Rockets and Horses
One of my favorite bits of Internet lore says the rocket boosters on the Space Shuttle were sized by the back ends of two Roman war horses. As told, it doesn’t hold up. The true version is better. Here it is, one stop at a time: back along the track to Rome, then forward to the launch pad.
-
United States, today
Standard gauge
The legend: English expatriates built America's railroads, so they used England's gauge.
Close. Early lines bought English engines, like the John Bull, and laid track to fit them. Plenty of others didn't, and it took a fight to settle. Congress overruled Lincoln's 5 feet in 1863. In 1886 the South moved one rail on 11,000 miles of track in two days, to 4 feet 9 inches, near enough to share cars.
-
England, 1825–1846
Stephenson's gauge
George Stephenson laid his railways to the gauge of the coal wagonways he grew up around, 4 feet 8 inches, then added half an inch of play. In 1829 the Rocket, built by his son Robert, won the trials to run on one of them. Brunel tried 7 feet on the Great Western, and lost. In 1846 Parliament made Stephenson's gauge the rule for new lines.
-
Northumberland, the 1700s
Wagonways
Before steam, horses hauled coal from the pits down to the river in wagons on wooden rails. Each wagon was built for one horse to pull, and each colliery chose its own spacing. Killingworth, where Stephenson worked, used 4 feet 8 inches.
-
England, the Middle Ages
Ruts
The legend: Any other spacing would snap an axle in the old Roman ruts.
Wheels did cut ruts into dirt roads, and a cart that didn't fit them had a rough ride. But no rut was standard. A cart was as wide as its horse and its builder made it, and the width changed from county to county.
-
The Roman Empire
Roman ruts
The legend: The ruts were made by Roman war chariots.
Roman wheels did wear grooves into stone. You can still see them in Pompeii's streets, roughly as far apart as modern rails. Carts and wagons hauling goods cut them, not chariots. The legions walked, and most of the roads Rome built in Britain were gravel.
-
Rome
Chariots
The legend: The chariots were sized for the back ends of two war horses, and everything after copied them.
Rome's army didn't fight from chariots; Rome raced them, in front of 150,000 people or more at the Circus Maximus. The big races ran four horses, the junior ones two. Horses set the width, not Romans. Anything built for a horse or two to pull ends up about this wide, in Rome, in Northumberland, anywhere.
-
Utah, the 1980s
The booster
The legend: A tunnel only a little wider than the track set the boosters' size.
Forward 2,000 years. Thiokol built the Shuttle's solid rocket boosters in Utah and sent them to Florida by train, in segments about 12 feet across. Rail is a big part of why they came in pieces. The joints between those pieces held the O-rings that failed on Challenger in 1986.
-
Florida, 1981–2011
Liftoff
The Shuttle flew 135 times. Every flight rode on boosters that had come from Utah by rail, in pieces made to fit the trip, on rails first laid out for horse-drawn coal wagons.
So no, a horse didn’t design the Space Shuttle. But rails laid out for horse-drawn coal wagons became the standard for most of the world, and the Shuttle’s boosters were built in pieces so they could ride those rails.
The headline in my mind is that design decisions propagate.
Stephenson’s choice of “whatever the coal wagons used” got passed along over and over, simply because each new line had to connect to the last. Changing it got more expensive every year. The American South needed an army of track crews and two days to do it. Britain needed an Act of Parliament.
At each stage in the design tree from horse to spaceship, no one said “welp, gotta keep this to the original horse spec.” They kept it because it was already built. No need to redo the work. What gets my attention is the boosters. Rail travel is a big part of why they came in segments, and a joint between segments is where Challenger’s O-rings failed in 1986. The Rogers Commission blamed the joint’s design, the cold, and the decision to launch, not the railroad. Still, a constraint nobody chose for rockets ended up built into one.
Back down on a micro level, the decisions we make can propagate farther than we anticipated. Writing new backend code for your company’s web app? Whatever comes next is probably going to build on top of it. Designing a database structure? Whatever comes next is probably going to build on top of it.
All of this is my long-winded way of saying: when you’re making something, take the time to understand the constraints that might be piggybacking on your work, and whether they’re holding you back.
I don’t remember where I first heard the booster and horse story, but I came across it most recently in Joe Celko’s “SQL for Smarties”, which I highly recommend.
Tangentially related: Richard Feynman explains why train axles don’t need differentials.