Phillips Screw

Phillips specifies flat-bottomed grooves whose geometry allows the recess to be made by a simple punching operation, which also work-hardens the head. Manufacture is claimed, not just shape.

Hold only the screwdriver and try to start a slotted screw into softwood. It falls off. Time how long it takes to get it biting.

Repeat one-handed with a Phillips screw and driver. It seats and stays. Record both times — the difference is the invention's entire commercial case.

Bring the driver in at an angle and off to one side. The tapered recess walls drag the tip to the middle. This is self-centring, and no slot does it.

Run twenty slotted and twenty cross-recess screws. Count how many times the driver slipped out of each.

Look at the wood beside the slotted screws. A blade escaping a slot travels sideways at speed and scores the workpiece — the cost is not only lost time.

Keep turning past seating. At some torque the driver rises out of the recess and spins. This is cam-out, and it is the most argued-about property of the design.

Keep driving a slotted screw. It does not cam out; it either drives on, twists off, or the blade destroys the slot. Note which failure you get.

Use a rounded, worn tip. Cam-out happens far earlier and chews the recess. Most of the design's bad reputation comes from worn drivers, not from screws.

Hammer a cross-shaped punch into a scrap of soft metal or thick foil. One blow, full shape, no cutting and no swarf.

Saw a slot in a second piece. It takes longer, removes material and leaves a weaker head. Punching displaces metal instead of removing it.

Scratch-test beside the punched recess and beside the sawn slot. The punched region resists more — cold working hardened it, which the patent claims as a benefit.

Drive one with a slotted screwdriver. It engages two arms only, cams out almost immediately and rounds the recess. The system is a matched pair, not a screw.

The patent. US 2,046,343, "Screw", filed 3 July 1934 and granted 7 July 1936 to Henry F. Phillips of Portland, Oregon.

Phillips was a promoter more than an originator, and that is not a criticism. Cross-recess screw heads had been proposed before; the widely repeated account is that Phillips acquired and substantially developed an earlier design rather than conceiving it from nothing, and his real achievement was making it manufacturable and then getting it adopted. Screw manufacturers initially refused it as too difficult to produce. The breakthrough was General Motors, which used the screws on the 1936 Cadillac and found the assembly-line savings decisive; the rest of the industry followed within a few years.

Why an assembly line cared so much. Three properties matter when a screw is driven by a machine or a worker doing it a thousand times a shift. It can be started with one hand, so the other hand holds the part. It self-centres, so a powered driver does not need a person aiming it. And it resists the driver wandering, so it does not score painted bodywork. Steps 3 to 6 measure exactly these, and the numbers are why a car company would re-tool for a fastener.

The cam-out argument, stated honestly. It is very commonly claimed that Phillips designed cam-out deliberately as a torque limiter, to stop early powered drivers over-tightening and stripping threads before torque-limiting clutches existed. This is disputed. Other accounts hold that cam-out is a side effect of a recess shaped for punching and self-centring, and that Phillips regarded it as a defect to be minimised — later designs such as Pozidriv exist specifically to reduce it. The patent text itself emphasises maximum bearing surface and manufacturability, not torque limiting. Both stories are repeated confidently; the document supports the second more than the first. Step 7 lets you form your own view, and step 9 shows how much of the effect is simply tool wear.

What came after. The cross recess is now one design among many — Pozidriv adds ribs to resist cam-out, Torx uses a six-lobe star with near-vertical walls to eliminate it almost entirely, Robertson's square socket predates Phillips and beats it on grip. Phillips won not by being the best transmitter of torque but by arriving at the moment mass production needed a self-centring fastener, and by being adopted by an industry large enough to make it a standard. Step 13's mismatch is the lasting lesson: a screw drive is a system, and half of it lives in your hand.