Unmanned aerial vehicles have proliferated in the last few decades, with applications that include military, commercial and recreational. Their size and typical flight velocities are characterized by moderate Reynolds numbers O(10^4-10^5). For such conditions, boundary layers can remain laminar and therefore are highly susceptible to separation and the generation of laminar separation bubbles (LSBs), when compared to more conventional aircraft. Extensive research has been done to study the influence of Reynolds number, angle of attack, sweep angle or freestream turbulence level on the nature of LSBs. However, the study of LSBs subject to unsteadiness is rather limited. This is especially relevant for small aircraft that typically fly in gusty environments such as cities. The problem is aggravated by the recent shift toward composite manufacturing, which allows more efficient high-aspect-ratio configurations, but that deform considerably more when subjected to unsteady loads.
The LSB that forms on the suction side of a modified NACA 64_3-618 airfoil at a chord-based Reynolds number of Re = 200k is studied in a series of wind tunnel experiments conducted at The University of Arizona. Three different flow measurement techniques are considered in the experiments to identify the bubble: surface pressure measurements, Particle Image Velocimetry and Infrared Thermography. The capabilities of the three techniques are first explored in a static characterization of the LSB over a range of angles of attack. For the conditions tested, excellent agreement between the techniques is obtained, showing an upstream shift of the bubble with increasing incidence. For the study of static LSBs, the infrared approach is superior, given its higher spatial resolution and experimental simplicity.
The complexity is then increased to study the influence of aerodynamic unsteadiness on the bubble. For this purpose, two different types of structural motion are imposed on the wind tunnel model. A first experiment considers a pitching-type motion, with reduced frequencies up to k = 0.25. While surface pressure measurements and PIV are not heavily affected by the change in experimental conditions, the infrared approach becomes limited by the thermal response of the surface. To overcome this limitation, an extension of the recently proposed Differential Infrared Thermography (DIT) method is considered. Even so, the unsteady behaviour of the bubble can be only partially detected with this method. All-three techniques considered indicate a hysteresis in bubble location between the pitch up and pitch down parts of the motion, caused by the effect of the aerodynamic unsteadiness on the adverse pressure gradient.
The second type of structural motion studied consists of a sinusoidal plunge, with an amplitude of h = 6% of the airfoil chord and a reduced frequency of k = 0.67. The surface pressure measurements and PIV still capture a hysteresis in bubble location along the cycle, expressed in terms of the effective angle of attack induced by the plunging motion. However, due to the increased frequency of the motion, the thermal response of the surface reduces and the infrared approach fails to detect the unsteady bubble. 

The technical advancements achieved during the last decade have permitted the implementation of low-fidelity numerical methods to analyze simplified aeroelastic problems. However, the available computational power is still incapable of dealing with more realistic events involving complex structural dynamics and flow regimes. For this reason, experimental wind tunnel campaigns are still required to characterize relevant fluid-structure interaction events and to assist in the development of more accurate computational models. Nevertheless, conventional measurement systems present several constraints that limit the complete description of the phenomena being analyzed. This thesis is part of the research effort to define alternatives capable of providing full-field and simultaneous information of both flow and structural quantities in a non-intrusive way.

A wind tunnel experimental campaign has been developed in the Open Jet Facility to evaluate the ability of Lagrangian Particle Tracking (LPT) to characterize aerodynamic loads acting on large-scale flexible bodies. Three different load reconstruction approaches have been considered for this purpose: integral momentum conservation equation in its classical formulation, integral momentum conservation equation with an alternative formulation based on vorticity, and Kutta-Joukowski theorem. The accuracy and precision of each of the forenamed procedures have been assessed in both steady and unsteady inflow conditions taking force balance measurements as a reference.

The results of this thesis have proven the capabilities of LPT to determine the aerodynamic loads in aeroelastic wind tunnel investigations with acceptable accuracy and precision. Finally, it has also been demonstrated that the range of information provided by these techniques is superior to that achievable with conventional measurement devices, which improves the overall characterization of the fluid-structure interaction phenomena.

A proportionate controller is investigated experimentally for unsteady load alleviation purposes on a 2D wing model with a trailing edge flap. This study is supposed to be a simplified version of a Smart Rotor, a mechanism which deals with the reduction of unsteady loads acting on a wind turbine and aims to understand the behavior of this type of control under dynamic stall conditions. The controller acts on the velocity of the flaps, and pressure sensors are used to detect the unsteady loads, which are generated by actuating the wing model in a sinusoidal motion. Two different regimes are considered: attached flow (oscillations around zero angle of attack, where little flow separation is expected) and dynamic stall (oscillations close to the static stall angle). The influence of actuation frequency and controller time lag is also studied, as these parameters were found to be crucial for such a concept during the literature review. A reduction of 87.5% in the standard deviation of the lift is obtained for a frequency of 0.2Hz and time lag in the control system of 12ms for attached flow conditions. The reduction of the standard deviation of the lift deteriorates for increased frequency and time lag. The proposed controller is also able to reduce the loads during dynamic stall, although the reduction is smaller, close to 40% in the reduction of the lift coefficient. A maximum reduction of 59% in the peak-to-peak loads experienced during dynamic stall is observed, for 0.2Hz and a time lag of 12ms. Time lag is again crucial for this reduction, and larger and faster actuation systems are needed to fully compensate for the larger drop in loads. However, the active load control of the loads during dynamic stall can negatively affect the aerodynamic damping of the model, since the flap actuation is dependent on the detachment of the leading-edge vortex from the surface of the wing. The flap actuation is also shown to delay the onset of dynamic stall, by increasing the static stall angle with respect to the case without flap deflection.