Belts and pulleys: rubber bands that move machines

A workshop drill press, a centrifugal pump, and a car engine accessory drive all use the same quiet trick: two shafts stand apart, yet one still manages to make the other turn. That trick is the belt drive. A flexible belt loops around pulleys, grips them by friction, and carries motion from a driving shaft to a driven shaft without any meshing teeth.

That is why belts and pulleys are everywhere. They are cheaper and quieter than many gearboxes, they tolerate some misalignment and shock, and they let one motor serve many speed combinations just by changing pulley diameters. The beginner version sounds almost too simple: "a rubber band moving a machine." The real engineering question is how that loop actually carries torque, how much tension it needs, how much it slips, and when it stops being a good choice.

The plain-language picture

Imagine trying to pull a heavy box using a rope wrapped around a pole. If the rope barely touches the pole, it slips. If you wrap more of the rope around the pole and pull harder, the grip improves. A belt drive works in the same way. The belt hugs the pulley over a contact angle called the wrap angle, and friction between belt and pulley lets the driving shaft drag the belt along.

One side of the moving belt becomes the tight side, where tension is higher because it is transmitting force. The returning side is the slack side, where tension is lower. The difference between those two tensions is what creates useful torque on the pulley. If the tension difference is too small, the belt slips. If the overall belt tension is too high, the bearings and shafts suffer even if the belt never slips.

Belts come in several families. Flat belts rely on surface friction across a wide band. V-belts wedge into pulley grooves, which increases grip and makes them very popular for industrial drives. Timing belts have teeth and act more like positive drives with almost no slip. This article stays focused on the beginner physics of ordinary friction belt drives, because that is where the main ideas become clear.

Figure 1: A simple belt drive lets a motor and a machine sit on separate shafts and still share power. The belt is flexible, but the transmitted torque comes from very real tension forces and pulley geometry.

What the pulleys really control

A pulley does more than guide the belt. Its effective diameter sets the shaft speed. If the driver pulley is smaller than the driven pulley, the output shaft turns more slowly but gains torque. If the driver pulley is larger, the output shaft speeds up. This is the same speed-torque trade that gears make, but belts do it through friction and diameter rather than through tooth counts.

That flexibility is the main reason engineers like belt drives. If a fan should run at half motor speed, a larger fan pulley may be enough. If a pump needs a softer start, the slight compliance of the belt helps cushion the shock. If the shafts must sit far apart or a cheap replacement ratio is needed later, swapping pulleys is often easier than redesigning a gear train.

The weakness is equally important: ordinary friction belts are not exact positive drives. Their speed ratio is close to the pulley ratio, but small slip and creep mean the output speed is not perfectly locked to the input speed. For fans, blowers, and many pumps that is acceptable. For cam timing or precise indexing, it may not be.

The governing physics

For a first estimate, the belt speed is the rim speed of the pulley:

v = pi x D x N / 60

where v is belt speed in m/s, D is pulley pitch diameter in m, and N is rotational speed in rpm.

If slip is neglected, the belt speed is the same on both pulleys, so:

pi x D1 x N1 = pi x D2 x N2

which simplifies to:

N2 = N1 x D1 / D2

That is the beginner speed-ratio equation. A smaller driving pulley feeding a larger driven pulley reduces speed.

Power transmission comes from the difference between tight-side and slack-side tension:

P = (T1 - T2) x v

where P is power in W, T1 is tight-side tension in N, T2 is slack-side tension in N, and v is belt speed in m/s.

The reason wrap angle matters is that friction can only support a limited tension ratio. For a flat belt, the classic capstan relation is:

T1 / T2 = e^(mu x theta)

where mu is friction coefficient and theta is wrap angle in radians. A bigger wrap angle or better friction lets the drive carry a larger tension difference before slipping. V-belts improve this further because the groove wedging action increases effective grip.

Figure 2: A V-belt does not mainly drive by touching the pulley bottom. It grips through the sidewalls of the groove, which is why groove shape, belt section, and correct tension all matter so much in service.

Worked example 1: finding the output speed

A workshop exhaust fan is driven by an electric motor running at 1450 rpm. The motor pulley diameter is 100 mm and the fan pulley diameter is 250 mm. Estimate the fan speed if slip is neglected.

Use the pulley-ratio equation:

N2 = N1 x D1 / D2

N2 = 1450 x 100 / 250 = 580 rpm

So the fan should rotate at about 580 rpm. The motor spins quickly, but the larger fan pulley slows the shaft while increasing available torque. This is why many fans and blowers use belt drives: the motor can stay near a standard speed while the fan sees the speed it actually wants.

We can also estimate belt speed from the motor pulley:

v = pi x 0.10 x 1450 / 60 = 7.59 m/s

That belt speed is moderate and realistic for a small workshop fan drive.

Worked example 2: tension difference needed for a power load

A belt-driven water pump needs 3.0 kW at operating speed. The driving pulley pitch diameter is 180 mm and the motor speed is 960 rpm. During the design estimate, the engineer expects the running tight-side to slack-side tension ratio to be about 2.5. Find the belt speed, the tension difference T1 - T2, and the approximate values of T1 and T2.

First calculate belt speed:

v = pi x 0.18 x 960 / 60 = 9.05 m/s

Now use the power equation:

P = (T1 - T2) x v

T1 - T2 = P / v = 3000 / 9.05 = 331.5 N

The assumed tension ratio gives:

T1 / T2 = 2.5

So:

2.5T2 - T2 = 331.5

1.5T2 = 331.5

T2 = 221.0 N

T1 = 552.5 N

The useful message is that the transmitted power comes from the difference between the two tensions, not from one side alone. The belt may look light and flexible, but it is carrying several hundred newtons of force while running. If the installer greatly over-tightens it to "be safe," the pump and motor bearings can be overloaded for no benefit.

Worked example 3: why slip changes the real output speed

A small conveyor uses a 125 mm motor pulley and a 375 mm driven pulley. The motor speed is 1750 rpm. The ideal ratio predicts one output speed, but the drive experiences about 3% total slip under load. Estimate the actual driven speed.

Ideal speed first:

N2,ideal = 1750 x 125 / 375 = 583.3 rpm

Now account for slip:

N2,actual = 583.3 x (1 - 0.03) = 565.8 rpm

The conveyor really sees about 566 rpm, not 583 rpm. That difference is small enough for many material-handling jobs, but it shows why friction belts are best for practical power transmission rather than precision synchronization.

Assumptions and their limits

The simple equations above assume the belt is flexible but not stretching much, the pulley diameters are the effective pitch diameters, slip is small, and the tensions stay steady. Real drives are less tidy. Belt material creeps, startup torque spikes, temperature changes the rubber stiffness, and pulley grooves wear. Even the way the motor base is aligned changes how evenly the belt loads each sidewall.

Another limit is that the textbook tension relation does not automatically include centrifugal effects, dynamic shock, or bending fatigue from running over very small pulleys. At high belt speed, part of the belt tension is "wasted" just keeping the belt curved around the pulley. At low wrap angle, the drive may slip even when the nominal power calculation looked acceptable. Real design therefore uses service factors, manufacturer ratings, and experience, not only one clean equation.

Common failure modes and what they look like

• Slip and polishing: the belt squeals, heats up, and leaves shiny glazed sidewalls because the available friction is too low for the load.
• Cracking and cord fatigue: the belt bends around pulleys every revolution, so undersized pulleys or old rubber eventually create transverse cracks and broken tension cords.
• Sidewall wear: misaligned V-belts rub unevenly in the groove and lose section shape, which reduces grip and tracking stability.
• Bearing overload: a belt can stop slipping after someone over-tightens it, but then the shaft bearings run hotter because the radial load is excessive.
• Pulley groove wear: worn sheaves change the effective belt seating and reduce wedging action, so a new belt may still perform poorly on an old pulley.
• Contamination: oil, dust, and coolant reduce friction or attack the belt compound, causing erratic behavior that looks like a sizing problem.

Figure 3: Belt-drive problems usually show up as a system issue rather than a single bad number. Tension setting, alignment, pulley wear, contamination, and duty cycle all change whether the drive runs quietly or starts eating belts.

Practical rules of thumb

• Use belt drives when shafts are separated, quiet running matters, and a small amount of slip is acceptable.
• Avoid very small pulleys because sharp bending shortens belt life and reduces transmitted power.
• Protect wrap angle on the smaller pulley. If the belt barely hugs the driver, grip disappears quickly.
• Do not solve every slipping problem by adding tension. Check load, pulley wear, alignment, and belt section first.
• Remember that belt tension becomes shaft and bearing load. A "good" belt drive can still be a bad bearing drive.
• If exact phase or speed must be maintained, move toward timing belts, chains, or gears instead of fighting friction-belt limitations.

How standards and design practice treat belts

Professional belt selection is usually based on manufacturer rating methods backed by standardized belt sections and pulley geometry. In practice, designers work with classical or narrow V-belt families, matched pulley groove dimensions, minimum recommended pulley diameters, and service factors for shock, daily hours, and start frequency. Standards bodies define the belt and sheave geometry language so belts from different suppliers can be selected and replaced consistently.

Good design practice also includes guarding and maintenance, not just power rating. A belt drive that is mechanically adequate still needs safe guarding around rotating parts, correct tensioning procedure, alignment checks, and inspection intervals for heat cracking, cord exposure, and groove wear. That standards mindset is useful for beginners because it shifts the question from "Will this belt spin?" to "Will this drive survive real duty safely and repeatably?"

Engineering judgment

The most important thing to understand about belts and pulleys is that they are flexible components solving a rigid power problem. They are excellent when you want low cost, quiet running, adjustable speed ratio, and some forgiveness in the drivetrain. They are poor when you need perfect synchronization, zero slip, or compact high-torque transfer in a very small envelope.

When reviewing a belt-drive concept, ask five questions in order. What speed ratio is needed? How much power must be carried? What wrap angle and pulley size are available on the small pulley? What belt tension will the shaft bearings see? And what happens if the drive slips during startup or overload? Those questions usually reveal whether the belt drive is a sensible engineering choice or just the easiest sketch on paper.

If you want to follow what the pulley is loading next, continue with Shafts: the spinning backbone of every machine, or browse the full EnggTools engineering article library.