Westinghouse Air Brake

Keep test pressures low, use fittings rated well above what you run, and never stand in line with a plugged end. A failed joint throws its fitting hard.

Westinghouse's 1869 patent is a straight air brake: air is sent to apply. The fail-safe inversion is not in this patent — knowing that is the point of the exercise.

Make a locomotive and two wagons that roll freely and couple together. The demonstration needs a TRAIN — a single braked vehicle shows none of the interesting behaviour.

Mount a shoe that presses a wheel when a lever is pushed. Spring it off so it releases cleanly when nothing acts on it.

Fit a small pneumatic cylinder per wagon, piston rod to brake lever. This is the brake cylinder, and it is the only part that touches the mechanism.

Join wagon to wagon with hose and quick couplings so the pipe is continuous and can be parted deliberately.

Connect the train pipe straight to every brake cylinder. Open the driver's valve and all three wagons brake together. It works, and it is a real improvement on brakemen.

Pull a coupling apart mid-application. Every brake releases instantly. The detached wagons roll away completely free — this is the flaw, and you have just reproduced it.

Fit a small air tank per vehicle. Each wagon now carries its own energy supply and no longer depends on the pipe for the force to brake.

Install a valve on each wagon that senses train-pipe pressure and does three things: charge the reservoir, hold the brake off, or connect reservoir to cylinder when pipe pressure FALLS.

Pressurise the pipe. The reservoirs fill and the brakes stay off. Air now holds the brake OFF instead of putting it ON — the whole logic has flipped.

Vent a little from the driver's valve. Every triple valve sees the drop and admits its own reservoir air to its own cylinder. Less pressure, more braking.

Part a coupling. Pressure collapses, every triple valve fires, and the runaway half brakes itself hard. That is the entire safety argument, demonstrated on a bench.

Watch how long the rearmost wagon takes to respond. The signal travels at the speed the pressure wave moves down the pipe — this lag is why long freight trains take so far to stop.

The patent. US 88,929, "Improvement in Steam-Power-Brake Devices", granted 13 April 1869 to George Westinghouse, Jr. of Schenectady, New York. He was twenty-two. The system it describes is the straight air brake: a compressor on the locomotive, a pipe running the length of the train, and air sent along it to apply the brakes on every vehicle simultaneously. Compared with brakemen running along carriage roofs turning handwheels, it was transformative — and it still had the defect that any leak, burst hose or parted coupling released every brake at once.

The inversion is the real invention. Westinghouse's automatic air brake, developed in the early 1870s, reverses the sense of the signal. The train pipe is kept charged, and that pressure is what holds the brakes OFF while also keeping a reservoir on each vehicle topped up. To brake, the driver REDUCES pipe pressure; each wagon's triple valve detects the drop and connects its local reservoir to its local brake cylinder. A break in the pipe therefore causes a full emergency application on every vehicle, including any that have separated. This is fail-safe design in its clearest form: the failure mode of the control system is made identical to the safe command.

The triple valve is where the intelligence lives. It is a purely pneumatic device that compares train-pipe pressure against reservoir pressure and selects one of three states — charge, hold, or apply. No electricity, no wires, no central computer; each vehicle decides for itself using a signal that is broadcast by absence. Distributed control implemented in valve geometry, and it has been running the world's railways for a century and a half.

Why it had to be made compulsory. Railroads resisted the cost. In the United States the Railroad Safety Appliance Act of 1893 mandated air brakes and automatic couplers, and casualty rates fell sharply thereafter. The physics has one stubborn consequence you measured in step 14: because the command propagates as a pressure wave down the pipe, rear vehicles apply later than front ones. Long freight trains therefore brake progressively rather than instantly, which is why modern systems add electronically controlled pneumatic braking to signal every wagon at once — an improvement to the messenger, not to Westinghouse's logic.