External CNC Internal CI Grinding: How Micron Accuracy Is Actually Made

• 1
  What External CNC Internal CI Grinding Actually Means
• 2
  Wheel Contact Geometry and Why Bore Size Changes Everything
• 3
  Spark-Out, Thermal Drift and the Real Micron Limit
• 4
  Choosing Between Composite Grinding and Ordinary CNC Turning
• 5
  In-Process Gauging and How the Machine Holds Size
• 6
  Composite Grinding vs Turning vs Separate Grinding

An external CNC internal CI machine grinds cylindrical surfaces from two directions. The internal spindle reaches into a bore with a small wheel. The external wheel rides the outside diameter. CI here stands for the composite internal arrangement: both heads sit on one bed, share one control, and index the same workhead. That shared structure is the whole point.

The practical gain is datum continuity. When a bore and its mating outer diameter are ground without re-chucking, the relationship between them stays inside the machine's own repeatability instead of depending on fixture transfer. For a hydraulic sleeve with a 40 mm bore and a 55 mm outer diameter, concentricity is no longer a stack-up of two operations. It is one measurement.

That is what the phrase precise to microns means in real shop language. It is not a claim about the best number a machine can hit on a good day. It is a statement about how little the number moves across a batch of 500 parts.

Composite grinding is not a universal upgrade. It pays off on parts where the bore and the outer diameter must stay aligned, and it costs setup time on parts where only one surface matters. Sorting that out early saves money later.

An internal wheel is small. On a 30 mm bore, the wheel may be Ø20 mm or less. A small wheel has a low surface speed at the same spindle rpm, so the machine pushes rpm higher to keep cutting speed in range. Typical vitrified CBN runs at 60–80 m/s for internal work, and aluminum oxide sits lower, around 30–45 m/s.

Contact arc length is the other half of the story. The internal wheel touches the bore over a short, highly curved arc, so the same infeed removes more material per unit of contact area and generates heat faster. External grinding spreads the load over a longer, flatter arc. Two surfaces, two thermal behaviors, one bed.

Wheel diameter shrinks as it wears. On an internal spindle a 10 mm reduction is a large percentage change in surface speed, so the control has to compensate or the finish drifts across the run. External wheels tolerate the same absolute wear with far less effect.

This is why an external CNC internal CI machine is not simply two grinders bolted together. The internal head carries the tighter process window, and the control strategy has to respect that asymmetry.

The last few microns come off during spark-out, when infeed stops and the wheel keeps passing the surface. Spring-back in the wheel, spindle and workpiece relaxes gradually. Cut spark-out short and the part springs back oversize; run it too long and cycle time climbs for no gain.

Thermal drift is the harder problem. Grinding puts 70–90% of its energy into the workpiece as heat. Coolant carries most of it away, but not all. As the part warms it grows, so the machine cuts a bore that measures correct when hot and shrinks when it cools.

The usual fix is to let the machine reach thermal steady state before the first finish pass, then hold that state. A cold machine and a warm machine do not produce the same bore. This is the single most common reason a process that worked in the sample shop drifts in production.

On our own cells we hold ±0.005 mm (±0.0002 in) and finishes between Ra 0.2–0.8 μm on ground cylindrical features. Those numbers come from a settled thermal state, not from a cold start.

Grinding is a finishing process. It removes small amounts of material at low stock allowance, typically 0.1–0.3 mm on diameter for a hardened internal bore. If a part arrives with 2 mm of stock on the bore, the grinder is the wrong tool. Turn or mill it first.

Hardness is the usual deciding factor. Above roughly 45 HRC, turning becomes slow and tool life collapses, so grinding wins. Below that, a mill-turn center with a good boring bar can often hold the required tolerance on its own and skip a whole operation.

Surface finish matters too. If the drawing calls for Ra 1.6–3.2 μm and a bore tolerance of ±0.02 mm, turning is usually enough. When the callout drops to Ra 0.8–1.6 μm or tighter, with a roundness requirement attached, grinding is the safer route.

Geometry decides the rest. A short bore with a large diameter is friendly to a boring bar. A long, small-diameter bore with a tight roundness callout is grinding territory, because a boring bar will deflect and the grinder will not.

A modern composite grinder does not grind to a position and hope. It measures. In-process gauging rides the surface during the cycle and feeds the actual size back to the control, which adjusts infeed before the final passes.

This changes the nature of the tolerance. Instead of holding size by controlling wheel position, the machine holds size by controlling the measured result. Wheel wear, thermal growth and dressing shifts get absorbed rather than accumulated.

Gauging has limits. It needs a clean, stable surface to read, and it needs the part to be rigid enough that the measurement is not picking up vibration. On thin-wall sleeves, contact gauging can push the wall, so the process needs lower gauging force or a different strategy.

For the parts we run, every batch ends with 100% inspection before shipment, and dimensional reports are available on request. The in-process gauge keeps the process centered; the final inspection proves it stayed there.