PSI - Issue 84

Diego Esposito et al. / Procedia Structural Integrity 84 (2026) 1198–1205

1204

overall stiffness does not seem to produce the desired effects, particularly regarding the rotation trend of the midspan section, where (for spans C1 and C2) an additional increase in angle is observed up to configuration 3B, while all theoretical models predict a reduction after configuration 2B (Figure 9). This observation suggests that a more accurate simulation of the test should differentiate stiffness levels between the two girders of the deck, applying greater flexibility to the more loaded girder. This is further justified by the fact that the deck lacks transverse diaphragms except at the supports; therefore, the two girders exhibit minimal mutual interaction. Consequently, two additional models were developed: the first (FEM1) assigns 50% stiffness to girder 1 and 70% to girder 2; the second (FEM2) further reduces girder 1 stiffness to 35% while for girder 2 the stiffness unchanged. A response curve was then constructed by successive model approximation as follows: up to load configuration 2A, the EF model with stiffness E70% applies; for load configuration 3A, FEM1 applies; and for the subsequent three load steps (3B–4A–4B), FEM2 applies. The graphs below demonstrate the effectiveness of this approach, particularly regarding midspan section rotation (Figure 10).

Figure 9. Comparison of Deflections

Figure 10. Comparison of the Rotation of the Midspan Section of the Deck

Finally, at load step 3A, the prestressed beam exceeds its decompression stress and the crack opening carries out a significant stiffness loss. The previous load configuration, 2B, can therefore be considered the serviceability limit condition for the bridge. In fact, the deflection-load graph remains linear until this configuration, because the concrete stresses are within values of compression stress or low tensile. Up this load level the structural response begins to deviate significantly from the linear elastic behavior, that is the fundamental design hypothesis for a prestressed concrete structure. Conclusions The paper presents the application of a combined diagnostic methodology and the development of a numerical model for an existing post-tensioned reinforced concrete bridge characterized by significant and widespread defects.