Master's degree in Numerical Methods in Engineering (Pla 2012)http://hdl.handle.net/2099.1/75402021-07-31T00:49:54Z2021-07-31T00:49:54ZExploring Lagrangian MPM in application to CFDBoniquet Aparicio, Marcoshttp://hdl.handle.net/2117/3461842021-05-27T08:21:49Z2021-05-27T08:16:25ZExploring Lagrangian MPM in application to CFD
Boniquet Aparicio, Marcos
The Material Point Method (MPM) [Sulsky et al. 1995] [1] combines a La-grangian description of a domain using a set of particles with a computational mesh used to solve numerically the system of governing equations. At each time step, the governing equations are solved on the nodes of the computa-tional grid, while history dependent variables and material information are saved on the particles for the entire deformation process.
The Finite Element Method (FEM) has a counterpart which has beendeveloped in the last decades, the so called Meshless Methods (MMs) . Asa start, this project studies one of these methods, the Moving Least Squares(MLS) method, a previous step to combine it with the MPM, through Python programming language.
First, it is tested in depth the MLS method for a 2D domain and its exact approximation cases for different consistency orders with different functions.It is considered either a structured or a random particle distribution, analyz-ing the optimal weight functions to be used and different dilation parameters.
Next, MLS is to be applied as a solving solution for Partial Differen-tial Equation (PDE) systems for a 10x10m square domain submitted to self weight. Two cases are studied, depending on the constitutive model, one for Plane Stress and one for Stokes, being the latter a particular considera-tion of Navier-Stokes system of equations in which the convective terms areneglected in front of the viscous terms, thus for this Computational Fluid Dynamics (CFD) simulation high viscosity is implied for the momentum conservation equation.
Plane Stress case is compared to the FEM case, for lateral fixed and base fixed cases, while Stokes case is only computed through MLS, being considered steady and transient cases, both for based fixed. A stabilization method for first and also for second order derivative terms is applied through Galerkin Least Squares (GLS) method.
In addition, the software Comsol Multiphysics®is used as a tool to compare results obtained to an industry well-established FEM program andas a guide for the project
2021-05-27T08:16:25ZBoniquet Aparicio, MarcosThe Material Point Method (MPM) [Sulsky et al. 1995] [1] combines a La-grangian description of a domain using a set of particles with a computational mesh used to solve numerically the system of governing equations. At each time step, the governing equations are solved on the nodes of the computa-tional grid, while history dependent variables and material information are saved on the particles for the entire deformation process.
The Finite Element Method (FEM) has a counterpart which has beendeveloped in the last decades, the so called Meshless Methods (MMs) . Asa start, this project studies one of these methods, the Moving Least Squares(MLS) method, a previous step to combine it with the MPM, through Python programming language.
First, it is tested in depth the MLS method for a 2D domain and its exact approximation cases for different consistency orders with different functions.It is considered either a structured or a random particle distribution, analyz-ing the optimal weight functions to be used and different dilation parameters.
Next, MLS is to be applied as a solving solution for Partial Differen-tial Equation (PDE) systems for a 10x10m square domain submitted to self weight. Two cases are studied, depending on the constitutive model, one for Plane Stress and one for Stokes, being the latter a particular considera-tion of Navier-Stokes system of equations in which the convective terms areneglected in front of the viscous terms, thus for this Computational Fluid Dynamics (CFD) simulation high viscosity is implied for the momentum conservation equation.
Plane Stress case is compared to the FEM case, for lateral fixed and base fixed cases, while Stokes case is only computed through MLS, being considered steady and transient cases, both for based fixed. A stabilization method for first and also for second order derivative terms is applied through Galerkin Least Squares (GLS) method.
In addition, the software Comsol Multiphysics®is used as a tool to compare results obtained to an industry well-established FEM program andas a guide for the projectComparison and coupling of high-order continuous and hybridzable discontinunous galerkin methods for elliptic problemsDing, Zichenhttp://hdl.handle.net/2117/3434602021-04-09T10:50:47Z2021-04-09T10:49:06ZComparison and coupling of high-order continuous and hybridzable discontinunous galerkin methods for elliptic problems
Ding, Zichen
2021-04-09T10:49:06ZDing, ZichenMachine learning to build reduced order models of solid mechanic models with uncertaintyFelipe Ramudo, Valeria Agustinahttp://hdl.handle.net/2117/3431762021-04-06T14:50:42Z2021-04-06T14:46:34ZMachine learning to build reduced order models of solid mechanic models with uncertainty
Felipe Ramudo, Valeria Agustina
In civil engineering, mining and the petroleum industry in-situ stress state knowledge is fundamental. To estimate the stress state in-situ methods are applied, these methods are not sufficient to fully estimate the stress field in 3D due to the sparsity of observations and the paucity of stress measurements. In the current master’s thesis, a framework is developed through Reduced-Order Models (ROM) to predict the vertical displacement field, and consequently the stress state, when variations in the material and geometric properties of the problem are introduced. To achieve this, four different methods are tested in four different examples, one of them with uncertainty in the properties of the material and the last three with geometric uncertainties. The four methods tested are Artificial Neural Network with Levenberg Marquardt training algorithm (LM), Artificial Neural Network with Gradi-ent Descent with Momentum training algorithm (GDM), Proper Orthogonal Decom-position (POD) and Encapsulated Proper Generalized Decomposition (Encapsulated PGD) The developed framework consists of three parts. First, the data generator module creates a series of samples through Latin Hypercube Sampling varying the geometric or material properties of the problem and then obtains the vertical displacement field by the Finite Element Method (FEM) simulations. The second module applies any of the four methods to build a reduced-order model and finally, the third module is applied, if necessary, to reconstruct the vertical displacement field.
2021-04-06T14:46:34ZFelipe Ramudo, Valeria AgustinaIn civil engineering, mining and the petroleum industry in-situ stress state knowledge is fundamental. To estimate the stress state in-situ methods are applied, these methods are not sufficient to fully estimate the stress field in 3D due to the sparsity of observations and the paucity of stress measurements. In the current master’s thesis, a framework is developed through Reduced-Order Models (ROM) to predict the vertical displacement field, and consequently the stress state, when variations in the material and geometric properties of the problem are introduced. To achieve this, four different methods are tested in four different examples, one of them with uncertainty in the properties of the material and the last three with geometric uncertainties. The four methods tested are Artificial Neural Network with Levenberg Marquardt training algorithm (LM), Artificial Neural Network with Gradi-ent Descent with Momentum training algorithm (GDM), Proper Orthogonal Decom-position (POD) and Encapsulated Proper Generalized Decomposition (Encapsulated PGD) The developed framework consists of three parts. First, the data generator module creates a series of samples through Latin Hypercube Sampling varying the geometric or material properties of the problem and then obtains the vertical displacement field by the Finite Element Method (FEM) simulations. The second module applies any of the four methods to build a reduced-order model and finally, the third module is applied, if necessary, to reconstruct the vertical displacement field.Petrov-Galerkin Proper Generalized Decomposition strategies for convection-diffusion problemsPerelló i Ribas, Rafelhttp://hdl.handle.net/2117/3431602021-04-06T11:30:58Z2021-04-06T11:21:36ZPetrov-Galerkin Proper Generalized Decomposition strategies for convection-diffusion problems
Perelló i Ribas, Rafel
The present thesis explores the capabilities of a dual based Petrov-Galerkin Proper Generalised Decom-position (PGD) method to solve non self-adjoint parametric partial differential equations (PDE). The(consistent) Galerkin PGD method is used as reference. The Petrov-Galerkin PGD method is formulated asa Galerkin PGD making possible a non-intrusive implementation.Then, different problems are introduced and solved using the Petrov-Galerkin PGD methodology byseparating the domain in space and time. In particular, a transient advection-diffusion problem is solved usinga streamline upwind Petrov-Galerkin (SUPG) stabilisation. We also study a transient advection-diffusionproblem where the Dirichlet/Neumann splitting of the boundary conditions (BC) is time depending.The traditional fixed-point algorithm to solve the rank-one problem lacks of convergence for certainconditions. We introduce several alternative approaches to solve this inconvenient.Finally, we present a few examples of transient parametric PDE solved using the introduced methodology
2021-04-06T11:21:36ZPerelló i Ribas, RafelThe present thesis explores the capabilities of a dual based Petrov-Galerkin Proper Generalised Decom-position (PGD) method to solve non self-adjoint parametric partial differential equations (PDE). The(consistent) Galerkin PGD method is used as reference. The Petrov-Galerkin PGD method is formulated asa Galerkin PGD making possible a non-intrusive implementation.Then, different problems are introduced and solved using the Petrov-Galerkin PGD methodology byseparating the domain in space and time. In particular, a transient advection-diffusion problem is solved usinga streamline upwind Petrov-Galerkin (SUPG) stabilisation. We also study a transient advection-diffusionproblem where the Dirichlet/Neumann splitting of the boundary conditions (BC) is time depending.The traditional fixed-point algorithm to solve the rank-one problem lacks of convergence for certainconditions. We introduce several alternative approaches to solve this inconvenient.Finally, we present a few examples of transient parametric PDE solved using the introduced methodologyAerodynamics of Cronuz with Kratos multiphysicsCall Piñol, Oriolhttp://hdl.handle.net/2117/3423542021-04-13T11:31:59Z2021-03-24T13:12:39ZAerodynamics of Cronuz with Kratos multiphysics
Call Piñol, Oriol
Electric vehicles are becoming an increasing need since the traditional and pollutant
combustion engines are about to be extinguished. The future, in the present moment,
is the design of better, more efficient and less pollutant cars that allow the mobility
taking into consideration the global warming. This is why IDIADA presented the
Cronuz, a concept car that aims at the future being one of the most efficient
aerodynamic cars ever designed.
The goal of this Project is the aerodynamics study of Cronuz car developed by IDIADA.
The analysis is going to be performed with GiD for pre and post processing and Kratos
Multiphysics as solver and give conclusions to the work carried out
2021-03-24T13:12:39ZCall Piñol, OriolElectric vehicles are becoming an increasing need since the traditional and pollutant
combustion engines are about to be extinguished. The future, in the present moment,
is the design of better, more efficient and less pollutant cars that allow the mobility
taking into consideration the global warming. This is why IDIADA presented the
Cronuz, a concept car that aims at the future being one of the most efficient
aerodynamic cars ever designed.
The goal of this Project is the aerodynamics study of Cronuz car developed by IDIADA.
The analysis is going to be performed with GiD for pre and post processing and Kratos
Multiphysics as solver and give conclusions to the work carried outNumerical simulation of externally reinforced concrete wall with composite materialsIberico Leonardo, Juan Diegohttp://hdl.handle.net/2117/1844062021-04-13T11:32:01Z2020-04-23T07:36:44ZNumerical simulation of externally reinforced concrete wall with composite materials
Iberico Leonardo, Juan Diego
2020-04-23T07:36:44ZIberico Leonardo, Juan DiegoAproximació numèrica de problemes d'interacció fluid-estructura amb fluids viscoelàsticsCattoni Correa, Domingo Eugeniohttp://hdl.handle.net/2117/1798132021-04-13T11:32:04Z2020-03-12T13:55:32ZAproximació numèrica de problemes d'interacció fluid-estructura amb fluids viscoelàstics
Cattoni Correa, Domingo Eugenio
El problema d'interacció fluid-estructura és un tema clàssic en enginyeria, amb aplicacions tant diverses com l'aerodinàmica de ponts o el flux en conductes deformables. Un exemple d'aquest segon grup és el de flux de sang en artèries, que conté dos dels ingredients que es volen estudiar en aquest treball: un sòlid elàstic en grans deformacions de gruix petit que conté el flux d'un líquid que es comporta com un material viscoelàstic.
2020-03-12T13:55:32ZCattoni Correa, Domingo EugenioEl problema d'interacció fluid-estructura és un tema clàssic en enginyeria, amb aplicacions tant diverses com l'aerodinàmica de ponts o el flux en conductes deformables. Un exemple d'aquest segon grup és el de flux de sang en artèries, que conté dos dels ingredients que es volen estudiar en aquest treball: un sòlid elàstic en grans deformacions de gruix petit que conté el flux d'un líquid que es comporta com un material viscoelàstic.Advanced modelling of adhesive bonds in automotive components by the finite element methodAlameda González, Alfredohttp://hdl.handle.net/2117/1723812021-04-13T11:32:04Z2019-11-14T10:25:58ZAdvanced modelling of adhesive bonds in automotive components by the finite element method
Alameda González, Alfredo
In the design of new components for the automotive framework, simulation methodologies are applied in order to reduce experimental costs. The inclusion of new materials leads to the need of consider effects that were not needed for the design in classical materials, as cracks in High strength steels, or crushing damage in composite materials.
2019-11-14T10:25:58ZAlameda González, AlfredoIn the design of new components for the automotive framework, simulation methodologies are applied in order to reduce experimental costs. The inclusion of new materials leads to the need of consider effects that were not needed for the design in classical materials, as cracks in High strength steels, or crushing damage in composite materials.Numerical Analysis of Interior Aircraft Structures made of Composite MaterialsPan, Leihttp://hdl.handle.net/2117/1720182021-04-13T11:32:04Z2019-11-09T09:33:17ZNumerical Analysis of Interior Aircraft Structures made of Composite Materials
Pan, Lei
This master thesis focuses on the simulation of interior aircraft structures made of composite materials. These structures are made of a sandwich material consisting of two glass fiber skins around a honeycomb core. Different strategies developed to simulate composite materials will be used for their mechanical characterization. The composites that will be used in this project are bio-based. Developing this project the student will learn basic concepts of structural analysis using FEM and will familiarize with formulations to characterize composite materials.
2019-11-09T09:33:17ZPan, LeiThis master thesis focuses on the simulation of interior aircraft structures made of composite materials. These structures are made of a sandwich material consisting of two glass fiber skins around a honeycomb core. Different strategies developed to simulate composite materials will be used for their mechanical characterization. The composites that will be used in this project are bio-based. Developing this project the student will learn basic concepts of structural analysis using FEM and will familiarize with formulations to characterize composite materials.Numerical Simulation of AM processes by SLM technologyWiener Rocca, Reinaldohttp://hdl.handle.net/2117/1306842021-04-13T11:32:02Z2019-03-21T08:01:29ZNumerical Simulation of AM processes by SLM technology
Wiener Rocca, Reinaldo
The objective of this work is the Numerical Simulation of AM processes by SLM technology. The Shrinkake method as well as the Inherent strain method will be explored. The characterization of the material data as well as the sensitivity to process parameteres will be explored. The numerical results obtained will be compared with the experimental evidence.
2019-03-21T08:01:29ZWiener Rocca, ReinaldoThe objective of this work is the Numerical Simulation of AM processes by SLM technology. The Shrinkake method as well as the Inherent strain method will be explored. The characterization of the material data as well as the sensitivity to process parameteres will be explored. The numerical results obtained will be compared with the experimental evidence.