An eddy current brake is a non-contact braking device that slows a moving conductor by inducing circulating currents in it with a magnetic field, producing a drag force opposing motion. Modern truck retarders like the Telma Focal 3000 generate up to 3,000 Nm of braking torque with zero pad wear. The brake exists to dump kinetic energy as heat without friction surfaces, which is why you see it on ICE roller coasters, the Shanghai Maglev, and Class 411 tilting trains where mechanical brakes would fade or wear too quickly.
Eddy Current Brake Interactive Calculator
Vary rotor speed and air gap to see how normalized eddy-current braking torque collapses as magnetic flux falls.
Equation Used
This calculator follows the article relationship that eddy-current braking torque scales with magnetic flux squared. Using the stated air-gap rule, flux is modeled as proportional to 1/gap^3, so torque changes with the sixth power of the air-gap ratio. A simple normalized speed curve is included to show the rise to peak braking torque near the selected peak RPM.
- Torque is normalized to rated torque at the specified air gap and peak RPM.
- Air-gap flux density follows the article approximation B proportional to 1/g^3.
- Because eddy-current torque is proportional to B^2, the gap factor is (g_spec/g)^6.
- Speed behavior is represented by a simple peak curve that rises at low speed and falls after rpm_peak.
The Eddy Current Brake in Action
When a conductive disc or rail moves through a stationary magnetic field, the changing flux through the conductor induces circulating loops of current — eddy currents. By Lenz's law these currents create their own magnetic field that opposes the motion that produced them, so you feel a drag force on the moving part. No contact, no friction surfaces, no pads. Just induction, resistance, and heat. The conductor warms up because the eddy currents dissipate as I²R losses in the metal itself.
The drag force scales linearly with velocity at low speeds, peaks at a critical velocity, then falls off at very high speeds — this is the classic eddy-current torque-speed curve. Designers tune the air gap, conductor thickness, and magnet strength to put that peak where it's useful. On a Telma rail retarder the peak sits around 1,000 RPM driveshaft speed. On an ICE roller coaster fin brake the peak is set for the train's full-speed pass. If the air gap is too wide — say 8 mm instead of the spec'd 3 mm — flux density drops with the cube of distance and braking torque collapses to a fraction of rated. Conversely, if the disc is too thin the eddy current loops can't develop enough cross-sectional path and resistive losses dominate over induced field strength.
Common failure modes are conductor overheating (aluminium discs above 300 °C lose conductivity fast and torque sags mid-stop), magnet demagnetisation on permanent-magnet variants if temperature exceeds the Curie point of the NdFeB grade used, and air-gap drift from worn bearings letting the disc wobble into the pole faces. The brake doesn't squeak or fade audibly — it just quietly under-performs, which is why retarder manufacturers spec disc thickness to ±0.2 mm and air gap to ±0.5 mm.
Key Components
- Conductive Rotor or Rail: Usually copper, aluminium, or mild steel. The eddy currents form here. Aluminium gives high conductivity and low mass but limits operating temperature to around 300 °C. Steel rotors handle 600 °C+ but produce less peak torque per unit field strength.
- Field Magnets: Either electromagnets (variable, switchable) or permanent NdFeB magnets. Field strength typically 0.8–1.4 Tesla at the pole face. Electromagnet coils on Telma retarders draw 40–60 A at 24 V to reach full field.
- Pole Pieces: Soft iron or laminated steel that channels flux across the air gap into the conductor. Lamination thickness is typically 0.35–0.5 mm to suppress eddy losses inside the pole pieces themselves — you only want eddies in the rotor, not in the stator.
- Air Gap: Typically 1–5 mm for rotary retarders, up to 7 mm for linear track brakes. Tolerance is usually ±0.5 mm because flux density falls roughly with the square to cube of gap distance, and braking torque follows the square of flux density.
- Heat Sink Fins or Forced Cooling: The rotor is also the heat sink. Cast cooling fins on truck retarders dissipate 200–400 kW continuously. Without adequate cooling, the rotor reaches thermal saturation in 30–60 seconds of full braking and torque collapses.
Who Uses the Eddy Current Brake
Eddy current brakes earn their place anywhere you need wear-free, fade-free, controllable retardation — and where it's acceptable that the brake cannot hold a stationary load (zero velocity means zero induced current means zero force). You'll find them as truck and bus retarders, on rail vehicles, on roller coasters, in gym equipment, and as damping elements in precision instruments and tools.
- Heavy Trucks & Buses: Telma Focal driveshaft retarders fitted to Volvo and Scania coaches, providing up to 3,000 Nm of supplementary braking on long descents to spare the service brakes.
- Rail: ICE 3 high-speed trains use linear eddy current track brakes generating up to 150 kN per bogie at 300 km/h, acting on the rail itself with no mechanical contact.
- Amusement Rides: Intamin and B&M roller coasters use fin brakes with copper or aluminium fins passing between magnet arrays — Kingda Ka at Six Flags relies on them for its 206 km/h launch return.
- Fitness Equipment: Concept2 RowErg rowing machines use a flywheel and adjustable magnet for resistance — though that's air-braked, the Schwinn Airdyne and most spin bikes use eddy current resistance with a moving magnet over an aluminium flywheel.
- Tools & Safety: SawStop-style fall-arrest devices and zip-line stoppers use eddy current damping to limit terminal speed without abrupt jerk.
- Precision Instruments: Analogue energy meters, watt-hour meters, and laboratory balances use a small aluminium disc passing between a magnet pair to provide critically damped needle motion.
The Formula Behind the Eddy Current Brake
The braking force on a thin conductive plate moving through a uniform magnetic field follows a clean closed-form expression at low to moderate speeds, before saturation effects kick in. At the low end of the typical operating range the force grows linearly with velocity — useful for damping but weak at parking speeds. Around the design sweet spot the brake delivers near-peak retardation. Push past the critical velocity and the induced field starts to shield the conductor's interior, force flattens out, and continued speed gives diminishing return. Knowing where your operating point sits on that curve is the whole game.
Variables
| Symbol | Meaning | Unit (SI) | Unit (Imperial) |
|---|---|---|---|
| Fbrake | Eddy current drag force opposing motion | N | lbf |
| σ | Electrical conductivity of the rotor material | S/m | S/in |
| t | Thickness of the conductive plate or disc | m | in |
| A | Effective pole-face area covering the conductor | m2 | in2 |
| B | Magnetic flux density across the air gap | T | G (×10-4 T) |
| v | Relative velocity between conductor and magnet | m/s | ft/s |
Worked Example: Eddy Current Brake in a vertical zip-line descent retarder
You are sizing the eddy current retarder on a 90 kg climber's vertical zip-line auto-belay descender, similar in concept to a Head Rush TRUBLUE unit. The carrier passes a 4 mm thick aluminium fin between two NdFeB magnet stacks producing 1.0 T across a 3 mm air gap. The pole-face area covering the fin is 0.004 m². You need to confirm the brake limits the climber's free-fall to a controlled descent across 1, 2, and 4 m/s — the realistic range from a slow controlled lower to a panic drop.
Given
- σ (aluminium) = 3.5 × 107 S/m
- t = 0.004 m
- A = 0.004 m2
- B = 1.0 T
- Climber mass = 90 kg
- Weight to retard = 883 N
Solution
Step 1 — compute the velocity-independent coefficient k = σ × t × A × B2:
Step 2 — at the nominal design speed of 2 m/s (a controlled descent pace, roughly walking speed downward):
That comfortably exceeds the 883 N weight of the climber, so at 2 m/s the brake decelerates them. Good.
Step 3 — at the low end of the typical operating range, 1 m/s (a slow lower, near the foot of the descent):
That's only 63% of the climber's weight, so at 1 m/s the climber is still accelerating downward — the brake hasn't grabbed enough yet. This is fine because the system relies on velocity feedback: as speed builds, force builds, until equilibrium near the terminal velocity.
Step 4 — at the high end, 4 m/s (a panic drop after slipping off the holds):
That's 2.5 × body weight of retardation, giving a deceleration of about 1.5 g — sharp but survivable, exactly what an auto-belay is meant to deliver. In practice the linear formula starts to under-predict above 5–6 m/s on this geometry because the induced counter-field begins shielding the fin interior, and real TRUBLUE-class units bake in a soft saturation around 4 m/s by design.
Result
Nominal braking force at 2 m/s is 1,120 N, with terminal velocity for the 90 kg climber settling near 1. 6 m/s where induced drag exactly matches body weight. Across the operating range the force scales linearly from 560 N at 1 m/s to 2,240 N at 4 m/s — meaning a slow climber lowers gently while a panic drop gets caught firmly without ever locking up. If you measure terminal velocity above 2 m/s on a real build, three failure modes top the list: (1) the air gap has crept above the spec'd 3 mm because the fin carrier bushings have worn, dropping flux density with the cube of distance, (2) the aluminium fin has heated above 200 °C from repeated drops and conductivity has dropped 15–20%, or (3) the NdFeB magnets have partially demagnetised because someone arc-welded near the unit or it saw temperatures above the N42 grade's 80 °C limit.
When to Use a Eddy Current Brake and When Not To
Eddy current brakes are not a universal replacement for friction brakes — they shine in specific operating envelopes and fail outright in others. Compare them against the two brakes they most often displace: a conventional friction disc brake and a hydraulic retarder.
| Property | Eddy Current Brake | Friction Disc Brake | Hydraulic Retarder |
|---|---|---|---|
| Holding torque at zero speed | Zero — cannot hold a parked load | Full rated torque | Zero — also speed-dependent |
| Wear & maintenance interval | No wear surfaces, 1M+ cycles typical | Pads replaced every 50,000–150,000 km on trucks | Fluid change every 200,000 km |
| Peak braking torque | Up to 3,000 Nm (Telma Focal 3000) | 10,000+ Nm achievable on heavy trucks | Up to 4,000 Nm (Voith VR3250) |
| Fade resistance | Excellent until thermal saturation (~60 s full load) | Fades sharply above 600 °C disc temp | Excellent with adequate cooling |
| Response time | 10–50 ms (electromagnet excitation) | 100–300 ms (hydraulic line + caliper) | 200–500 ms (fluid fill time) |
| Cost & complexity | Mid — coils, controller, heavy rotor | Low — mature, commodity parts | High — pump, valves, cooler circuit |
| Best application fit | Continuous retardation, no parking | Stopping and parking duty | Continuous retardation on heavy vehicles |
Frequently Asked Questions About Eddy Current Brake
Because force is proportional to velocity in the linear regime — at 0.1 m/s you get 1/20th the force you'd get at 2 m/s with the same magnets. There is no induced current without changing flux, and no changing flux without motion. This is also why eddy current brakes cannot hold a parked vehicle. If you need low-speed retardation, increase the magnet count or pole-face area rather than chasing stronger magnets — doubling A doubles low-speed force, while doubling B2 requires roughly 1.4× field strength, which is expensive in NdFeB.
Permanent magnet (NdFeB) versions are simpler, lighter, and need no power — but the brake is always on whenever there's relative motion, which is parasitic drag you can't switch off. Electromagnet versions let you modulate or fully release the field, which is essential for vehicles where you want the retarder off during cruise. The trade is coil power (40–60 A at 24 V on truck retarders) and added control electronics. Rule of thumb: roller coasters and zip-line descenders use permanent magnets because always-on is a feature; trucks, buses and trains use electromagnets because they need a switchable retarder.
Check the rotor material certification. A surprising number of "aluminium" plates from low-cost suppliers are 3000-series alloys with high silicon content and conductivity around 2.4 × 107 S/m, not the 3.5 × 107 S/m of pure or 6061-T6 aluminium. That single substitution drops braking force by 30%+ and looks identical visually. A four-point conductivity probe or a known reference resistance test on an offcut confirms it in two minutes. Copper C110 plates have the opposite problem — they're heavier than the design assumed and add rotational inertia.
Thermal saturation. The rotor is the heat sink, and at full duty an aluminium disc can easily reach 250–300 °C in under a minute. Conductivity of aluminium drops about 0.4% per °C, so by the time the disc is at 300 °C it has lost roughly 30% of its room-temperature conductivity, and braking force scales linearly with σ. The fix is forced air cooling, a larger rotor, or duty cycling — not stronger magnets. On Telma units this is precisely why the rotors are cast with deep radial fins and why the manufacturer publishes a continuous vs intermittent torque curve.
Not directly. The simple F = σ × t × A × B2 × v expression assumes the conductor is thin compared to the skin depth at the relevant excitation frequency. Steel has high permeability and low conductivity, so the eddy currents concentrate in a thin surface layer and the effective t is the skin depth, not the physical thickness. You also pick up significant hysteresis losses on top of the I2R losses, which the formula doesn't account for. For steel rotors, finite-element simulation in a tool like FEMM is the practical route — empirical correlations exist but they're geometry-specific.
Partly, but the bigger effect is usually NdFeB demagnetisation if the magnet stacks see direct sun on the brake elements. Standard N42 grade loses about 0.12% remanence per °C reversibly, and once you cross the maximum operating temperature (around 80 °C for N42, 150 °C for N42SH) the loss becomes permanent. Force scales with B2, so a 5% drop in remanence is a 10% drop in braking force. Either upgrade to an SH or UH grade, shade the magnet array, or add convective cooling. Fin thermal expansion only changes air gap by tens of microns and is rarely the dominant cause.
They serve different goals. Regenerative braking captures energy back into a battery or capacitor and is bounded by the inverter and battery acceptance rate — typically 50–150 kW on a passenger EV. Eddy current braking dumps energy as heat with no acceptance limit, which matters when the battery is full or cold and can't take regen. Many hybrid trucks and trains use both: regen until the battery saturates, eddy current retardation for the remainder. The eddy brake is also a fail-safe — it works with no electronics functioning beyond the field excitation circuit.
References & Further Reading
- Wikipedia contributors. Eddy current brake. Wikipedia
Building or designing a mechanism like this?
Explore the precision-engineered motion control hardware used by mechanical engineers, makers, and product designers.
