PSI - Issue 78

Antonio Cibelli et al. / Procedia Structural Integrity 78 (2026) 1221–1228

1225

once near the column (to maximize its displacement), and once at mid-span (to emphasize effects on the truss). A total of 36 scenarios were considered, of which 29 involve multiple vehicles. The entire set of scenarios, along with their corresponding intensity measures (energy released and peak heat release rate), is reported in Table 1.

Table 1. Fire scenarios considered for a vehicle storage facility.

q fd [GJ]

HRR peak [kW]

q fd [GJ]

HRR peak [kW]

ID Involved vehicles

ID Involved vehicles

1 Moto 2 Car3 3 HGV 4 Bus 5 Tanker 6 1/2 HGV

1.0 1324

19

(Car3 + Car2 + Moto)y

7.4

4880

5.0

4395

20 (3Moto + Bus)y

13.0 10724

240.0 206500 21 (Car3 + Car2 + Car3)x

10.6 10.8

7499 5253

52.0

39500

22 (2Car3 + 2 x Car2)y 23 (2Car3 + Car2 + Moto)y

472.0

296430

10.2 5301

120.0 103250 24 (3Moto + Car2)y

1.9 1.4

1729 1743

7 Car2

40.0 1473 7.17 4542 5.36 5407

25 (4 x Moto)y

8 9

(Car3 + Car2)x (Car3 + Moto)x

26 (Bus + HGV + Car2)y

228.6 163909 229.7 164284

27 (Bus + HGV)y

10 (Moto + Bus)x 11 (Car3 + HGV)x 12 (Moto + Tanker)x

61.6

39500

28 29

(Bus + HGV + Bus)y

261.6 199.5 409.2 733.2

177639 100652 261444 1817308

179.6 832.9

156364

(Bus + 1/2 HGV + 1/2 HGV)y

222475 30 (2HGV)y

13 (4Moto)x

1.4 1743 1.9 1729

31 (2HGV + 1/2 HGV)y

14 (3Moto + Car2)x

32 (3 x HGV)y

1060.7 3634615 1155.3 3659757 1184.4 3684899 1061.4 3634936

15 (Car3 + Car2 + Moto)x 16 (Moto + Bus + Car3)x 17 (Car3 + Moto + 1/2 HGV)x

7.4

4880

33 (3HGV + 1/2 HGV)y

65.4

42748

34 (4HGV)y

64.1 53114 45.0 21107

35 (3HGV + Car2)y 36 (3HGV + Bus)y

18 1082.2 3638152 The thermal modelling was performed using the LOCAFI model (Tondini et al., 2019) to estimate the thermal impact on vertical elements and the HASEMI model (Kamikawa et al., 2002) for the trusses, both suitable for simu lating localized fires in contexts such as open parking lots or large-span environments (Heskestad, 1984) . 3.3. Modelling of passive protection The thermo-mechanical analyses were carried out on two configurations: (i) unprotected structure and (ii) structure protected with lightweight sprayed gypsum-based plaster (SFRM). To accurately simulate the effectiveness of the protection, the physical and thermal properties of the material (thermal conductivity, specific heat, density) were de fined as functions of steel temperature, calibrating the numerical curves against the experimental data obtained from cone calorimeter tests. The simulations were performed using SAFIR, replicating tests conducted on steel plates ex posed to non-standard heating conditions (Venezia et al., 2025). (Car3 + Car2 + Bus)y

Fig. 2. Calibration of the thermal properties for the analysed protective materials. (a) thermal conductivity vs. steel temperature for the sprayed plaster, (b) experimental vs. numerical heating curves for steel samples protected through sprayed plaster with three different section factors.