Post-Liquefaction Soil-Structure Interaction of Pipelines Buried in Sand

Abstract

The starting point of this thesis lies in the area on Groningen, Netherlands. More specifically, induced earthquake activity is observed in this area due to gas extraction activities, which reduce underground gas pressure. In combination with the location’s geomorphology, liquefaction can occur due to earthquake excitations. Furthermore, the area of Groningen is a location with urban activity and infrastructure, such as pipelines, present. The post-liquefaction effect of these pipelines is their uplift, which can lead to failure, along with all its negative consequences. Furthermore, the creation of two different models, capable of describing the soil-structure interaction of pipelines and post-liquefied soil was the main objective of this thesis. The means by which these models were created was the implementation of the spring and dashpot method. More specifically, Kelvin-Voigt models were used for the ground’s behaviour to be modelled. These models required two main parameters to be specified in each case, the spring coefficient (k) and the dashpot coefficient (c). In order for the spring coefficient to be determined, the creation and usage of post-liquefaction p-y curves was deemed necessary while the problem specific dashpot coefficients used, had previously been determined by means of physical modelling. The ground profile consisted of fully saturated loose sand and the pipes investigated had diameters of 110, 160 and 200mm. The first model was a single degree of freedom one, with that being the vertical. As for the second one, it was a multi degree of freedom one. More specifically, it consisted on two degrees of freedom, the vertical and horizontal ones. The results of these models were validated by comparison to previous studies and further investigation of affecting parameters was carried out. More specifically, the way and the degree up to which several parameters of both ground and structure characteristics affect their interaction were investigated. More specifically, the first parameters considered were the pipelines’ diameter as well as their weight, resulting from their geometry and transported material. Next, the effect of the pipelines’ burial depth was considered and then, the effect of the soil’s initial stiffness. Finally, the effect of the dashpot coefficient was investigated for the multi degree of freedom models. A comparison between single and multi-degree of freedom solution results was also carried out. The results of this study were validated by comparison to previous research findings. Most results and conclusions were in complete agreement with the so far existing bibliography. However, what was found to be inaccurate was the previously determined dashpot coefficient for 110mm diameter pipelines and hence, these were left out of scope of study.