What each connects to. A clutch connects two rotating members to each other; a brake connects a rotating member to something stationary.

That one difference drives everything else. A clutch transfers torque between a driving and a driven shaft, and during engagement the driven side accelerates up to the driving speed. A brake removes energy from a rotating mass and brings it to rest, or holds it there against an applied torque.

The hardware is often similar - friction faces, a means of pressing them together, and a way of commanding it - which is why the two are sold together and why clutch-brake combinations exist as single units.

But the selection questions differ. A clutch is sized on transmitted torque and on the heat of accelerating the driven inertia. A brake is sized on stopping torque, on the heat of decelerating the moving inertia, and frequently on the static torque it must hold indefinitely once stopped.

On two independent criteria: the torque it must transmit, and the heat it must dissipate at the required cycle rate.

Torque is the obvious one. The clutch must transmit the driven machine's torque demand with a service factor for shock and duty, or it will slip in normal running.

The thermal criterion is the one that is missed. Every engagement slips briefly while the driven side accelerates, and the energy of that speed mismatch becomes heat in the friction faces. Multiply that energy by engagements per hour and you have a continuous heat load that the clutch has to shed.

A clutch adequate on torque and inadequate on thermal capacity glazes its friction faces, loses torque capacity, slips more, generates more heat and fails - a failure that looks like an undersized clutch but is really an over-cycled one.

Provide the driven inertia, the speed difference at engagement and the cycles per hour, and the supplier can check both.

The one that matches the control you have available and the force the application needs.

Electromagnetic actuation suits fast, frequent, electrically controlled cycling with modest torque - packaging machinery, paper handling, conveyors with indexing. It needs only a DC supply and a signal.

Pneumatic or hydraulic actuation provides much greater engagement force, so it handles high torque and large inertia, and it can be modulated to control how quickly engagement occurs. It needs an air or hydraulic supply and valving.

Wrap-spring actuation suits single-revolution and indexing duty, where the requirement is precisely one turn per command.

Centrifugal needs no control at all and is used where the engagement should simply follow engine speed.

Start from what the machine already has - a plant with no compressed air near the drive is a poor place for a pneumatic clutch - then check the torque and thermal figures.

It is how long the clutch slips between the command and full lock-up, and it governs both the heat generated and the machine's productivity.

A longer engagement time means the driven side accelerates gently, which is kind to the machine and to the product, but it also means the clutch slips for longer and therefore generates more heat per engagement.

A shorter engagement is thermally kinder and faster in cycle terms, but it applies a sharper acceleration to the driven machine, which can damage product, snatch a web or shock the drive train.

Electromagnetic clutches engage quickly, in fractions of a second. Pneumatic and hydraulic units can be tuned by controlling how fast pressure builds, which is one of their advantages.

On a high-cycle machine, engagement time is worth optimising deliberately rather than accepting whatever the clutch does - it is usually the lever that resolves a thermal problem without changing the clutch size.

In engagements rather than in hours, because wear is caused by the slipping during engagement and hardly at all by running engaged.

Once a friction clutch is locked up there is no relative motion between the faces and effectively no wear. All the wear happens during the brief slip at each engagement, and the amount depends on the energy dissipated in that slip.

So the meaningful life figure is total engagements, or total energy dissipated, not running hours. A clutch that engages twice a shift will last for years; the same clutch engaging twice a minute may need relining within months.

That is also why reducing the energy per engagement extends life directly - lowering the speed difference at engagement, or reducing the inertia being accelerated, pays back more than buying a larger clutch.

Manufacturers publish life in engagements at a stated energy per engagement. Compare on that basis, and count your actual cycles rather than estimating.

Generally no - a friction clutch transmits torque between shafts, and holding a static load is a brake's job.

A clutch engaged and stationary is transmitting torque with no relative motion, which it can do. But it has no connection to ground: if the driving side is not itself held, the whole assembly rotates. A clutch cannot stop anything; it can only connect it to something else.

Where a load must be held - a hoist, a lift, an inclined conveyor - the requirement is a brake, and specifically one that holds when power is removed.

Clutch-brake combinations exist precisely for this: a single unit that disconnects the drive and simultaneously applies a brake to the driven side, so the machine stops rather than coasting. On any application where the driven machine must stop promptly and stay stopped, that combination is usually the correct product rather than a clutch alone.

Over-cycling for the thermal rating, contamination of the friction faces, and incorrect air gap or adjustment - roughly in that order.

Over-cycling is the commonest. The heat from too many engagements per hour glazes the friction material, which reduces its coefficient of friction, which causes more slipping, which generates more heat. Once that cycle starts the clutch degrades quickly.

Contamination is the next. Oil or grease reaching a dry friction face - from an adjacent leaking bearing or seal, or from over-lubrication during maintenance - destroys its torque capacity immediately. A clutch that suddenly slips with no change in duty is usually contaminated rather than worn.

Adjustment matters on types with a wearing air gap. As the friction material wears the gap grows, engagement becomes slower and less positive, and heat rises. Units that require periodic gap adjustment need it on schedule; ignoring it produces a failure that looks like wear-out well before the rated life.