PSI - Issue 84

Fabio Mazza et al. / Procedia Structural Integrity 84 (2026) 944–951

950

(a) 2D model. (b) 3D model. Fig. 8. Time-temperature curves at different points of a central HDRB at half isolator height.

Afterwards, the serviceability limit state for normal stress of a central PC beam is checked in terms of the Guyon’s zone and resultant tendon by means of the 500 °C isotherm method. Results for the A.SV2 and A.SV4 fire scenarios reported in Fig. 9, without considering the RC slab, highlight that the verification is not satisfied towards the centre of the PC beamin either case. Further results, omitted for brevity, confirm that the serviceability condition for the A.SV2 fire scenario is verified when the contribution of the RC slab is taken into account; moreover, only for the A.SV4.4 fire scenario the ultimate bending moment verification is not satisfied considering the exceptional load combination that does not consider the traffic loads.

(a) A.SV2.

(b) A.SV4. Fig. 9. Guyon’s zone and elastic core of a central PC beam exposed to simple fire scenarios.

Finally, the cloud method is adopted for the fire fragility analysis of HDRBs, based on fitting a linear regression model in the logarithmic scale to the thirty-one pairs of the ratio between maximum temperature and its critical value (i.e. 150°C), assumed as engineering demand parameter representative of the structural response of HDRBs, versus the corresponding fire load, assumed as intensity measure. To this end, Robust Fragility software, based on the cloud data described above, is used in order to obtain fragility curves providing information on the probability of exceeding the isotherm 150°C conditioned on the maximum temperature attained during the fire (Jalayer et al., 2015). Fragility curves of a central HDRB are reported in Fig. 10 with reference to the fire location B, assuming five points in the proximity of the end steel plate (Fig. 10a) and at half isolator height (Fig. 10b). As shown, HDRBs are significantly vulnerable to fire damage at least up to one third of the diameter (i.e. i=3), and this behaviour is more evident for the points placed at half isolator height (Fig. 10b) while comparable fragility is observed for the interior points (i≥5).