Screw thread acceptance can be checked three fundamentally different ways — by functional pass/fail attribute gauging, by optical profile comparison, or by full dimensional/geometric measurement on a CMM. Each answers a different question, and most quality programs end up using more than one.

1. Go/No-Go (Functional) Gauging

How it works: A GO gauge (ring for external threads, plug for internal) must thread fully onto/into the part by hand under its own weight or light torque, over the full length of engagement. A NO-GO (or “HI”) gauge must not engage more than a turn or so. This implements the Taylor principle: the GO member checks the thread’s maximum-material (functional/virtual) envelope over its full length, while the NO-GO member checks that no single point exceeds the minimum-material limit.

• Measures: Nothing numeric — it’s an attribute pass/fail against the functional envelope, not a dimensional readout.
• Calibration: Working gauges are periodically checked against master setting plugs/rings, or ring gauges are verified indirectly using the three-wire method.
• Strengths: Fast, cheap, requires minimal training, directly answers “will this part assemble and function” — which is arguably the real engineering question for a screw thread.
• Weaknesses: No trend data, no magnitude of deviation, gauge wear silently erodes accuracy over time, provides no diagnostic insight into why a thread failed, and a separate gauge set is needed per size/class.

2. Optical Comparators (Shadowgraph / Profile Projectors)

How it works: Collimated light projects a magnified shadow of the thread profile onto a screen, where it’s compared to a standard overlay chart (10x–100x) or measured directly with crosshairs/digital software on video-based comparators.

• Measures: Actual values — pitch diameter, major/minor diameter, flank angle, crest/root condition, visible form defects, and localized lead error.
• Strengths: Combines visual defect inspection with real dimensional data; excellent for diagnosing why a part failed (flank angle error, chipped crest, form irregularity); non-contact, so delicate or soft threads aren’t damaged; good for first-article inspection and root-cause work; moderate cost.
• Weaknesses: Fundamentally a 2D silhouette at one rotational position (multiple readings needed for a fuller picture), operator-dependent alignment to the overlay chart, awkward for internal threads (often needs special fixturing or a dedicated optical thread comparator), and slower per-part than go/no-go.

3. CMM Thread Measurement

How it works: A touch-trigger or scanning probe collects coordinate points on the flanks, crest, and root; thread-analysis software fits a helical model to the point cloud to extract the full parameter set.

• Measures: Major/minor diameter, simple and virtual pitch diameter, lead/pitch error over the full engaged length, flank angle on each side independently, taper, out-of-roundness, and total form deviation — for both internal and external threads.
• Strengths: Most complete and traceable dataset available; true variable data suitable for SPC; single setup can relate thread geometry to other part datums; highest achievable accuracy (sub-micron on a good machine in a controlled environment).
• Weaknesses: Highest capital cost, slowest cycle time of the three (dedicated thread-measuring CMMs narrow this gap but add their own cost), needs skilled programming, probe access can be limited in fine-pitch or deep threads, best accuracy requires a temperature-controlled lab, and results still need engineering interpretation to judge functional fit — a part can be “in spec” on paper and still be assessed against Taylor’s principle separately.

Side-by-Side Comparison

	Go/No-Go	Optical Comparator	CMM
Output type	Pass/fail (attribute)	Dimensional + visual	Full dimensional/geometric (variable)
Speed	Seconds	Minutes	Minutes–tens of minutes
Relative cost	Low	Moderate	High
Operator skill	Minimal	Moderate	High (programming)
Internal threads	Yes (plug gauge)	Difficult	Yes
Diagnostic value	None	Good	Best
Typical setting	Shop floor, 100% screening	Incoming/first-article, root cause	Lab, final certification, R&D
Governing standards	ASME B1.2, ISO 1502	—	ASME B89.1.x, internal thread-fit software standards

Practical Guidance

In most manufacturing environments these methods are complementary rather than competing:

• Go/No-Go gauges handle high-volume production screening at the machine, because they’re fast and directly validate functional assembly.
• Optical comparators come in when a go/no-go gauge rejects a part and you need to know why — or for first-article and periodic in-process checks where some dimensional feedback is useful without full CMM cycle time.
• CMM measurement is reserved for final certification, critical/aerospace-grade threads, new-product qualification, calibrating the go/no-go gauges themselves, and any situation requiring statistically trackable, traceable data.

A common setup is: go/no-go for 100% floor inspection, optical comparator for first-piece and troubleshooting, and CMM for periodic capability studies and gauge calibration — giving speed, diagnostic insight, and traceability without paying CMM cycle times on every part.