PSI - Issue 84

Donatella de Silva et al. / Procedia Structural Integrity 84 (2026) 1353–1360

1358

a)

b)

c)

700

time [min]

10 15 20 25 30 35 40 45

unprotected s= 1 cm s= 1.4 cm

0

30

60

90

120

150

180

0

600

s= 2 cm s= 3 cm

500

-0,1

400

-0,2

300

-0,3

200

HRR [MW]

Temperature [ C]

-0,4

100

-0,5

unprotected s= 1 cm s= 1.4 cm

0 5

0

Vertical displacement [m]

0

30

60

90

120

150

180

-0,6

0

20

40

60

80

100 120

s= 2 cm s= 3 cm

time [min]

time [min]

-0,7

Fig. 1. Temperature in the bottom flange of steel main profile for unprotected configuration and for the protected ones; fire scenario S2; vertical displacement for the fire scenario S2 in function of the unprotected or protected configurations.

The thermo-mechanical analyses allowed evaluation of the structural response of the bridge under the combined action of thermal and mechanical loads, highlighting the role of passive fire protection on both global and local structural behaviour. In the unprotected configuration, the rapid temperature increase in the steel components led to a severe reduction in stiffness and load-carrying capacity. This resulted in large vertical deflections, local instability phenomena, and, in some fire scenarios, structural collapse, confirming the pronounced vulnerability of unprotected steel members when exposed to intense thermal actions. An example of the mid-span vertical displacement response for fire scenario S2 is reported in Fig. 1c. By contrast, the protected configurations showed a markedly improved response. The presence of the insulating layer delayed heat penetration into the steel cross-sections, allowing the mechanical properties to be retained for a longer time. Consequently, stiffness degradation occurred more gradually, and vertical displacements were significantly reduced throughout the fire exposure. The effectiveness of the protection increased with the insulation thickness; however, consistent with the thermal analyses, the incremental benefit became less pronounced for thicknesses exceeding approximately 2 cm. Finally, the protective layer also influenced the cooling phase. Compared to the unprotected case, heat dissipation occurred more slowly, leading to reduced thermal gradients within the cross-section but prolonging the duration of elevated temperatures. The fragility response of the steel-concrete composite (SCC) bridge is examined by comparing the unprotected configuration documented in previous studies with the protected configurations analysed in the present work. Fragility curves are developed using both the peak heat release rate (HRR) and the total fire load (Q) as intensity measures, with particular attention to performance levels III and IV, see Fig. 2. a) b)

Fig. 2. Fragility curve for PLIII: a) HRR peak as IM; b) Q as IM.

For the unprotected bridge, reference data show a high probability of exceeding the considered performance levels for most fire scenarios, with SC1 generally emerging as the governing sub-criterion. When passive fire protection is introduced in the present analyses, a significant reduction in thermal demand on the steel-concrete elements is observed, leading to consistently lower DCR values across all scenarios. Therefore, the fragility curves of the protected configurations shift toward higher fire intensity levels, indicating a substantial decrease in the probability of exceeding