Here's the thing about marrying an architect — you think you understand buildings until she says something like "concrete has terrible tensile strength" over dinner, and suddenly you realize you've spent your entire life not knowing what a building is actually doing. Hannah sent Daniel down a rabbit hole, and Daniel sent it to us. Here's what he wrote.
"One of my favorite things about talking to my wife Hannah, who's an architect, is that every so often she'll mention a concept that people in her profession take completely for granted, and I'll realize I don't actually understand it at all. Recently she was explaining why concrete is incredibly strong in compression but surprisingly weak in tension, and why that's such a fundamental idea in structural engineering. We got into talking about stone construction, reinforced concrete, and how buildings have evolved over the centuries, but I quickly realized I was missing the basic intuition behind what these different kinds of strength actually mean."
He goes on — "Could you give me a proper for-dummies explanation of tensile strength versus compressive strength? Start with those two concepts, then expand into the other important forces and material properties — shear, bending, torsion, ductility, brittleness, and anything else that's essential to understanding how structures stand up. Why is concrete so good in compression but poor in tension? Why does embedding steel reinforcement transform its capabilities? Before reinforced concrete existed, how did builders using stone work around these limitations? Is that why we see arches, domes, vaults, flying buttresses, and massive masonry walls throughout history?"
And then the part I really like — "More generally, I'd love to understand how engineers and architects think about forces. Once you have that mental model, what are some everyday examples — from bridges and skyscrapers to shelves, chairs, ladders, and even tree branches — where you can immediately recognize which parts are under tension, which are under compression, and why different materials are chosen for each? The goal isn't to learn the equations, but to come away seeing the built world through a completely different lens."
That's a prompt that earns its keep. So today we're going to give you that lens — and by the end, you'll never look at a bridge, a bookshelf, or even a tree branch the same way again.
The question hiding inside all of this is deceptively simple. Why does a material that can hold up a skyscraper crumble when you try to hang something from it? That's not intuitive. If something is strong, it should be strong, right?
That was my assumption. Strong is strong.
And it's wrong. Strength is directional. The same concrete column supporting thirty floors above it would snap if you tried to use it as a beam across a gap — same material, same dimensions, completely different outcome. So let's start with what's actually happening inside the material.
There are really only two fundamental ways to load something. Tension is when you pull the atoms apart — the material lengthens, stretches, the atoms are trying to get away from each other. Compression is when you push them together — the material shortens, the atoms are being squeezed closer than they'd like to be. At the atomic level, atoms resist both. They have an equilibrium spacing they want to maintain, and any force that tries to change that spacing — in either direction — they push back.
So it's not that atoms prefer being squeezed to being pulled. They dislike both.
They dislike both, but here's the asymmetry. Tension tends to be self-correcting for small deflections — if a rod under tension starts to bow sideways, the pulling force actually straightens it back out. Compression does the opposite. If a column under compression starts to bow even slightly, the compressive force amplifies that bow. It pushes the deflection further and further sideways until — buckle. The column folds. And that happens long before the material itself is crushed.
Wait. So buckling isn't the material failing — it's the shape failing?
The material is fine. The geometry gave up. A long, slender column under compression will buckle at a load far below what the material could actually handle if you could keep it perfectly straight. That's why a drinking straw crumples when you push the ends together, but if you pull it, it stays straight until the plastic itself tears. Same straw.
Okay, so that's the physics of tension and compression. Now concrete.
Concrete is a mixture of aggregate — gravel, sand — bound together by cement paste. When you add water, the cement hydrates, which means it forms microscopic crystal lattices that grow and interlock, gripping the aggregate particles. Those crystal lattices are extremely good at resisting being crushed. They lock together like a three-dimensional jigsaw puzzle where every piece is braced against its neighbors.
And when you pull?
The lattice has almost no ability to resist being pulled apart. Those interlocking crystals separate, cracks form and propagate, and the whole thing comes apart. The numbers tell the story. Typical concrete has a compressive strength around twenty-eight megapascals — that's about four thousand pounds per square inch. Its tensile strength? Two to five megapascals. That's an order of magnitude weaker.
An order of magnitude. So for every ten pounds it can hold in compression, it can hold about one pound in tension.
Roughly, yes. And that single number — that ten-to-one ratio — explains more about the history of architecture than almost anything else I can think of.
Before we get to the history, Daniel also asked about the other forces. Shear, bending, torsion.
Shear is the simplest to picture. Imagine a stack of papers on a desk. Push sideways on the top sheet. It slides relative to the one below it. That sliding force across a plane is shear. In a beam, shear is what tries to slice it vertically — like a giant pair of scissors. Bolted connections in steel frames have to resist shear constantly.
Torsion is twisting. Take a wet towel and wring it out — that's torsion. Drive shafts in cars experience it, and tall buildings have to resist wind trying to twist them. It's less common in everyday buildings than bending, but engineers design for it.
And bending — bending is where the real insight lives. When you load a beam, the top of the beam gets squeezed. The bottom gets stretched. Simultaneously. The same piece of material is in compression on top and tension on the bottom, with a neutral plane somewhere in the middle that experiences neither.
This is the hidden tension trap Daniel's talking about.
It's the thing most people get wrong. They think a beam is "holding weight" as if it's one uniform job. It's not. The top half is doing compression work, which concrete loves. The bottom half is doing tension work, which concrete hates. That's why a plain concrete beam, with no reinforcement, will crack on its underside while the top looks perfectly fine. The tension side failed.
And that's where rebar lives.
Near the bottom of the beam, exactly. Because that's where the tension is. The rebar isn't randomly scattered through the concrete — its position is a map of where the tension stresses are highest.
Before we get to rebar, I want to sit with bending for another second. You said the top is compressed and the bottom is stretched. Is that true for any horizontal span, regardless of material?
Any horizontal span under vertical load. A wooden shelf, a ladder rung, a bridge deck, a tree branch. The top fibers are being pushed together, the bottom fibers are being pulled apart. If you've ever seen a wooden shelf sag over time and crack from underneath — that's the tension side giving up. Wood is actually stronger in tension along the grain than in compression, so wooden beams tend to fail in compression first, which is the opposite of concrete. But the pattern is the same — one side fails before the other because the two sides are doing different jobs.
Okay, so now ductility and brittleness. Daniel asked about those.
This is about how materials fail, not just when they fail. A brittle material — concrete, glass, ceramic, stone — reaches its strength limit and fractures suddenly. There's almost no warning. The stress-strain curve is basically a straight line up, then snap. A ductile material — steel, most metals — does something different. It reaches its yield point, and then instead of breaking, it starts to deform plastically. It stretches, it necks down, it visibly changes shape before it finally breaks.
So steel waves a flag before it goes.
That's exactly the right way to think about it. Steel yields and deforms visibly — you can see the beam sagging, you can see cracks in the plaster, you get warning. Concrete gives you nothing. One moment it's fine, the next moment it's rubble. That's why ductility is a safety feature. Engineers care as much about failure mode as failure load.
And this connects back to the numbers you gave earlier. Steel's tensile strength is what, compared to concrete?
Structural steel — the standard A36 grade — yields around two hundred fifty megapascals and has an ultimate tensile strength of four hundred to five hundred fifty megapascals. Concrete's tensile strength is two to five megapascals. We're talking about a factor of a hundred or more. Steel isn't just stronger in tension — it's in a completely different league. And it's ductile on top of that.
So concrete is a compression specialist that's brittle and weak in tension. Steel is a tension specialist that's ductile and gives warning before failure. And someone figured out you could marry them.
Now we're at the heart of it. Reinforced concrete is a composite material. The concrete handles the compression, the steel rebar handles the tension, and together they do what neither can do alone. But for that marriage to work, several things have to be true. The steel has to bond well to the concrete so they act as one unit — if the steel slips, the composite fails. They have to expand and contract at similar rates with temperature changes, which they do — that's one of the lucky coincidences that makes this whole thing possible. And the steel has to survive in the alkaline environment inside concrete, which actually protects it from corrosion — the high pH passivates the steel surface.
Passivates?
Forms a protective oxide layer. The concrete doesn't just hide the steel from water — it chemically protects it. Until, that is, cracks let in water and carbon dioxide over decades, the concrete carbonates, the pH drops, and the steel starts rusting. That's spalling — when the rebar rusts, it expands, and chunks of concrete pop off the surface. But in a well-designed, well-maintained structure, that takes a very long time.
So how did we figure this out? The history.
The first person to build with iron-reinforced concrete was a Frenchman named François Coignet. Eighteen fifty-three to eighteen fifty-five, he built a four-story house in Saint-Denis, just outside Paris, using iron embedded in concrete. Around the same time, an English builder named William Wilkinson reinforced the roof and floors of a two-story house, and his placement of the iron showed he actually understood where the tensile stresses were.
But the name everyone associates with early reinforced concrete is Monier.
Joseph Monier, a French gardener. He wanted better flowerpots. In eighteen sixty-seven he patented reinforced concrete flowerpots — wire mesh embedded in a mortar shell. Flowerpots. One of the most important structural innovations of the modern world started because a gardener was annoyed that his pots kept cracking.
That's genuinely beautiful.
It gets better. Monier wasn't an engineer — he just knew the wire mesh stopped the cracking. He kept patenting: pipes, panels, bridges. By eighteen seventy-seven he had a patent for iron rods arranged in a grid. He understood what worked even if he didn't have the full theory. The theory came from people like Thaddeus Hyatt, who ran systematic experiments in the eighteen seventies, and Ernest Ransome, who invented twisted rebar — the deformations on the bar surface that improve the bond with the concrete.
Twisted rebar is a bonding innovation?
The ridges and deformations on modern rebar exist to create mechanical interlock with the concrete. Smooth bars can slip. Twisted or deformed bars can't. Ransome figured that out, and his buildings survived the nineteen-oh-six San Francisco earthquake.
Which brings us to the earthquake proof.
Julia Morgan built El Campanil, a seventy-two-foot reinforced concrete bell tower at Mills College, in nineteen-oh-four. When the nineteen-oh-six earthquake hit San Francisco, it stood undamaged. That tower became a poster child for reinforced concrete. Around the same time, the Ingalls Building went up in Cincinnati — sixteen stories, one of the first reinforced concrete skyscrapers. Meanwhile, the Bixby Hotel in Long Beach collapsed during construction that same year, nineteen-oh-six, killing ten workers. That disaster and the earthquake together forced the industry to get serious about codes and standards.
So in the span of about fifty years, we went from flowerpots to skyscrapers.
And the whole time, the fundamental insight was the same: concrete carries compression, steel carries tension, and together they're more capable than either alone. Now here's where it connects to the much older story. Because reinforced concrete is only about a hundred and seventy years old. Humans have been building monumental structures for five thousand years. How did they do it with just stone?
This is the part of Daniel's question about arches, domes, vaults, and flying buttresses.
Stone has the exact same problem as concrete. Compressive strength up to about a hundred megapascals — that's enormous. Tensile strength effectively zero. In masonry design, engineers literally assume stone can handle zero tension. So pre-modern builders couldn't use stone the way we use steel beams. They couldn't span a gap with a flat stone beam and expect it to hold — the underside would be in tension, and it would crack.
So they changed the shape instead of the material.
That's the entire story of pre-modern architecture in one sentence. The arch is the key invention. An arch takes vertical loads and resolves them into forces that follow the curve of the arch down to the supports — mostly compression all the way through. The bending moment in any segment of an arch is tiny compared to a flat beam of the same span. Everything stays squeezed.
But there's a catch.
The catch is thrust. An arch doesn't just push down — it pushes outward at the base. Those horizontal forces have to be resisted, or the arch spreads and collapses. Roman builders solved this with mass — enormous abutments, thick walls, triumphal arches that are basically piles of stone with a hole in the middle. In an arcade, a row of arches, each arch's outward thrust is balanced by its neighbors, except at the ends where you need something heavy.
And flying buttresses?
The Gothic cathedrals wanted something different. They wanted height and light — thin walls, enormous windows. A thin wall can't resist the outward thrust of a stone vaulted ceiling. So they externalized the resistance. A flying buttress is a ramp of stone that reaches from the upper wall, where the vault thrusts outward, down to a massive pier planted in the ground some distance away. The thrust travels along that ramp — in compression — to the pier, which is heavy enough to absorb it. Notre-Dame de Paris had these built around eleven eighty.
It's a compression channel. They're routing the force somewhere that can handle it.
And the whole thing stays in compression. The flying buttress is an arch turned on its side, doing the same job — carrying force along a curve to a place where it can be safely grounded.
Now you mentioned Robert Hooke in your notes. The hanging chain thing.
This is my favorite part. In sixteen seventy-six, Robert Hooke published a Latin anagram that, when decoded, read: "As hangs the flexible line, so but inverted will stand the rigid arch." Hang a chain from two points. Gravity pulls it into a curve — a catenary. That chain is in pure tension. Every link is being pulled. Flip that shape upside down, and you have an arch that's in pure compression. Every stone is being squeezed.
So every arch is secretly an upside-down suspension bridge.
And every suspension bridge is secretly an arch flipped on its head. They're structural mirror images. The ideal arch shape for a given load pattern is called the funicular curve — the shape a hanging cable would take under that same load. If you build an arch to that shape, there's almost no bending anywhere. Pure compression, which is exactly what stone and concrete want.
That's the kind of thing that, once you hear it, you can't unsee it.
And it connects the two eras directly. Pre-modern builders solved the tension problem with geometry — arches, domes, vaults, buttresses, shapes that route forces into compression paths. Modern engineers solved it with material science — embed steel where the tension lives. Reinforced concrete is the arch, but made of two materials instead of one shape.
So now let's do what Daniel asked. Give me the everyday lens. I'm walking down the street, I see a building, a bridge, a lamppost. What am I looking for?
Two questions. Which parts are being pulled? Which parts are being squeezed? And then: is the material good at that job?
Look at a suspension bridge. The main cables drape between the towers in that catenary curve. They're in pure tension — every strand of steel is being stretched. Steel is phenomenal in tension, so that works. The towers those cables hang from? They're being compressed — the cables pull down on them, the weight of the whole bridge is squeezing them into their foundations. Towers are often concrete or stone — compression specialists. The bridge deck hangs from vertical suspender cables, which are also in tension. The whole structure is a map of forces expressed in material choices.
What about something smaller? A chair.
Sit in a chair. Your weight pushes down on the seat. The legs are in compression — they're being squeezed between the floor and the seat. If the legs are wood, that's fine — wood is decent in compression along the grain. But now look at the joints where the legs meet the seat. When you lean back, the front legs want to pull away from the seat. That's tension at the joint. If it's just glue, the glue is doing tension work. If there's a screw or a bolt, the metal fastener is carrying the tension. This is why chairs get wobbly — the tension connections fail first.
A ladder.
Rungs are beams. Your weight on a rung compresses the top and tensions the bottom. Wooden ladder rungs sometimes split from underneath — tension failure. The side rails of the ladder are mostly in compression when you're climbing, but if the ladder is leaning, there's bending in the rails too. Extension ladders have a cable or a rope — that's in pure tension when the ladder is extended, holding the sections together.
Tree branches.
A horizontal branch is a cantilever beam — fixed at one end, free at the other. The top of the branch is in tension, the bottom in compression. Wait — I said that backwards. A cantilever is the opposite of a simply supported beam. The top is in tension, the bottom is in compression. Trees actually respond to this — they grow "tension wood" on the upper side of branches and "compression wood" on the lower side. The tree knows which fibers are being pulled and which are being squeezed, and it reinforces accordingly.
Trees are structural engineers.
Trees have been doing this for three hundred million years. We're the ones who showed up late.
The material selection rule is: if a part is being pulled, you need a ductile material with high tensile strength — steel, or wood with the grain running along the pull. If a part is being squeezed, you can use brittle materials — concrete, stone, brick, cast iron.
If a part is doing both — like a beam — you either use a material that's decent at both, like steel or wood, or you use a composite like reinforced concrete where each material handles the job it's good at. That's why you don't see pure concrete beams spanning any significant distance without rebar. They'd snap.
What's the single most common wrong belief people have about this whole topic?
That concrete is just "strong." It's strong in one direction and an order of magnitude weaker in the other. Calling concrete strong without specifying the direction is like saying a fish is a fast animal. In water, yes. On land, no.
And arches being decorative.
Right. People see arches in old buildings and think they're an aesthetic choice. They're a structural necessity. The Romans weren't putting arches in aqueducts because they looked nice. They were solving the tension problem with the only tool they had — geometry.
The other one is ductility. People assume it's about strength — that ductile materials are just tougher. But it's really about how failure happens. Steel announces itself. Concrete just quits.
In a clinical sense, it's the difference between a condition that gives you symptoms before a crisis and one that doesn't. You want the warning. That's why building codes require minimum ductility in seismic zones — you want the structure to deform and crack visibly before it collapses, so people can get out.
The framework Daniel can take away is this. Look at any structure. Ask: what's being pulled, what's being squeezed? Then check: is the material good at that job? If you see concrete or stone doing tension work, something else must be carrying it — rebar inside, or a clever shape routing the forces into compression. If you see steel cables, they're almost certainly in pure tension. If you see a massive masonry wall, it's absorbing compression and resisting thrust.
If you see an arch, flip it upside down in your head. The hanging chain that matches it tells you whether the shape is right for the loads it's carrying. That's the Hooke test — four hundred years old and still the most elegant diagnostic tool in structural engineering.
Daniel, I suspect Hannah will be pleased. You'll now be able to look at a building and see the forces moving through it instead of just the facade.
One thing I keep thinking about. We've been talking about buildings, but this applies to everything. Your own skeleton is a compression-tension system. Bones are compression elements — they're strong in compression, weaker in tension. Your tendons are pure tension cables. Your muscles pull on tendons, tendons pull on bones, bones carry the compressive load of your body weight. You're a tensegrity structure.
The built world and the biological world are solving the same problems with the same principles. That's a nice place to land.
If you have a weird prompt you want us to explore, send it in. Show at my weird prompts dot com. The weirder the better.
Thanks to our producer Hilbert Flumingtop. This has been My Weird Prompts.
We'll be back soon.
