PSI - Issue 78

70 Dario De Domenico et al. / Procedia Structural Integrity 78 (2026) 65–72 yielding in the steel diaphragm. This yielding load is approximately estimated at 20 kN, corresponding to a mid span bending moment in the transverse beam of about = × =5.5 kNm, where =275 MPa (steel grade S275, used for the IPE80) and =20 000mm 3 is the plastic section modulus of the profile. By applying a safety factor, the elastic limit load is conservatively set at = 15 kN. Consequently, during service load testing, the live load should never exceed this 15 kN threshold. In contrast, for ultimate load testing, the prestressed concrete beam is tested independently — without the transverse steel beam — until failure. 3. Selected experimental results and discussion For the sake of brevity, only some selected experimental results relevant to the tests in serviceability conditions are presented in this paper. In particular, the results of the tested specimens, in their undamaged and damaged configurations, are compared in terms of natural vibration frequencies and displacement modal shapes. The damage scenario examined in this experimental campaign is reproduced by means of four cuts of the prestressing strands, namely two cuts within each inspection hole (that, as said above, are located at 23 cm from each other). In the undamaged configuration and after each single cut the PC beam is systematically excited by means of loading-unloading cycles of amplitude 15 kN as sketched in Fig. 4, so as to reproduce damage conditions in SLS. OMA techniques, namely enhanced frequency domain decomposition (EFDD) and stochastic subspace identification – covariance (SSI-COV) methods are implemented within customized, ad-hoc developed Python routines and used to determine modal characteristics of the beam by considering a time-history segment of vibrational response triggered by a hammer shot applied to the middle of the beam, in both the vertical and horizontal direction.

dynamic tests

dynamic tests

dynamic tests

dynamic tests

dynamic tests

15kN loading & unloading, dynamic test

dynamic test

first inspection hole

Fig. 4: Sketch of the damage scenario (and loading configuration at SLS) of one of the tested PC beams with two strand cuts within each inspection hole (four cuts in total) at one quarter of the clear span (beam R03_2).

The identified frequencies of the beam R03_2 (prestressing strands cut at one-quarter of the clear span) are listed in Table 1: the removal of two out of five strands in two beam sections (which represents a local damage that should affect the beam behavior for just 50 cm on both sides of the inspection hole, where 50 cm is the transfer length of the prestressing strands) do not produce any macroscopic modification of the overall beam stiffness and, consequently,