Skip to content

Drawing No. EH–CA–014 // Engineering Case Study

Tacoma Narrows Bridge Flutter Case Study

A four-month-old suspension bridge failed in a moderate wind after its familiar vertical oscillation—historically described as “galloping”—changed abruptly into destructive torsional motion. This interactive case study reconstructs the documented sequence, illustrates the mainstream torsional-flutter interpretation while noting remaining debate, and explains why Tacoma Narrows changed long-span bridge design. [1][3][6]

Prepared by EngineerHub · Technical reconstruction reviewed 11 August 2026

Main span: 2,800 ft / 853 mDeck width: 39 ft / 11.9 mGirder depth: 8 ft / 2.44 mWind observed: 42 mph / 18.8 m/sMajor collapse: 11:02 a.m.

01 // Interactive failure replay

Select a stage or run the event.

Tacoma Narrows — deck motion reconstructionNorth and south deck edges are animated separately to distinguish vertical/heaving motion from torsional motion.
Motion is schematic and visually amplified. Historical amplitudes, times and wind speeds are shown numerically where documented.
Animated Tacoma Narrows Bridge deck motionSuspension bridge schematic showing stage-dependent vertical and torsional motion, wind direction and progressive loss of deck sections.TACOMA NARROWSDECK CROSS-SECTIONtilt ≈ 0°
Observed structural statePre-event flexibility
VERTICAL RESPONSE
Animated deck edges
North edge
South edge
Crosswind
Damage / failed deck

Bridge status summary

Main span2,800 ft
Deck width39 ft
Girder depth8 ft
Span / depth≈350
Span / width≈72
Maximum reported roadway angleup to 45°

Observation console — latest stage first

02 // Aeroelastic energy explorer

Why self-excited flutter can grow instead of damping out.

0°
25%
Aerodynamic work — normalized
Structural dissipation — normalized
Net per cycle
Energy tendencyGrowing
Physical interpretationnegative damping
NORMALIZED SINGLE-MODE ENERGY EXCHANGEaerodynamic momentangular velocitypositive work per cycle → amplitude grows
Educational model. For an idealized harmonic torsional mode, average aerodynamic work depends on the component of aerodynamic moment in phase with angular velocity. If positive aerodynamic work exceeds dissipation, amplitude can grow. The sliders are dimensionless teaching controls; they do not reproduce Tacoma’s measured flutter derivatives, critical flutter speed, structural damping, or nonlinear post-flutter response.

03 // Cause-and-effect chain

The failure is best understood as a feedback process, not a single impact load.

01 / SLENDER DECKA 2,800 ft span used 39 ft cable spacing with only an 8 ft solid plate girder. Historical comparisons show exceptionally low relative torsional rigidity.
02 / WIND-INDUCED MOTIONThe bluff plate-girder section interacted strongly with crosswind. Large vertical/heaving oscillations—historically called “galloping”—were observed during construction and operation.
03 / TORSIONAL TRANSITIONOn 7 November the familiar vertical motion changed abruptly into large twisting. The 1941 investigation associated this transition with slippage of the north-cable midspan cable band; later studies have continued to examine its role.
04 / SELF-EXCITED FEEDBACKUnder the mainstream flutter interpretation, deck rotation altered the aerodynamic forces and moments so that net aerodynamic work sustained and amplified the torsional motion.
05 / PROGRESSIVE FAILURERepeated large cyclic rotations were followed by damage to the roadway, floor system, suspenders and girders, culminating in loss of major deck sections.

04 // Event figures

Historical measurements with clearly marked reconstruction.

Observed wind and illustrative event severity

Wind speed, mphIllustrative severity index — not measured
View timeline data
TimeWindObserved condition
07:3038 mphPronounced vertical wave motion (historically described as “galloping”)
09:3042 mphSeveral vertical waves, about 2–5 ft high
10:03Not separately reportedLateral twisting begins
10:07No separate exact measurement citedExtreme torsion: roadway angle up to about 45°; WSDOT describes the roadway tilting as much as 28 ft on one side then the other
10:30Large concrete section falls
11:02About 600 ft of roadway breaks free
11:08Final major roadway section falls
The blue points at 07:30 and 09:30 are documented wind measurements. The orange series is a deliberately illustrative stage-severity index created for this page; it is not a measured engineering quantity and must not be used as a physical time history.

Deck slenderness — contemporary comparison

Span / girder depthSpan / deck width
View comparison data
BridgeCenter spanGirder depthDeck widthReported depth ratioReported width ratio
George Washington3,500 ft36 ft106 ftSource inconsistency*1:33
Golden Gate4,200 ft25 ft90 ft1:1681:47
Bronx–Whitestone2,300 ft11 ft74 ft1:2091:31
Tacoma Narrows2,800 ft8 ft39 ft1:3501:72
Ratios and dimensions follow WSDOT’s historical comparison tables. *WSDOT lists the George Washington Bridge as 3,500 ft center span, 36 ft girder depth and 1:120 span/depth; 3,500/36 is approximately 97, so the published values are internally inconsistent. The orange George Washington depth-ratio bar is omitted rather than silently resolving the discrepancy. A higher ratio denotes a more slender system.

05 // What actually happened?

The collapse is not adequately explained as simple forced resonance alone. Historical observations and later analyses distinguish the earlier heaving response from the destructive torsional mode. WSDOT identifies torsional flutter as the primary explanation, while peer-reviewed work has continued to examine the roles of vortex-induced motion, cable-band slip, structural nonlinearities and aerodynamic feedback. [3][7][8][9][10]

Vertical/heaving oscillation came first

Vertical wave motion was noticed while the bridge was still being completed and continued after opening. Historical accounts called this behavior “galloping,” but modern wind-engineering literature often describes the pre-torsional response as heaving or vortex-induced oscillation. Engineers tried hydraulic devices, tie-down cables and other measures to suppress the visible motion; these measures did not eliminate the bridge’s aerodynamic vulnerability. [1][4]

The critical transition was torsional

WSDOT’s historical reconstruction records a sudden lateral twisting motion at 10:03 a.m. on 7 November 1940. By 10:07, WSDOT describes the roadway tilting as much as 28 ft on one side and then the other, at an angle up to about 45 degrees, with the violent twisting cycle repeating about every five seconds. [1][4]

Why “flutter” matters

In the classical aeroelastic-flutter interpretation, aerodynamic forces and moments are coupled to the bridge motion itself. If their net work over a cycle exceeds structural and aerodynamic dissipation, oscillation amplitude grows without requiring an external periodic forcing to remain fixed at exactly one structural natural frequency.

A note on uncertainty

The precise details of the Tacoma failure remain debated in the engineering literature. WSDOT explicitly states that no explanation has unanimous acceptance, although it presents torsional flutter as the primary explanation. This page follows that mainstream aeroelastic interpretation while also noting alternative or complementary analyses and clearly separating documented observations from educational visualization. [3][7][8][9]

07 // Design choice: deep open truss vs shallow solid girder

The cross-section changed the bridge’s stiffness and aerodynamic behavior. [2]

Clark Eldridge concept — 25 ft deep stiffening truss

OPEN TRUSS — AIR CAN PASS THROUGH25 ft
Proposed main span2,600 ft
Truss depth25 ft
Characterdeep / open

Built 1940 bridge — 8 ft solid plate girder

SOLID BLUFF GIRDER — FLOW SEPARATES8 ft
Built main span2,800 ft
Girder depth8 ft
Charactershallow / solid

08 // Engineering lessons learned

Tacoma Narrows changed what engineers consider part of “structural design.”

Aerodynamics is a structural problemStrength under static wind pressure is not enough. Long flexible structures need stability checks for coupled aerodynamic motion.
Mode shape mattersA bridge that appears acceptable in vertical bending can still be vulnerable in torsion. Multiple modes and coupled modes must be investigated.
Cross-section shape mattersA bluff solid girder can create separated flow and destabilizing aerodynamic moments. Stiffness, venting and streamlining all affect stability.
Observe anomalies as system warningsRepeated large wind-induced oscillations before opening were more than a comfort issue. Persistent unexpected dynamics deserve root-cause investigation before normal operation.
Test beyond accepted theoryWSDOT notes that the Moisseiff design met accepted theory of the period. The failure exposed inadequate treatment of aerodynamic stability and accelerated wind-tunnel testing of long-span bridges.
Design trends need historical memoryTacoma Narrows culminated a trend toward longer, lighter and more slender suspension spans. The failure showed the risk of extending a design trend beyond the range in which its governing physics are well understood.

Defense in depth: what the 1950 replacement changed

StructuralDeeper and wider stiffening truss

The replacement used a much deeper, more torsionally resistant open truss instead of the 8 ft solid plate girder.

AerodynamicDeck venting

Longitudinal openings and steel grating allowed air to pass through the deck rather than forcing all flow around a bluff solid section.

DampingHydraulic dampers

Additional damping devices were incorporated to reduce motion and provide another layer of dynamic control.

VerificationWind-tunnel model testing

Post-collapse research established aerodynamic testing as a central part of long-span bridge design practice. [5][6]

09 // Historical chronology

Date / timeEvent
23 Nov 1938Construction starts.
May 1940Vertical deck wave motion is noticed while the floor system is being completed.
1 Jul 1940Bridge opens to traffic.
Oct 1940Temporary tie-down cables and other corrective measures are used while wind-tunnel studies continue.
7 Nov, 07:30Wind measured at 38 mph; pronounced vertical wave motion is present.
09:30Wind measured at 42 mph; several 2–5 ft vertical waves are observed.
10:03Sudden lateral twisting motion begins.
10:07Torsion becomes extreme; roadway tilts up to about 45°.
10:30WSDOT’s timeline records a large concrete section falling from the center-span area; some WSDOT photo captions indicate smaller concrete loss earlier in the torsional sequence.
11:02About 600 ft of roadway in the eastern half of the main span falls.
11:08Final major roadway section falls.
Mar 1941Carmody Board reports excessive flexibility, aerodynamic action of the solid girder/deck, and the need for model testing.
1950Replacement bridge opens with a deep open stiffening truss, venting and damping measures.

11 // Frequently asked questions

Common misconceptions and practical lessons.