Greek researchers found that wind turbines standing still can tremble like a tuning fork and built a model to catch it before they break
Wind turbines are built to spin. But some of their most dangerous moments happen when they’re standing perfectly still.
During installation, commissioning, or periods when a turbine sits idle without its rotor assembly, the bare cylindrical tower can begin to shake — not from mechanical failure, but from the wind itself. As air flows around the tower, it sheds rhythmic vortices that can trigger large crosswind oscillations within seconds. For manufacturers, predicting when and how severely this will happen has remained a stubborn engineering challenge.
The hidden danger in a standing turbine
When wind flows around a cylindrical tower, it doesn’t pass through cleanly. Instead, it peels off in alternating vortices — first one side, then the other — creating a rhythmic sideways push on the structure. These are vortex-induced vibrations, or VIV, and they’re most dangerous not when a turbine is running, but when it’s stopped.
A Greek research team based at the National Technical University of Athens (NTUA) turned to the Hartlen–Currie lift-oscillator model as their predictive framework.
Turbines are especially vulnerable during installation, before the rotor-nacelle assembly is in place, and during idling or parked periods. Without the rotor spinning, the tower stands as a bare cylinder with almost no aerodynamic damping in the crosswind direction.
The most hazardous scenario is called lock-in. It occurs when vortex shedding frequency synchronizes with the tower’s natural bending frequency, at which point oscillation amplitudes can grow rapidly — within seconds — accelerating fatigue damage and compromising structural integrity over time.
Why taller towers face greater risk
The wind industry has been building taller. At low-wind sites, towers up to 50% taller than standard configurations are used to reach stronger winds at higher altitudes, and that design choice carries a real tradeoff.
On a taller tower without its rotor assembly, lock-in can occur at wind speeds between 15 and 20 m/s — significantly higher than the 5–15 m/s range typical for fully assembled turbines. At those speeds, aerodynamic forces are much larger, making vibrations more severe. The crosswind direction stays the most exposed, since the drag force that damps longitudinal motion offers no equivalent protection against lateral oscillations.
A smarter model for predicting the shake
A Greek research team based at the National Technical University of Athens (NTUA) turned to the Hartlen–Currie lift-oscillator model as their predictive framework. Unlike simpler single-equation approaches, this formulation uses two coupled equations: one describing the tower’s structural motion, and a second governing the aerodynamic lift force itself.
Single-equation models can estimate peak vibration amplitude but struggle to capture how amplitudes evolve across the full lock-in range. The Hartlen–Currie model handles this more naturally, even with constant parameter values.
The team calibrated two free parameters — G′ and H′ — using wind tunnel data from rigid, elastically mounted cylinders tested at different Scruton numbers, varying them linearly and tapering toward zero at the edges of the lock-in zone to better reflect how vortex-induced oscillations build and decay.
From wind tunnel to real tower: validation results
With the model calibrated, the team integrated it into hGAST, NTUA’s in-house multibody aeroelastic solver, then tested it against wind tunnel measurements of a cantilevered flexible cylinder designed to mimic a real turbine tower.
For two of the four damping conditions tested — Scruton numbers of 5.18 and 7.02 — the model accurately reproduced maximum tip displacements. It outperformed both the Eurocode standard, which tends to overpredict VIV response, and the Blevins model, which tends to underpredict it.
At the lowest damping condition, a Scruton number of approximately 2.94, the model overpredicted displacement by about 55%. The researchers attribute this to calibration relying on rigid-cylinder data and say that incorporating cantilever-specific calibration data can close this gap.
What comes next for wind turbine safety
The model is designed to slot directly into existing aeroelastic design workflows, giving manufacturers a practical tool for assessing fatigue loads and structural integrity during the phases when towers are most exposed.
One clear next step is recalibration using higher-Reynolds-number data. Both the calibration and validation work relied on low-Reynolds-number wind tunnel experiments — conditions well below those encountered by full-scale turbines in the field. The authors note this isn’t a fundamental limitation; the model is ready to be updated as soon as relevant datasets become available.
The current study also sets aside VIV in higher-order bending modes and the influence of ambient turbulence on lock-in behavior, both flagged as priorities for future research. As those gaps close, the model’s accuracy in real-world conditions should improve, potentially reducing costly structural failures and extending the working life of turbines.
Reference: Vlastos, D., Tzimi, N., Trampa, G., Riziotis, V., Manolas, D., and Spyropoulos, N.: Calibration and validation of an engineering model for vortex-induced vibration prediction in wind turbine towers, Wind Energ. Sci., 11, 2585–2600, https://doi.org/10.5194/wes-11-2585-2026, 2026
Daniel García is an Editor-in-Chief with strong expertise in structural work and engineering principles. He combines this technical foundation with deep knowledge of energy, spatial design, and emerging technologies, bringing a forward-thinking and analytical approach to editorial leadership.
