PSI - Issue 78

Noemi Di Rienzo et al. / Procedia Structural Integrity 78 (2026) 1474–1481

1477

Connecticut Bridge (Connecticut, USA) (Liu & DeWolf, 2007) West Street On-Ramp (WSOR) Bridge (Anaheim, USA) (Gomez et al., 2011) Lavic Road Overcrossing (California, USA) (Choi et al., 2004)

Continuous beam

3

234.20 58.00 88.10 88.10 151.30 45.80 60.10 45.40

17.30 18.34

41.15∙10 24

1.52

2.11

-

Continuous beam

3

10.80 7.43

6.54∙10 24

2.33

-

-

Continuous beam

2

73,50 36.00 37.50

10.40 4.98

3.47∙10 24

3.19

4.62

-

(a): Value of Elastic Modulus obtained from in-situ tests on core samples taken from the structure. (b): Value of Elastic Modulus derived from design documents or standard concrete grade specified in drawings.

The dataset covers a wide spectrum of reinforced concrete bridges. In particular, the span lengths range from 12 m (Béni Khiar Bridge) to 88 m (Connecticut Bridge). Also, different length-to-width ratios were considered, from almost square decks (Béni Khiar Bridge, SL / DW = 1.21) to very elongated ones (WSOR Bridge, SL / DW = 5.56). As far as continuous bridge decks are concerned, the number of spans varied from 2 to 4, with samples having equal span lengths (Forg Bridge) or significantly variable ones (Voigt Drive). The experimental setup is also very variable in the dataset. Some of the tests were performed according to EMA approaches, which involve the use of controlled vibrations on the bridge; the majority used ambient vibrations (OMA). In one case (WSOR) a number of earthquakes were recorded and their effects used for identification, while for Lavic Road Overcrossing a series of identification were performed with the last one occurring after extensive damage due to a seismic event between September and October 1999. Different sensor types and number were also used, spanning from uni-/bi-/three-axial cabled piezo-electric devices to wireless sensors. Finally, post-processing was carried out by simple frequency analysis of the single signals to more complex Frequency Domain Decomposition or automatic Stochastic Subspace Identification. This large variability also increases the quality variance of the results, with cases where the second flexural frequency could not be identified. The experimental values of the first flexural vibrational frequencies collected are reported in Fig. 2 in relation to the span length of the decks (for continuous decks the average span length was considered). As might be expected, in both structural schemes, the natural frequencies decrease with increasing length, exhibiting a similar inverse dependency. The analysis of the literature showed that while the first frequency is generally associated to the first flexural mode, very seldom the second flexural mode follows, and often torsional or plate-like modes are characterised by similar low frequencies. This may be particularly evident in the case of bridges with low SL / DW ratios. This is also exacerbated if the bridge is horizontally curved (WSOR). The type of static scheme plays a significant role. Seen in Fig. 1, simply supported bridges seem to be characterized by higher frequencies (for both first and second modes) than continuous beams in the short-span (10-30 m) range, on average. Above the threshold of about 30 m, the two clouds tend to overlap, but the highest values are still only found in simply supported decks. With reference to span length, simply supported bridges with short spans (i.e. less than 25 m) have first-mode frequencies ranging from 7.2 to 9.4 Hz, while the two continuous supported bridges available in this span class have a frequency of approximately 5 Hz. In the second mode, the simply supported bridges reach 22-27 Hz, compared to 6-12 Hz for continuous static scheme. In the range 30-45 m the two patterns seem to converge to similar values: the first mode oscillates between 1.9 and 4.5 Hz, with the l owest recorded value (≈1.9 Hz) observed at the Sabarmati bridge (42.5 m, simple support), and values close to 4.5 Hz belonging to the Nine Wells (continuous beam, spans 27 32 m). Second-mode frequencies are between 7.7 and 12.7 Hz in single-supported bridges and between 3.5 and 4.3 Hz in continuous viaducts. Finally, the Connecticut bridge (a continuous bridge with spans ranging from 58 to 88 m) exhibits frequencies of approximately 1.5 Hz ( f 1 ) and 2.1 Hz ( f 2 ).