Go no go gauges: how a plain limit gauge judges a size

A gauge that carries the limits

A plain limit gauge is a fixed-value checking tool with no scale: its measuring faces are made to the two limit sizes of the hole or shaft, and it controls the boundary conditions of the feature. It can only say yes or no — it never tells you the size. For a single shaft on a bench that sounds like a loss; on a pallet of two hundred identical pins it is the fastest honest answer there is.

GB/T 1957-2006《光滑极限量规 技术条件》governs the terminology, tolerances, requirements, inspection, marking and packaging of these gauges, for hole and shaft sizes up to 500 mm in tolerance grades IT6 to IT16. The forms and dimensions are standardised separately in GB/T 10920-2008.

Go end and no-go end

A gauge for a hole is a plug gauge: its measuring face is an outside cylinder, and the cylinder whose diameter equals the hole’s minimum limit of size is the go plug, while the one that equals the hole’s maximum limit of size is the no-go plug. A gauge for a shaft is a ring gauge or a snap gauge (caliper gauge): the ring at the shaft’s maximum limit of size is the go end, the one at the minimum limit of size is the no-go end. The go end and no-go end are used as a pair, marked T (通) and Z (止).

The judgment is fixed: if the go end passes and the no-go end does not, the workpiece is acceptable. If the no-go end goes in — even part-way, even once — the workpiece is rejected. There is no “almost”.

Taylor principle: what each end really checks

Form errors make one number insufficient. A shaft that is thicker in the middle than at its ends can have every cross-section inside the limits and still not fit a straight bore. The limit size judgment principle, called Taylor’s principle, closes that hole with two conditions: the mating effect size must not cross the maximum-material boundary — a shaft must not exceed it, a hole must not fall below it — and the local actual size at any position must not cross the least-material boundary. For a shaft: d_fe ≤ d_max and d_a ≥ d_min; for a hole: D_fe ≥ D_min and D_a ≤ D_max.

The two ends of the gauge divide the work between them. The go end checks the effect size, so in theory it is a full-form gauge — a complete cylinder as long as the mating length — at the maximum material size. The no-go end checks the local actual size, so in theory it is a two-point contact, not a full form, at the least material size.

Real gauges may depart from the ideal when the form error cannot affect the fit. Large holes and shafts are commonly checked with non-full-form go gauges simply because a full-form gauge of 300 mm would be a handful of steel nobody could hold straight.

The usual forms

Plug gauges come as full-form plugs, non-full-form plugs, sheet (flat) plugs and ball-end rod gauges; shaft gauges come as ring gauges and snap (caliper) gauges. GB/T 1957-2006 Annex B recommends forms by size: for example, the go end for a hole up to 18 mm is a full-form plug, for a shaft up to 18 mm it is a ring gauge (snap gauge second), and for shafts over 18 to 100 mm a snap gauge; ball-end rod gauges serve holes above 315 mm.

Gauge tolerances T and Z, wear, master gauges

A gauge is itself made with tolerances, set so that the gauge cannot accept a bad part or reject a good one more often than intended. Its manufacturing tolerance T and its position element Z are taken from the workpiece tolerance, sized and placed inside the workpiece tolerance zone; the values are read from the tables of GB/T 1957-2006. Two placements matter in daily work:

• The go gauge’s tolerance band sits inside the workpiece tolerance zone, offset from the maximum material size by the position element Z, and it carries a wear allowance: the wear limit coincides with the workpiece’s maximum material size. A go gauge worn to that size is done.
• The no-go gauge’s tolerance band starts at the least material size and goes into the workpiece tolerance zone. It gets no wear allowance — only a reject part ever passes it, so it hardly wears.

Because a snap or ring gauge cannot be measured with an ordinary caliper in a meaningful way, shaft gauges are checked with master gauges: TT (master-go) and ZT (master-no-go) control the sizes of a new working gauge, and TS (master-wear) checks a working go gauge against its wear limit — if the TS master goes on, the working gauge is worn out. A master gauge is made to half the manufacturing tolerance of the gauge it masters.

By use there are working gauges (T and Z, new or little worn, for the operator) and acceptance gauges (worn working gauges, for the inspection department) — the current standard does not set a separate tolerance band for acceptance gauges — and the user’s representative should gauge with a go end close to the maximum material size and a no-go end close to the least material size.

Use it without faking the answer

A limit gauge under force tells you nothing. The check is done under the gauge’s own weight: lay the snap gauge over the shaft and let it slide down onto the diameter; guide the plug into the hole squarely and let it sink. Do not push hard, do not screw or rock the gauge through, and never tap it — a no-go end forced past a shoulder proves the strength of your wrist, not the size of the hole. Only when the gauge itself is very light, or when checking a horizontal feature, is a little finger pressure allowed.

If two qualified gauges disagree on one workpiece — one accepts, one rejects — the workpiece counts as accepted. The dispute rule sounds strange until you notice who owns each gauge: acceptance gauges are deliberately worn toward the material limits, so they are the ones a cautious buyer trusts.

Worked example: an Ø10 h6 pin

The drawing says Ø10 h6. From the limit deviation tables of GB/T 1800.2-2020, grade 6 over 6 to 10 mm gives 0/−9 µm, so the limits of size are 10.000 and 9.991 mm; the shaft’s maximum material size is 10.000 mm and its least material size 9.991 mm.

These are the nominal sizes: a real gauge is not cut exactly at the limits. The go end is offset by the position element Z and given its manufacturing tolerance T and a wear allowance, the no-go end its T — all from the GB/T 1957-2006 tables. The pin in the KELIBRON scene measures 9.996 mm: the go end slides over it (9.996 < 10.000), the no-go end stops (9.996 > 9.991), and the pin is accepted.

Gauging or measuring?

A limit gauge is the extreme case of the acceptance rules for plain sizes. Its go wear limit sits at the maximum material size and its no-go band starts at the least material size, so its decision boundary is the limits of size — equivalent to “way b” of GB/T 3177-2009, with the gauge tolerance system acting as the built-in guard. When you measure with a caliper instead, the reading is uncertain and the standard answers with the safety margin A and the instrument uncertainty u₁; see acceptance limits. And if the tolerance is tighter than about a tenth of what a caliper can do, the choice between a micrometer and a gauge is a question of quantity: one piece, measure it; ten thousand, gauge them (choosing the instrument).

In KELIBRON

The lab does not have plain limit gauges yet — the family enters with batch G3 of the gauge catalog, and this lesson stays planned until then. The judgment logic it prepares is already practised with instruments: the lab judges a size by comparing the (corrected) reading with the closed tolerance interval, and refuses a verdict when the band is narrower than twice the resolution or the reading is closer than half a division to a limit — the same spirit as the no-go end that must never be argued with. The reject scenarios shaft-reject and blind-depth-reject show what a failed judgment looks like while the gauges themselves are on the bench.