Sep 29, 2026 Leave a message

Why Chain Drives Need 17 Teeth: Chordal Action Explained

Why Chain Drives Need 17 Teeth: Chordal Action Explained

news-600-600

Sprockets in stock and custom tooth counts

Target keyword: chain drive chordal action

Secondary: minimum sprocket teeth, sprocket tooth count selection, polygonal effect chain

Meta description: Chordal action causes 6% speed variation on a 9-tooth sprocket and 1.7% on 17 teeth. The maths behind the 17-tooth rule, plus maximum driven tooth count and why it exists.

Chain does not wrap around a sprocket in a circle. It wraps in a polygon.

That single geometric fact produces speed variation, vibration, noise and accelerated wear, and it is the reason every chain drive guide tells you to use at least 17 teeth on the driver. Understanding the mechanism lets you decide intelligently when to break the rule.

What chordal action is

Each chain link is a rigid straight segment. When the chain engages a sprocket, the links form a polygon inscribed in the pitch circle, not a smooth arc.

As the sprocket turns, the effective radius at which the chain leaves the sprocket changes continuously. It is maximum when a roller sits at the top of the sprocket, and minimum half a tooth later when the chain spans a chord between two rollers.

Since linear speed is angular velocity times radius, a varying radius at constant rpm gives varying chain speed. Every single tooth engagement, forever.

The standard formulation:

Even when the sprockets rotate at constant speed, chain speed is not steady, because the effective sprocket radius changes through each tooth engagement. Chordal action is governed by the number of teeth:

>

Ratio of speed change = (V max − V min) / V max = 1 − cos (180°/N)

Expanded, the same relation reads:

v_max = ωR

v_min = ωR · cos(180°/N)

Δv / v_avg = 1 − cos(180°/N)

A note on formulas: a variant, δv = (1/cos(π/z) − 1) × 100%, normalises against minimum rather than maximum speed. The two diverge slightly at low tooth counts. We use the first form throughout.

The numbers

Published data points from the sources above:

news-1843-1306

Figure 1. Chain speed variation against sprocket tooth count.

Teeth

Speed variation

9

≈ 6 %

17

≈ 1.7 %

19

≈ 1.38 %

25

≈ 0.8 %

Evaluating the formula across the range gives the shape of the curve. (These intermediate rows are our own arithmetic on the published formula, not copied from a manufacturer table - the endpoints agree with the published values above.)

Teeth

180°/z

1 − cos(180°/z)

Speed variation

9

20.000°

0.06031

6.03 %

11

16.364°

0.04051

4.05 %

13

13.846°

0.02906

2.91 %

15

12.000°

0.02185

2.19 %

17

10.588°

0.01703

1.70 %

19

9.474°

0.01364

1.36 %

21

8.571°

0.01117

1.12 %

25

7.200°

0.00789

0.79 %

30

6.000°

0.00548

0.55 %

The curve falls steeply and then flattens. Going from 9 to 17 teeth cuts variation by a factor of 3.5. Going from 17 to 30 teeth gains you another factor of 3 - real, but with diminishing returns against the cost of a much larger sprocket.

Seventeen teeth is where the curve turns. Below it you pay heavily; above it you pay progressively more sprocket for progressively less benefit.

What 6% speed variation actually does

On a 9-tooth sprocket the chain accelerates and decelerates by 6% of its mean speed, several times per revolution. Three consequences follow.

Inertial loading. Accelerating the chain mass adds tension beyond the transmitted load. The force scales with the square of speed, so a fast drive with low tooth count is punished twice.

Vibration and noise. At 9 teeth the result is significant vibration. Below 17 teeth the variation exceeds roughly 3%, which is generally described as objectionable vibration and noise.

Articulation wear. Fewer teeth means each link rotates through a bigger angle as it enters and leaves the sprocket. With very few teeth, each link undergoes a larger angular articulation per engagement cycle, which accelerates pin-bushing wear.

That third one is the quiet killer. Pin-bushing wear is what elongates chain, and chain elongation is what eventually ends its life.

Minimum tooth count: what the sources say

Recommendation

Basis

Driver sprocket 17 teeth or more; driven sprocket under 120 teeth

General drive design practice

ANSI B29.1 specifies a minimum of 17 teeth on the driver for standard drives, though 11–13 teeth are permissible for very slow, lightly loaded applications

ANSI B29.1

ISO 606 recommends ≥ 17 driver teeth, with an odd count to distribute roller-seat wear evenly

ISO 606

An absolute floor of 15 teeth, with 13 as the practical minimum in constrained designs

Chain manufacturer guidance

The most useful version graduates the minimum by drive speed:

news-1797-1230

Figure 2. Minimum tooth count recommendations by drive speed class.

- Slow speed drive (drip type lubrication) - 12 teeth

- Medium speed drive (oil bath type lubrication) - 17 teeth

- High speed drive (pumped lubrication) - 25 teeth

This is the answer to "can I use fewer than 17 teeth?" Yes - at low speed. The inertial penalty from chordal action scales with speed, so a slow drive tolerates variation that would destroy a fast one. A 12-tooth sprocket on a slow conveyor is a reasonable engineering choice. The same sprocket at 1,000 rpm is not.

Use an odd number of teeth

ISO 606 and general drive practice both recommend odd tooth counts on the driver.

Most drives use an even number of pitches in the chain. Pairing that with an odd number of driver teeth ensures wear distributes uniformly over both the chain and the sprocket teeth.

Chain lengths are almost always even (an odd length needs an offset link, which is weaker). With an even chain and an even sprocket, the same chain links repeatedly meet the same sprocket teeth, and wear concentrates. An odd tooth count breaks the repeat cycle so wear distributes around both components.

It costs nothing. Choose 17, 19, 21, 23, 25 rather than 18, 20, 22.

Maximum driven tooth count - and why it exists

The published caps differ slightly:

114 maximum on any driven sprocket

fewer than 120

not more than 120

The reasoning is the same in all three, and it is about worn chain, not new chain:

For a given elongation of chain due to wear, the working pitch diameter of the chain on the sprocket increases relative to the nominal pitch diameter - the chain assumes a higher position on the sprocket tooth. Allowable safe chain wear is generally taken as around 2% elongation over nominal length."

As chain wears, pitch grows. A longer pitch cannot sit in the tooth pockets at the original radius, so the chain rides progressively higher up the teeth. On a small sprocket this is tolerable. On a very large one, the accumulated error around the wrap is enough for the chain to climb out entirely.

Quantified: a 2% elongated chain on a 120-tooth sprocket has already drifted by 1.2 pitches relative to the sprocket circumference.

One formula unifies the different published caps:

A simple formula for determining how much chain elongation a sprocket can accommodate is 200/N, expressed as a percentage, where N is the number of teeth on the largest sprocket."

Check it: 200/114 = 1.75%, 200/120 = 1.67%. Both land near the 2% allowable wear figure. The "cap" is simply whichever tooth count still tolerates the elongation you plan to allow before replacement.

This also means the limit is yours to set. If you replace chain at 1% elongation, 200/N = 1 gives N = 200 teeth. If you run to 2%, stay at 100.

Two more rules worth knowing

Sum of teeth ≥ 50. Good practice is to keep the sum of teeth at 50 or above where both sprockets run on the same chain - on a 1:1 drive, that means 25 teeth each. A 1:1 drive with two 17-tooth sprockets satisfies the minimum on each but sums to 34, so both ends suffer chordal action simultaneously.

Maximum ratio 7:1 per stage. The maximum recommended ratio for a single reduction is 7:1; above that, a double reduction is preferred. With shock loading the figure tightens to about 5:1.

Centre distance interacts with chordal action. From CEMA's conveyor chain material: "if the shaft center distance is a common multiple of the chain pitch, chordal action is small. But if the shaft center distance is multiple of the chain pitch +0.5 pitch, chordal action increases." Setting centre distance to a whole number of pitches lets the two sprockets' chordal variation partly cancel; a half-pitch offset makes them add.

Selection sequence

Ratio - keep within 7:1 per stage (5:1 with shock)

Driver teeth - 17 minimum for general drives; 12 acceptable at low speed with drip lubrication; 25 for high speed with pumped lubrication

Odd count for the driver

Driven teeth - check against 200/N for your elongation replacement threshold

Sum of teeth ≥ 50

Centre distance at a whole number of pitches

FAQ

Can I use a 15-tooth sprocket?

At low speed, yes - 15 is widely accepted and 13 is workable in constrained designs. Speed variation at 15 teeth is 2.19%, versus 1.70% at 17. Judge it against your speed and lubrication method.

Does chordal action affect the large sprocket too?

Yes, but far less - variation falls with tooth count, so a 60-tooth sprocket contributes almost nothing. The small sprocket dominates.

Why is the driven sprocket capped at all?

Worn chain rides higher on the teeth. Past a certain tooth count the accumulated drift lets the chain climb off. Use 200/N to set your own limit based on your replacement policy.

Do toothed belts have chordal action?

Yes, but much smaller. Published figures put it generally under 0.6%, though combined with pulley deflection and pitch errors it can reach 2–3%.

Does a bigger sprocket always give longer chain life?

Up to the elongation limit, generally yes - lower chordal action, lower articulation angle, lower tooth loading. The trade is cost, weight and space.

Tianjin OuNaiDa Transmissions Technology manufactures sprockets to ANSI and ISO standards in stock and custom tooth counts, with hardened teeth where the application needs them. Send us your ratio, speed and centre distance and we will work the tooth counts with you.

Send Inquiry

whatsapp

Phone

VK

Inquiry