Here is the thing almost nobody outside structural engineering knows: a code-minimum building is not designed to come through a big earthquake intact. It's designed to do one job — not collapse, long enough for everyone to walk out. After the shaking stops it may be cracked, leaning, condemned, and bulldozed — and still be a complete success, because the lobby emptied before the frame gave up.
The hero of that job is the moment frame: beams and columns welded into rigid joints that are built to bend and take damage on purpose. The frame survives the quake by absorbing its energy — yielding at the joints like a boxer rolling with a punch instead of standing stiff and shattering. The bent, ugly, ruined frame is the one that saved everybody. The pretty one just got lucky.
Below: shake one. Push the intensity, stack the stories, and choose the connection. Watch the joints go plastic, watch P-delta — the building's own weight leaning through the drift — try to run away with it. Win condition isn't "no damage." It's still standing at the moment the last person clears the door.
Drift ratio is how far a floor leans sideways divided by its height — codes hold the life-safety limit near 2%. θ is the P-delta stability coefficient; as it climbs toward 1.0 the building's own weight starts doing the earthquake's work for it, and the lean feeds itself. Switch to the pre-Northridge weld and watch the joints fracture early instead of yielding — the failure that surprised an entire profession in 1994.
🐧 NULL watched the frame bend until it nearly lay down, and watched the last door close behind the last person. NULL said nothing. NULL marked it: passed.
Code-minimum seismic design has three performance levels, and "survive intact" is not the everyday one. A standard building is meant to reach Immediate Occupancy in a small, frequent quake (walk back in tomorrow); Life Safety in the design-level quake (damaged, but everyone gets out); and Collapse Prevention in the rare maximum quake (a wreck, still standing, occupants escaped). The official goal for the big one is not to save the building. It's to not kill the people inside it on the way down.
That's why this lab's win condition is an evacuation clock, not a damage meter. A modern special moment frame earns that time by being ductile — the joints are detailed to yield and absorb energy through dozens of cycles without losing their grip on gravity. It drifts, it cracks, it groans — and it holds. Long enough.
P-delta is the villain. Once a building leans, its own weight is now pulling sideways through that lean, adding to the push. Past a certain drift the effect compounds itself faster than the frame can resist, and the lean runs away to collapse. Taller buildings carry more weight up high, so they hit that runaway sooner — which is why stacking stories in the sim drops the margin.
And the connection matters more than anyone believed. Switch the toggle to pre-Northridge and the joints fracture early and brittle instead of yielding — the building loses its capacity to buy time, and people don't get out. That's not a hypothetical; it's the lesson of 1994, in the About tab.
This is the structural-engineering half of The Core Echo: that lab asks what the ground does; this one asks whether what we built can take it long enough to matter. The QuakeSimulator sits between them — if it shook here, who feels it, and does the frame hold.
The performance philosophy is real. The intended seismic performance of an ordinary building under the codes — Immediate Occupancy / Life Safety / Collapse Prevention across frequent / design / maximum quakes — is set out in ASCE 7 (Minimum Design Loads) and its companion ASCE 41 (Seismic Evaluation & Retrofit). The plain version: design so the building doesn't collapse under the maximum considered earthquake, giving occupants time to escape — damage is accepted, not prevented.
Ductile detailing is real. Special moment frames are governed by AISC 341 (Seismic Provisions for Structural Steel Buildings) — the strong-column / weak-beam rule, where beam ends are made the sacrificial fuse so the columns keep holding up gravity while the frame yields.
The Northridge lesson is real — and it's the heart of this lab. In the 1994 Northridge earthquake, welded steel moment connections that everyone trusted fractured brittlely at the beam-to-column weld, some at very low demand, a few while the structure was still essentially elastic. It was a profession-wide shock. FEMA stood up the SAC Joint Venture and, in under a year on the connection-repair side, produced the guidance that rewrote the detail: FEMA-350, Recommended Seismic Design Criteria for New Steel Moment-Frame Buildings (2000). See also FEMA's plain-language case study on the connection fix. The pre-1994 detail is no longer accepted for seismic use. The pretty connection that looked fine on paper was the one that failed.
Stylized here: the frame, the drift, the hinges, and the P-delta runaway are an illustrative teaching model tuned for legibility — not a finite-element solver. Drift ratios, story counts, and timing are chosen to make the mechanism visible, not to certify a building. The numbers quoted (the ~2% life-safety drift, the θ stability idea, the three performance levels) are real; nothing on the canvas is computed to code. For the labs that do solve real beam math, see the Static Beam and Live Beam.
The pedagogy: the building isn't the patient — the people are. Built to teach, not to scare.