Repositório RCAAP

The present report describes the research work carried-out with the aim of developing, implementing and evaluating a three-dimensional modeling procedure of fluid-structure phenomena involving slender structures, such as beams, bars and cables. The novel approach adopted consists of the combination of the Cosserat theory applied to slender beams, which accounts for geometrical nonlinearity, and the Immersed Boundary methodology, which is used to represent the interactions between the structural and fluidic domains. The study is included in the scope of Vortex-Induced Vibrations, which is a topic of great interest in the oil industry, especially as related to the modeling of risers used for oil exploitation in deep seas. According to the Cosserat theory, the deformed configuration of the structure is described in terms of the displacement vector of the curved formed by the cross-sections center of area, and the orientation of a vector bases fixed to each cross-section, with respect to an inertial reference frame. The main advantage of this theory is that is geometrically exact. Finite element is employed for discretization of the equations of motion. According to the Immersed Boundary methodology, the solid-fluid interface forces are calculated by enforcing momentum balance to the fluid particles over the interface. The main features of the proposed methodology are evaluated by means of a number of numerical simulations, both in static and dynamic regimes, regarding the structural model, in a first step and the complete fluid-structure model, in a second step. The results obtained enable to evaluate the accuracy and the main advantages and shortcomings of the methodology, especially regarding the numerical aspects. Also, they enabled to put in evidence some relevant phenomenological aspects related to the dynamic behavior of cylindrical structures with various levels of bending flexibility, subjected to transverse flows characterized by different values of the Reynolds number.

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior

Conselho Nacional de Desenvolvimento Científico e Tecnológico

The mathematical modeling of turbulent flows around complex moving geometries has always been an extensive area of practical applications and therefore takes place in the recent engineering research. A possible numerical method proposed to handle these problems is the so called Immersed Boundary Methods. This methodology is still under development and consists in separating the problem in two domains, a fixed Eulerian domain used to discretize the fluid equations and a Lagragian domain used to represent the solid/fluid interface. Since there is no geometric dependence between these two meshes, the immersed boundary method can easily handle moving or deformable bodies immersed in the fluid flow. This work presents an extension of the Virtual Physical Model, an immersed boundary methodology developed at the LTCM/UFU, to simulate fluid flows at high Reynolds numbers around moving or deformable bodies. The model was used in the simulation of immersed deformable bodies in laminar flows and was applied in shape optimization problems. Simulations of the turbulent flow past a pitching NACA 0012 airfoil was also presented. A brief comparative studied of three turbulence methodologies implemented with immersed boundary methods is also presented in this work. The results were compared with the experimental and numerical results available from literature, and a good physical coherence was obtained.

Conselho Nacional de Desenvolvimento Científico e Tecnológico

Conselho Nacional de Desenvolvimento Científico e Tecnológico

Conselho Nacional de Desenvolvimento Científico e Tecnológico

Fundação de Amparo a Pesquisa do Estado de Minas Gerais

Conselho Nacional de Desenvolvimento Científico e Tecnológico

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior

The search for numerical methods aiming at the solution of solve the Navier-Stokes equations accurately and with a high convergence rate represents a major area of interest for CFD researchers. Such a quest is motivated by physical phenomena, like turbulence in fluids, in which only accurate methodologies with high convergence rate may allow to satisfactorily obtain a physical solution that characterizes the problem. One of the main difficulties in obtaining high accuracy in the numerical solution of the Navier-Stokes equations is the high computational cost. In order to circumvent such a drawback, in the MFLab, a new methodology has began to be developed with the works of Mariano (2007), Moreira (2007), Mariano (2011). Such a methodology is based on the coupling of the Fourier pseudo-spectral (FPSM) and immersed boundary methodologies (IBM). A main characteristic of the FPSM is the high numerical accuracy associated to a relative low computational cost, since solving the linear system of the pressure-velocity coupling is not necessary. However, the FPSM methodology has its applicability restricted only to problems with periodic boundary conditions. In order to expanding the range of problems to which the FPSM can be applied more complex problems of the Computational Fluid Dynamics (CFD), the FPSM methodology is coupled with the IBM. The IBM is a methodology developed to simulate flows over complex, moving geometries, based on a cartesian-type grid. Normally, the IBM presents a low accuracy in regions near the immersed interface. The present work proposes to continue the developments of the hybrid FPSM-IBM methodology, focusing in the solution of the Navier-Stokes equations in its isothermal and incompressible form. The present work, as an evolution of the developments of Mariano (2011) , solves the Navier-Stokes equations in its three-dimensional form. The numerical code developed is capable of Large Eddy Simulations also. The method of manufactured solutions was used in the verification of the numerical code developed, and a spectral convergence rate is achieved. In the process of validation of the FPSM methodology, a isotropic turbulence flow was simulated and the results obtained are consistent from a physical point of view. Finally, the FPSM-IBM methodology was used in the simulation of turbulent jet in spatial development. The results obtained are compared with experimental data, and present a good agreement, both qualitative and quantitatively. It was observed that the FPSM-IBM implemented allows obtaining high accuracy, spectral convergence rates, computational efficiency, and allows the integrally Fourier-spectral solution of non-periodic problems.

The present work is devoted to the development and implementation of a computational framework to perform numerical simulations of low Mach number turbulent reactive ows. The numerical algorithm designed for solving the transport equations relies on a fully implicit predictor-corrector integration scheme. A physically consistent constraint is retained to ensure that the velocity eld is solved correctly, and the numerical solver is extensively veried using the Method of Manufactured Solutions (MMS) in both incompressible and variable-density situations. The nal computational model relies on a hybrid Large Eddy Simulation / transported Probability Density Function (LES-PDF) framework. Two dierent turbulence closures are implemented to represent the residual stresses: the classical and the dynamic Smagorinsky models. The specication of realistic turbulent in ow boundary conditions is also addressed in details, and three distinct methodologies are implemented. The crucial importance of this issue with respect to both inert and reactive high delity numerical simulations is unambiguously assessed. The in uence of residual sub-grid scale scalar uctuations on the ltered chemical reaction rate is taken into account within the Lagrangian PDF framework. The corresponding PDF model makes use of a Monte Carlo technique: Stochastic Dierential Equations (SDE) equivalent to the Fokker-Planck equations are solved for the progress variable of chemical reactions. With the objective of performing LES of turbulent reactive ows in complex geometries, the use of distributed computing is mandatory, and the retained domain decomposition algorithm displays very satisfactory levels of speed-up and eficiency. Finally, the capabilities of the resulting computational model are illustrated on two distinct experimental test cases: the rst is a two-dimensional highly turbulent premixed ame established between two streams of fresh reactants and hot burnt gases which is stabilized in a square cross section channel ow. The second is an unconned high velocity turbulent jet of premixed reactants stabilized by a large co- owing stream of burned products.

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior

Fundação de Amparo a Pesquisa do Estado de Minas Gerais

Conselho Nacional de Desenvolvimento Científico e Tecnológico

Conselho Nacional de Desenvolvimento Científico e Tecnológico

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior