چكيده به لاتين
Due to high uncertainties in the marine environment, design based on deterministic analyses and the use of experimental formula to design coastal structures such as breakwaters have high uncertainty and will lead to an unknown level of structural safety during its life time. To consider the uncertainties in the marine environment, the use of probabilistic analysis can be a good approach to evaluate and design coastal and marine structures. Using reliability method that can evaluate the probability of failure or reliability index of a structure under different failure modes by considering uncertainties, has been used in this research for probabilistic evaluation of the Shahid Beheshti port berm breakwater in Chabahar.
The results of this research include four parts. In the first part, the probabilistic evaluation of various models of recession in berm breakwaters is discussed. In probabilistic comparison of the six models of recession, Andersen-Burcharth model 2010 and Tørum model have the highest and lowest probability of berm failure with the values of 0.86 and 0.23 in the present climate, respectively. Andersen model 2014, which were modified model of Anderson- Burcharth, and the Van der Meer- Sigurdarson model, which have high reliability and have been developed using a large amount of data, have a failure probability of 0.4 and 0.6, respectively. Furthermore, the effects of climate change in all models except Van der Meer-Sigurdarson model increase the probability of failure by a maximum of 2 to 3%. The effects of change of mean and coefficient of variation of wave height, effects of water depth, berm freeboard and model uncertainty in calculating the probability of failure have been evaluated. Based on the sensitivity analysis, the parameters of wave height and wave period are the most important load parameters in all models and the parameters of armor mass and its density are the most important resistance parameters.
In the second part, the probabilistic comparison of wave overtopping models in two states of vertical waves and oblique waves is discussed. In the first section, assuming waves perpendicular to the structure (inaccurate and conservative assumption), the probability of failure in seven models was investigated. By reliability analysis of seven wave overtopping models, the results show that the Pillai et al. and Andersen models have the highest probability of failure among the models with the values of about 0.64 and 0.6, respectively. After that, Van der Meer-Sigurdarson models of 2016 and 2012 and EurOtop 2018 model have the probability of failure relatively close to the previous two models with values of 0.55, 0.51 and 0.48, respectively. Furthermore, the two models of EurOtop 2007 and Van der Meer- Janssen with the values of 0.19 and 0.11 predict the lowest probability of failure. Most breakwaters designed in the past have used the Van der Meer-Johnson formula to control the overtopping, which, compared to newer models, has led to results with very low overtopping estimates. Based on the sensitivity analysis performed in different models, the wave height parameter is the most important load parameter. Furthermore, the effect of changes in the mean and coefficient of variation of wave height, the effect of crest freeboard and the allowable overtopping rate on the probability of overtopping failure have been evaluated. In the second part, the probabilistic comparison of wave overtopping models with the assumption of oblique waves (correct assumption) and the effect of climate change for five models that consider the effects of oblique waves, is done. In this section, the effect of oblique waves reduces the probability of failure in Van der Meer-Sigurdarson 2012, 2016 and EurOtop 2018 models by about 30%. Furthermore, the effect of climate change in these three models can increase the probability of failure by up to about 7%. In all the above analysis, comparing the types of reliability methods with Monte Carlo Sampling (MCS) method, it can be stated that Importance Sampling(IS) method has the least error and First Order Reliability Method(FORM) has an acceptable error in most models and Second Order Reliability Method(SORM) in some models with additional curvature have a relatively significant error.
In the third part, the system reliability analysis is performed based on two sets of formula and considering three failure mode of wave overtopping failure, berm failure and reshaping of breakwater assuming the series system. According to the results, the first system has the probability of failure of 0.636 and the second system has the probability of failure of 0.487, which shows a difference of about 0.14. Furthermore, the effect of climate change in the second system has increased the probability of failure by 1.5%, while in the first system, this increase is very small, and it can be concluded that the effect of climate change in the second system is greater.
In the final part, the application of Bayesian networks in updating the probability of failure by observing random variables and considering six scenarios of wave height, wave period and water level has been attempted to update the probability of failure. By evaluating different failure modes, in all update scenarios, the probability of berm failure is more critical than the other two modes. Furthermore, the trend of changes in the probability of failure in the overtopping mode is very fast and due to the simultaneous effect of three parameters including wave height, wave period and water level, the probability of failure in the fourth scenario onwards is greatly reduced. The probability of failure of the reshaping mode at a value less than the wave height of 4.5 meters tends to zero, and this failure mode is only critical for severe waves.
