Reading a Gear Inspection Report: Deviation Categories, Where Tolerances Come From, and Three Judgement Points Often Missed
Gear inspection reports show pitch (fpt, Fp), profile (Fα, ffα, fHα) and helix (Fβ, ffβ, fHβ) deviations per GB/T 10095.1 (ISO 1328-1) or DIN 3962; tolerance values depend on module, diameter, face width and grade.
Quick answer: a gear inspection report carries three families of values — pitch deviation (fpt single pitch, Fp total cumulative pitch), profile deviation (Fα / ffα / fHα) and helix (lead) deviation (Fβ / ffβ / fHβ). Accuracy grades are judged to GB/T 10095.1 (the same document as ISO 1328-1) or DIN 3962, and the lower the grade number, the tighter the accuracy. Work through it in this sequence: verify first that the measuring section and the datum agree with the drawing, then check each allowable value against module, reference diameter, face width and grade. Indexing accuracy is governed by Fp, mesh smoothness and noise by fHα, and load distribution over the face width by Fβ.
1. Allowable values are looked up, not fixed — and that is why two suppliers differ
No individual deviation has one universal limit. The permitted value is read from standard tables, and the inputs are module, reference diameter, face width and accuracy grade. DIN 3962 lists the individual tolerances covering pitch, profile and helix. GB/T 10095.1 is technically identical to ISO 1328-1: it defines the deviations taken on the same tooth flank, together with the values allowed for each. Tooth thickness and backlash sit outside this system, under DIN 3967. An approximate cross-system grade comparison is given in Gear Accuracy Standards Explainedtech-library/33/).
Helix deviation is a useful illustration. Across grades 5~7 the usual limit falls somewhere around 6~16 µm, and where it lands inside that window depends on helix angle and face width — a wider face accumulates more error from an identical angular slope, so its limit is set looser. When two reports disagree, the first question is whether both parties measured the same measuring point on the same datum; only after that is it worth arguing about whether the part is out of tolerance.
2. Deviation families, and the gear behaviour each one predicts
A gear measuring center works tooth by tooth through a finished gear, records pitch, profile and helix deviations, and compares them with a standard accuracy grade; the resulting record is the inspection report. Every family on it maps to a distinct performance characteristic:
| Deviation family | Symbols | Meaning | Behaviour affected |
|---|---|---|---|
| Pitch deviation | fpt / Fp / fpk | Adjacent pitch deviation, cumulative deviation over one revolution, cumulative deviation over k pitches | Indexing accuracy, uniformity of transmission ratio |
| Profile deviation | Fα / ffα / fHα | Total profile deviation, form deviation, slope deviation | Mesh smoothness, noise, contact stress |
| Helix (lead) deviation | Fβ / ffβ / fHβ | Total deviation across the face width, form, slope | Load distribution across face width, edge loading, tooth-end chipping |
| Radial runout | Fr | Maximum radial variation of the probe during one revolution of the gear | Backlash fluctuation, noise |
That table accounts for half of what a report contains. The remaining half is knowing how tight these values ought to be held.
3. Three details that are easy to miss when judging a report
Fp and fpt describe different problems. Two neighboring teeth are what single pitch deviation fpt looks at, whereas total cumulative pitch deviation Fp is produced by error building up across a complete revolution. During hobbing, indexing error in the table worm gear pair accumulates into Fp and appears as uneven noise at particular angular positions. In applications demanding a uniform transmission ratio — indexing mechanisms, multi-station rotary tables — Fp is frequently the tighter requirement, and a report listing fpt alone hides the cumulative error entirely.
Profile slope deviation fHα can be restated as a pressure angle error. It measures how far the flank is sloped away from the theoretical involute, and that quantity is convertible into a pressure angle deviation. When the tip and the root both depart from nominal, one outward and one inward, the flank takes on a crowned or hollowed form; the value that captures this is the form deviation ffα, which is a direct reflection of how the grinding wheel was dressed. Bringing these two quantities back under control is the central task of post-heat-treatment grinding.
Datum and measuring section have to be declared. Take one gear and measure it from the bore, then measure it from the journal — the two sets of results may land in different grades. Reports ordinarily state where the measuring section sits; for a helical gear the normal section at mid-face-width is the usual choice. Where the drawing calls out transverse tooth thickness and the report quotes normal, a numerical mismatch between them is expected — convert one to the other before drawing any conclusion.
4. What the drawing and the purchase order must carry
- Fix the grade: DIN 4-5 for high-speed precision drives, 7~8 for general industrial gearboxes, roughly grade 9 where speeds are low and loads are heavy.
- Call out every item: pitch, profile and helix all need to be stated; naming just one leaves the other two uncontrolled.
- Declare the datum: state whether measurement references the bore or the journal, and state the measuring section, so that buyer and supplier read the drawing the same way.
- Handle tooth thickness separately: its tolerance, together with backlash, follows DIN 3967 and is written as its own requirement rather than folded into the accuracy grade.
- Settle how inspection will be done: gear measuring center or mesh inspection, report supplied with the shipment or issued by a nominated third party — agree this in the technical requirements before the order.
- Put heat treatment into the process first: distortion from carburizing and quenching changes the final accuracy, so leave grinding allowance and measure after grinding; never assume the pre-heat-treatment grade survives to the finished part.
5. How much a report is worth depends on the measuring equipment
Report credibility is bounded by inspection capability: an inaccurate measurement cannot support an accurate assessment. A gear measuring center may carry measurement uncertainty down to 0.5 µm, yet the profile allowance for a DIN grade 3 gear is itself measured in microns — where equipment is not capable enough, a recorded “pass” carries no weight. CMMs, hardness testers, mesh testers and roundness testers each cover one part of the job; the division of labour between them is set out in Inspection Methods and Instruments for Precision Gearstech-library/16/.
6. PinyiGear capabilities
PinyiGear cuts spur and helical cylindrical gears to drawing, with module range, outside diameter and ground accuracy per the values stated on the drawing; bevel gears are produced to the GB/T 11365 grade stated on the drawing, worm gear sets follow the tooth forms specified on the drawing, and the materials we work in most often are 20CrMnTi carburized and quenched, 42CrMo quenched and tempered, and 38CrMoAl nitrided, with hardness per the drawing. Our gear measuring centers, CMMs, hardness testers and mesh testers allow us to check pitch, profile, helix and hardness and to release reports — see Custom Spur Gears, Inspection and Quality System and Machining Capabilities Overview.
Share your drawings and your operating conditions — module, tooth count, accuracy grade, heat treatment requirements, application — and once the drawing review is complete, our engineering team will come back with the process route, the inspection plan and a quotation. Contact PinyiGearget-a-gear-quote/) or go to the PinyiGear website). An inspection report only has something meaningful to compare against when the drawing itself states the grade and the datum without ambiguity.