#800 1920 · Albert R. Thompson (Anderson-Barngrover Company) · Food processing / canning
Every valve that let cans into the pressure cooker had to open, so Thompson built one that never fully did
the problem
Canning required batch cooking in sealed retorts that couldn't run continuously
background
Into the 1910s, sterilizing canned food meant loading cans by hand into mesh baskets, locking them into a sealed retort filled with pressurized steam, cooking, then unloading — a start-stop batch process needing as many as 15 men per line, with cans sitting immobile long enough for heat to fully penetrate their centers since nothing agitated them during cooking. Everyone in the canning industry could see the fix in outline: make the process continuous, feeding cans in and pulling them out without ever stopping the line — but a genuinely hard mechanical problem stood in the way.
The obstacle was physical, not conceptual: any door or valve wide enough to let a can pass into a pressurized steam chamber would also let the pressure and steam escape the instant it opened. Anderson-Barngrover's own first attempt, unveiled in 1915, used sliding valves — resembling the sliding breech-lock of a bolt-action rifle — that opened and closed to admit each can. It worked, but the valves cycled too slowly, capping the whole system at roughly 50 cans a minute, cumbersome enough that engineers went back to the drawing board.
what everyone would do
The engineering instinct, and the one Anderson-Barngrover tried first in 1915, was to build a valve that opens and closes like a door: a sliding gate that retracts to let a can pass, then reseals behind it, resembling the sliding breech-lock of a bolt-action rifle. That design worked, in that it did let cans in and out of a pressurized chamber, but every open-then-close cycle takes real time, and cycling a heavy steam-tight gate fast enough for a real production line proved too slow, capping throughput at only about 50 cans a minute.
what they saw
Thompson saw that the problem wasn't how to make a valve open and close faster, it was that opening and closing at all was the wrong shape for the job. A valve that is always in transition — continuously rotating rather than snapping between two static states — never has to fully open, because at any given instant a different, still-sealed section of the same rotating wall is what's actually against the pressure boundary while a pocket elsewhere in its rotation receives or releases a can. The pressure seal and the can-transfer mechanism could be the same continuously moving part instead of two separate states one valve had to alternate between.
the move
Chief engineer Albert R. Thompson replaced the sliding valve with what ASME's own account calls "an ingenious rotary pocket valve, something like a revolving door on its side": a continuously spinning valve with pockets around its circumference, each pocket receiving one can from the feed line and, as the valve kept turning, delivering it through the pressure wall into the cooking reel — without ever opening a static gap between the pressurized chamber and the outside.
why it works
Because the rotary valve's pockets pass the feed point and the pressure-chamber wall in one continuous motion rather than pausing to open, the chamber is never exposed to atmosphere the way a sliding gate exposes it during its open stroke — containment and can transfer happen simultaneously rather than sequentially. That is what let throughput jump roughly eightfold over the sliding-valve design, since the system's speed limit was no longer how fast a heavy gate could physically cycle open and shut, but simply how fast the valve could rotate while still sealing cleanly.
the payoff
Because the valve was never actually "open" the way a door is open — at any instant, some solid part of the rotating wall sealed the boundary while a different pocket presented itself to the feed line — cans could pass continuously in and out of a fully pressurized steam chamber without a single break in containment. Throughput jumped roughly eightfold, from about 50 cans a minute with the sliding-valve design to as many as 400 cans a minute with the rotary version introduced in 1920, while cutting cook-room labor from as many as 15 workers to one and reducing steam consumption by roughly half.
where it breaks
The mechanism depends on manufacturing the rotating valve and its housing to extremely tight tolerances, since the seal's integrity rests entirely on how closely the moving wall fits its housing at every point in the rotation — any wear or misalignment that opens even a small gap defeats the whole principle, unlike a sliding gate whose seal can be inspected and adjusted while stationary. It also only helps when the item being transferred can survive being physically carried through the valve's rotation without damage or spillage; a continuously rotating lock is the wrong tool for anything that needs to sit still, or that a moving pocket would crush or disturb.
what came after
Anderson-Barngrover (later folded into FMC Corporation) made high-quality, safe, low-cost canned food available at industrial scale, and Thompson's rotary pocket valve — moving a continuous stream of product across a sealed pressure boundary via a constantly rotating lock rather than a periodically opened one — was designated a Historic Mechanical Engineering Landmark by both ASME (1982) and the American Society of Agricultural Engineers (1992). The same principle remains the basic mechanism behind continuous pressure sterilizers used in food processing today.
references
- [1]The FMC Rotary Pressure Sterilizer, Introduced in 1920 (International Historic Mechanical Engineering Landmark)American Society of Mechanical Engineers, 1982asme.org
- [2]FMC SterilizerInvention & Technology Magazine, 1992inventionandtech.com