Everyone Says the Phillips Screw Was Designed to Slip. It Wasn't.

A popular story says the Phillips head was built to slip out at a set torque, so factory workers could not overtighten screws. It is a good story with no good evidence behind it. The real selling point was a socket that centers the driver by itself.

By Marcus Feld
Everyone Says the Phillips Screw Was Designed to Slip. It Wasn't. (patent US2046343)

US2046343 was filed on July 3, 1934, and granted to Henry F. Phillips of Portland, Oregon, on June 30, 1936. The title is simply "Screw." It covers the cross-recess screw head that carries his name and sits in most of the furniture, appliances, and cars built in the last ninety years.

Phillips did not invent it. He bought it.

The design came from an Oregon inventor named John P. Thompson, who had patented a cross-recessed screw in the early 1930s and could not persuade any manufacturer to make it. Thompson sold the rights to Phillips, who was a businessman rather than a machinist. Phillips refined the geometry, filed his own patents, and then did the thing Thompson could not, which was to get an industry to adopt it.

Why slots were failing

The flat-blade screw had a specific problem on an assembly line. A screwdriver in a slot wants to slide sideways out of it, and the harder you push, the more likely the blade is to skate off and gouge the workpiece. Slotted screws also cannot be started easily with one hand, because nothing holds the driver in place.

Detroit was building cars faster every year and wanted power screwdrivers. Slotted heads and power drivers are a bad combination. The blade cams out of the slot at speed and damages the paint, the panel, or the operator.

The cam-out story

Phillips solved the alignment problem with a tapered socket. The four-point recess is self-centering, so a driver dropped near the head falls into position, and it holds the screw well enough to start one-handed.

The taper has a side effect. Turn hard without pushing hard and the driver rides up and out of the recess. That behavior is called cam-out, and a popular story holds that it was the point: with no cheap torque clutch for production screwdrivers, the screw itself would limit the torque, popping the driver out before a worker could strip a thread or crack a casting.

It is a satisfying explanation, and there is no good evidence for it. The original patents do not present cam-out as a feature; what they describe is the shape of the recess and how it guides the driver. Cam-out was a byproduct of the taper that production lines learned to live with.

General Motors tried the screws on its Cadillacs in the mid-1930s and adopted them widely. Most of the American auto industry followed within a few years.

Why it feels broken now

Modern power tools have adjustable clutches, so even the imagined benefit is gone. What remains is a screw head that slips under load, chews itself into a useless cone, and frustrates anyone trying to remove an old fixture.

Later designs fixed it by refusing to cam out at all. Pozidriv adds a second set of ribs. Robertson, the square socket used widely in Canada, grips without slipping and predates Phillips by decades. Torx uses a six-lobe star that transmits high torque without pushing the driver out.

All of them are better fasteners. Phillips still dominates, mostly because the tooling, the drivers, and the habits are everywhere and switching costs more than it saves.

See the original

The full text and figures of US2046343 are on Google Patents.

Patent drawing by Henry F. Phillips for "Screw", U.S. Patent 2,046,343 (1936)
Sheet 1 of 1 of the drawings for U.S. Patent 2,046,343, Screw, by Henry F. Phillips, issued July 7, 1936. Public domain. Read the full patent on Google Patents →

Related stories