Where the setting is permanent or near-permanent, and where the joint must resist axial movement as well as rotation.

The deep penetration gives the highest holding power of the common point styles.

Seated in a spot-drilled dimple it locates positively, so the hub cannot creep along the shaft under thrust.

That suits collars, stops and factory-set adjustments that should not change in service.

It is a poor choice wherever the assembly is dismantled regularly, because the shaft accumulates deep pits.

For frequently adjusted joints, a flat or dog point is the appropriate style.

Not to make it work, but a matching dimple is what turns the hold into a positive location.

Seated in a prepared recess the point bears on a cone rather than a point, so contact area is larger and stress on the screw is lower.

It also removes any guesswork on reassembly, since the screw finds the same place every time.

The drill angle should match the screw's included angle, or the point contacts on an edge and the benefit is lost.

On a plain shaft the screw makes its own pit, which holds well and cannot be positioned accurately.

Spot drilling through the tapped hole with the hub in place is the usual way to get the two features aligned.

More than any other point style - a deep conical pit that in practice stays for the shaft's life.

Dressing the pit out means removing material well below the original surface, which changes the shaft diameter locally.

The raised material around the pit will stop a hub sliding past until it is filed down.

Several pits along a seating length weaken the section and make disassembly awkward.

Where the shaft must survive many assemblies, the damage argues for a different point.

Where the fixing is permanent, the depth of the pit is precisely what makes the joint secure.

Whatever matches the recess it seats in - and standards differ, so the two should be specified together.

Blunter angles spread the load, penetrate less and are more robust.

Sharper angles penetrate more and hold better, at the cost of a more fragile point.

Different standards specify different nominal angles for the same nominal screw, so assuming a match between a spot drill and a screw from another source is risky.

A mismatch means the point sits on the edge of the recess rather than in it, which concentrates load on the screw tip.

Where both are being specified, using one manufacturer's stated pairing removes the question.

Seated in a dimple, yes - and for many designers that is the deciding argument for it.

The point sitting inside a conical recess resists movement in every direction, not just rotation.

A cup or flat point resists rotation well and axial creep poorly, because nothing stops the hub sliding.

On a plain shaft the self-made pit gives some of the same benefit, though it is shallower and less symmetrical.

For serious thrust loads a shoulder, circlip or locating collar is the correct feature, with the set screw preventing rotation.

Relying on a set screw alone to carry thrust is a common cause of slow creep in service.

The screw itself comes out normally; the difficulty is getting the hub off afterwards.

The raised metal around the pit acts as a stop against the hub bore.

Filing or stoning the burr down is usually enough to free it, and should be done before forcing anything.

On a heavily pitted shaft several marks may need dressing before the component will pass.

The screw's socket is the other risk - a cone point is often tightened hard, and a rounded socket on a headless screw is difficult to extract.

Applying penetrating oil and using a correctly sized key fully seated prevents most removal problems.

Hardened, and harder than the shaft - the point is doing the cutting and it is the most highly stressed feature.

Standard set screws are supplied through-hardened, and the hardness is stated by the manufacturer.

A soft screw against a hard shaft flattens the point on first tightening and holds by friction instead.

Stainless set screws are softer than hardened alloy steel ones, so their holding power is lower and they are chosen for corrosion resistance rather than grip.

In corrosive environments the trade-off between grip and material is real and worth deciding deliberately.

Hardened points are also brittle, so shock loading can chip a cone point where a cup point would deform.