Because the safe condition - brake on - is the condition that exists when nothing is working.

The springs are always trying to clamp the brake. Power is used only to hold them off. So the brake engages not just on a deliberate command but on any failure that removes the release signal: a power cut, a blown fuse, a broken wire, a failed contactor, an emergency stop, a tripped protective device, or a controller fault.

None of that depends on the control system doing the right thing. The safe state is the default physical state of the mechanism, and it requires energy to leave it.

That is why standards for lifting equipment, for vertical axes and for machine safety functions call for this construction rather than a power-applied brake with a battery backup or a well-designed control circuit - both of which can fail in ways the spring cannot.

From the static load torque multiplied by a safety factor set by the governing standard - not from the stopping torque and not from the motor rating.

The brake must hold the worst-case load torque at the shaft it acts on, indefinitely, without creeping. That torque comes from the suspended or gravity load referred through the drive train to the brake location.

The safety factor is prescribed rather than chosen. Lifting standards typically require the brake to develop substantially more torque than the maximum load torque, so that wear, contamination, temperature and manufacturing tolerance cannot bring it below what is needed.

Check which shaft the brake acts on and apply the gear ratio correctly - a brake on the motor shaft sees the load torque divided by the ratio, which makes it smaller and cheaper but leaves it dependent on the integrity of everything between it and the load.

Also confirm whether the standard requires a second, independent brake acting directly on the load.

That depends on the application, and it is worth deciding explicitly because it changes the sizing.

In normal operation many machines stop under drive control, with the spring-applied brake engaging only once motion has ceased. Used that way the brake does almost no work, dissipates almost no energy, and its lining lasts a very long time - it is purely a holding brake.

But on an emergency stop or a power failure the drive control is gone, and the brake becomes the stopping device, absorbing the full kinetic energy of the load. It must be capable of that even if it happens rarely.

So the selection has two parts: static torque with the standard's safety factor for holding, and enough thermal and mechanical capability to survive emergency stops from full speed with the load. If emergency stops are frequent - as on some machines they are - the thermal duty becomes a real design consideration rather than an exceptional case.

Because a brake that holds without power also traps the load, and recovery requires releasing it deliberately.

If a hoist stops with a load suspended and the supply is gone, the brake is doing exactly its job - and nothing can be lowered until it can be opened. A manual release lets an operator or engineer free the brake in a controlled way so the load can be lowered and the machine recovered.

The critical design requirement is that it cannot be left released. A brake propped open by a release lever provides no protection at all, and someone will do it if the mechanism allows. Releases are therefore commonly spring-returning, or require continuous hand pressure, or use a removable lever that cannot be stowed in place.

On passenger-carrying and lifting equipment the arrangements are prescribed by the governing standard, including how lowering speed is controlled during manual release. Follow that rather than treating it as a convenience feature.

Contamination, wear beyond the adjustment range, weakened or broken springs, and a coil that fails to fully de-energise.

Contamination is the most abrupt. Oil or grease reaching the friction faces - from an adjacent leaking motor bearing seal, or from over-greasing during maintenance - collapses the torque immediately. A brake that suddenly creeps is usually contaminated rather than worn.

Wear widens the air gap. On an electromagnetic release, that reduces the force the coil can produce, which affects release rather than holding - but excessive wear eventually takes the mechanism outside its working range.

Springs weaken or break, and because they are the thing producing the torque, a broken spring is a direct loss of capacity that nothing else reveals.

A coil that stays partially energised - through a leaking suppression component or a stuck contactor - holds the brake partly released, so it neither holds properly nor is obviously failed. Testing holding torque periodically is the only way to find several of these.

By applying a known torque and confirming the brake does not slip, at intervals set by the governing standard and with the results recorded.

On lifting equipment this is normally a specified periodic test: the brake is required to hold a defined overload - typically a test load exceeding the rated load - without creeping, and the test is documented as part of the statutory examination regime.

On other machinery, a practical check is to confirm the brake holds the maximum load without movement and to observe whether any creep occurs over time.

What should not be relied on is the absence of complaints. A brake that has lost part of its torque may hold the everyday load perfectly while no longer having the margin the standard requires, and nothing about its behaviour reveals that.

Record the test results over time. A trend of decreasing margin is far more useful than a single pass, and it gives warning before a failure.

Wherever a power failure could allow a load to fall, run back or move dangerously - and in most of those cases a standard makes it mandatory.

The clear cases are hoists, cranes and winches; lifts and vertical conveyors; vertical and cantilevered machine axes; robot arms and manipulators; inclined conveyors carrying material; and any stored-energy system that would release on power loss.

In each, gravity or stored energy continues to act when the electricity stops, and the brake is the only thing that will.

Beyond lifting, they are also used as part of machine safety functions where a guard interlock must bring a hazardous motion to a stop and keep it stopped - the brake forming the final element of a safety-related control function, with its own performance and diagnostic requirements.

Check the standard that applies to the machine type and market. It will usually specify not only that the brake is spring-applied but the torque factor, the testing regime and whether a second independent brake is needed.