This paper presents a numerical based study on the undrained load capacity of a typical torpedo anchor embedded in a purely cohesive isotropic soil using a three-dimensional nonlinear finite element model. In this model, the soil is simulated with solid elements capable of representing its nonlinear physical behavior and the large deformations involved. The torpedo anchor is also modeled with solid elements, and its geometry is represented in detail. Moreover, the anchor-soil interaction is addressed with contact finite elements that allow relative sliding with friction between the surfaces in contact. A number of analyses are conducted in order to understand the response of this type of anchor when different soil undrained shear strengths, load directions, and number and width of flukes are considered. The results obtained indicate two different failure mechanisms: The first one involves significant plastic deformation before collapse and, consequently, mobilizes a great amount of soil; the second is associated with the development of a limited shear zone near the edge of the anchor and mobilizes a small amount of soil. The total contact area of the anchor seems to be an important parameter in the determination of its load capacity, and, consequently, the increase in the undrained shear strength and the number of flukes and/or their width significantly increases the load capacity of the anchor.

RESUMOO presente relatório técnico trata de um dos projetos realizados no Laboratório de Controle -CONTROLAB do Núcleo de Computação Eletrônica -NCE da Universidade Federal do Rio de Janeiro -UFRJ, nos anos de 2004 e 2005. Este projeto é voltado para a implementação de um robô autônomo AGV (Autonomous Vehicle Guide) para navegar em ambientes internos, tais como hospitais, escritórios, residências e etc. Especializado na área de visão computacional, O trabalho apresenta um novo algoritmo 3D para visão robótica, utilizado para identificação automática do estado de portas, sem quaisquer informações prévias sobre o ambiente, tais como mapas, localização de obstáculos, iluminação e etc. As próprias características das portas que influenciariam na decisão do robô sobre a identificação de seu estado, também não ignoradas, tais como textura, cor, dimensões. Publicado no 8th IFAC International Sysmposium on Robot Control (SYROCO 2006), o trabalho apresentou a aplicação da Transformada de Hougth em imagens capturada pela visão robótica. As coordenadas de Hougth apresentavam então a localização das arestas das portas e a análise de corelação cruzada das quinas encontradas, nos davam a configuração do estado da porta (aberta, fechada, entre-aberta). Com os resultados obtidos, a navegação do robô autônomo se tornava mais segura, vez que os riscos de colisão eram minimizados. A navegação utilizando visão 3D, com processamento em tempo real, também reduzia o risco de colisão com objetos dinamicamente posicionados no ambiente. Na primeira parte deste documento, apresentamos o artigo publicado em (SYROCO 2006). Na segunda parte, apresentamos alguns dos resultados experimentais obtidos durante a realização do projeto.

This paper presents a numerical based study on the undrained load capacity of a typical torpedo anchor embedded in a purely cohesive isotropic soil using a three-dimensional nonlinear finite element (FE) model. In this model, the soil is simulated with solid elements capable of representing its nonlinear physical behavior as well as the large deformations involved. The torpedo anchor is also modeled with solid elements and its complex geometry is represented. Moreover, the anchor-soil interaction is addressed with contact finite elements that allow relative sliding with friction between the surfaces in contact. Various analyses are conducted in order to understand the response of this type of anchor when different soil undrained shear strengths, load directions as well as number and width of flukes are considered. The obtained results point to two different failure mechanisms: one that mobilizes a great amount of soil and is directly related to its lateral resistance; and a second one that mobilizes a small amount of soil and is related to the vertical resistance of the soil. Besides, the total contact area of the anchor seems to be an important parameter in the determination of its load capacity and, consequently, the increase of the undrained shear strength and the number of flukes and/or their width significantly increases the load capacity of the anchor.

As exploitation activities moves into fields located in deep water, the industry has been addressing studies aiming at concepts of offshore systems that reduce the influence of environmental loads on risers. The Buoy Support Riser (BSR) system is one of these new proposed concepts. The BSR is composed by a subsurface tethered buoy, where flexible jumpers connect the Floating Production Unit (FPU) to the BSR and Steel Catenary Risers (SCRs). Due to its complexity and non-linearity, this offshore system requires a highly refined finite element model for dynamic analysis, which demands a high computational cost. In order to increase feasibility of the analysis it is proposed a low computational cost methodology based on Artificial Neural Networks (ANN). This work aims to develop a program to train an ANN to predict the jumpers’ dynamic tension from FPU motions without running the finite element model for every time step. In this way, the purpose is to find results as reliable as those achieved in a dynamic analysis with a finite element model. Statistical parameters will be used for this comparison.

Torpedo anchors have proven to be one of the most important alternatives for mooring systems, especially in Brazilian offshore fields. Their installation involves many hydrodynamic and soil-structure interaction aspects and, therefore, their vertical stability cannot be always assured. This often results, after their installation, in an inclination between the anchor and a vertical axis called tilt angle. The traditional approach to assess the holding capacity of torpedo anchors relies on the hypothesis that the relative angle between the load direction and the axis of the anchor may be used in conjunction with a finite element model in which a perfectly vertical anchor is considered. In this work, this assumption is discussed. A parametric study in which four different tilt angles are combined to various load inclinations was conducted. The study relied on a previously proposed FE model. The results obtained indicate that the use of the traditional approach is safe for relative angles higher than 40° and lower than 140° For other angles, the approach leads to values slightly higher than those predicted with the direct consideration of the tilt angle.