Gear Manufacturing Knowledge Base

What this page is

PinyiGear (东莞市品一机电科技有限公司) is a gear and transmission component manufacturer in Humen, Dongguan, China. We machine to customer drawings: spur, helical, bevel, worm, rack, sprocket, gear shaft, magnetic drive and custom profiles.

This page collects the engineering knowledge our team uses day to day — how to pick a gear type, which steel goes with which heat treatment, what the different accuracy systems mean, how to read an inspection report, and how a drawing becomes a delivered batch. It is written for the engineer on the other side of the purchase order, and for anyone looking up a specific tolerance or process question.

If a specific topic needs more depth, the technical library carries long-form articles on each subject, and the product pages list what we actually cut and grind.

Choosing a gear type

The choice is driven by shaft arrangement, load direction, ratio, noise limit and cost — in that order. Marketing pages usually start from the gear type; engineers start from the shaft layout.

Three practical notes that come up in almost every RFQ:

Noise. Helical teeth engage gradually, so the contact ratio exceeds 1 through the mesh. At the same module and load a helical pair typically runs several dB quieter than a spur pair. If the noise spec is tight and the shafts are parallel, helical is the default answer; if cost dominates and speed is low, spur wins.

Axial force. Helical gears generate thrust. That thrust has to go somewhere — a bearing, or a double-helical (herringbone) arrangement that cancels it. A customer who switches from spur to helical without changing the bearing arrangement has created a failure the drawing does not show.

Worm efficiency. A worm drive gives large reduction in one stage, but efficiency falls as the lead angle shrinks. Self-locking behaviour appears at small lead angles — useful for holding loads, expensive in power loss. For continuous running drives, check the efficiency budget before selecting.

Material and heat treatment

The heat treatment route decides the steel grade, not the other way round. A grade name alone says nothing about performance until it is paired with a treatment.

Three routes cover most of our work:

• Quench and temper (Q&T). Quench, then temper to tempered sorbite. HB 240–320 is the normal window. Ductility and strength together, adequate for slow and medium speed gears whose flanks do not need to be very hard.
• Carburize and quench. A low-carbon grade is held at 920–950 °C in a carbon-rich atmosphere, then quenched. The surface becomes martensitic while the core stays low-carbon and tough — a hard skin over a soft core. This is the route for gears under heavy contact stress.
• Nitride. At roughly 500–570 °C, ammonia-bearing gas drives nitrogen into the surface. Distortion stays low because the temperature is low, but the layer is shallow — 0.3–0.6 mm in normal practice — which limits it to small-module precision gears that need no final grinding.

Common grades and what they are for:

On an identical blank, a carburizing steel runs roughly 1.3–1.5× the cost of a Q&T steel. What that premium buys is contact fatigue life: at equal load, a carburized flank typically holds off pitting one grade better than a Q&T flank.

Distortion is the process risk, not the hardness. Carburizing at 920–950 °C followed by an oil quench leaves residual stress and dimensional movement. The countermeasures we apply are: uniform heating and controlled atmosphere, fixture quenching or press quenching for thin-walled parts, limiting case depth to what the contact stress actually requires, and where it matters, a stress-relief temper before finish grinding. Deep case is not automatically better — it lengthens the cycle, increases distortion and adds cost.

Case depth follows module. Effective case depth is normally specified as a fraction of module and checked on a companion test bar from the same charge. A drawing that states a hardness only, with no case depth and no measurement position, will produce disputes at incoming inspection.

Machining routes

Four processes dominate, and the choice is driven by accuracy class and volume.

The cost curve is not linear. Moving from hobbing to grinding roughly doubles the machining time per piece at small volume; the economics change at volume because grinding setup amortises. A gear specified at DIN 5 with no functional reason for it is paying for accuracy nobody measures.

Hobbing is a generating process — the hob and the blank rotate in a timed relationship, and the tooth form emerges from the envelope of cuts. Shaping uses a reciprocating cutter with the same generating relationship, which is why it handles internal teeth. Grinding after heat treatment is what recovers the distortion introduced by carburizing; a carburized gear that is not ground retains the distortion as an accuracy error.

Accuracy standards

Four systems appear in drawings we receive: ISO 1328, DIN 3962/3967, AGMA 2015 and GB/T 10095. They describe the same physical deviations with different grade numbering, and the grade numbering runs in opposite directions: DIN grade 1 is the tightest, AGMA grade 16 is the tightest.

DIN and ISO grades are numerically compatible for the range normally quoted in commercial work; the tolerance formulas behind them are close enough that a DIN 6 and an ISO 6 are treated as equivalent in practice. AGMA requires a conversion, and the conversion is approximate — it depends on module and diameter. A drawing that mixes systems without stating the conversion basis is a drawing that will be argued about.

The deviations that matter:

• Fp — total cumulative pitch deviation around the whole gear. Drives transmission error at low frequency.
• fpt — single pitch deviation. Affects load sharing between teeth.
• fα — profile form deviation. Governs how the tooth contacts along its height.
• fβ — helix (lead) deviation. Governs end-to-end contact along the face width; the most damaging of the four when it is out, because it concentrates load at one end of the tooth.

A gear can pass Fp and fail on fβ and still be unusable in a quiet drive. This is why an inspection report that lists only one or two parameters is not a report — it is a summary.

Reading an inspection report

Three judgement points are missed more often than any others.

Where the tolerance comes from. Every deviation limit on the report should trace to a grade in a named standard at a named diameter and module. If the report shows a number without saying which grade it was measured against, the number cannot be compared to the drawing.

Which flank was measured. Profile and lead deviations are measured on a specified number of teeth, at specified positions. A gear measured on three teeth out of forty is a sample, and the report should say so.

What the chart shows beyond the number. A total deviation inside tolerance can still show a shape problem — a pressure-angle error at the tip, a lead curve that is not straight. The number passes, the shape predicts noise. Any report worth reading includes the trace, not just the value.

Failure modes

Four modes account for most field failures, and each points back to a different decision.

Pitting. Sub-surface or surface-origin fatigue on the flank, from contact stress above what the material and its heat treatment can carry. Carburizing raises the limit; under-sizing the case depth or specifying a Q&T steel for a carburized duty brings it forward.

Scuffing. Metal-to-metal welding and tearing under high sliding velocity and marginal lubrication. Worm drives and high-speed helical pairs are exposed. It is a lubrication and surface-finish problem more often than a material problem.

Wear. Gradual material loss, accelerated by contaminated oil, inadequate viscosity or a filter that is not doing its job. Slow and steady, and it is the mode most easily prevented by maintenance.

Tooth breakage. Bending fatigue at the root, from overload or from a root fillet that was not cut clean. A single overload event leaves a different fracture face than fatigue; reading the fracture is how the two are separated.

The useful discipline is to name the mode before changing anything. A pitting failure answered with a harder material can simply move the failure to the root.

From drawing to delivered batch

The workflow we run on custom gear projects:

1. Drawing review. We read the drawing against process capability before quoting: module, accuracy grade, material and heat treatment, case depth, inspection basis, quantity. Where something is missing or inconsistent, the question goes back to the customer before a price is issued.
2. Process routing. Machining route, heat treatment route and inspection plan are fixed together, because the required accuracy grade determines whether grinding is in the route.
3. First article. A first piece is measured and the report goes to the customer for approval before the batch runs.
4. Production. Hobbing/shaping, heat treatment, grinding where specified, then dimensional and gear-specific inspection.
5. Inspection report. Each batch ships with measured values against the drawing’s stated grade — Fp, fpt, fα, fβ and any other parameters the drawing calls out.
6. Packing and shipping. Parts are protected against transport damage and packed to the agreed export terms.

A drawing is only as good as the three entries in its material block: grade, heat treatment, hardness range. Absent heat treatment, the supplier guesses and acceptance becomes a negotiation.

Frequently asked questions

What accuracy grade can you hold on a custom gear? Ground gears are held to DIN 5 and tighter where the design requires it; hobbed gears in the DIN 7–8 range. What we quote depends on the drawing’s stated grade, module and size, because the achievable grade falls as diameter and module grow.

Can you work from a sample instead of a drawing? We can, but a sample leaves the material, heat treatment and tolerance basis unstated. Reverse-engineering a sample means reconstructing those from measurement, and the result is only as reliable as the measurements. A drawing with the three material-block entries is always the better starting point.

What is the minimum order quantity? It depends on the process route and whether tooling is required. A standard module with no special cutter is a different proposition from a profile that needs a dedicated hob. Send the drawing and we will state the real number rather than a generic figure.

Do you supply inspection reports with every batch? Yes. Each batch carries measured values against the drawing’s grade, taken on specified teeth and stated positions.

Which steel should I specify for a gearbox gear under heavy load? Carburizing steel — 20CrMnTi for small and medium module, 17CrNiMo6 where the module is large and impact loading is expected. Decide the heat treatment route from the contact stress first, then pick the grade that supports it.

Why does my carburized gear come back out of tolerance? Because carburizing at 920–950 °C followed by quenching moves the part. Distortion is controlled by uniform heating, controlled atmosphere, fixture quenching, and limiting case depth — and it is corrected by grinding after heat treatment. A carburized gear with no grinding in the route keeps its distortion as an accuracy error.

What is the difference between DIN and AGMA grades? DIN grade 1 is the tightest and the number rises as tolerance loosens; AGMA grade 16 is the tightest and the number falls as tolerance loosens. A conversion between them is approximate and depends on module and diameter, so a drawing should state which system and which grade it means.

How do I know the inspection report is complete? It should name the standard and grade, state the diameter, module and number of teeth measured, and include the traces for Fp, fpt, fα and fβ — not just pass/fail values. A report missing the traces tells you the gear passed but not why.

Can you handle small prototype quantities? Yes, and the constraint is tooling rather than batch size. Where a standard cutter covers the profile, a small quantity is straightforward; where a dedicated hob or a shaped cutter is needed, the tooling cost dominates and is quoted separately.

What information do you need to quote? The drawing, the quantity, the inspection basis, and any marking or packaging requirement. Where a drawing is unclear on material, heat treatment or accuracy grade, we ask before pricing.

Do you export, and under what terms? Yes. We work with the usual export terms and can supply the inspection documentation that goes with the shipment. Specific terms are stated on the quotation.

How long does a custom gear project take? It depends on the route. A hobbed part with no heat treatment is short; a carburized and ground part adds heat treatment and grinding to the critical path. The schedule is quoted with the price, not after it.

Contact

Send a drawing for engineering review: get a gear quote. Product scope sits on the products pages; process detail on capabilities.

PinyiGear — 东莞市品一机电科技有限公司, Humen, Dongguan, Guangdong, China.