PhD defence of Clément Raimbault

19 November 2021

Clément Raimbault defends his PhD in Numerical Mechanics and Materials on Nov. 25, 2021

"Foaming of polyurethane injected on a porous material"

Clément Raimbault conducted his PhD under the supervision of Patrice Laure, Michel Vincent and Séverine Boyer. Je defends his PhD in Numerical Mechanics and Materials on Nov. 25th, 2021 in front of the following jury:

– Prof. Julien Bruchon, Ecole des Mines de Saint Etienne, SMS

– Prof. Boussad Abbes, UFR Sciences Exactes et Naturelles de Reims

– Prof. Christophe Binetruy, Ecole Centrale de Nantes

– Dr. Vincent Ladmiral, ICGM Montpellier

– M. François Choquart, Trèves PSI Reims

Abstract:

Polyurethane foams are sometimes combined with a porous layer to improve physical properties like sound and thermal insulation in the automotive industry. Such parts are manufactured by a reactive injection molding process, with a porous medium positioned beforehand on one of the cavity walls. A mixture composed of polyol, isocyanate and water is then injected in the closed cavity. The reaction between isocyanate and water creates carbon dioxide bubbles. Simultaneously, polymerization of polyurethane occurs resulting from the polyol and isocyanate reaction. The expanding foam fills the cavity and a part of the porous layer. The objective of this work is to develop a macroscopic model to simulate the influence of the porous medium on the mixture flow inside the cavity along with the evolution of temperature and pressure in the foam. Foam density, bubble size distribution and impregnation of the porous medium are also studied. First, foaming parameters are identified with the FOAMAT® device and experiments are performed to measure density, porosity and permeability values of a porous material. Then, foam injection tests are carried out on a stantard square mold geometry, with a porous layer on its lower wall. The cell size distribution is identified from scanning electron microscopy and micro-tomography image analyses. Numerical simulation is realized with REM3D® software. To model the foam flow in the porous medium, the Navier-Stokes-Brinkman model is coupled with the foaming model. After the implementation of this model, the upgraded REM3D® version is used to check the model by comparison with experimental data. Correlations are made between the size of bubbles, the density and temperature fields, the penetration depth of the foam in the porous medium and the permeability.

Keywords: injection, foam, polyurethane, porous, numerical simulation

PhD defence of Ramy Nemer

18 November 2021

Ramy Nemer defends his PhD on Nov. 23rd, 21

An adaptive immersed mesh method for fluid-structure interaction

Ramy Nemer conducted his PhD work in CFL team, under the supervision of Elie Hachem and Thierry Coupez in the framework of a project with Aqualung group. Ramy Nemer will defend his PhD in Computational Mathematics, High Performance Computing and Data, on November 23rd, 21 in front of the following jury:

– Prof. Thomas J.R. Hughes, The University of Texas at Austin, President of the jury

– Prof. Masud Arif, University of Illinois at Urbana-Champaign, reviewer

– Prof. Ramon Codina, Universitat Politècnica de Catalunya, reviewer

– Prof. Stefanie Elgeti, Vienna University of Technology

– Prof. Trund Kvamsdal, NTNU

– Dr Aurélien Larcher, Mines Paris -PSL

– Mr. Nicolas Peyron, Aqualung group, guest

– Mme Stéphanie Godier, Recherche et Avenir, guest

Abstract:

The need to simulate flexible, relatively thin structure is of growing interest with applications ranging from thin cylindrical sensors to membrane-like structures. These structures usually interact with their surroundings to accumulate data, or for a specific purpose. The inevitable interaction between the surrounding fluid and the solid is solved using a novel Fluid–Structure Interaction (FSI) coupling scheme. This thesis proposes a novel way to model the interaction between the fluid and solid. It consists of a hybrid method that combines both the traditional monolithic and partitioned approaches for Fluid–Structure Interaction (FSI). The solid mesh is immersed in a fluid-solid mesh at each time step, whilst having its own independent Lagrangian hyperelastic solver. The hyperelastic solver consists of a mixed formulation in both displacement and pressure, where the momentum equation of the continuum is complemented with a pressure equation that handles incompressibility inherently. It is obtained through the deviatoric and volumetric split of the stress that enables us to solve the problem in the incompressible limit. A linearization of the deviatoric part of the stress is implemented as well. The Eulerian mesh contains both the fluid and solid, and accommodates additional physical phenomena. Anisotropic mesh adaptation and the Level-Set methods are used for the interface coupling between the solid and fluid to better capture the interaction between them. All of the above components form the Adaptive Immersed Mesh Method (AIMM). The Variational Multi-Scale (VMS) method is used for both solvers to damp out any spurious oscillations that may arise for piece wise linear tetrahedral elements. The framework is constructed in 3D with parallel computing in mind. Extensive 2D and 3D test cases are presented that validate the hyperelastic Lagrangian solver, and the FSI AIMM framework. An application of the industrial partners was lastly tackled.

Keywords: Fluid–Structure Interaction, Finite elements, Variational Multi-Scale Methods, Anisotropic Mesh Adaptation, Hybrid method, Hyperelastic

PhD defence of David Xu

27 October 2021

David Xu defends his PhD in Numerical Mechanics and Materials on Nov. 4, 2021

Thermomechanical analysis of the Fused Filament Fabrication process: Experimental and numerical investigations

David Xu conducted his PhD work in CFL team, under the supervision of Franck Pigeonneau. David Xu defends his PhD in Numerical Mechanics and Materials on Nov. 4th, 2021 in front of the following jury:

– Dr BOYARD Nicolas (Lab. de Thermique et Energie de Nantes)

– Pr.U. FEREC Julien (Inst. de Rec. Dupuy de Lôme, Univ. Bretagne, Lorient)

– Dr BARRES Claire (Ingiénérie des Mat. Polymères, INSA Lyon, Villeurbanne)

– Pr.U. DUMONT Pierre (LaMCoS, INSA Lyon, Villeurbanne)

– Pr. SOULESTIN Jérémie (IMT Lille-Douai, Douai)

– Pr. TILLIER Yannick (MINES ParisTech-CEMEF, Sophia Antipolis)

– Dr ZHANG Yancheng (MINES ParisTech-CEMEF, Sophia Antipolis)

– Dr JOFFRE Thomas (CT-IPC, Bellignat)

Abstract:

This Ph.D. aims at the thermal and mechanical investigations of the additive manufacturing process of Fused Filament Fabrication (FFF). In this framework, the process is investigated at two different scales. Numerical models are developed to perform finite element simulations of the process. The numerical tools are supported by experimental observations and measurements. This is done to have a clear idea of the different phenomena occurring during the process as well as to assess the numerical models.

A first model focuses on the polymer fluid dynamics in the liquefier and its extrusion and deposition on a substrate. This model enables to study the influence of the nozzle geometry and printing parameters on the polymer deposition behaviour. An experimental protocol is implemented to measure the deposited strand cross section. An empirical equation is proposed to predict the strand dimensions based on the printing parameters.

A second model is developed to investigate the process on a larger scale. The first step aims to simulate the thermal behaviour of parts of several cm heights. Experimental measurements of the temperature via thermocouple are performed to assess the model. The thermal model is then adapted for the case of semi-crystalline polymers. The crystallisation kinetics is computed based on the cooling of the part. The results explain the degree of crystallinity measured experimentally. Solid mechanics is also implemented in the numerical model to account for part distortion upon cooling. An experimental protocol is developed to perform in-situ measurements of the strain during the printing of the part by digital image correlations. The combination of experimental and numerical investigations performed in this Ph.D. should provide the necessary information to better tailor the process depending on the needs of the user.

Numerical simulation of the deposition of multiple strands in the FFF process

Keywords: Fused filament fabrication, Thermomechanical analysis, Fluid dynamics, Numerical simulations, Polymers, Crystallisation

PhD defence of Juhi Sharma

14 October 2021

Juhi Sharma defends her PhD in Numerical Mechanics and Materials on October 22, 2021

Microstructural evolutions during hot forging of VDM Alloy 780: mechanisms, kinetics and mean field modelling

Juhi Sharma conducted her PhD work in MSR team, under the supervision of Nathalie Bozzolo et Charbel Moussa. She defends her PhD in Numerical Mechanics and Materials on Oct. 22nd, 2021 in front of the following jury:

Pr. RAE Catherine (Univ. of Cambridge – Materials Science and Metallurgy, Royaume-Uni)
Pr. KESTENS Leo (Ghent Univ. Technologiepark Zwijnaarde, Belgique)
Pr. BERNACKI Marc (MINES ParisTech-CEMEF, Sophia Antipolis)
I.R. GEHRMANN Bodo (VDM Metals international, Altena, Germany)

Abstract:

The demand to reduce the emission levels in aircraft engines has motivated the development of new high temperature alloys. VDM Alloy 780 is a new polycrystalline nickel-based superalloy, developed for turbine disc applications, with higher service temperatures up to 750°C. VDM Alloy 780 comprises of γ′ strengthening phase, in addition to the grain boundary plate-like precipitates which are identified as η/δ phase (mostly η phase but likely to include thin δ layers). A precise evaluation of the microstructural evolutions during the multistep industrial forging operations is crucial to the final properties of the alloy. Through a series of isothermal heat treatments, microstructural analyses, advanced EBSD techniques and electrical resistivity measurements, a detailed description of the precipitation behavior, associated precipitate shape and size evolutions as well as the grain growth kinetics were obtained. Hot compression tests were designed in accordance with the industrial forging conditions to determine the recrystallization mechanisms and kinetics in the supersolvus domain. The dynamic and post-dynamic recrystallization kinetics were established in function of the thermomechanical parameters such as strain, strain rate, temperature and post-deformation holding time. A mean field model was calibrated based on those experimental results and proved to be capable of correctly predicting the microstructural evolutions in the single-phase domain. In addition, the influence of the second phase particles on the recrystallization behavior in the subsolvus domain was investigated. This work provides guidelines to optimize the industrial forging conditions for this new superalloy to obtain a homogenous and fine-grained microstructure.

Keywords: Polycrystalline nickel-based superalloys, hot forging, recrystallization, grain growth, precipitation, mean field modelling

PhD defense of Gabriel Manzinali

16 July 2021

Gabriel Manzinali defends his PhD in Numerical Mechanics and Materials on July 16th, 2021

"Adaptive control for iterative solvers in an FE framework with mesh adaptation, for CFD simulations of industrial processes"

Gabriel Manzinali conducted his PhD work in CFL team, under the supervision of Elie Hachem, Youssef Mesri and Aurélien Larcher. Gabriel Manzinali defends his PhD in Numerical Mechanics and Materials on July &-th, 2021 in front of the following jury:

– Prof. Alvaro Coutinho

– Dr. Suzanne Michelle Shontz

– Erwan Liberge

Abstract:

The aim of this work is to propose a practical and general stopping criterion using an a posteriori approach, that relies on the error estimates available from the mesh adaptation procedure. This stopping criterion has to be robust and applicable to the different types of equations used to describe the complex physics involved in a conjugate heat transfer problem.

The final goal is to prove that with such stopping criterion is possible to drastically reduce the CPU time required for the solution of the linear system that stems from the Finite Element discretization.

Keywords:linear solver, Anisotropic adaptive mesh, error estimator, algebraic error

PhD defence of Jules Baton

4 June 2021

Jules Baton defends his PhD in Numerical Mechanics and Materials on June 4th, 2021

Dislocation structures in cold deformed pure tantalum: evolutions and influences on recovery and recrystallization

Jules Baton conducted his PhD work in MSR team, under the supervision of Nathalie Bozzolo et Charbel Moussa and in the framework of research project with CEA. Jules Baton defends his PhD in Numerical Mechanics and Materials on June 4th, 2021 in front of the following jury:

Brigitte BACROIX, Directeur de recherche CNRS, Université Paris 13

Javier SIGNORELLI, Professeur, Universidad Nacional de Rosario

Andrea TOMMASI, Directeur de recherche CNRS, Université de Montpellier

Vladimir ESIN, Chargé de recherche, MINES ParisTech – Université PSL

Charbel MOUSSA, Chargé de recherche, MINES ParisTech – Université PSL

Nathalie BOZZOLO, Professeur, MINES ParisTech – Université PSL

Wilfried GESLIN, Ingénieur, CEA Valduc

Abstract:

Thanks to its high ductility at room temperature, pure tantalum is ideal for the cold forming of parts with complex geometries with minimal rupture risks. This thesis aims at understanding and providing a description of the physical mechanisms taking place during a plastic deformation and then an annealing.

Various samples of pure tantalum were cold deformed by compression and rolling. The deformed microstructures were then characterized by scanning electron microscopy at grain and substructure scales. These characterizations revealed that the substructure development is strongly influenced by the crystallographic orientation of the grains and also by the texture of the initial state. In particular, γ-fiber grains form more substructures than θ-fiber grains. These differences cannot be correctly depicted only by dislocation density values. Parameters for the quantification of substructures have been proposed and their evolutions have been described by models.

Recrystallization is greatly impacted by the orientation dependence of the deformed state. Nucleation is promoted in γ-fiber grains due to the more advanced substructure development during deformation. This heterogeneity of behavior can be well described with stored energy estimated at substructure scale. Recovery has been studied directly and indirectly through its effects on recrystallization with various pre-recovery treatments. Two opposite effects on recrystallization were observed. A first non-contributive effect is related to the dislocation annihilation, which involves a decrease in the driving force for recrystallization. A second contributive effect is related to the enhancement of migration ability of subboundaries, which promotes nucleation. The balance between these two effects determines the overall effect of recovery on recrystallization. This balance varies depending on the deformation, the crystallographic orientation and the conditions of the pre-recovery treatment (time and temperature).

Keywords: Tantale, microstructure, dislocation structure, recovery, recrystallization