A coil creates a magnetic field that pulls a friction armature into contact with a rotor.

With the coil de-energised, the armature is held clear of the rotor by leaf springs, leaving a small air gap. Nothing touches, so the two shafts turn independently with only bearing drag between them.

Energising the coil creates a magnetic circuit through the rotor and armature. The armature is drawn across the gap and clamped against the rotor face, and friction between them transmits torque. The clamping force - and therefore the torque capacity - depends on the magnetic flux, which depends on the coil current and on the size of the gap it has to pull across.

Removing the coil current collapses the field and the springs pull the armature clear again.

The whole sequence takes milliseconds, which is what makes these clutches suited to rapid cycling under electronic control.

Very fast mechanically - many times per minute - with the real limit set by heat rather than by the mechanism.

Engagement and release each take only milliseconds, so the clutch itself imposes almost no restriction on cycle rate. Machines running several engagements per second are entirely normal.

What limits the rate is the energy dissipated at each engagement. Every time the clutch picks up the driven inertia it slips briefly, and that slip becomes heat. Double the cycle rate and you double the continuous heat load without changing anything else.

So the practical question is never 'how fast can it engage' but 'how much energy per engagement, times how many engagements per hour, against the unit's thermal rating'.

Where the thermal budget is exceeded, the answers are a larger unit, a lower speed difference at engagement, reduced driven inertia, or forced cooling - in roughly that order of cost.

Because it grows as the friction faces wear, and a wider gap weakens the magnetic pull that produces the torque.

The coil has to pull the armature across the gap. Magnetic force falls off sharply with distance, so a gap that has widened through wear produces less clamping force for the same coil current. Torque capacity drops, engagement becomes slower and less positive, and the clutch slips for longer at each engagement.

That extra slip generates more heat and more wear, which widens the gap further. It is a self-accelerating failure, and it presents as a clutch that gradually loses torque and then fails fairly suddenly.

Many designs are self-adjusting and need no attention. Those that are not have a specified gap range and a means of adjusting it, and they should be checked on a schedule based on engagement count rather than calendar time.

A clutch that has started slipping is worth checking for gap before being condemned as worn out.

Usually a low-voltage DC supply, switched by a relay or a purpose-made clutch controller.

The coil is a DC device, so an AC supply requires a rectifier. Standard coil voltages are low - commonly 24 V DC - which suits control panels and keeps the wiring simple and safe.

Switching matters more than it appears. A coil is an inductive load, and interrupting its current produces a voltage spike that can damage the switching device and generate electrical noise. Suppression - a diode, varistor or RC network across the coil - is normally required and is sometimes built into the clutch.

However, a simple suppression diode also slows release, because it lets the collapsing field circulate current. Where fast release matters, a controller designed to force the field down quickly is used instead.

On fast-cycling machines, over-excitation controllers apply a higher voltage briefly at engagement for speed, then drop to a holding level - which improves response and reduces heat.

On machinery that engages and releases frequently under electronic control, with modest torque.

Packaging and labelling machines, printing and paper handling, converting equipment, textile machinery, conveyor indexing and diverting, office and vending equipment, and machine tool auxiliary drives all use them heavily.

What those share is the pattern the clutch suits: a control system that produces an electrical signal, a requirement to engage quickly and precisely, high cycle counts, and torque within the range a magnetically-generated clamping force can produce.

They are less appropriate where the torque required is very large, since magnetic force has practical limits and a pneumatic or hydraulic clutch generates far more clamping force for its size. They are also not the choice where no electrical supply is available at the drive, or where the environment makes a coil and its wiring impractical.

Yes - toothed and hysteresis types exist, and each solves a problem the friction type does not.

Toothed or positive electromagnetic clutches engage interlocking teeth rather than friction surfaces. They transmit high torque in a small package, cannot slip once engaged, and have no friction wear at all. The limitation is that they can only engage at or near zero speed difference, since engaging teeth at speed would be violent - so they suit indexing rather than picking up a moving load.

Hysteresis clutches transmit torque magnetically with no mechanical contact, producing a smooth torque that is independent of speed and fully adjustable with coil current. They are used for tension control in web handling and for test equipment, where controlled slip is the purpose rather than a byproduct, and they have essentially no wear.

Friction types remain the default where a moving load has to be picked up on demand.

Air gap checking where the design requires it, protection from contamination, and attention to the electrical connection - not much else.

There is nothing to lubricate on a dry friction clutch, and lubricating one is actively harmful. Oil or grease reaching the friction face destroys its torque capacity, and the commonest source is over-enthusiastic greasing of an adjacent bearing.

Check the air gap on units that need it, at intervals based on engagement count.

Electrically, check the coil resistance if performance changes - a partially shorted coil produces reduced field and weak engagement. Inspect the slip ring and brushes on rotating-coil designs, and the connection and suppression components on all types.

Otherwise, monitor the symptom that matters: a clutch that takes longer to lock up than it used to is telling you something, and investigating it then is much cheaper than after it fails.