The Involute Gear Tooth Profile

Why the involute profile matters

Two shafts have to run at a rock-steady speed ratio. If the tooth flanks are shaped arbitrarily, the driven gear speeds up and slows down within every tooth engagement — the meshing throws off a torque ripple and a whine that destroys precision machinery and hammers bearings. The demand that the ratio stay perfectly constant while teeth roll and slide against each other is the fundamental law of gearing: the common normal to the two flanks at the contact point must always pass through a single fixed point on the line of centers, the pitch point. Any curve that satisfies this is called a conjugate profile.

The involute is the conjugate curve that engineering converged on, and it beats the alternatives (cycloidal, circular-arc/Novikov) on the properties that matter in a factory:

• Constant velocity ratio. The line of action is a fixed straight line, so torque direction and speed ratio never fluctuate through a mesh cycle.
• Center-distance tolerance. Move the shafts apart by wear, thermal growth, or sloppy assembly and the ratio does not change — only backlash grows. No other common profile does this.
• Single cutter per module. A straight-sided rack or hob of the right module and pressure angle generates every gear in that family — 12 teeth to 200 teeth — because the involute is generated by rolling that straight rack. This is why involute tooling is cheap and universal.
• Interchangeability. Any two involute gears of the same module and pressure angle mesh correctly, regardless of tooth count.
• Straight-line tooth force. The load always acts along the line of action at the fixed pressure angle, which simplifies bearing and shaft force analysis.

The result: the involute is the tooth form in virtually all automotive transmissions, industrial gearboxes, wind-turbine drivetrains, robotics reducers, and clocks. Cycloidal teeth survive mainly in horology and cycloidal-drive reducers, where their specific advantages (low sliding, high contact ratio) outweigh their intolerance of center-distance error.

How the involute is generated, step by step

Picture a spool of radius r_b — the base circle — with a string wound around it and a pencil tied to the free end. Hold the string taut and unwind it. The pencil traces the involute of that base circle. Every point on that curve has two defining properties, and both are the source of the involute's magic:

1. The taut string is always tangent to the base circle. At any instant the unwound string is a straight line touching the base circle at one point.
2. The string is always perpendicular to the curve it draws. Because the pencil momentarily rotates about the tangent point, the string is the normal to the involute at the pencil's position, and its length equals the arc of base circle already unwound.

Now mesh two involute gears. The unwound "strings" of both gears are the same straight line — a common internal tangent to the two base circles. This line is the line of action. Follow the mesh:

1. The contact point between two teeth always lies on this line of action.
2. Because each tooth's normal is the line of action, the common normal at contact is fixed in space — it never wobbles as the gears turn.
3. A fixed common normal crosses the line of centers at one fixed point, the pitch point. The fundamental law of gearing is satisfied automatically, for the whole engagement, not just at one instant.
4. The two pitch circles (radii r₁, r₂) are the imaginary circles through the pitch point that roll on each other without slipping. Their radii set the velocity ratio.
5. The angle between the line of action and the common tangent to the pitch circles is the pressure angle φ. It is fixed by the geometry: r_b = r · cos φ.

Because the pitch point is fixed, the velocity ratio is

i = ω₁ / ω₂ = r₂ / r₁ = r_b2 / r_b1 = z₂ / z₁

where ω is angular velocity (rad/s), r is pitch radius, r_b is base-circle radius, and z is tooth count. Every one of those quantities is a fixed constant of the cut gears, so the ratio cannot fluctuate. That is conjugate action, expressed in one line.

The parametric equation of the involute

In polar-ish form, with the string unwound through an angle t (radians) from the base circle of radius r_b, the traced point is:

x = r_b(cos t + t·sin t)  ,  y = r_b(sin t − t·cos t)

Here t is the roll (unwind) angle in radians, r_b the base-circle radius (mm), and (x, y) the Cartesian coordinates (mm) of the flank point. The distance from the origin is r = r_b·√(1 + t²), and the corresponding pressure angle at that radius is α = arccos(r_b/r). The involute function inv(α) = tan α − α is the workhorse for tooth-thickness and span-measurement calculations.

Worked example: sizing a 20° spur pair

Design a reduction with pinion z₁ = 20 teeth, gear z₂ = 60 teeth, module m = 4 mm, pressure angle φ = 20°. Standard full-depth proportions.

Quantity	Formula	Pinion (z=20)	Gear (z=60)
Pitch diameter d	d = m·z	80.0 mm	240.0 mm
Base diameter db	db = d·cos φ	75.18 mm	225.53 mm
Addendum a	a = 1·m	4.0 mm	4.0 mm
Dedendum b	b = 1.25·m	5.0 mm	5.0 mm
Outside (tip) dia. da	da = d + 2m	88.0 mm	248.0 mm
Circular pitch p	p = π·m	12.566 mm	12.566 mm
Base pitch pb	pb = π·m·cos φ	11.809 mm	11.809 mm

The center distance is the sum of pitch radii:

C = (d₁ + d₂) / 2 = m(z₁ + z₂) / 2 = 4·(20 + 60)/2 = 160 mm

The velocity ratio is i = z₂/z₁ = 60/20 = 3:1 exactly. Drive the pinion at 1500 rpm and the gear turns at 500 rpm — constant, all the way around. The transmitted tangential force on each tooth is F_t = 2T/d, and the total tooth normal force acts along the line of action: F_n = F_t/cos φ, with a radial separating component F_r = F_t·tan φ that tries to push the shafts apart. At φ = 20°, F_r ≈ 0.364·F_t; raising φ to 25° raises that separating force to ≈ 0.466·F_t — the trade you accept for stronger teeth.

Contact ratio check

The contact ratio ε must exceed 1.0 so a fresh tooth pair engages before the old one leaves, giving continuous drive. It is the length of the path of contact Z divided by the base pitch p_b:

ε = Z / p_b = Z / (π·m·cos φ)

where Z is the distance along the line of action between the two tooth-tip intersection points (mm). For this 20/60 pair at 20° full depth, ε ≈ 1.67 — comfortably above 1.0, so at least one and often two tooth pairs share the load. Spur gears typically fall in the 1.4–1.8 range; helical gears add a face (overlap) contact ratio on top, which is why they run quieter.

Involute vs. cycloidal profiles

Property	Involute	Cycloidal
Generating curve	Unwinding string from base circle	Point on a rolling circle (epi/hypocycloid)
Line of action	Straight, fixed, tangent to base circles	Curved, changes through the mesh
Pressure angle	Constant (20° std)	Varies from 0° at pitch point
Center-distance error	Ratio unaffected; only backlash grows	Loses conjugate action; can jam
Tooling	One straight-sided rack/hob per module	Different cutter per tooth count
Sliding / wear	Sliding through mesh, rolling only at pitch point	Lower sliding near pitch; less pitting
Where used	~99% of power gearing	Clocks, watches, cycloidal-drive reducers

Common misconceptions and failure modes

• "Pitch circles are physical." They are not machined features — they are the imaginary rolling circles set by the operating center distance. The real geometric anchor is the base circle. There is no involute below the base circle at all.
• "The teeth roll on each other." Pure rolling happens only at the pitch point. Everywhere else the flanks both roll and slide, and the sliding velocity (largest at the tooth tip and root) drives scuffing and heat — which is why lubrication and surface finish matter.
• "Any two gears of the same module mesh." They must match in both module and pressure angle. A 20° and a 25° gear of identical module will not run conjugately.
• Undercutting. Cut a 20° pinion with fewer than ~17 teeth and the generating rack gouges the flank below the base circle, removing active profile and weakening the root. Fix it with positive profile shift (addendum modification) or a higher pressure angle.
• Interference. If the tooth tip of one gear tries to contact the non-involute region of the mate below its base circle, the tips dig in. Tip relief, profile shift, or fewer-tooth limits prevent it.
• Excess backlash from center-distance drift. The involute's tolerance is a gift, but push the centers too far and backlash grows enough to cause impact loading and noise on load reversals; the ratio stays perfect but the lash does not.