Direct Numerical Simulation of Vertical-Axis Wind Turbine Near-Wake Dynamics

This study employs geometrically-resolved Direct Numerical Simulations to reveal that increasing the blade count in vertical-axis wind turbines accelerates the breakdown of dynamic stall vortices through blade-vortex interactions, causing the near-wake to transition more rapidly to bluff-body dynamics and demonstrating that blade number, rather than tip-speed ratio, is the primary factor governing this transition and downstream inflow characteristics.

The Gist

Imagine a vertical-axis wind turbine (VAWT) not as a giant machine, but as a spinning pinwheel standing upright in a river of air. This paper is like a high-definition, slow-motion movie that zooms in so closely on the blades that we can see the invisible "air currents" swirling around them. The researchers used a super-computer to simulate exactly how the air behaves right behind the spinning blades, a place called the "near-wake."

Here is the story of what they found, explained simply:

The Big Picture: Why Look at the "Near-Wake"?

Most wind turbines spin like a propeller on a plane (horizontal-axis). But these vertical ones spin like a salad spinner. The researchers wanted to know: What happens to the air immediately after it passes the blades?

They found that the air doesn't just flow smoothly behind the turbine. Instead, it gets chaotic. The blades chop the air, creating giant, swirling tornadoes called Dynamic Stall Vortices (DSVs). Think of these like giant, invisible whirlpools that get kicked up by the blades and trail behind the turbine.

The Main Discovery: The "Blade Count" Matters More Than Speed

The team tested turbines with one blade, two blades, and three blades. They also changed how fast the turbines spun.

Here is the surprising twist: How many blades the turbine has matters much more than how fast it spins.

• The One-Blade Turbine (The Soloist):
  Imagine a lone dancer spinning in a room. When they spin, they create one giant, powerful whirlpool of air behind them. This giant whirlpool (the DSV) stays strong and intact for a long time. It's like a heavy, slow-moving cloud that takes a long time to dissipate. Because this giant cloud lingers, the air behind the turbine stays "messy" and chaotic for a long distance.

• The Three-Blade Turbine (The Trio):
  Now imagine three dancers spinning close together. As one dancer spins, the air they just disturbed (the whirlpool) is immediately bumped into by the next dancer.
  The researchers discovered a new mechanism they call "Vortex Impingement."

  • The Analogy: Think of the giant whirlpool as a soap bubble. On the one-blade turbine, the bubble floats away whole. On the three-blade turbine, the next blade acts like a pin, popping the bubble prematurely.
  • The Result: The giant, messy whirlpool gets smashed into tiny, harmless bubbles (smaller vortices) before it can travel far. The air behind the three-blade turbine becomes "calm" and orderly much faster than the one-blade version.

The "Traffic Jam" Effect

The paper also explains that having more blades creates a bit of a "traffic jam" for the wind.

• With more blades, there is less open space for the wind to flow through the center of the turbine.
• This forces the wind to go around the turbine, like water flowing around a rock in a stream.
• This changes the behavior of the wake. Instead of being dominated by the giant spinning whirlpools (which happen at low blade counts), the wake starts to behave like the wake behind a simple, solid rock (a "bluff body").
• Why this is good: The "rock-like" wake recovers (becomes smooth air again) much faster than the "whirlpool" wake.

The "Self-Similarity" Test

The researchers wanted to know exactly when the messy air behind the turbine turns back into smooth, predictable air. They used a mathematical trick called "self-similarity analysis."

• The Analogy: Imagine trying to predict the shape of a smoke plume. At first, the smoke is a chaotic mess of swirls. But eventually, it settles into a predictable, bell-curve shape.
• They found that the three-blade turbines settle into this predictable shape much sooner than the one-blade turbines. The "messiness" disappears faster.

What This Means for the Future (According to the Paper)

The paper specifically mentions that these findings matter for wind farms where turbines are placed very close together.

• If you put a second turbine right behind a first one, the air it "breathes in" depends heavily on how many blades the first turbine has.
• If the first turbine has few blades, the second turbine gets hit by giant, chaotic whirlpools.
• If the first turbine has many blades, the air has already "healed" and become smoother by the time it reaches the second turbine.

In summary: This paper used a super-powerful computer simulation to show that adding more blades to a vertical wind turbine acts like a "whirlpool breaker." It smashes the giant, messy air vortices into tiny pieces, allowing the wind to recover its smooth flow much faster. This is crucial for designing wind farms where turbines are packed closely together.

Technical Summary

Technical Summary: Direct Numerical Simulation of Vertical-Axis Wind Turbine Near-Wake Dynamics

Problem Statement
While Horizontal-Axis Wind Turbines (HAWT) dominate current research, Vertical-Axis Wind Turbines (VAWT) offer distinct advantages for offshore, urban, and remote applications, including lower centers of gravity and omnidirectional insensitivity. A key theoretical benefit of VAWTs is their potential for higher power densities in closely-spaced arrays due to shorter wake regions and enhanced wake recovery. However, the near-wake of a VAWT is dominated by complex flow features associated with dynamic stall, specifically the formation and shedding of large-scale Dynamic Stall Vortices (DSVs). Unlike the far-wake, which behaves like a bluff-body wake, the near-wake is highly dependent on geometric and operational parameters such as the tip-speed ratio ($\lambda$) and static solidity ($\sigma$). Existing experimental data often lacks resolution near the geometry or in three dimensions, while high-fidelity Large-Eddy Simulations (LES) coupled with the Actuator Line Method (ALM) fail to geometrically resolve the blade boundary layer. Consequently, ALM cannot capture the laminar separation bubble (LSB) dynamics, the precise formation of the DSV, or the critical blade-vortex interactions (BVI) that modulate these features.

Methodology
To address these limitations, the authors present geometrically-resolved Direct Numerical Simulations (DNS) of straight-bladed Darrieus VAWTs using the spectral/hp element method framework, Nektar++. The study investigates three turbine configurations with one, two, and three blades (NACA 0018 airfoils) across a range of tip-speed ratios ($\lambda = 1.5, 2.5, 4.5$). This parameter space corresponds to static solidities ($\sigma$) of 0.2, 0.4, and 0.6, and a dynamic solidity ($\sigma_D$) ranging from 0.47 to 0.94.

The simulations employ a Moving Reference Frame (MRF) formulation to handle the rotating geometry without dynamic remeshing. The flow is modeled as incompressible Navier-Stokes equations at a chord-based Reynolds number of $Re_c = 10^4$, a value chosen to represent urban/extraterrestrial environments and ensure numerical feasibility while maintaining relevance to fundamental physics. The mesh resolution satisfies DNS requirements ($\Delta y^+ \lesssim 1$) with 32 Fourier planes in the spanwise direction. Spectral Vanishing Viscosity (SVV) is applied for numerical stability in under-resolved regions. The methodology is validated against experimental data from Battisti et al., showing strong agreement in time-averaged streamwise velocity profiles at near-wake locations.

Key Contributions and Results

1. Resolution of Dynamic Stall and BVI:
   The DNS successfully captures the full dynamic stall cycle, including boundary layer roll-up, DSV formation, separation, and convection. A novel finding is the identification of a vortex-impingement mechanism driven by BVI.

   • One-Bladed Configuration: The DSV forms, separates, and convects into the near-wake as a coherent large-scale structure, causing significant flow acceleration on the leeward side.
   • Two-Bladed Configuration: Increased solidity reduces the convection of the DSV, causing it to diffuse and break down into smaller structures before reaching the near-wake.
   • Three-Bladed Configuration: Reduced blade spacing introduces BVI at the onset of the upwind passage. These interactions occur during the early phase of dynamic stall (laminar separation bubble development), perturbing the LSB and reducing the eventual size of the DSV. Furthermore, as the blade approaches the downwind zone, direct vortex impingement from upstream wakes causes the premature breakup of the DSV into multiple distinct cores. This mechanism accelerates the transition from coherent vortex dynamics to bluff-body dynamics.

2. Wake Recovery and Transition to Bluff-Body Dynamics:
   The study quantifies the downstream evolution of the wake. Higher solidity (more blades) results in more severe initial velocity deficits but initiates recovery earlier and more rapidly.

   • Low Solidity ($N_{blades}=1$): Recovery is governed by cross-stream mixing induced by large DSVs, leading to slower recovery and persistent wake asymmetry.
   • High Solidity ($N_{blades}=3$): The BVI-driven breakdown of DSVs removes the coherent structures that distinguish the VAWT wake from a bluff body. Consequently, the wake transitions more rapidly to shear-layer-associated recovery mechanisms, which are inherently faster.

3. Self-Similarity Analysis:
   Extending self-similarity analysis into the near-wake (previously limited to the far-wake in ALM studies), the authors quantify the rate at which the wake loses dependence on blade-generated coherent structures and collapses onto a self-similar Gaussian solution.

   • The analysis reveals that blade number is more influential than tip-speed ratio in setting the rate of this transition.
   • Higher blade counts lead to a faster onset of self-similarity. Interestingly, the rate of transition peaks at $\lambda=2.5$ rather than the highest $\lambda$, suggesting a plateau effect beyond a certain dynamic solidity.

Significance and Claims
The paper claims that the geometrically-resolved DNS approach is essential for identifying the specific mechanisms by which blade number influences VAWT aerodynamics. The primary significance lies in the discovery of the BVI-induced premature DSV breakup, a mechanism that accelerates the transition to bluff-body wake dynamics independently of blockage effects.

The authors assert that these findings have direct implications for the design of closely-spaced turbine arrays and coupled-pair configurations. Because the inflow experienced by a downstream rotor is fundamentally dependent on the blade number of the upstream turbine, low-blade-count pairs will encounter large, coherent DSVs, whereas high-blade-count pairs will encounter a more diffuse, disturbed field. The paper concludes that DNS of paired configurations is warranted to fully understand how these distinct inflow conditions affect downstream performance, noting that the faster wake recovery and earlier self-similarity onset at higher solidities may allow for more accurate analytical modeling of such arrays at shorter downstream distances.