Quality Inspection

What to Measure in a Deep Hole

Five parameters plus the bore profile decide whether the hole works and survives. Each is measured with a different instrument and governed by a different standard.

Typical Tolerances by Process

Know the process capability before you choose the gage — the tolerance band decides which method is even measurable.

Surface Roughness Measurement

Contact profilometers (Taylor Hobson, Mitutoyo) traverse a diamond stylus along the bore axis and record the surface profile. ISO 4287 defines Ra (arithmetic mean deviation) and Rz (maximum height of profile). In deep hole drilling, Rz is often the more critical parameter because it correlates with fatigue initiation sites — for aerospace and medical parts, monitor both Ra and Rz.

Reaching into the hole

The challenge is reach. Standard stylus units measure only near the mouth; portable instruments offer small-bore sensors for bores down to about 2 mm, deep-groove and curved-surface sensors, and extension rods (typically 50 mm each, up to two) that push the traverse deeper. A reference shop-floor tester quotes 0.001 μm accuracy over a 0.005–16 μm range.

Non-contact alternatives

• Optical profilers: non-contact, for softer materials or fragile burnished surfaces where a stylus would scratch.
• Fiber-optic laser scattering: a probe with several sensors and right-angle prisms is pulled through the hole (feed under 5 mm/s) and measures roughness on multiple lines of the bore simultaneously at full depth; eccentricity points are filtered out.
• MEMS cantilever probes: quantitative roughness inside micro bores (down to ~50 μm, e.g., diesel injector spray holes) where destructive sectioning used to be the only option.

Straightness — The Hardest Parameter in Deep Holes

Straightness is the most critical deep-hole quality parameter because deviation grows with depth. Formally, straightness error is the minimum distance between two concentric cylinders that contain the line joining all cross-section center points. Typical tolerance is ≤ 0.05 mm per meter of depth; NASA JSC-67701 calls out 0.025–0.076 mm TIR for precision holes.

Roundness, Lobing, and Cylindricity

Roundness (circularity) measures cross-section variation only; cylindricity combines roundness and straightness, controlling the entire hole surface against a perfect cylinder. In BTA drilling, lobe formation is a known phenomenon — 2 to 8 lobes form on the bore surface from process dynamics between the guided head and the rotating part.

Why two-point gauging misses lobing

A two-point bore gage measures an average diameter across two opposite points — it catches ovality (2-lobe) but is blind to 3-, 5-, and 7-lobe patterns, which are common in rotary deep hole drilling. Multi-point air plugs with an odd number of jets detect lobing; a precision roundness tester is the definitive check.

Instruments

• Roundness testers (Taylor Hobson Talyrond class): precision spindle + stylus tracing the circumference; highest accuracy (±0.02 μm); lab grade.
• Air plugs: 2 jets for two-point out-of-round; odd jet counts for lobing; multi-circuit plugs measure several diameters and runout in one pass.
• Bore gages (dial / digital): shop-floor alternative, ±0.005 mm, good for routine diameter and ovality checks.
• CMM / form tester: full cylindricity over multiple sections down the bore.

Non-Contact Air Gauging

Air (pneumatic) gauging is the workhorse of production bore measurement. An air plug with one or more orifices is inserted into the bore; regulated air escapes through the gap between the orifice and the bore wall. As the bore grows, leakage increases and back-pressure drops — the change in pressure or flow is read on a dial or an electronic display. It is a comparison instrument: the plug is calibrated against master/setting rings and deviations are read as ± from nominal.

What air plugs measure

• Diameter — internal and external, including deep and interrupted bores
• Out-of-round — 2-jet plugs detect two-point ovality; odd jet counts detect lobing
• Straightness / bow — custom plugs traversed along the bore length reveal tight and loose sections (classic for rifle barrels)
• Squareness — nozzles arranged in a “Z” pattern measure bore-to-face squareness
• Averaging — multiple equally spaced nozzles average the size of thin-walled or out-of-round parts
• Taper — measured as the difference of two diameters
• Honing control — the gage senses when the bore reaches size and stops the operation automatically
• Multi-circuit plugs — measure several diameters and runout conditions simultaneously

Bore Gages and the CMM

Shop-floor bore gages

Dial and digital bore gages (±0.005 mm) are the everyday tool for diameter and ovality. Telescoping gages are cheap but involve many manual steps — each one an operator-dependent error source. For deeper work, depth-indexed diametric probes measure bore diameter continuously as a function of depth; rotating the probe 90° and re-passing reveals roundness/ovality, and the approach cuts measurement time from roughly 90 seconds to about 4 seconds per hole while removing operator variation.

CMM strengths and hard limits

A CMM provides full 3D geometry — diameter, position, roundness, and cylindricity across multiple sections, typically ±0.002 to 0.010 mm. Star probes and extended styluses reach deep bores, but with real limits:

• Depth reach: practical probe reach is on the order of a few hundred millimeters; one on-machine apparatus was chosen precisely because a CMM stopped at 300 mm depth in a 700 mm blind hole.
• Stylus deflection: cantilevered styli flex under gravity and contact force, distorting data in high length-to-diameter holes.
• Surface damage: contact force can leave indentations on finely finished bore walls.
• Segmentation: CMM probing is effectively linear interpolation between points — it can miss local shape and positional drift in the middle of a long bore.

When the CMM needs help

For holes deeper than probe reach, treat the CMM as a section-sample check, not full-length proof. Options include mirror-based structured-light scanning that reaches bores down to about 4 mm and stitches internal and external scans into one coordinate system without cutting the part open, and on-machine probes that run the full bore. Blind-hole depth by CMM relies on the drill-point cone angle — a ±5° tool tolerance can shift the calculated depth by roughly ±0.073 mm, so verify the actual cone angle before trusting the trig correction.

Borescope Inspection

Rigid or flexible borescopes provide the only direct visual of the bore wall. Use them to detect:

• Surface tears or scoring from chip evacuation
• Guide pad burnish marks — uneven patterns indicate misalignment or vibration
• Built-up edge deposits on the bore surface
• Cracks or heat discoloration in the transition zone of hardened materials

Perform borescope inspection as part of first-article inspection and whenever a quality issue is suspected. Digital borescopes with recording capability allow trend analysis over time — compare a new bore against a known-good reference image. For small-diameter work, reflective endoscopic panoramic systems inspect bores of 7–17 mm at length-to-diameter ratios beyond 10:1 with micrometer-level defect resolution.

Laser Systems for Bore Geometry

Laser methods measure bore geometry without touching the wall, and several reach depths a CMM cannot. Two families matter in practice.

Laser bore alignment systems

Commercial systems (Easy-Laser XT950 class) align bore centerlines, bearing journals, and drivelines. A reference unit measures bores of 80–500 mm with 0.001 mm resolution over working distances up to 40 m, detects out-of-roundness, ovality, and taper from uneven wear, and documents results for acceptance reports — far faster and more repeatable than string-and-wire methods.

Laser-guided deep-hole probes

• Autocollimation probes (Kyushu University class): a stylus-equipped probe with CCD cameras detects its own position and inclination inside the bore while scanning circumferentially at depth intervals; designed for 100–600 mm holes, it measures diameter, roundness, cylindricity, and straightness in a single scan — resolution ~0.076 μm over a 0.25 mm range, with roundness results tracking a commercial roundness tester within about 1 μm.
• Laser interferometer + pentaprism probe: for small deep holes (17–21 mm up to 1000 mm long), measuring hole shape, deviation, roundness, and roughness.
• Dynamic Abbe-principle diameter measurement: an ultra-precision approach for large-diameter deep bores.

Choosing the Right Method

Inline vs. Post-Process — and Trusting the Gage

Inline inspection

Inline methods catch drift before it scrapes a part: air gages that auto-stop honing at size, on-machine laser and confocal probes that profile the bore and feed error compensation back to the machine, and indirect signals — spindle torque and coolant pressure — that flag chip clogging or tool breakage in real time. Inline data feeds statistical process control (SPC) and turns rework from a firefight into a memory.

Post-process inspection

Offline metrology — CMM, roundness tester, profilometer, full laser scans — gives the most complete geometry but only after value has been added, so a late failure is expensive. Use it for first articles, machine acceptance, periodic audits, and final certification.

Gage R&R — can you trust the number?

Every measurement is Y = T + e: what the gage reads equals the true value plus measurement error. Gage repeatability and reproducibility (GR&R) separates that error into repeatability (same operator, same part, repeated readings) and reproducibility (between operators). Acceptance rule: <10% is excellent, 10–30% marginal, >30% unacceptable — the gage is eating the tolerance. The number of distinct categories (ndc) must be at least 5 for the gage to discriminate parts at all.

• A dial bore gage measuring an Ø18.1 mm bore produced a 17.4% GR&R — marginally acceptable — until analysis showed the contacts were touching a conical surface instead of the bore; switching to an internal micrometer improved R&R to 7.33%.
• Air gauges on identical transmission bores: three of four units passed with GR&R under 5%, while the fourth failed at 81.4% — a worn gauge, not a bad process.
• Gage design is geometry-dependent: a bore gage that passes R&R on a 50 mm bore can be unsatisfactory on a 200 mm bore; non-tipping two-point contact designs cut operator dependence.

Inspection Plan, Documentation & Common Errors

Build the inspection frequency

Sampling logic

Base sample size on process capability and AQL, not convenience. Precision and fatigue-critical holes get 100% inspection; stable general production can move to rational sampling — but keep trend-based monitoring on every hole, because roughness and diameter drift announce themselves long before they exceed tolerance.

Documenting results

• Record instrument, serial number, calibration date, and part/operator for traceability
• Record measured values per depth — mouth, mid-bore, and exit at minimum — not a single “diameter”
• Archive digital borescope images for trend comparison
• Feed every measurement into SPC and compute Cp/Cpk per parameter per feature
• For certified work (AS9102 first articles, NADCAP special process), keep the full package: parameters, calibration, inspection results, and sign-off

Common inspection errors

• Measuring only mouth and exit — misses taper, barrel, and mid-bore drift
• Two-point gage on odd-lobed bores — blind to 3-, 5-, and 7-lobe patterns from BTA dynamics
• Calibrating on a different diameter or depth than the part — invalidates R&R
• Stylus contact on soft or freshly burnished surfaces — scratches the part and reads a false number
• Ignoring temperature — a 1 m steel part changes roughly 11 μm per °C; control part and gage temperature
• One instrument for two tolerance classes — the rule of ten breaks silently
• Borescope without recording — no baseline, no trend data
• Skipping gage R&R until a capability study fails — an expensive way to find out

Keep Reading