PSI - Issue 5

H Hammoum et al. / Procedia Structural Integrity 5 (2017) 69–76 H Hammoum et al./ StructuralIntegrity Procedia 00 (2017) 000 – 000

74 6

10 20 50

39,3 47,6 57,5 65,5 90,5 116

100

1000

10000

4.2. Number of simulations

To ensure the accuracy of the failure probability P f obtained with Monte Carlo simulations, convergence tests were performed for different limit state functions as shown in Figure 4. These results show that the convergence and the stability of failure probability values are obtained from a number of simulations equal to 20 000. Therefore, the number 32 000 is used to generate Monte Carlo samples.

Fig. 4. Convergence and stability of failure probability as function of the number of simulations.

4.3. Results interpretation

We present in Table 2 the evaluation of the failure probability obtained for the different frequential flood flows and for the five failure state functions of the Reno mattress. We notice in Figure 5 that the failure probability to the velocity test increases with the flood flow rate. The failure probability is admissible only for the decennial return period, considering that for the civil engineering structures, the failure probability must be included between the following admissible values (10 -3

Table 2. Failure probability of the different failure state functions. Return P f to P f to traction P f to traction

P f to bottom

P f to banks

P f to