Sena Kizildemir’s physics-driven simulations let architects preview a skyscraper’s collapse in a truck-bomb scenario weeks before the first beam is welded—turning post-9/11 “what-ifs” into build-it-right-the-first-time blueprints.
The Day That Broke the Code
No building code in 2001 accounted for a 200-ton jetliner slicing through 110 stories of steel. When the World Trade Center pancaked in 102 minutes, the collapse pattern became a macabre textbook that every structural engineer now studies. IEEE Spectrum’s forensic timeline confirms that severed columns plus roaring jet-fuel fires pushed floor trusses past their creep limit, triggering sequential failure.
The takeaway: if you can’t simulate it, you can’t defend against it.
Enter Kizildemir’s Virtual Crash Lab
Sena Kizildemir, now a senior project engineer at Thornton Tomasetti’s Applied Science practice, runs what amounts to a digital crash-test dummy for buildings. She meshes 50-million-element finite-element models in Abaqus, then slams them with F-4 Phantom-level kinetic energy, truck-bomb over-pressure, or rogue vehicle impact—whatever the threat matrix demands.
Her deliverable: color-coded strain maps that tell architects exactly where rebar needs wrapping, where column splices need armor, and where a sacrificial “fuse” can let a tower shed energy without killing occupants.
From Rail Cracks to Skyscraper Skin
Kizildemir’s mastery of fracture mechanics began at Lehigh University while she studied subsurface rail flaws for the U.S. Federal Railroad Administration. Phase-one findings she co-authored are already tightening inspection intervals for high-speed rail. Same physics, different scale: a micro-crack in rail steel behaves like a column shear in a tower—stress concentrates, material softens, catastrophe propagates.
How the Sim Actually Works
- Threat definition: client picks impact velocity, explosive yield, or fire curve.
- Model seeding: laser-scanned point cloud of the real site becomes a voxel mesh.
- Material cards: every alloy, concrete grade, and glass pane gets its own stress-strain law calibrated in the lab.
- Run: 4,000-core cloud cluster chews through 48 hours of wall time to compress 20 seconds of real collapse into solvable time steps.
- Post-mortem: engineers overlay injury-risk curves on floor-plate heat maps to see if occupants can escape before progressive collapse initiates.
Codes Are Always Playing Catch-Up
Current ASCE 7 standards treat blast, impact, and progressive collapse as “exceptional events” with prescriptive fixes—extra column ties, minimum tie strength, alternate load paths. Kizildemir’s models expose the blind spots: a prescriptive 2-percent tie force may be irrelevant if the initial breach ejects that very tie. She feeds her results to code committees, pushing performance-based language that quantifies residual capacity instead of mandating bar diameter.
Bottom Line for Tenants, Developers, and Cities
- Tenants: You get egress routes proven to stay intact 30 minutes longer—potentially the margin between life and death.
- Developers: Up-front $500k in simulation can shave $5M off retrofits later and unlock insurance discounts up to 15 percent.
- Cities: Faster permitting when you show the building authority a validated digital twin instead of a static checklist.
What’s Next: From One-Off Models to a Shared Library
Kizildemir is pushing Thornton Tomasetti to containerize its best-practice meshes—think open-source Lego blocks any firm can drop into a project. Pair that with IEEE’s Collabratec platform and you get a living repository where each new blast test refines everyone’s next tower. The goal: collapse-proof skylines, not just collapse-proof single buildings.
Her mantra, borrowed from her employer: “When others say no, we say ‘Here’s how.’” For anyone who watched 9/11 unfold, that promise is as steel-solid as the beams she models.
Stay ahead of the next code cycle and the next disaster scenario—bookmark onlytrustedinfo.com for the fastest, most authoritative tech intelligence on the planet.
