Reliability of flexible dolphins

Abstract

The subject of this master thesis is introduced by the lack of regulations for the design of flexible dolphins. Currently, the design of flexible dolphins is based on the Eurocode, which does not cover all aspects which are of importance in dolphin design. Consequently, different countries and associations developed their own guidelines, which often do not coincide and are used interchangeably by different advisers. Nowadays, there is no uniform design approach for the design of flexible dolphins. The partial safety factors used in dolphin design are therefore not consistent and often subject of debate. This master thesis starts with a literature review on the available design models which are nowadays most often used in flexible dolphin design. The oldest model which is still frequently used in the preliminary design stage, is the empirical method of Blum. The simplicity of this model also introduces several limitations. Some more advanced models are D-Sheet Piling and D-Pile Group, which approach the behaviour of the soil by bi-linear and non-linear springs along the pile. However, research reveals that the Finite Element Method approaches the behaviour of laterally loaded piles best. With Plaxis 3D it is possible to assign a different soil model to the each individual layer of soil, resulting in a relatively accurate approximation of the soil behaviour. However, because of the complexity of Plaxis 3D and the lack of probabilistic tools, this model is not considered to be the most appropriate for this master thesis. In order to determine which of the two springs models, D-Sheet Piling or D-Pile Group, is most appropriate for the derivation of the partial safety factors, a comparison between the models was made based on the results obtained from performed tests. It was demonstrated that both models tend to overestimate the bending moments which develop in laterally loaded piles, whereas D-Pile Group seems to give a better approximation of the deformations. However, it was found that under some conditions, D-Pile Group overestimates the reliability of the structure. Therefore, D-Sheet Piling was considered to be most suited for the determination of partial factors for dolphin design. To determine the partial safety factors for flexible dolphins, a reliability analysis was carried out with the probabilistic toolbox Prob2B. With this toolbox a FORM-analysis was performed on a flexible mooring dolphin from practice. The evaluations were performed with respect to three different limit states, namely structural failure of the cross-section of the pile, excessive deformations and soil mechanical failure. After the evaluation of the initial mooring dolphin design, modifications were introduced to the soil structure in order to examine the sensitivity of the influence factors and partial safety factors to changes. From the probabilistic evaluations with respect to structural failure, it was concluded that the reliability of the structure is mainly defined by the exerted mooring load and the structural parameters of the pile. The soil hardly has any influence on the reliability of the structure. Only the most upper soil layer prevents the development of the bending moment in the structure, and an increasing strength of this layer therefore introduces a higher reliability. With respect to the deformations of the structure, it was concluded that again the mooring load strongly contributes to the reliability of the dolphin. Furthermore, the strength of the upper soil layers is of importance. The influence of the soil on the reliability of the structure appears to decrease with increasing depth. Moreover, it was concluded that the influence of the different variables is most sensitive to changes in strength and layer thickness. The final limit state which was evaluated, was soil mechanical failure. As a result of the lateral load exerted on the pile, a passive wedge develops at the rear side of the structure. At a certain point, this wedge is fully developed and soil mechanical failure of the structure is induced. From the probabilistic analysis it was found that the mooring load also has a large, but smaller, influence on the reliability of flexible dolphins with respect to this limit state. Furthermore, it could be concluded that the soil parameters which have most influence on the reliability are the weight of the soil and the angle of internal friction. Finally, a recommendation has been made for a set of partial safety factors for flexible mooring dolphin design. It was proposed to apply partial factors equal to 1.30 on the characteristic mooring load and the stiffness of the soil. Furthermore, it was recommended to apply a partial safety factor equal to 1.15 on the characteristic value of the friction angle of the soil and to apply a factor equal to 1.05 on the mean wall thickness of the pile. For the other parameters, a partial safety factor equal to 1.00 was recommended.