It releases - which is the defining characteristic and the reason it must not be used where a load could fall.

The coil produces the clamping force. No current means no magnetic field, no clamping and no braking torque, so on loss of supply the shaft is free to turn.

On a machine where free coasting is harmless - a horizontal conveyor, a rotating table, a spindle - that is perfectly acceptable, and it has the advantage that the machine can be turned by hand for setting and maintenance without any release mechanism.

On a hoist, a lift, an inclined conveyor or a vertical axis it is unacceptable, because a power failure would let a suspended or gravity-loaded mass run away.

The test is simple: if losing power right now could allow something to fall or run back, this is the wrong type of brake and a spring-applied fail-safe unit is required. Many machines use both, for different purposes.

By the braking torque against the total inertia, plus the response delay before the torque appears.

Once the brake is fully applied, deceleration is the braking torque divided by the total moment of inertia referred to the brake shaft. Stopping time follows from that deceleration and the initial speed, and stopping distance from the average speed during the stop.

The part frequently forgotten is the delay. Between the command and full braking torque there is the controller's switching time, the time for current to build in an inductive coil, and the time for the armature to travel the air gap. During all of it the machine is still moving at full speed, and that distance adds directly to the total. On a fast machine it can exceed the braking distance itself.

Where stopping distance is a safety requirement - guarding interlocks, for instance - the delay must be measured on the actual machine rather than calculated, and re-measured as the brake wears.

Because each stop dumps the machine's kinetic energy into the friction faces, and that energy has to leave before the next stop arrives.

The brake converts motion into heat. One stop raises the friction face temperature by an amount set by the energy absorbed and the thermal mass of the brake. If the next stop comes before that heat has dissipated, the temperature ratchets upward stop by stop.

Above a certain temperature the friction material glazes and its coefficient of friction falls. Torque drops, so stopping takes longer, so more energy is dissipated per stop, so it gets hotter still.

The cure is thermal headroom rather than more torque: a larger brake with more mass and surface area, forced cooling, a lower speed at the moment of braking, or fewer stops.

Calculate the continuous thermal load as energy per stop times stops per hour and compare it against the brake's published thermal rating. It is a separate check from torque, and it is usually the one that governs.

It grows as the lining wears, and a wider gap means a weaker magnetic pull and a slower, later stop.

The coil must draw the armature across the gap. Magnetic force falls steeply with distance, so as wear widens the gap the same current produces less clamping force. Braking torque falls and, just as importantly, the armature takes longer to travel, which lengthens the response delay.

The machine's stopping distance therefore increases gradually over the brake's life. Where that distance is part of a safety function, the increase is a safety degradation, not merely a performance one.

Some designs are self-adjusting and maintain the gap automatically. Those that are not have a specified gap and a means of resetting it, and it should be checked on a schedule based on the number of stops rather than on elapsed time.

If a machine's stopping distance has crept up, check the gap before concluding the brake is worn out.

As the other half of a start-stop cycle, and the two are usually interlocked so they are never engaged together.

On a cycling machine, the clutch connects the driven side to a continuously running drive to start the motion, and the brake stops it at the end. If both were engaged simultaneously the drive would be fighting the brake, wasting energy and wearing both.

Control therefore releases one before applying the other, with a short overlap-free interval. On fast machines that timing is critical to cycle rate and is usually handled by a purpose-made clutch-brake controller rather than by separate relays.

When both functions sit on the same shaft, a clutch-brake combination unit is generally the better answer than two separate components: it is more compact, the interlock is inherent in the design, and the timing is fixed mechanically rather than depending on control logic being correct.

Low-voltage DC coils, switched with suppression, and often with over-excitation on fast-cycling machines.

The coil is a DC device, commonly at a low control voltage, so an AC supply requires a rectifier. Because it is an inductive load, switching it produces voltage spikes that must be suppressed to protect the switching device.

How the coil is switched affects performance directly. Simple diode suppression across the coil slows the collapse of the field and therefore the release; where fast release matters, a controller that forces the field down quickly is used instead.

Over-excitation applies a higher voltage briefly at engagement to build current faster, then drops to a lower holding level. That shortens the response delay - reducing stopping distance - while limiting the heat dissipated in the coil.

On safety-related stopping functions, the switching arrangement is part of the safety function and must be designed accordingly, not treated as an ordinary control circuit.

On cycling machinery where the brake defines the end of a motion and no load has to be held.

Packaging and labelling machines, converting and paper handling equipment, indexing tables and conveyors, machine tool auxiliary axes, textile machinery and office equipment all use them - typically alongside an electromagnetic clutch on the same shaft.

What those applications share is fast repeated cycling, modest inertia, horizontal or otherwise non-gravity-loaded motion, and a requirement for repeatable stopping position rather than long-term holding.

They are not used where a load must be held on power failure, where the inertia and speed produce more heat than a compact brake can shed, or where the required torque exceeds what a magnetic circuit can generate - all of which point to spring-applied, caliper or thruster brakes instead.