Advanced FEM for Strongly Coupled Fluid-Structure Interactions in Offshore Applications Including Low and Zero Mass Ratios

点击翻译按钮获取中文摘要

Fluid-Structure Interactions (FSI) are critical to the design of many engineering systems. From sea-worthiness to airplane maneuverability, nearly all mechanical systems involve a certain degree of coupling between structures and fluid flows. At the Ecole Polytechnique de Montréal, we have developed effective monolithic fully coupled algorithms for generic FSI with flexible structures for 3D unsteady configurations. Our work has generated sufficient interest that several companies have funded extensions of our research (Gaz Transport & Technigaz, Commissariat à l’Énergie Atomique, TOTAL SA).The research we propose is a natural extension of our previous research on simulation of FSI. It will focus on fundamental formulation and computational issues (contact, spectral element method) for accurate and consistent finite element simulation of structures undergoing large displacements and deformations induced by the flow of incompressible fluids.Our goal is to extend FSI formulations to tackle outstanding while typical challenges in the offshore industry: flow induced vibrations of offshore structures with very low (zero) mass ratios, in-line with the flow arrangements of risers and galloping of riser towers. The usual decoupled algorithms fail miserably on these problems. Also, the current trend to drill in ever deeper waters implies that riser lines carrying the oil from the ocean floor to the surface now span more than 2000 m which requires that new buoyancy devices and new riser concepts be designed. Risers, buoys and riser towers dynamical behaviors result from strong Flow Induced Vibrations that are poorly handled by traditional loosely coupled algorithms. We believe that our fully coupled approach can overcome many of the difficulties presented by these new extreme FSI. To achieve our goal we have identified the following objectives:1. REFINE our fully coupled monolithic computational method to improve its accuracy and computational efficiency. We will extend our FE Formulation to isoparametric, higher order spectral elements to improve spatial accuracy and benefit from their higher computational efficiency by combining time-step and mesh adaptation strategies. This constitutes a natural extension of our methodology.2. IMPLEMENT contact algorithms for rigid-body dynamics that will preserve our high-order temporal accuracy method. We favor using Lagrange multipliers for inequalities since they are free from approximations introduced by penalty methods for instance.3. APPLY the formulation to problems with very low mass ratios that are relevant to offshore applications. Validation through experimental collaborative work between the low Reynolds number chanel of the Laboratoire de Mécanique des Fluides and Fluid-Structure Interaction Laboratory of the École Polytechnique de Montréal. We have identified the following problems of practical relevance:a. Riser tower torsional and translational galloping,b. Behavior of buoyancy can,c. Riser assemblies converging to an offshore platform,This generic research is innovative, as it will lead to the development and validation of computational models to shed insight into flow-­induced vibration problems typical to offshore mechanics. The focus on low mass ratio constitutes a real added value since it is an identified weakness of present CFD/CSD softwares. Our work is motivated by the need to develop the tools required by the industry to operate safely future oil offshore field developments in Canadian ultra-deep waters. Our approach offers excellent perspectives of success because it is a natural extension of existing engineering design procedures.