*3.1. Wind Tunnel Tests of the Turbine*

The research on the turbine performance was carried out in the wind tunnel of CRIACIV, which is an Inter-University Research Consortium for Building Aerodynamics and Wind Engineering clustering eight Italian Universities. The laboratory hosting the open-circuit, boundary layer wind tunnel (Figure 4a) is located in Prato (Italy). The tunnel has a total length of about 22 m, including a nozzle at the inlet (characterized by a contraction ratio of 4.2 after the honeycomb) and a T-diffuser at the outlet. The test section has a rectangular shape and it is 2.42 m wide and 1.60 m high. Thanks to the possibility of changing both the speed of the motor driving the fan and the pitch of its blades, the flow velocity inside the section can be varied almost continuously up to approximately 30 m/s. The normal free-stream turbulence intensity in the tunnel is around 0.7% [34]. In preparation to the tests carried out for the present study, the growth of the freestream turbulence was obtained by means of squared-mesh wooden grids placed in development zone of the wind tunnel (see Figure 4a). To verify the flow conditions during the acquisitions with the turbine installed in the tunnel (Figure 4b), a single-component hot-wire probe was used to determine the spectral properties of the turbulence induced by the wooden grid and the homogeneity of the flow. The model used for the study is

a two-blade, H-type Darrieus turbine having a diameter *D* = 0.5 m and a frontal area *A* = 0.4 m2. Blades are made of a single NACA0018 airfoil and feature a constant chord *c* = 0.05 m; they are connected to the rotating shaft by two inclined struts per blade (see Figure 4b). Due to the small size of the rotor, revolution speeds (and thus centrifugal loads) were high (around 1200 rpm) in order to achieve significant Reynolds numbers.

**Figure 4.** (**a**) Wind tunnel with a wooden grid for turbulence generation installed; (**b**) analyzed wind turbine.

As a consequence, the rotor was manufactured using carbon-epoxy composite [35] to ensure high-quality mechanical properties. The chosen turbulent conditions for the experimental campaign are presented in Table 1. These values represent a compromise between the levels found in urban environments [4,5] and the size limitations of the wind tunnel, as explained in [34]. A uniform wind speed profile with a mean of 9 m/s was chosen as a benchmark as it allowed the widest range of rotational speeds in the turbine prototype available and a clear comparison between the different levels of turbulence.

**Table 1.** Tests conditions reproduced during experiments.

