Abhishek S.
Shipping in public. Listening in private.

Abhishek

I lead women’s Indo-Western & Premium at Max Fashion. I also wrote the AI that runs the buying floor.

Rare profile. Category operator who ships production code.

Senior Buying Leader · Max Fashion Women’s Indo-Western & Premium · 530+ India stores NIFT ’12 · Twelve years on the floor

abhishek@bengaluru ~ %
>role: senior buying lead
>dept: women’s indo-western + premium
>floor: 530+ stores india

Aeroelastic Flutter

On November 7, 1940, a four-month-old suspension bridge at Tacoma Narrows twisted itself apart in a 42 mph wind. For sixty years, physics textbooks blamed resonance. They were wrong. The actual mechanism is a self-excited coupling between airflow and structural motion, where the wind does not push the structure at its natural frequency. The structure's own motion reshapes the airflow, which then pumps energy back into the motion. Damping goes negative. Amplitudes grow until something yields.

What's actually happening

A solid structure in moving fluid has at least two oscillation modes that matter: bending (up-down) and torsion (twisting). At low wind speeds, these modes are independent and air drag bleeds energy out of both. Above a critical speed, the phase relationship between them shifts. Lift produced by the twisting motion now does positive work on the bending motion, and vice versa. The system extracts energy from the steady airflow at a rate faster than damping can dissipate it.

The math was worked out for aircraft wings by Theodore Theodorsen at NACA in 1935, five years before Tacoma. His complex-valued lift function (C(k)) gave engineers the first usable flutter prediction equations. The bridge community did not read the aviation literature. They built Tacoma Narrows with a solid plate girder eight feet deep and 39 feet wide, a section so bluff it shed vortices at frequencies that locked onto the deck's 0.2 Hz torsional mode.

The collapse film, shot by Barney Elliott, became the most-watched engineering footage of the 20th century. Robert Scanlan and John Tomko's 1971 reanalysis confirmed it was flutter, not resonance. The distinction matters: resonance needs an external driver at a matching frequency, flutter creates its own driver from the steady wind.

Where it shows up

What's contested

The Tacoma Narrows pedagogy fight is still active. Several introductory physics textbooks (as of 2020 surveys) still teach the bridge as a resonance example. K. Yusuf Billah and Robert Scanlan's 1991 American Journal of Physics paper "Resonance, Tacoma Narrows Bridge Failure, and Undergraduate Physics Textbooks" tried to settle it. The myth persists because resonance is a clean Chapter 4 topic and flutter requires fluid-structure coupling that doesn't fit into a one-week unit.

Among aeroelasticians, the open question is nonlinear flutter prediction in transonic flow, where shock waves move with the structure. Limit-cycle oscillations (LCO) — bounded flutter that doesn't grow to destruction but fatigues components — remain hard to predict from first principles. The F-16, F/A-18, and F-35 have all encountered LCO in flight test that pre-flight CFD missed.

Why this matters to other realms

Flutter is a control-theory problem dressed as a fluids problem. The structure is a plant, the airflow is a feedback loop, and the question is whether the closed-loop poles cross into the right half-plane. This is the same math that governs whether an concept-economic-bubbles grows or stabilizes, why concept-feedback-loops runs away in biological systems, and why aircraft flight control laws need active damping. The engineering insight — that a system can extract energy from a steady environment and convert it into destructive oscillation — generalizes far beyond bridges and wings.

It also explains concept-winglets on the Boeing 737-800. Wingtip vortices interact with the wing's first bending mode at high subsonic speeds. The blended winglet, introduced in 2001, raised the flutter margin by ~7% while cutting induced drag. Two problems solved by one piece of bent aluminum.

An open question

If flutter is fundamentally a problem of energy extraction from steady flow, can it be inverted? Bio-inspired energy harvesters using flutter for piezoelectric power generation are a live research area. The numbers are still small — milliwatts per square centimeter as of 2024 — but the physics says nothing about why you couldn't build a kite that flutters productively at 100 W/m². What would it look like?

Key sources

Further reading

Abhishek's take

I see the same failure shape when a reorder rule starts feeding on its own sales signal. A black kurta that sells out in 10 stores can look like demand, when the first stockout is already bending the data. I treat that as a damping problem: slow the loop, add a floor check, and make the tool prove the signal before the next buy.

See Also