Pasar al contenido principal

MODELOS NUMERICOS AVANZADOS PARA INTERACCION FLUIDO ESTRUCTURA Y OPTIMIZACION DE FUTUROS DISEÑOS DE SISTEMAS DE TURBINAS EOLICAS OFFSHORE

PID2024-160181OB-I00

•
Nombre agencia financiadora Agencia Estatal de Investigación
Acrónimo agencia financiadora AEI
Programa Programa Estatal para la Investigación y el Desarrollo Experimental
Subprograma Subprograma Estatal de Generación de Conocimiento Científico-Técnico y Desarrollo Experimental
Convocatoria Proyectos de I+D+i (Generación de Conocimiento y Retos Investigación)
Año convocatoria 2024
Plan Estatal 2024-2027 Plan Estatal de Investigación Científica y Técnica y de Innovación
Centro beneficiario UNIVERSIDADE DA CORUÑA

Resultados relacionados

Resultados totales (Incluyendo duplicados): 3
Encontrada(s) 1 página(s)

RUC. Repositorio da Universidade da Coruña
Publicación Tesis doctoral (DoctoralThesis)

Estimation of Fatigue Damage of Offshore Wind Turbine Jackets Using Numerical Simulations and Artificial Neural Networks

  • Towaiq, Ahmad
[Abstract] This thesis studies fatigue damage in welded joints of jacket substructures for offshore wind turbines (OWTs) and proposes a practical way to estimate junction-level damage at low computational cost during early design. The motivation is straightforward: jackets are attractive in the 40–60 m water-depth range, but the large number of welded tubular joints makes fatigue a design driver. Conventional time-domain simulations followed by joint-level fatigue checks are accurate but expensive when many jacket alternatives must be screened. The goal here is to keep the physics and standard practice in view while reducing computational cost to make broad design exploration feasible. The work is structured in three parts. First, a finite-element (FE) model is built for a reference OWT with a jacket substructure. The structure is represented as a 3D frame of circular steel tubes with Bernoulli beam elements. Environmental loading is parameterized compactly with four quantities: wind direction, annual mean wind speed, a wind–wave factor linking wind to wave height, and a wind non-uniformity factor that controls how short-interval signals compound to one year. Wave forces are computed with Airy kinematics and Morison’s equation. Damping is handled by small modal ratios appropriate for fixed steel structures. Modal analysis is performed, and the system is solved using the Method of Modal Superposition. The model is checked against the OC4 jacket; the first modes and frequencies agree well, which gives confidence in the structural modeling. A fatigue computation workflow is implemented. Nominal FE stresses are converted to hot-spot stresses using Efthymiou’s parametric equations for stress concentration factors in tubular joints. The hot-spot stress time series are rainflow-counted to obtain stress-range cycles, which are then evaluated using an appropriate S–N curve. Damage is accumulated by Miner’s rule. The analyses are repeated for numerous structural scenarios involving different geometrical and loading combinations. The resulting damage distribution exhibits a strong skew toward very small damages. A supervised artificial neural network (ANN) is trained as a metamodel to approximate the physics-based pipeline at junction level. Each junction is encoded by 30 features that summarize (i) global jacket geometry, (ii) global loading parameters, and (iii) local joint information. The model is a fully connected multilayer perceptron with ReLU activations and a linear output; training uses Adam and mean-squared error with Early Stopping. Hyperparameter grids are run over depth, width, and batch size., This work has been partially supported by Grant #ED431C 2022/06 of the Consellería de Educación, Universidade e Formación Profesional of the Xunta de Galicia. This work also acknowledges the support provided by Grant PID2021-125447OB-I00 funded by MCIN/AEI/10.13039/501100011033 and by “ERDF A way of making Europe” and the funds by Grant TED2021-129805B-I00 funded by MCIN/AEI/10.13039/501100011033 and by the “European Union NextGenerationEU/PRTR”, and by Grant PID2024- 160181OB-I00 of MCIN/AEI., Xunta de Galicia; ED431C 2022/06
RUC. Repositorio da Universidade da Coruña
Publicación Artículo científico (JournalArticle)

Data-driven Riemann solvers: A neural network approach and a hybrid solver

  • Nogueira, Xesús
  • Ramos, Lucía
  • Seijo Conchado, Sonia
  • Couceiro, Iván
  • Khelladi, Sofiane
  • Ramírez, Luis
[Abstract] The accurate and efficient numerical solution of the Riemann problem is the basis of Godunov-type schemes. Approximate Riemann solvers are widely used for their efficiency, although they exhibit inaccuracies and instabilities in challenging regimes such as strong rarefactions or near-vacuum conditions. This work explores the use of deep neural networks (NNs) to address these limitations. We present two distinct data-driven frameworks: first, a NN-based solver trained to predict the exact solution of the Riemann problem, and second, a high-performance hybrid scheme. The hybrid approach uses the standard Harten–Lax–van Leer-contact (HLLC) Riemann solver as the main solver, enhanced with a computationally inexpensive, physics-based detector that identifies interfaces where the HLLC solution is likely to be inaccurate or to fail. At these interfaces, the scheme selectively uses the pretrained NN to ensure a more accurate solution. Through a series of benchmark tests, we show that the NN solver accurately reproduces the exact solution of the Riemann problem, but at a significant computational cost. In contrast, the proposed hybrid solver achieves a comparable level of accuracy to the NN solver, while it requires nearly the same computational cost as the standard HLLC solver., X. Nogueira and L. Ramírez acknowledge the support provided by the [Grant PID2021- 125447OB-I00] funded by MCIN/AEI/ 10.13039/501100011033 and by “ERDF A way of making Europe” and the funds by [Grant TED2021-129805B-I00] funded by MCIN/AEI/ 10.13039/501100011033 and by the “European Union NextGenerationEU/PRTR”. L. Ram´ırez and I. Couceiro acknowledge the funds by MCIU/AEI [Grant PID2024-160181OB-I00]. They also acknowledge the funding provided by the Xunta de Galicia (grant #ED431C 2022/06). L. Ramos acknowledges the support from the Xunta de Galicia through (grant #ED431C 2024/33)., Xunta de Galicia; ED431C 2022/06, Xunta de Galicia; ED431C 2024/33
RUC. Repositorio da Universidade da Coruña
Publicación Tesis doctoral (DoctoralThesis)

The Molecular Element Method (MEM): A FEM-based Formulation for Linear and Non-Linear Molecular Elasticity

  • Fernández-San Miguel, Andrés
[Abstract] Since its beginnings, humanity has always been characterized by building. From shelters to protect us from wind and rain, to pyramids erected to honor the pharaohs, to stone cathedrals and today’s slender cable-stayed bridges, we need knowledge about materials to make these structures possible. Until the end of the 20th century, the design and analysis of complex structures was particularly limited, but the advent of the Finite Element Method (FEM) has made it possible today to analyze them in different elasticity regimes, incorporating nonlinear effects and to design new and larger structures. On another level, the nanoscale has not been left without its advances and, despite the size, has claimed its relevance too. The discovery of materials with extraordinary and promising properties such as graphene and carbon nanotubes, or the devastating effects of the SARS-CoV-2 pandemic, has highlighted the impact that, even the smallest things, might have in our lives. However, unlike what happens at the macroscale with the FEM, only small systems of the order of hundreds of atoms can be analyzed at the nanoescale with techniques such as Density Functional Theory (DFT). This thesis proposes a new formulation coined the Molecular Element Method (MEM) that enables elastic analysis of molecular structures, from small molecules to proteins with thousands of atoms. The method relies on force fields developed in recent years by combining experimental and ab initio techniques, which have been exploited in Molecular Dynamics (MD). However, unlike MD, the proposed formulation is not restricted to dynamic analysis using explicit integration schemes. The proposed method allows to obtain linearized responses in first-order theory rigorously, to perform instability analyses such as buckling, and include nonlinear effects, exploiting typical FEM algorithms superior in order of convergence. The formulation permits working directly on geometries and potential energy at the atomic level, making it particularly versatile and competitive against laborious processes such as the generation of approximated continous models for specific problems. To perform first-order theory analysis, it has been proven that stiffness matrices only require the calculation of geometrical descriptor gradients. So, in an innovative way, the different gradients have been calculated exactly, treating possible singularities as in the case of aligned molecules or graphene sheets. The application of the formulation in this regime has made it possible to analyze different geometries and degrees of localized phenomena: from the precise determination of molecular vibration frequencies to the prediction of the mechanical properties of graphene and carbon nanotubes. The accurate determination of stiffness matrices has facilitated, unlike Molecular Dynamics where a correct selection of temporal discretization is essential to ensure simulation stability, the use of the Newmark method in its unconditionally stable version. As an example, the dispersion of mechanical waves in a nanotube was successfully analyzed with results equivalent to Molecular Dynamics. This work also analyzed the vibrations of protein systems, such as the spikes of SARS-CoV-2. The precise assembly of the stiffness matrices makes it feasible to determine the vibration frequency required to disable the proteins by resonance, and suggests the possibility of building an ultrasonic device for this purpose. Additionally, based on the results established in first-order theory, an extension to the nonlinear regime has been carried out. Consistently, the tangent stiffness matrices are determined exactly, which creates the possibility of using Newton algorithms or the arc length method. These algorithms are used to simulate elastic instabilities with experimental evidence such as single-walled and multi-walled nanotubes kink formation and the indentation of fullerenes. In turn, given that the extension makes it feasible to determine exactly the geometric stiffness matrices of nanostructures such as nanotubes, their critical loads have been calculated under different boundary conditions. This has enabled the calibration of a consistent continuous model using only geometry and force field parameterization as data, yielding mechanical properties that are in agreement with experimental results. Thefore, the Molecular Element Method allows to obtain, directly from the force fields the structural response, either static, dynamic or even non-linear of any molecular structure. In this sense, the method developed is a bridge between energy and structural response., This work has been partially supported by Grant #ED431C 2022/06 of the Consellería de Educación, Universidade e Formación Profesional of the Xunta de Galicia. This work also acknowledges the support provided by Grant PID2021-125447OB-I00 funded by MCIN/AEI/10.13039/501100011033 and by “ERDF A way of making Europe” and the funds by Grant TED2021-129805B-I00 funded by MCIN/AEI/10.13039/50110 0011033 and by the “European Union NextGenerationEU/PRTR”, and by Grant PID2024-160181OB-I00 of MCIN/AEI. We would also like to thank the funding provided by the “Programa de axudas para realizar estadías dirixido ao persoal investigador adscrito ao CITEEC” by the Centro de Innovación Tecnolóxica en Edificación e Enxeñaría Civil and by Grants PID2023-146793OB-I00 and PID2024-160181OB-I00 of MCIN/AEI., Xunta de Galicia; ED431C 2022/06