Engineering Community, Scottsdale, AZ (2026)

06/23/2026

People say it constantly: they do not build them like they used to. Look at a modern car after a bad crash and your gut says the carmaker failed. The front end is gone, the hood is folded, the whole thing looks like a crushed can.

They are right. They build them better now. And that wreck is the reason the people inside walked away.

For the first decades of car design, everyone believed the same thing. A safe car is a strong car. Build it stiff and solid and it will protect you. It feels obvious. It is also wrong.

A Mercedes-Benz engineer named Béla Barényi worked out why. In a crash, your car carries a certain amount of energy that has to go somewhere. If the car is rigid and stops almost instantly, all that energy is dumped out in a fraction of a second, and it passes straight through the structure into the people inside.

The car survives. The people do not.

His answer, patented in 1952, was to stop resisting the crash and start absorbing it. He split the car into three parts: a rigid cell around the passengers that must never deform, and softer zones front and rear designed to fold and crush.

When the front crumples, it stretches the crash out over time. Instead of stopping instantly, the car takes a few tenths of a second longer to stop.

That sounds like nothing. It is everything. The force your body feels depends on how fast you are stopped, and those few tenths of a second can turn a fatal crash into one you walk away from.

The crushed metal is not failure. It is the car spending the energy on itself so your body does not have to. Every fold is energy that never reached you.

The idea was so against the grain it took years to catch on. Today it is built into virtually every passenger car, and Barényi, who held more than a thousand patents, is one of the most important engineers most people have never heard of.

06/19/2026

America built ships faster than anyone in history. To beat the U-boats sinking cargo across the Atlantic, the Liberty ship program turned out more than 2,700 vessels, welding them together instead of riveting, sometimes launching one just days after laying the keel. It worked. And then the ships began to crack.

Across the wartime welded fleet there were nearly 1,500 serious brittle-fracture incidents. A few ships broke completely in two.

The most dramatic case is the one in the photo above: a sister tanker, the Schenectady. On 16 January 1943, fresh from her sea trials, she sat tied up at the dock in calm weather. Without warning, her hull split almost in half with a bang heard a mile away. The bow and stern sagged toward the riverbed while the middle rose out of the water. Nobody had touched her.

The blame fell on the shipyards. The work had been done fast, by crews with little experience, using welding techniques that were still new. It looked like bad workmanship.

It was not. A metallurgist at Cambridge, Constance Tipper, one of the first women to take the Natural Sciences Tripos there, ran the tests that changed the answer. She showed the cracks did not start in the welds.

The steel itself was the Problem.

Below a certain temperature, the steel stopped behaving like a tough, bendable metal and became brittle, the way glass is brittle. It could snap with almost no warning. The North Atlantic in winter was cold enough to cross that line.

This is the ductile-to-brittle transition, and in the 1940s almost no one designing ships understood it. The welds were not blameless. A riveted hull stops a crack at the edge of each plate. A welded hull is continuous, so once a crack starts, it can run the length of the ship. The square corners of the deck hatches gave those cracks a place to begin.

Tipper's work led to a test for the temperature at which steel turns brittle, and that test is why the steel in ships, bridges, and pipelines today is chosen to stay tough in the cold. The ships were never the failure. The understanding had not caught up to the engineering yet. She made it catch up.

We will go deeper into the materials science behind this in our newsletter, exploring why metals turn brittle, what fracture toughness is, and how engineers test for it. Live within 24 hours. We will put the link in the comments.

One last thing, from us. Thank you for making it all the way down here. These take us a while. We go down rabbit holes in old reports and original sources, argue over whether a detail is really true, and throw out the lines we cannot quite stand behind, because we would rather tell you less and have it be right.

The people in these stories worked hard to figure things out. The least we can do is get them down honestly. If something here taught you something, or made you look twice at a thing you walk past every day, that is the whole point of this. And if
we got something wrong, or left out a piece you happen to know, please tell us. We genuinely mean that. Some of the best things we have learned came from someone in the comments who knew more than we did. Mostly, thank you for spending a bit of your time here, on this thing we all quietly love. We do not take it for granted.

06/18/2026

In 1952, Britain was ahead of the world. The de Havilland Comet was the first jet airliner, fast and smooth, flying above the weather while everyone else flew propellers. Then it started coming apart in the sky.

On 10 January 1954, a Comet broke up at 27,000 feet over Elba, in Italy. All 35 aboard died. The fleet was inspected, modified, and put back in the air. Three months later, a second Comet disintegrated near Naples. All 21 died.

The world's most advanced airliner was grounded, and the cause of the crashes remained unknown.

The investigation, led by Sir Arnold Hall at Farnborough, did something nobody had tried. Most of the wreckage was on the seabed, so the team took a complete Comet fuselage, sealed it inside a giant water tank, and pumped it up and emptied it over and over, recreating the pressure of every climb and descent. After the equivalent of a few thousand flights, the fuselage tore open.

The cause was metal fatigue, barely understood back then. Every time the cabin pressurized for altitude, the skin stretched. Every descent, it relaxed. Do that enough times and tiny cracks form, and they form where stress builds up: at sharp corners.

The usual telling gets one thing wrong. It is normally told as "the square passenger windows." But investigators traced the critical cracking to sharp corners in small cutouts in the roof, openings for the navigation aerials, not the
big cabin windows everyone remembers. And the shape was only part of it.

Those panels were supposed to be glued to the structure. That proved hard in production, so they were riveted instead, with holes punched rather than drilled, leaving rough edges where cracks could start.

Three ordinary decisions, a sharp corner, glue swapped for rivets, a quicker way to punch a hole, stacked into a flaw no one knew to look for.

The Comet never got its lead back. Boeing watched, learned, and built the 707. But every rounded window and every fatigue test on every airliner since comes from this. The first jet age was paid for in advance, by people who had no idea the metal around them was counting.

06/15/2026

You have pushed one a thousand times without looking. A horizontal bar across a door that opens the instant you lean on it. It is one of the most quietly engineered objects in any public building, and it exists because of how people
die in crowds.

On 16 June 1883, more than a thousand children packed the gallery of Victoria Hall in Sunderland, England, for a variety show. As gifts were handed out below, the children rushed downstairs. At the bottom, a door had been bolted to leave a gap of about twenty inches, wide enough for one child at a time, so staff could check tickets. The children at the front were pushed into that gap by the weight of the crowd behind them. The bolt was on the far side. The adults could not reach it.

183 children, aged three to fourteen, were crushed to death.

Britain understood the engineering lesson immediately. A door that a crowd is pushing against cannot open inward, and it cannot rely on a lock someone has to reach.

By 1892, Robert Briggs had patented a horizontal bar that unlatched a door under the pressure of the crowd itself.

The thing pushing the door became the thing that opened it.

But the lesson did not cross the Atlantic.

On 30 December 1903, the Iroquois Theatre in Chicago, two weeks old and advertised as fireproof, caught fire during a sold-out show. Many exit doors opened inward. Some had been locked to stop people slipping in without tickets. More than 1,700 people were inside.

602 did not get out. It remains one of the deadliest building fires in American history.

Carl Prinzler, a hardware manager, had planned to be at that show. His plans changed, and he spent the next years building the answer with an architect named Henry DuPont: a single bar at waist height that stays locked from outside but
opens from inside the moment a body presses against it. They filed the patents in 1908 and sold it under a name stitched from three surnames, Von Duprin.

The first one went on a school door in Indianapolis.

The bar is the part you can see. The harder question is how engineers get an entire crowd out of a building alive: occupant loads, exit widths, why doors must swing outward, how far you can ever be from a way out. That is in this week's
newsletter.

06/13/2026

In 1967, the Soviet Union revealed the MiG-25. Western analysts saw giant wings and twin engines and concluded it was a superfighter: near Mach 3, with the agility to out-turn anything.

They were designing against a ghost.

The truth came out in 1976, when a Soviet lieutenant defected to Japan with his MiG-25 and engineers took one apart. We told the Foxbat's half of this story already. The short version: a heavy steel interceptor built for straight-line speed, redlined around Mach 2.8, that could barely pull a hard turn. The big wings existed to lift its weight, not to turn.

This is the other half. The jet the West built because of it.

But in 1968, nobody knew that. So the US Air Force wrote the requirements for a new fighter against the airplane it feared, not the one that existed.

An analyst named John Boyd shaped the brief. His energy-maneuverability theory, a way of scoring a fighter by how fast it gains and sheds energy, showed the early 60,000-pound concepts were far too heavy. The requirement came down to a 40,000-pound class machine built to out-accelerate and out-turn everything. Air superiority and nothing else. The rule became famous: not a pound for air-to-ground. McDonnell Douglas won the contract two days before Christmas, 1969.

That one rule shaped everything. A thrust-to-weight ratio above one, so the jet could accelerate while climbing straight up. Two engines, each making about eight pounds of thrust per pound of weight. A wing so large that each square foot carries less load, which is what lets the F-15 turn hard without bleeding away its speed. A bubble canopy with full 360-degree vision. A nose built around a radar big enough to spot the enemy first.

The record followed the discipline. Across five decades and three air forces, the F-15 has never been confirmed lost in air-to-air combat.

The strange part is what the misread produced. The superfighter never existed. But because the engineers built against the threat as imagined, with one ruthless rule, they put so much margin into the airframe that it is still in production today, more than 50 years later.

And the irony: when the Air Force needed its best strike aircraft, it picked an Eagle variant, the F-15E. The jet built under "not a pound for air-to-ground" became one of the best air-to-ground aircraft ever made. The rule was never about the bombs; it was about refusing to compromise.

06/05/2026

Most aircraft get faster as they are redesigned. The B-1 got slower, and the reason is one of the better engineering stories in Cold War aviation.

In the 1970s, the B-1A was built around a simple idea: outrun the threat. Fly high, dash across Soviet airspace at Mach 2.2, and cross the defended zone faster than anything could react.

To hit that speed, the engineers gave it variable-geometry engine intakes. The ramps inside the inlets physically moved to reshape the airflow and manage the shockwaves that form at supersonic speed.

Then the ground changed underneath it.

By the late 1970s, Soviet radar and surface-to-air missiles had improved enough that flying high and fast was no longer a way to survive.

Look-down radar could spot a bomber against the ground. Missiles could reach it at altitude. Speed had stopped being protection.

The B-1A was cancelled in 1977.

When the program came back in the 1980s as the B-1B, the entire theory of survival had flipped.

Instead of flying high and fast, the new plan was to fly low, hugging the terrain under the radar horizon. And instead of being too quick to catch, be too faint to see.

That second goal forced a hard trade. Those moving intake ramps were also strong radar reflectors, bouncing signals straight back from the engine faces.

So the engineers ripped them out and replaced them with simple fixed intakes. The aircraft's radar signature dropped to roughly a fiftieth of the B-52's, about the size of a small fighter. But fixed intakes cannot manage supersonic shockwaves the way the moving ones did.

The price was speed. The B-1B can no longer reach Mach 2. It tops out around Mach 1.2 to 1.25.

They gave up the airplane's headline number to make it harder to detect, because survival no longer came from speed. It came from not being seen.

To be clear, the B-1B is not a stealth aircraft like the B-2. It is a large bomber with a deliberately reduced radar signature.

And that is the whole story in one airframe. The wing was built to go fast. The fixed intakes are what replaced that speed when speed stopped keeping bombers alive.

06/04/2026

A fighter that flies off an aircraft carrier has to do two things that fight each other.

To land on a carrier deck, it has to fly slow. That needs a big, wide wing that makes a lot of lift at low speed. To intercept a Soviet bomber before it can launch missiles at the fleet, it has to fly at twice the speed of sound. That needs the opposite wing: thin and sharply swept, to cut drag.

One wing shape is good for landing. The other is good for speed. No single fixed wing does both well.

Grumman's answer, in the F-14 Tomcat, was a wing that changes shape in flight.

The wings pivot. Forward to 20 degrees for takeoff, landing, and slow flight. Back to 68 degrees for supersonic speed. As the wings sweep back, the aircraft's wingspan shrinks from 64 feet to 38 feet.

Spread wide, the Tomcat is stable and controllable at the low speeds a carrier approach demands.

Swept back, it becomes a clean supersonic dart. The same aircraft, two different wings, chosen by the moment.

This is where it gets harder than it sounds. The correct sweep angle changes constantly with speed and altitude. Doing it by hand would be a full-time job for the pilot, and a slow one, at the exact moments the pilot has no attention to spare.

So the F-14 did it automatically. A computer in the nose, the Central Air Data Computer, read the aircraft's Mach number and altitude many times a second and drove the wings to the right angle on its own.

The F-14 was one of the few aircraft anywhere with fully automatic, computer-controlled wing sweep. The pilot could override it, but most of the time the machine flew its own wings.

The wing is the famous part. The computer that flew it is not. That computer was built around one of the earliest microprocessors ever made.

Two engineers at Garrett AiResearch, Steve Geller and Ray Holt, led a 25-person team that designed it for the F-14, and it was running in the jet by 1970, more than a year before the Intel 4004, the chip usually called the first microprocessor, reached the public. The Navy classified the design. It stayed secret until 1998.

So the wing is the icon. Every photo of a Tomcat shows it. But the wing only flew itself because of a quiet computing breakthrough hidden in the nose, kept out of the textbooks for three decades.

We saw the same pattern in our post on AlexNet and CUDA. The breakthrough everyone remembers, the swing wing, the neural network, rode on a leap in computing that almost nobody was looking at.

06/03/2026

In 1913, the inventor of the diesel engine vanished from a ship and was never seen alive again.

On the night of September 29, Rudolf Diesel boarded the steamer Dresden from Antwerp toward England. He had dinner, went to his cabin, and laid out his nightclothes. His bed was never slept in. His coat was found folded by the deck
rail. Ten days later, a body was recovered at sea.

The headlines reached for murder. Big Oil. German spies. The truth was quieter and sadder. He was nearly broke.

But the real story is what he built, and what happened to it. In 1897, Diesel ran an engine that converted about 26% of its fuel into work,
when the best steam engines wasted around 90% of theirs. No spark plug. It compressed air until it was hot enough to ignite the fuel by itself.

He designed it to burn almost anything. And deliberately, that included vegetable and peanut oil. At the 1900 Paris World Exposition, one of his engines ran on peanut oil. He imagined farmers and countries with no coal and no oil
growing their own fuel in a field. In 1912 he said vegetable oils might one day become "as important as petroleum."

Then the petroleum industry found a cheaper path. A byproduct of refining could power his engine, and it was easier to store than crop oil. That byproduct is what we now call diesel fuel. The engine built to free people from oil became one
of its biggest customers.

Diesel never saw how it played out. His patents had expired. His investments had failed. Before he left, he handed his wife a bag and told her not to open it for a week. Inside was 20,000 marks and papers showing their debt.

The engine kept his name. His actual vision was buried for a century. Today biodiesel is a multi-billion-gallon industry, running in ordinary diesel engines.

Why his engine could burn peanut oil when a gasoline engine never could is an engineering story in itself. That one is in our newsletter: https://engineers.beehiiv.com/p/the-engine-that-was-built-to-burn-anything

06/01/2026

In 1965, Pan Am president Juan Trippe asked Boeing for an airliner that could carry 400 passengers, cross any ocean non-stop, and cut the per-seat cost by 30%. Boeing's CEO Bill Allen agreed on a handshake.

"If you build it, I'll buy it," Trippe said.

"If you buy it, I'll build it," Allen replied.

Then Boeing had to figure out how.

The job went to Joe Sutter, a 44-year-old engineer Boeing pulled off the 737 program. He had 29 months from concept to rollout. He had to design an aircraft 2.5 times the size of anything Boeing had ever built. And he had to do it in a factory that did not yet exist.

The first disagreement was about the shape. Trippe wanted a double-decker. Essentially two 707 fuselages stacked. It was the obvious move. Single-aisle was the only proven commercial layout in 1965.

Sutter saw the problem. The FAA required all passengers to evacuate in 90 seconds. Getting people off an upper deck through slides that long was not possible with 1960s engineering.

So Sutter pushed back.

He proposed a single deck with two aisles. A fuselage wider than anything in the air. Nobody had built a twin-aisle wide-body airliner before. His team built two full plywood mockups side-by-side inside a Boeing hangar. Trippe walked through both. When he came out, he said: "You made the right decision."

The next disagreement was about the cockpit. Boeing and Pan Am both believed supersonic jets would replace the 747 within a decade. So they designed it to convert into a freighter. To load cargo through the nose, the entire nose had to hinge open. The cockpit could not stay there. They moved it up onto a bump on the upper fuselage. Engineers then extended the bump aft into a teardrop to reduce drag. That teardrop became the most recognizable silhouette in aviation.

Sutter's team of 4,500 engineers worked nights and weekends through the 29-month sprint. Boeing nicknamed them The Incredibles. The first 747 rolled out of the new Everett factory, then the largest building in the world by volume, on September 30, 1968. It entered service with Pan Am on January 22, 1970.

The supersonic jets that were supposed to replace it never did. The Boeing 2707 was cancelled. Concorde flew for 27 years and stopped. The 747 stayed in production for 55 years. Boeing delivered the last one on January 31, 2023. They built 1,574 in total.

The wide-body twin-aisle layout Sutter fought for in 1965 is now the default for every long-haul aircraft in the sky.

Engineering Community

06/23/2026

06/19/2026

06/18/2026

06/15/2026

06/13/2026

06/05/2026

06/04/2026

06/03/2026

06/01/2026

Address

Website

Alerts

Shortcuts

Share