PSI - Issue 5

M.H. Hebdon et al. / Procedia Structural Integrity 5 (2017) 1027–1034 Liu et al. / Structural Integrity Procedia 00 (2017) 000 – 000

1030

4

a

b

Joint at the top chord

Refined part

Rigid region

Joint at the bottom chord

y

Coupling interface

x

z

Fig. 4. (a) Global FE model of HLB; (b) Refined connection.

3. Fatigue provisions in Eurocode

3.1. Fatigue load

Load generated by traffic is generally very complex for the variability of truck geometry and uncertainty of traffic, inducing different degrees of fatigue damage on the details in bridges. To treat the complexity of loading in the fatigue evaluation, five FLMs have been defined in EN 1991-2 (CEN, 2003) to predict the actual loading conditions accurately through an equivalent fatigue damage at specific details, which can be categorized into two design concepts: unlimited fatigue life (such as FLM-1 and 2) and limited fatigue life (such as FLM-3, 4 and 5). The FLM-1 is composed of double-axle concentrated loads (called the Tandem System) applied in conjunction with a uniformly distributed load. The FLM-2 consists of a set of frequent lorries defined by detail geometries and axle loads. The FLLM-1 and 2 are intended to be used to determine the maximum and minimum stresses when designing for an unlimited fatigue life. The FLM-3 consists of a four axles fatigue truck, each axle having 120 kN (26.98 kip), and each axle consisting of two wheels with a square contact surface on each side of 0.40 m (15.75 ″ ), as shown in Fig. 5. The FLM-4 is a comprehensive mode including five types of trucks with different configurations and different distribution ratios in each lane. The FLM-5 is based on recorded road traffic and a direct application of measured traffic data. The FLM-3, 4 and 5 are intended to be used for fatigue life assessment by reference to fatigue strength curves defined in EN1992 to EN1999. The fatigue truck specified in Eurocode includes the dynamic load amplification appropriate for pavements of good quality (except for an additional amplification factor Δφ fat considered for the section near the expansion joint). To define the fatigue and fracture limit state and assess the finite fatigue life, load should be specified along with the frequency of load occurrence. In EN 1991-2, the number of heavy vehicles (maximum gross vehicle weight more than 100 kN) is defined and symbolized with N obs per slow lane (i.e. a lane used predominately by trucks) every year. Indicative values for N obs are assumed to be 2.0 × 10 6 for roads and motorways with 2 or more lanes per direction with high flow rates of lorries (i. e. about 5479 lorries per lane every day). Although the HLB in the middle of the city, it is very special for located between the mainland and the island and few ways are available between the two places. Besides, the average daily truck traffic is 6104 (Rossigali et al., 2015) according to the statistical result of traffic in Brazil. Hence, the above traffic category specified in Eurocode is suitable. 10% of N obs may be taken into account for the fast lanes (i.e. lane used predominantly by cars). 3.2. Frequency of the fatigue load

3.3. S-N curve

In Eurocode, 14 S-N curves, which are equally spaced in log scale, as shown in Fig. 6. Each curve is characterized by the detail category Δ σ C (i.e. value of the fatigue strength limit at 2 million cycles, expressed in MPa). The spacing between each curve corresponds to a difference in stress range of about 12%. The slope coefficient m is equal to 3 for stress ranges above the constant amplitude fatigue limit (CAFL), Δ σ D , at 5 million cycles, and equal to 5 for stress ranges between the CAFL and cut-off limit, Δ σ L , at 10 million cycles. The CAFL is about 74% of Δ σ C and cut-off