A rigorous treatment of unsteady aerodynamics and aeroelasticity are essential to reducing design uncertainty, improving lifetime assessment, and ultimately contributing to lowering the cost of wind energy for next-generation assets. Wind turbines (WTs), in both horizontal- (HAWT) and vertical-axis (VAWT) configurations, are routinely exposed to unsteady inflow and loading arising from an array of individual/collective sources, all of which impose varying intensity and characteristics on the blade(s) response. Aeroelastic simulations of utility-scale HAWTs on floating platforms indicate that induced surge–pitch motions can advect stall from the blade root towards and above mid-span, with non-negligible fatigue implications due to the combined effect of stronger elastic response and increased wind shear in these sections. Dynamic-stall mechanisms are long documented on Darrieus-type WTs (DWTs) at low tip-speed ratios. Because both configurations share the same class of unsteady physics and are routinely modelled with Beddoes–Leishman-type dynamic stall closures, a model-scale Darrieus rotor provides an experimentally accessible analogue for stress-testing and calibrating stall models intended for multi-MW rotors. This dissertation addresses these issues through two complementary strands that span from bladelevel flow physics to system-level aeroelastic response. First, a time-resolved Particle Image Velocimetry (PIV) campaign is conducted on a two-bladed H-type Darrieus rotor operating in a stall-dominated regime. Using a host of experimental methodologies, the statistically converged blade-level quantities are extracted and the life cycle of vortices across azimuths are documented. The vortex tracking framework quantifies convection and morphology, enabling a regime-resolved description of vortices in the attached flow, stall onset, transition to deep stall, and deep stall. These measurements provide blade-resolved, high-temporalresolution reference data for validation purposes. Against this dataset, a Double-Multiple Streamtube (DMST) model is implemented and tested. While DMST reproduces attached-operation aerodynamics with small deviations in angle of attack and relative velocity, its predictive capability degrades during stall onset and transition and in the downwind region, highlighting the need for physics-informed unsteady closures. The second strand implements and assesses an in-house engineering-fidelity aeroelastic framework, AEOLIAN, which couples Blade Element Momentum Theory with a Lumped Parameter structural model. The solver is validated against NREL Phase VI, NREL 5 MW and IEA 15 MW reference WTs under realistic inflow conditions (sheared, yawed) and, for floating configurations, under individual/combined surge and pitch motions. Comparative analyses show that AEOLIAN faithfully reproduces higher-fidelity outcomes in most operating regimes, while revealing systematic discrepancies in root and tip regions where separation and induction modelling are most challenging. Coupled aeroelastic analyses emphasize the role of blade torsion in shaping local angle of attack and loads, especially in the outboard span of more flexible blades. The Beddoes–Leishman dynamic stall model is also revisited, with its theoretical structure clarified and its calibration examined on canonical pitching-airfoil datasets. Physics-informed tuning improves its performance in thicker profiles, but also delineates its remaining limitations in deep stall. The dissertation shows that carefully designed experiments, disciplined blade-level metrics, and judicious engineering-fidelity models can be brought into quantitative agreement where it matters for design. It delivers a high-quality experimental benchmark for stall-dominated DWT aerodynamics, a validated modular aeroelastic framework for utility-scale HAWTs in fixed and floating operation, and refined guidance on the use and calibration of a dynamic-stall closure.
La tesi indaga l’aerodinamica e l’aeroelasticità in condizioni non stazionarie di turbine eoliche ad asse verticale (VAWT) e orizzontale (HAWT), focalizzandosi sul fenomeno dello stallo dinamico, mostrando come esso sia presente in entrambe le configurazioni geometriche indipendentemente dalla scala dimensionale. Da un punto di vista sperimentale, sono state condotte delle misure di velocimetria ad alta risoluzione temporale per una turbina ad asse verticale di tipo Darrieus con rotore ad H a due pale, in regime di stallo, tramite la tecnica time-resolved Particle Image Velocimetry. Quest’ultima consente di ottenere i campi istantanei e medi in fase di velocità, con i quali è possibile descrivere l’evoluzione dinamica dei vortici e ricavarne i parametri del ciclo di vita. Vengono, inoltre, ricostruite altre metriche rilevanti per il confronto fra i risultati dei modelli numerici e quelli sperimentali. In particolare, è stato implementato un modello DMST con correzioni high-thrust e sottomodello di stallo dinamico, i cui risultati sono stati confrontati con quelli ricavati dalle analisi PIV, con altri dati sperimentali presenti in letteratura e con una simulazione bidimensionale Reynolds Average Navier-Stokes (RANS), evidenziando il ruolo, potenzialità e limiti dei modelli d’ordine ridotto nelle condizioni di stallo. Per le HAWT, il codice aeroelastico AEOLIAN (BEMT + multibody a masse concentrate) viene validato impiegando una turbina NREL Phase VI (UAE) ed una NREL 5 MW, simulando diverse condizioni: un flusso a monte del rotore uniforme, la presenza di uno strato limite atmosferico, una direzione del vento non ortogonale al rotore, e condizioni galleggianti-offshore. L’analisi è estesa anche alla IEA 15 MW realizzando un confronto fra diverse classi di modelli quali: Blade Element Momentum Theory (BEMT), actuator-line, free-vortex wake e RANS risolte alla scala delle pale. Le risultanze mostrano che la deformazione torsionale della pala governa le variazioni locali dell’angolo d’attacco e dei carichi, soprattutto nella porzione esterna della pala (outboard). Infine, una revisione sistematica e una calibrazione guidata dalla fisica del modello di Beddoes–Leishman su profili oscillanti ampliano significativamente il suo campo predittivo. Contributi principali: (i) Valutazione sperimentale–numerica blade-level di una Darrieus in regime di stallo. (ii) Procedura per la determinazione della posizione ed orientamento della pala in ogni immagine PIV, seguita da una ricostruzione dei campi medi in fase di velocità assoluta, indotta e relativa, della circolazione, e delle metriche necessarie alla validazione dei modelli numerici nello specifico regime di flusso. (iii) Sviluppo e verifica di AEOLIAN per HAWT on/offshore. (iv) Analisi aeroelastica e verifica di Reference Wind Turbines (RWTs) industriali (NREL 5 MW, IEA 15 MW) in condizioni complesse. (v) Rivalutazione e ricalibrazione del modello di Beddoes–Leishman. Nel complesso, i risultati di questo studio consentono lo sviluppo e la validazione di nuovi modelli industriali ad alta fedeltà per le turbine eoliche di nuova generazione.
Integrated experimental and numerical investigation of unsteady aerodynamics and aeroelasticity in horizontal and vertical axis wind turbines
ARYAN, NAVID
2026
Abstract
A rigorous treatment of unsteady aerodynamics and aeroelasticity are essential to reducing design uncertainty, improving lifetime assessment, and ultimately contributing to lowering the cost of wind energy for next-generation assets. Wind turbines (WTs), in both horizontal- (HAWT) and vertical-axis (VAWT) configurations, are routinely exposed to unsteady inflow and loading arising from an array of individual/collective sources, all of which impose varying intensity and characteristics on the blade(s) response. Aeroelastic simulations of utility-scale HAWTs on floating platforms indicate that induced surge–pitch motions can advect stall from the blade root towards and above mid-span, with non-negligible fatigue implications due to the combined effect of stronger elastic response and increased wind shear in these sections. Dynamic-stall mechanisms are long documented on Darrieus-type WTs (DWTs) at low tip-speed ratios. Because both configurations share the same class of unsteady physics and are routinely modelled with Beddoes–Leishman-type dynamic stall closures, a model-scale Darrieus rotor provides an experimentally accessible analogue for stress-testing and calibrating stall models intended for multi-MW rotors. This dissertation addresses these issues through two complementary strands that span from bladelevel flow physics to system-level aeroelastic response. First, a time-resolved Particle Image Velocimetry (PIV) campaign is conducted on a two-bladed H-type Darrieus rotor operating in a stall-dominated regime. Using a host of experimental methodologies, the statistically converged blade-level quantities are extracted and the life cycle of vortices across azimuths are documented. The vortex tracking framework quantifies convection and morphology, enabling a regime-resolved description of vortices in the attached flow, stall onset, transition to deep stall, and deep stall. These measurements provide blade-resolved, high-temporalresolution reference data for validation purposes. Against this dataset, a Double-Multiple Streamtube (DMST) model is implemented and tested. While DMST reproduces attached-operation aerodynamics with small deviations in angle of attack and relative velocity, its predictive capability degrades during stall onset and transition and in the downwind region, highlighting the need for physics-informed unsteady closures. The second strand implements and assesses an in-house engineering-fidelity aeroelastic framework, AEOLIAN, which couples Blade Element Momentum Theory with a Lumped Parameter structural model. The solver is validated against NREL Phase VI, NREL 5 MW and IEA 15 MW reference WTs under realistic inflow conditions (sheared, yawed) and, for floating configurations, under individual/combined surge and pitch motions. Comparative analyses show that AEOLIAN faithfully reproduces higher-fidelity outcomes in most operating regimes, while revealing systematic discrepancies in root and tip regions where separation and induction modelling are most challenging. Coupled aeroelastic analyses emphasize the role of blade torsion in shaping local angle of attack and loads, especially in the outboard span of more flexible blades. The Beddoes–Leishman dynamic stall model is also revisited, with its theoretical structure clarified and its calibration examined on canonical pitching-airfoil datasets. Physics-informed tuning improves its performance in thicker profiles, but also delineates its remaining limitations in deep stall. The dissertation shows that carefully designed experiments, disciplined blade-level metrics, and judicious engineering-fidelity models can be brought into quantitative agreement where it matters for design. It delivers a high-quality experimental benchmark for stall-dominated DWT aerodynamics, a validated modular aeroelastic framework for utility-scale HAWTs in fixed and floating operation, and refined guidance on the use and calibration of a dynamic-stall closure.| File | Dimensione | Formato | |
|---|---|---|---|
|
Tesi_dottorato_Aryan.pdf
accesso aperto
Licenza:
Creative Commons
Dimensione
66.96 MB
Formato
Adobe PDF
|
66.96 MB | Adobe PDF | Visualizza/Apri |
I documenti in UNITESI sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/20.500.14242/357154
URN:NBN:IT:UNIROMA1-357154