Clóvis de Arruda Martins

An Investigation on the Effect of Tension Variation on VIV of Risers

This paper aims at investigating the effect of vertical motion (or equivalently the effect of variable tensioning) of the floating unit on the vortex-induced vibrations of vertical risers. This is done using a numerical procedure, based on modeling assumptions, which, though simple, succeeded in describing some expected dynamic behaviors. The model simulates the riser dynamics using a finite element model coupled to a wake-oscillator model, of the van der Pol type, used to emulate the fluid dynamics. Vertical motion (or dynamic tension) is directly imposed to the top. The transverse amplitudes at each section feed the wake-oscillator, which responds with a transverse force that is applied to the riser. The rigidity matrix is updated at each time integration step. The analysis is also carried out with a commercial simulation code dedicated to riser analysis, with a similar wake-oscillator VIV module. Amplitude envelopes are extracted from the time series, showing response mode jumps. The application of the Hilbert-Huang spectral analysis technique helps distinguishing mode jumps by tracking frequency responses in time. The results of the two different dynamic models are compared with very good agreement.Copyright

Flexible Pipes: Influence of the Pressure Armor in the Wet Collapse Resistance

When submitted to high external pressure, flexible pipes may collapse. If the external sheath is damaged, all the external pressure is directly applied on the internal polymeric layer that transmits the loading to the carcass layer, which can fail due to this effect, leading to wet collapse. This failure mode must be taken into account in a flexible pipe design. A model can be set up neglecting the influence of the pressure armor, but this assumption may underestimate the wet collapse pressure value. This work aims to include the pressure armor effect in the numerical prediction of wet collapse. The main contribution of the pressure armor to the flexible pipe resistance to collapse is to be a constraint to the radial displacement of the carcass and the internal polymeric layers. Two models were developed to find the wet collapse pressure in flexible pipes. A first study was done using a ring approximation three-dimensional (3D) finite element method (FEM) model. Comparisons are made with more simplified models using a 3D FEM equivalent ring approximation. The aim is to clarify the mechanical behavior of the pressure armor in the wet collapse scenario. Parametric studies of initial ovalization of carcass and initial gaps and interference between polymeric layer and pressure armor are made and discussed.

Structural Behavior of Flexible Pipe Carcass During Launching

The interlocked carcass of a flexible pipe, besides being designed to withstand hydrostatic and squeezing loads applied by the tensioned armor layers, must resist to localized loads applied by caterpillars during the launching operation. Load distribution is dependent on the number of tensioners and on their geometry, as well as on the properties of external plastic layers and cross section design. Being the winding pitch of an interlocked carcass small enough and considering that the plastic layers extrusion process provides some confinement to this structure, one may consider, at least as a first approach, an equivalent pipe to model the structural behavior for radial loads. Any alternative approach, as a finite element model, would depend on a difficult assessment of parameters, as those related to the internal contact problem. Recovering classical results from the theory of elasticity, the structural behavior is formulated, for a general radial load distribution. The instability problem is also addressed. Despite the simplicity of such a model, results agree quite well with experiments conducted under controlled conditions. Applied to a typical carcass, the model predicts a safe behavior regarding instability and recovers elastic deformation as a function of a general dimensionless load parameter.Copyright

Crushing and Wet Collapse of Flowline Carcasses: A Theoretical-Experimental Approach

The present paper brings together theoretical predictions and experimental results, comparing crushing tests results as well as carcass wet collapse tests. The theoretical models are of two kinds: (i) numerical (FE) and (ii) analytical. The first kind is a restricted 3D version of a finite element model. The second kind is based on classic assumptions of equivalent ring behavior. Discussion is made on the real yield stress value to be adopted, as well as on the pertinence of geometric hypotheses. Sensitivity analyses, regarding ovalization and helical pitch are also presented.© 2010 ASME

Bending Stiffener Design Through Structural Optimization

Bending stiffeners are very important ancillary equipments of umbilicals or flexible risers, since they protect the lines from overbending. Their design however is a complex task, since many load cases must be taken into account; the structure itself has a section that is variable with curvilinear coordinate. To aid the designer in this task, optimization algorithms can be used to automate the search for the best design. In this work an optimization algorithm is applied to the design of the bending stiffener. First, a bending stiffener model is created, which is capable of simulating different load case conditions and provide, as output, results of interest such as maximum curvature, deformation along the stiffener, shear forces and so on. Then, a bending stiffener design procedure is written as an optimization problem and, for that, objective function, restrictions and design variables defined. Study cases were performed, comparing a regular design with its optimized counterpart, under varying conditions.Copyright

Simplified Finite Element Models to Study the Dry Collapse of Straight and Curved Flexible Pipes

Dry collapse is one of the possible failure modes of flexible pipes. It refers to the situation in which no damage occurs in the flexible pipe external sheath. In this scenario, all layers of the pipe withstand the external pressure loading in a deep-water application. Such a situation is addressed in this work, which proposes some simplified modeling techniques to represent straight and curved flexible pipes subjected to external pressure, undergoing dry collapse during simulation procedure. The results of the proposed models are compared to other reference results, from a fully three-dimensional (3D) finite element model. Good agreement has been got, even with the proposed simplifications with a large reduction in computational cost when compared to full 3D model.

Loop Formation in Catenary Risers on Installation Conditions: A Comparison of Statics and Dynamics

Loop formation may occur in cables, ropes, and also risers — flexible pipes or umbilical cables — used for offshore exploitation. The phenomenon occurs when there is enough torsion moment in the line and also a low tension condition or, even in the cases of cable slack (almost zero tension) with some torsion moment. Many references discuss about the loop formation prediction in different conditions, but mostly for the cases of an initial straight cable (rope or riser) which is usually represented by a beam. The problem of the loop formation in a catenary riser including the nonlinear contact with the seabed is very important, once during the riser installation in some conditions a low tension combined with torsion moment can lead to the loop formation, which is undesirable. There is a necessity of a better understanding of this theme. The present work explores the loop formation in catenary risers comparing the results of the static and the dynamic predictions, showing an asymptotic tendency from the dynamics to the statics. For that study, geometrically-exact nonlinear beam models were employed in the statics and in dynamics, leading to very realistic loop formation predictions.Copyright

A Numerical Method to Solve the Static Problem of a Catenary Riser

The static configuration of a catenary riser can be obtained, with a good approximation, by a perfectly flexible cable model. However, such a model cannot deal with all the boundary conditions, as for an ideal cable there is no continuity of curvature at the touchdown point, at the top and at the points where there is change in the submerged weight. At the touchdown region, for instance, the cable model overestimates the maximum curvature. For real risers, the bending stiffness effect is relevant only at small boundary layers around the points where the cable model cannot represent well the curvature continuity. This represents a big problem in the numerical integration of the differential equation of the riser, as the leading order term is very small. One approach that can be adopted is to use firstly a perfect cable model and correct later the results with analytical expressions obtained from a boundary layer method. For a two-dimensional formulation it was already shown that this approach is very good. For a three-dimensional formulation, however, such expressions are very difficult to derive and the problem must be solved numerically. This work presents a numerical method to solve the differential equation of a catenary riser, including the bending stiffness. The results obtained are compared to analytical boundary layer solutions, for a two-dimensional case, and to a full nonlinear well-known commercial computer code.Copyright

Riser-soil interaction: local dynamics at TDP and a discussion on the eigenvalue problem

The eigenvalue problem of risers is of outmost importance, particularly if vortex-induced vibration (VIV) is concerned. Design procedures rely on the determination of eigenvalues and eigenmodes. Natural frequencies are not too sensitive to the proper consideration of boundary condition, within a certain extent where dynamics at the Touch down Area (TDA) may be modeled as dominated by the suspended part dynamics. Nevertheless, eigenmodes may be strongly affected in this region, as, strictly speaking, this is a nonlinear one-sided (contact type) boundary condition. Actually, we should consider a nonlinear eigenvalue problem. Locally, at TDA, riser flexural rigidity and soil interaction play important roles, affecting the dynamic curvature. Extending and jointing together former analytical solutions on TDP dynamics and on the eigenvalue problem, obtained through asymptotic and perturbation methods, the present work critically address soil and bending stiffness effects a little further.Copyright

Hydrostatic Pressure Load in Pipes Modeled Using Beam Finite Elements: Theoretical Discussions and Applications

AbstractThis work presents a derivation of equivalent loads, coming from the integration of hydrostatic pressure fields on the internal and external walls of a curved pipe, which is modeled as a Eu...