PSI - Issue 84

Gian Felice Giaccu et al. / Procedia Structural Integrity 84 (2026) 1055–1062

1058

The experiment has three main objectives: (1) to assess the feasibility of the proposed stabilizer through the design of a wind-tunnel-scale unit, (2) to evaluate its effectiveness via wind tunnel testing, and (3) to validate the numerical model by comparing predicted and experimentally observed flutter onset. In particular, the numerical model is used to analyze the flutter response and critical velocity of the wind tunnel model as a function of the stabilizer gyricity Ω . Northeastern University’s wind tunnel features a test section with a square cross-section of 559 mm × 559 mm and can generate steady airflow up to 18 m/s with turbulence intensity below 2%. Tests are conducted under smooth flow conditions, using an aeroelastic balance. The bridge model is suspended by eight strategically arranged extensional springs that permit vertical motion ( ℎ ) and torsional rotation ( ) of the 2DOF model. The experimental setup is illustrated in Fig. 3, and further details on the aeroelastic balance are provided in Rizzo and Caracoglia (2018). Two sectional models, denoted as Model 1 and Model 2, are designed, fabricated, and tested in the wind tunnel. Both models share a geometric scale of 1:170, with a deck width equal to =164 mm, a longitudinal span length ℓ=527 mm, and an aspect ratio ℓ/ = 3.2: 1 . The primary distinction between the two models lies in the installation of two gyroscopic devices with different dynamic properties. The gyroscopic unit is installed inside a sealed plexiglass container (Fig. 3) that, in this preliminary design, exceeds the dimensions of the deck trusses. The main characteristics of the section models and the gyroscopic devices are reported in Tables 1 and 2, respectively. Table 1. Main structural, modal properties of the wind tunnel section models, replicating the behavior of a truss-type deck similar to the Golden Gate Bridge. Mode Frequency [Hz] Damping ratio g [%] Mass/inertia of the model, found experimentally Section model 1 1 st vertical ( v ) 4.3699 0.10 2.9483 [kg] 1 st torsional ( t ) 7.1016 1.56 0.00358 [kg×m 2 ] Section model 2 1 st vertical ( v ) 4.3699 0.10 2.7933 [kg] 1 st torsional ( t ) 7.1016 1.56 0.00318 [kg×m 2 ] Table 2. Basic properties of the gyroscopic devices, installed at deck cross section in the wind tunnel. Model D [cm] s Ω [cm] M Ω [g] Ω,11 = Ω,22 [g×cm 2 ] Ω, [g×cm 2 ] [g×cm 2 ] k α [N×cm/rad] [%] [rad/s] Ω [%] Ω [%] 1 5.08 1.158 205.0 330.65 661.3 0 1727.14 23.325 20.0 36.75 6.95 2.08 2 2.54 1.130 50.0 20.16 40.32 1416.66 23.325 20.0 40.57 1.79 0.13 In Table 2, Ω , Ω , and Ω denote the diameter, thickness, and mass of each rotating disk, respectively. The quantities Ω,11 and Ω,22 are the mass moments of inertia about the local 1–1 and 2–2 axes, while Ω, is the polar moment of inertia about the local 3–3 axis. The parameters and represent the damping and torsional stiffness of each gyroscopic device, associated with the mechanism enabling the rotation about the vertical axis. The corresponding damping ratio and angular frequency , either estimated or experimentally measured, are also reported in Table 2. To assess the effectiveness of the proposed stabilizer, two dimensionless parameters are introduced: the mass ratio Ω and the polar inertia ratio Ω , M M M     = (2) (3) Here, Ω is the mass of a single gyroscopic device, is the participating mass of the section model’s deck, Ω, is the polar moment of inertia of the gyroscopic device, and is the generalized torsional inertia of the section model’s deck. Fig. 3 presents four photographs documenting the assembly and installation stages, including the plexiglass container used to protect the gyroscopic device beneath the deck. , p J  t J J   =