A bore measures 25.012 mm on a two-point bore gauge. The drawing calls 25H7, which under ISO 286 (IS 919 in India) is a band of 0 to +21 µm, so the reading sits comfortably mid-tolerance. The GO plug will not enter. The operator concludes the gauge is wrong and finds a different one, which also will not enter.
The gauge is usually right. A GO plug tests the whole feature at once, which is the gauging practice IS 3455 describes, and a two-point instrument tests one diameter at a time. When they disagree, the disagreement itself is the finding: something about the bore's form is outside what the envelope allows, even though every diameter you sampled is inside the size limits.
Work through the causes in this order, because it moves from the cheapest check to the most expensive.
Check this first, always. Swarf, a raised burr at the entry, or dried coolant will stop a gauge in a bore that is otherwise perfect. Wipe the bore and the gauge, deburr the entry, and try again before you conclude anything about the part.
A gauge that has been in someone's hand and a part fresh off the machine are not at the same temperature. Steel moves about 11.5 µm per metre per °C, so 25 mm of diameter shifts roughly 0.3 µm per °C. A 10 °C difference between part and gauge is worth about 3 µm, which is 14% of an H7 band at this size. This alone rarely stops a GO gauge, but stacked on top of a form error it decides the outcome.
Two-point measurement finds ovality if the operator rotates the instrument, and misses odd-numbered lobing almost entirely. A three-lobed bore, common from a three-jaw chuck clamping too hard, can read a constant diameter at every angle and still refuse a plug, because its inscribed circle is smaller than any diameter you measured.
Rotating the bore gauge and watching for a swing tells you about ovality. A constant reading with a gauge that will not enter points at lobing.
A boring bar deflects, so many bores come out slightly larger at the mouth and tighter at the far end. If your bore gauge was used near the entry, which is where it is easiest to use, you measured the largest part of the hole.
Measure at three depths: near the mouth, mid-length, and as deep as the instrument reaches. A difference of a few micrometres along the length explains a GO gauge that starts, goes in 8 or 10 mm, and stops.
| Symptom | Likely cause | Check |
|---|---|---|
| Gauge will not start at all | Burr, dirt, undersize throughout | Clean, deburr, measure at the mouth |
| Enters then stops part way | Taper or bend | Measure at three depths |
| Enters, tight, no single stop point | Roughness or lobing | Compare surface finish, rotate the bore gauge |
| Varies with rotation | Ovality | Rotate and record the swing |
| Fine in the morning, tight by afternoon | Temperature | Record part and gauge temperature |
A gauge rides on the peaks of a surface, and a two-point instrument with a small contact area sits partly between them. On a bore turned to a coarse finish, the peak-to-valley height consumes real clearance: an Rz around 25 µm on a 21 µm tolerance band is not a finish problem, it is a size problem wearing a different name.
If a bore passes the gauge only after honing without any change in nominal size, roughness was the cause the whole time.
Once you know which of the five it is, the fix belongs to the process rather than the inspection. Bellmouth points at bar deflection or a worn spindle. Lobing points at clamping. Roughness points at feed, speed or tool condition. Repeated ovality on thin-walled parts points at fixturing.
Recording which cause it was, rather than only that the part failed, is what turns a rejection into a process correction. A gauge register that notes "GO stopped at 10 mm depth, taper 6 µm over 25 mm" is worth reading six months later. One that notes "reject" is not.
If a GO gauge is refusing parts you believe are good, our calibration service will tell you whether the gauge has drifted, and gauge repair and reconditioning can bring a worn one back to class where the size allows it.