Because a pin physically passes through the nut and the bolt - nothing about it depends on friction.

Every other locking nut resists rotation by friction, deformation or bite, all of which can be overcome.

A cotter pin through aligned slots and a drilled hole mechanically prevents rotation.

The lock does not weaken with vibration cycles, temperature or reuse.

It is also visible: an inspector can see at a glance whether the joint is secured.

That inspectability is why safety-critical applications specify it.

A castle nut has a raised crown with the slots cut into it; a slotted nut is cut into a standard-height body.

The crown preserves full thread engagement below the slots, so the nut develops its full strength.

A slotted nut is simpler and cheaper but has slightly less engaged thread below the cuts.

Both work identically once the pin is fitted.

Where the joint is highly loaded, the castle form is generally specified.

Terminology varies between markets, so the dimensions matter more than the name.

Tighten to the specified torque and then advance to the next slot - never back off to reach one.

The two features will rarely coincide exactly at the target torque.

Advancing slightly increases preload, which is safe; retreating leaves the joint under-tightened, which is not.

There is normally a stated limit on how far the nut may be advanced.

If the next slot is out of reach within that limit, a washer of different thickness changes the relationship.

Backing off to find a slot is the classic misuse and defeats the point of specifying a torque.

Yes - some wheel bearing arrangements specify a low torque followed by alignment, because the bearing must not be preloaded.

In those joints the nut is retaining rather than clamping, and over-tightening damages the bearing.

The manufacturer's procedure often specifies seating the bearing at a higher torque and then backing off to a low figure before aligning.

That is one of the few cases where backing off is correct, and it is correct only because the procedure says so.

Applying a general rule instead of the specific procedure damages the assembly.

Anything involving bearings, steering or suspension should follow the vehicle or equipment manufacturer's instruction exactly.

No - it work-hardens when bent, straightened and re-bent, and can fail.

A pin is deformed permanently when its legs are opened over the nut.

Straightening it to remove it and bending it again fatigues the material at exactly the point that carries load.

Pins cost almost nothing against the consequence of the joint they secure.

The correct size matters as well: a pin that is loose in the hole allows movement and wears both parts.

Stainless pins are used where corrosion would otherwise seize or weaken the pin.

Then a castle nut is the wrong fastener - the hole is part of the system, not an optional extra.

Bolts intended for castellated nuts are supplied with a cross-drilled hole in the correct position.

Drilling a hole in an ordinary bolt reduces its section at a stressed point and is not a substitution to make casually.

Hole position has to suit the joint's grip length so the slots can align.

Where a positive lock is wanted on a plain bolt, wire locking or a different mechanism is used instead.

The bolt, nut and pin are specified together in designed joints.

Wire locking, tab washers and lock plates all achieve a positive lock by different means.

Safety wire threaded through drilled fastener heads is standard in aerospace and motorsport.

Tab washers are bent up against a nut flat after tightening and are common in machinery and gearboxes.

Lock plates secure several fasteners at once and are used in heavy plant.

All share the castle nut's advantages: no dependence on friction and visible confirmation.

All share its disadvantage too, in that installation takes longer and requires a prepared fastener.