hard materialsTitanium and Inconel need slower feeds and more passes.
• 4
  Thin wallsBelow 1 mm in aluminium needs light radial engagement.

Every time a part leaves the fixture, the shop pays twice: once for the setup labor, and once for the tolerance budget lost at the new datum. Historically, that is why complex parts were expensive. The machine could reach the geometry, but only in pieces.

Five-axis work collapses that. A single fixture, a single datum, and the tool reaches five sides. The trade is programming time and a fixture that will not vibrate when the table tilts. On a part with six features on four faces, one 5-axis setup usually beats four 3-axis setups on both cost and consistency.

Fixtures matter more than machine specs on repeat work. Soft jaws bored in place, a self-centering vise, or a modular plate with dowel pins will hold position better than a generic clamp. For 10,000+ part runs, a dedicated fixture pays back within the first batch.

The counter-case is real. If the part is a flat plate with holes, three axes plus a vacuum table is faster than a 5-axis program. Axis count is a tool, not a ranking.

• 1
  One setupBetter tolerance stack-up, less handling.
• 2
  Soft jawsBored in place for round or irregular parts.
• 3
  Modular platesDowel pins and stops for repeat batches.
• 4
  Simple plates3-axis with vacuum is often the fastest route.