Analyses of Capacity and Response of Plate Anchors: 2-D and 3-D Modeling with Advanced Constitutive Soil Models

Evaluation of the uplift capacity of plate anchors in saturated soils is an important aspect of the offshore foundation of various structures. These anchors are attached to the structures through mooring lines and embedded to a sufficient depth to provide the needed pullout resistance. In most of the literature reviewed, a constitutive model such as Tresca or Mohr- Coulomb has been used in characterizing the anchor response under loading. Previous studies show that analytical and numerical methods (using Mohr-Coulomb and Tresca models) overestimate the capacity of plate anchors. There exists a need to study the anchors pull out capacity using various soil models to understand better the pullout capacity and failure mode development with the progress of deformation level. The first two parts of this study present the results of finite element simulation of a plate anchor in saturated soils (clays and sands) using different constitutive models and soil properties. Mohr-Coulomb, Modified Cam-Clay, and Soft- Soil models are used to represent the clay. Mohr-Coulomb and Hardening Soil Models are used to represent sands. The dilatancy and size effects for the plate anchors in saturated sands are investigated and discussed. The capacity factors (Nc for clays and Nq for sands) of the plate are assessed through the application of displacement control approach using the computer program PLAXIS 2D. The third part of this study investigated the failure mechanism of plate anchors using PLAXIS 3D. Different soil properties and embedment depths are considered, and the results are compared with the available experimental observations in the literature. Also, the effect of a plate's shape on the failure mechanism and break out factor is investigated. Shape factors are presented for square and strip plates at different embedment depths. The results of the 2D and 3D finite element analyses are compared, and preciseness of 2D analysis for this problem is investigated. The last part of this study investigates the effect of bad keying process of Suction Embedded Plate Anchors (SEPLA). SEPLA is a type of plate anchors, which is embedded in the seabed by using a suction caisson. The results show that, Mohr-Coulomb model considers a constant mean effective stress during shearing and yields a higher pullout capacity in comparison with Modified Cam-Clay, and Soft- Soil models in undrained clays. Modified Cam-Clay and Soft- Soil models provide predictions that are in good agreement with each other as well as experimental data, especially for the axisymmetric condition. In sand, the results of the large deformation analyses using Hardening Soil model are in a good agreement with existing experimental data. A very good agreement is found between the results from Hardening Soil model and those from Mohr-Coulomb model at the shallow depths (H/D<4). At larger depths, Mohr-Coulomb model provides lower capacities in saturated sands. The increase in dilatancy angle increases the capacity by extending the failure zone. The effect of dilatancy angle in the shallow embedment depths and loose sands are negligible. However, this effect becomes significant at larger embedment depths and dense sands. The results show that the size effect is negligible when the friction angle is independent of stress level. When friction angle is changing (decreasing) with the stress level, the larger plate provides smaller breakout factor at a given embedment ratio. The results from the 3-D modeling indicated that the failure mechanism in sands depends on relative density and rigidity of the soil and depth to width ratio of the foundation. In the case where the plate anchor is not rotated ideally during SEPLA keying, a reduction in the final breakout factor is computed, and found to be a function of rotation of the plate.