PSI - Issue 84

Alessandro Lipari et al. / Procedia Structural Integrity 84 (2026) 615–622

619

The factors in Table 1 apply only to the load models in the NTC, such as the LM1. Partial safety factors for the transitable condition (2), i.e., when using vehicles from the CdS, depend on the level of control enforced on the road. Three levels are provided: Level 1 imposes only sample weight checks ( γ CdS,1 = 1.60), Level 2 has continuous weight checks with diversion of overloaded vehicles ( γ CdS,2 = 1.35) and Level 3 is the most restrictive with continuous weight checks with guaranteed diversion of overloaded vehicles ( γ CdS,3 = 1.10). No variations for different Consequence Classes are specified. It is worth noting that consequence factors k F , ranging from 0.9 to 1.1, have been introduced in the Eurocodes. These have to be applied to the partial safety factors for permanent and variable actions (European Committee for Standardization, 2023b) and effectively vary safety factors only for bridges in CC1 ( k F = 0.9) and in the upper class of CC3 ( k F = 1.1, applicable only to cases specified by the relevant authority or agreed for a specific project by the relevant parties). Therefore, in most practical cases k F = 1. The use of site-specific traffic data is certainly most useful when checking for operational conditions. Indeed, if a bridge failed the operational condition verification, knowledge of the actual traffic crossing the bridge is likely to reduce the characteristic values to be used in the structural verifications. Site-specific traffic data may also be used for the adequate condition verifications, by applying to the traffic data the same safety levels required at the design stage ( T = 1000 years and full γ Q as in Table 1); however, the guidelines do not currently mention any potential reduction of the LM1, e.g., through the adjustment factors α (Fig. 1), but state that the 1000-year return period may be reduced, provided that traffic is monitored. In theory, site-specific traffic data may also be used when checking for transitable conditions. However, the benefit of using site-specific traffic data is precisely to avoid any restrictions or limitations to the bridge. Moreover, traffic data could be collected anyway for enforcing the weight restrictions. Therefore, it seems more sensible to collect traffic data with a WiM system for a representative period and verify the bridge for operational conditions. 5. Traffic load modelling In general, the process of traffic load modelling based on site-specific data consists of (Lipari, 2016): 1. Traffic data collection 2. Generation of a database 3. Simulation of load effects 4. Extrapolation 5. Model calibration For assessing existing structures, steps 1 to 4 generally suffice, as the main objective is to derive a site-specific characteristic value. Step 5 is required only when a reliability-based design or assessment is carried out. Step 1. Since it is unfeasible to collect the large amount of data needed to identify extreme loading scenarios, data is collected for a representative period, and then the traffic characteristics are extracted for further processing, e.g., by forming a histogram of GVWs and fitting a distribution (OBrien et al., 2005). Bi-normal distributions are commonly found for GVWs, but care has to be taken in the fitting of the histogram tail. Step 2. An extended garage of vehicles is generated using common Monte Carlo techniques based on the collected data. Step 3. The generated traffic database is passed over the existing bridge and the load effects are typically computed through influence lines or surfaces. Step 4. As it may be still computationally demanding to simulate traffic for the very long periods required to identify extreme loading scenarios, the simulated load effects are statistically extrapolated. The characteristic value z k corresponding to the set probability of non-exceedance F* , or return period T*, can be found by inversions of Eqs. (4) and (3) for Type I, and Type II and III GEV distributions, respectively: = − (− ( ∗ )) (5) = + [(− ( ∗ )) − −1] (6)