The institute of Mechanics, Materials and Civil Engineering (iMMC) organizes a series of seminars about several themes :
- Energy, thermodynamics, and chemical engineering
- Mechatronics and biomechanics
- Processing, materials, assembly, and structures
- Fluid mechanics, environmental engineering, and geomechanics
- Numerical and computational methods
Here follows the list of seminars.
Register for the newsletter
Seminars to come
|24/05/2013 (13h00) [Lieu : Reaumur D382]|
Hosni Idrissi (iMMC)
"Study of nanoscale plasticity mechanisms in crystalline materials using advanced ex-situ and in-situ transmission
|31/05/2013 (13h00 -> 15h00) [Lieu : Reaumur D382]|
Audrey Favache (iMMC)
"Introduction à matlab"
|17/05/2013(13h00) [Lieu : Reaumur D382]|
"Essais mécaniques sur résines et composites : tour d'horizon des capacités/spécialités UCL"
|17/05/2013(16:15) [Lieu : Stévin, b-145]|
Olivier Cartiaux (UCL/IREC/CARS)
"Toward next-generation ISO-based technologies for computer-assisted orthopaedic surgery"
|The project aims at formulating and validating the first ISO-based methodology for evaluating the quality of bone cutting and osseous assembly in orthopaedic surgery. Specifically, the project has the objective to develop new surgical quality models and new intraoperative measuring methods (and associated instrumentation) for orthopaedic interventions, by using the ISO1101 standard commonly used since the 1980s in mechanical engineering for specifying and assessing cutting and assembly processes. The expected results may contribute to improve the objective and quantitative assessment of the value-added of computer and robot assistance technologies for bone-cutting and osseous assembly. This is the first time a research project tends to implement the recent concept of quantitative surgery by specifically using the ISO standards and guidelines. This is highly in line with the activities of CARS (Computer Assisted and Robotic Surgery) research group of UCL-Brussels. The achievements of the project will be the basis for further research works within CARS including (1) the application of the new ISO-based specification tools to design the first intraoperative surgical assistance technologies complying with the ISO1101 standard, and (2) the development of a new preoperative ISO-based planning software for modeling, simulating and optimizing bone-cutting and osseous assembly processes in orthopaedic surgery.|
|17/05/2013(10:45) [Lieu : BARB10]|
Quentin Van Overmeere(UCL/IMMC/IMAP)
"Energy storage in ultra-thin solid oxide fuel cells"
|Energy storage in ultra-thin solid oxide fuel cells"
Solid oxide fuel cells (SOFCs) are promising clean and highly efficient
power sources. Compared to other types of fuel cells, SOFCs do not
require precious metal catalysts, are relatively insensitive to
impurities in the gas streams and can generate electricity from almost
any fuel. This is due to their high operating temperature, typically 800
to 1000°C. These high temperatures however place severe constraints on
the design of SOFCs and has so far limited their use to stationary
applications such as auxiliary power sources in power plants. By
combining microfabrication and thin film deposition techniques it is now
possible to obtain SOFCs that can achieve practical power densities at
temperatures as low as 500°C. Lower operating temperatures allow using
materials that are unstable at high temperature. A multifunctional
vanadium oxide anode for thin film SOFCs will be presented capable of
both energy generation and storage.
|03/05/2013(13h00) [Lieu : Reaumur D382]|
"Characterisation of 'thin' copper
films performed by a new PVD process"
|26/04/2013(13h00) [Lieu : Reaumur D126]|
"Matériaux architecturés élaborés par friction malaxage"
|19/04/2013(13h00) [Lieu : Reaumur D382]|
Quentin de Radiguès(iMMC)
"Electrochemical reactors with porous
electrodes: study existing technology and design of new technology"
|19/04/2013(16:15) [Lieu : Stévin, b-145]|
Virginia Ruiz Garate(UCL/iMMC)
"Inverse dynamics problem of human gait - Investigation for robotics application"
|Conventional methods for inverse dynamic analysis of human gait often use recursive Newton-Euler algorithm to find the joint torques out from kinematic data. Anyway, this is performed from the ground to the torso link using force plate or insoles direct measurements as an input, so using some dynamic data also. Although this method gives good results, is of course highly limited by the need to use force plates or insoles, which are not always available. Therefore, a study for determining such torques from purely kinematic data is performed. This algorithm seeks simplicity, is easy to use and provides the basis for further developments. A more complex version will lead to less error, but at the cost of increasing complexity.
Once the forces and torques are computed, a comparison will be made between these results and direct measured ground reaction forces and torques. These errors may vary for different planes and locations, different sources or errors will be commented.
After, based on these errors, different simplifications of the total configuration and their influence in the errors will be presented. This gives an overview of the most important movements necessary for biped walking.
Finally, a similar study based on insole information and the use of Robotran modules for the Cyberlegs project will be presented so as to give another approach more focus on the computation of joint torques.|
|04/04/2013(15h00) [Lieu : Stevin B-145]|
Benoit Rosa(Institut des Systèmes Intelligents et de Robotique - Université Pierre et Marie Curie - CNRS, Paris)
"Robotisation de la réalisation de mosaïques dimages endomicroscopiques"
|L'endomicroscopie confocale laser fibrée est une modalité d'imagerie qui permet d'obtenir des images d'un tissu en temps-réel, avec une résolution micrométrique, lorsque l'on place une sonde d'imagerie au contact de celui-ci. En chirurgie digestive, cette technologie pourrait être utilisée pour remplacer les biopsies extemporanées par des biopsies optiques, sans prélèvement de tissu. Ceci permettrait de réduire le temps opératoire et l'invasivité de l'opération.
Si la résolution des images endomicroscopiques est très bonne, leur champ de vision est trop étroit pour permettre au clinicien de poser un diagnostic. Une solution possible est de déplacer la sonde sur la surface du tissu afin de collecter une suite d'images que l'on peut ensuite recaler grâce à un algorithme de mosaicing afin de reconstruire une image grand champ.
On s'intéresse dans ce travail à la conception et au développement d'instruments minimalement invasifs robotisés permettant de réaliser des biopsies optiques dans l'abdomen. Deux prototypes d'instruments ont été développés. Ceux-ci intègrent un système de compensation des mouvements physiologiques permettant de se stabiliser sur la zone d'intérêt, ainsi qu'un micro-actionnement à l'extrémité distale. Par ailleurs, un algorithme d'asservissement visuel permet de contrôler les mouvements de la sonde par rapport au tissu avec une bonne précision malgré les déformations des tissus mous au contact de la sonde. Des essais ex vivo et in vivo ont été menés avec succès afin de tester la précision et la robustesse des systèmes et algorithmes de contrôle proposés.|
|29/03/2013(16:15) [Lieu : Stévin, b-145]|
"Large-dimension cable-driven robots: static equilibrium, simulation and design"
|Parallel cable-driven robots mainly consists of a mobile platform connected in parallel to a base by cables. Each cable is wound and unwound by actuated winches which are generally fixed to the base. The winches control the position and orientation of the platform by modifying the length of their respective cable. Since the overall mass and inertia of the moving parts is reduced, cable robots exhibit interesting properties compared to rigid-link robots. According to their size and practical realization, they may be less expensive and easier to build, transport and reconfigure.
Another useful property of cable robots is their scalability. Indeed, cables with small to very large lengths are easily stored on drums permitting the realization of parallel cable-driven robots with a workspace of global dimension ranging from a few ten or so centimetres to hundred of meters. In this research project, we are interested in very large cable robots (e.g. 20m), possibly carrying heavy to very heavy loads (e.g. 100 kg). In practice, such manipulators may be used in aircraft or shipbuilding industries to improve safety and cut the costs of some operations such as welding or painting over huge distances.
This seminar will be divided into three parts. In the first part, the problem of finding the static equilibrium of a body suspended by cables will be investigated using global optimization techniques. In the second part, a multibody approach to simulate the effect of cable mass will presented. Finally, the third part will develop a two-phase strategy to design the best geometry of such cable-robots for pick-and-place applications.|
|08/03/2013(16:15) [Lieu : Stévin, b.044]|
"Robotic Locomotor Assistance"
|Robotic technology to replace or assist human legs has made major advances in recent years. However, clinical experience still reveals limitations in terms of functionality and therapeutic outcome. One reason is that current designs often limit user autonomy, due to kinematic constraints and undesired interaction forces between the user and the robotic device.
This talk will present hardware and control concepts for user-centered assistance of bipedal gait, ranging from minimalistic kinematics over the targeted use of compliance to collaborative balance control. Embodiments and experimental results will be shown for all concepts, for example a robot for overground gait training in rats, which enabled groundbreaking research on recovery after spinal cord injury, and a recent extension of the robotic principle to human scale. |
|22/02/2013(16:15) [Lieu : Stévin, b-145]|
"Modélisation détaillée et globale des convertisseurs électroniques de puissance fonctionnant en modulation de largeur d'impulsion "
|On montre en premier lieu comment construire un modèle dynamique détaillé des convertisseurs électroniques de puissance fonctionnant en MLI puis comment passer à un modèle global de ces convertisseurs.
On montre ensuite que la simulation de ces convertisseurs se fait aisément à partir des blocs de base de la librairie matlab-simulink en traitant le cas d'un hacheur et d'un onduleur monophasé.|
|21/02/2013(14:00 -> 18h15) [Lieu : BARB92 + Hall Vinci]|
à l'intention des étudiants en master à l'EPL Organisée par les chercheurs de l iMMC(institute of Mechanics,Materials and Civil Engineering)
"Journée d'information sur la recherche en mécanique, matériaux et génie civil""
14h-14h30 Auditoire Barb 92:
Information sur la carrière de chercheur.
14h45-16h00 Auditoire Barb 92:
10 mini présentations de sujets de recherche
abordés dans les cinq pôles de l'iMMC.
16h15-18h15 Hall Vinci:
Open space (rencontre avec les professeurs,
chercheurs, assistants et mémorants) + Drink
|08/02/2013(16:15) [Lieu : Stévin, b-145]|
"Piezoelectric materials: modelling and applications in actuated systems"
|Piezoelectric materials present interesting properties for mechatronic systems. They offer a direct electromechanical conversion, have a good frequency response and a high force density. Those characteristics make of them good candidates for integrated sensors and actuators very close from the driven system. However, the deformations are not sufficient for most applications. So an amplification system is often required. In this seminar, applications of piezoelectricity in various field will be presented. A review of piezoelectric actuators will be showed and one of the most common, the bender, will be presented more deeply. A modified beam model will be provided as well as the design of the ring shape recurve actuator. |
|25/01/2013(14h00) [Lieu : Stévin, b-145]|
Bram Vanderborght (VUB)
"Mechatronic design of robots that are used in close collaboration with humans"
|The long-term goal of robotics is that robots can intelligently and efficiently operate in dynamic and unstructured environments, including collaboration with humans. Using classical actuators (as servomotors) and control design (e.g. trajectory tracking), this will be a very daunting task and provides the motivation to start developing variable stiffness actuators. These soft actuators are used in the Robotics & Multibody Mechanics (R&MM) research group of the VUB as core actuation technology for robots for both physical human-robot interaction (pHRI) and cognitive HRI (cHRI). The presentation will cover the design of those actuators (pneumatic muscles, maccepa and series-parallel elastic actuation), the control of them for embodied intelligence principles as safety, energy efficiency and adaptability and the implementation in different applications. Most of the prototypes are designed and built in the lab. Their most well-known robots are the bipeds Lucy and Veronica, the step rehabilitation robots Knexo and ALTACRO and the robotic ankle-foot prosthesis AMPfoot, AMPfoot 2.0 and Hekta. The social robot Probo and the electronic device Probogotchi have been designed and deployed for cHRI and developing Robot Assisted Therapy (RAT) with children with special needs as target group.|
|21/01/2013(14:00) [Lieu : Stévin, b-145]|
Alexandra Zobova (Lomonosov Moscow State University)
"The dynamics of an omni-mobile vehicle."
|The motion of an omni-mobile vehicle on a horizontal plane is considered. Omni-wheels ( http://en.wikipedia.org/wiki/Omni_wheel ) were developed to construct manuvrable means of transport (such as, for example, mobile robots and wheelchairs). Several rollers are fixed onto the periphery of the disc of such a wheel, so that only one of the rollers is able to contact the supporting surface and this roller can freely rotate about a certain fixed axis in the disc of the wheel. The axis of rotation of the roller is directed either along the tangent of the periphery of the disc (an omni-wheel) or it is turned around the radius of the disc at an angle of 45° (a mecanum-wheel). This construction enables a wheel, supported on a roller and maintaining the orientation of its plane, to move easily along a straight line at a fixed non-zero angle to the plane of the wheel. The wheels are modelled by absolutely rigid discs. Slippage in a certain direction, which makes a constant non- zero angle with the plane of a disc, is possible at the contact point of a wheel and the plane, and the planes of the discs are fixed with respect to the platform of the vehicle. The dynamic equations of motion are obtained for vehicles of this type with an arbitrary number and arrangement of the wheels. A complete qualitative description of the inertial motion of a vehicle is given (there are no control actions and it is assumed that there is no friction in the axes). The result is presented in the form of a phase portrait of the system. The motion of a vehicle is then considered in the case when control moments are applied to the axes of the wheels. The stability and branching of a certain class of steady motions of the vehicle are investigated. The domain of parameters is separated out where Andronov-Hopf bifurcation occurs with the formation of unstable limit cycles. Also an optimal (in some sense) movement of such vehicle is discussed.|
|18/01/2013(10h45) [Lieu : Barb11]|
Graham L. W. Cross(Trinity College, Dublin)
"Polymer deformation during nanostructure fabrication by mechanical processing: size effects and novel flow mechanisms"
|Mechanics-based fabrication methods now provide a
comprehensive means to realize scalable nanoscale
device fabrication. Functional elements of the
mechanical assembly line required for the
massively parallel production of nanoscale
structures and devices, including mastering,
replication and transfer, have been demonstrated
and in some instances even commercialized. As
the efficiency and reliability of these methods
is improved, an increasing penetration to more
traditional fields of research is enabled. My
group seeks to make a connection between
fundamental materials physics and these emerging nanofabrication methods.
In this talk, I will discuss the mechanical
generation of shape and organization of structure
at the nanoscale, with an emphasis on the
deformation of thin film polymers in quasi-two
dimensional flow fields that arise during thermal
imprint. Molecular scale squeeze flow in this
process presents significant challenges to
understanding and controlling the mass transport
necessary for high fidelity replication of
patterned dies1. We discuss recent developments
in the use of a modified nanoindentation
technique2 to measure glassy forging and viscous
melt moulding flows in ultrathin polymer films,
on 10, 100, and 1000 nm length scales. A
surprising scaling of stress vs. strain
relationships as system size is reduced to
dimensions below the statistical size of the
polymer molecule is revealed3 and its connection
to chain network topology is discussed. In
addition, a brief review of a novel mass
transport mechanism4 for planar geometries
arising from the deliberate injection of small
amplitude oscillatory shear stress during forming
will be given. Connection and contrast to the
growing body of confined-molecule polymer physics
measurements and concepts are made.|
|11/01/2013(15h00) [Lieu : Stevin B-145]|
"Use of video to promote freshmen engineering students' physics
conceptual understanding, the case of APP0"
|21/12/2012(10h45) [Lieu : BARB00]|
Quentin Van Overmeere(Harvard University)
storage in ultra-thin solid oxide fuel cells"
|Solid oxide fuel cells (SOFCs) are promising
clean and highly efficient power sources.
Compared to other types of fuel cells, SOFCs do
not require precious metal catalysts, are
relatively insensitive to impurities in the gas
streams and can generate electricity from almost
any fuel. This is due to their high operating
temperature, typically 800 to 1000°C. These high
temperatures however place severe constraints on
the design of SOFCs and has so far limited their
use to stationary applications such as auxiliary
power sources in power plants. By combining
microfabrication and thin film deposition
techniques it is now possible to obtain SOFCs
that can achieve practical power densities at
temperatures as low as 500°C. Lower operating
temperatures allow using materials that are
unstable at high temperature. A multifunctional
vanadium oxide anode for thin film SOFCs will be
presented capable of both energy generation and storage.|
|07/12/2012(13h00) [Lieu : Reaumur D382]|
"Multifunctional hybrid for wide band electromagnetic absorption:
Sandwich panel and Metamaterial"
|05/12/2012(13h30) [Lieu : Euler A.002]|
Isaac Smith(Jackson School of Geosciences, University of Texas at Austin)
"Martian spiral troughs as aeolian cyclic steps: a framework for understanding
ice-atmosphere interactions at the north pole of Mars"
|It has long been recognized that the layers of ice and dust contained
within the North Polar Layered Deposits (NPLD) are fundamental to our
understanding of atmospheric deposition and thus climatic processes on
Mars. Until recently, the focus had been on outcrops of exposed layers
within the spiral troughs that dominate the ice cap surface. However,
modification of the troughs by wind affects depositing layers, leaving an
outcrop record that must be interpreted hand in hand with trough
morphology rather than independently.
The spiral troughs found NPLD of Mars are intriguing features that have
been the source of much literature and debate since their discovery
during the Mariner 9 mission. In order to fully explain the troughs'
features, a hypothetical model must explain many aspects. Here we
interpret the spiral troughs as belonging to a class of features called
cyclic steps, which have well understood characteristics. In this
context, we provide stratigraphical evidence from ground penetrating
radar for atmospheric transport of ice and for active surface evolution
from visual observations. We demonstrate that katabatic winds and
katabatic jumps in the flow, are sufficient to explain all geologic
aspects of troughs.
Cyclic steps require specific conditions to form and persist. The
governing principles are well understood, so the existence of cyclic
steps within the polar ice on Mars provides unique, scientifically useful
insights into the development of the ice cap and into Mars' recent
climate. We use this new interpretation to better characterize spiral
troughs and atmospheric conditions in Mars' recent past.
|16/11/2012(14:30) [Lieu : Stévin, b-145]|
Gabriel Abedrabbo (UCL/iMMC)
"Dynamic Analysis of Intervertebral Efforts in Scoliosis: a Multibody Modeling Approach"
|In idiopathic scoliotic patients, dynamical intervertebral efforts acting between vertebrae seem to be correlated with the spinal deformity. The quantification of these efforts, which is useful for the orthopedic surgeons to set up their surgical planning is the final objective of the present research. As a first step, we focus in this contribution on the geometrical reconstruction of the spine and especially on the rotation sequences between vertebrae in a multibody approach. The reconstruction process is performed in the standing position with possible bending, using an optimization process based on geometrical data reconstructed from radiographs. The obtained results will serve as input for the subsequent gait motion for which the limited set of geometrical information must be compensated.|
|02/11/2012(16:15) [Lieu : Stévin, b-145]|
Emmanuel De Jaeger(UCL/iMMC)
"Review of the technical challenges of the future electrical power systems and related R&D topics"
|The electrical power systems are presently facing big challenges in the context of the ambitious energy targets fixed by the EU for the coming 20 to 40 years. The main driver of this policy is the increase of electricity generation from renewable sources, expected to increase and reach a level close to 90% (and even more) - in some circumstances - in 2050. To cope with this evolution, complex technical problems will have to be tackled in order to guarantee the reliability, stability and quality of electricity supply. They will need significant R&D efforts targeting the generation units themselves (of all types, sizes and locations) as well as the power grids structure, architecture (including new components), planning and operation practices. Even the loads, being more and more active parts of the system, will be addressed. The whole value chain of electricity is thus concerned, from generation to end-use, including transmission, distribution and storage. The subject of this seminar is to briefly highlight the context, review the main technical issues facing the power systems of the coming 20 to 40 years and comment some already obtained results and foreseen possible solutions. The planned R&D activities at UCL for the coming years will be discussed as well.|
|30/10/2012(14h30) [Lieu : TFL seminar room - Stevin]|
Hernan Morales(Computational Imaging and Simulation Technologies in Biomedicine
(CISTIB), Barcelona, Spain)
"Modeling hemodynamics in treated image-based
|Around 2% of human population harbors cerebral aneurysms. A spontaneous
rupture of an aneurysm can lead to a subarachnoid hemorrhage, which is
lethal in around 50% of the cases. Endovascular techniques, such as coiling
or stenting, are a common option to treat cerebral aneurysms. They aims to
isolate the aneurysm from blood flow circulation by promoting fluid
cessation and triggering the coagulation cascade. Thereby, hemodynamics
plays a key role in the success of endovascular therapies. However, the
influence of the devices on local hemodynamics is not fully understood.
Computational tools can be use to model postoperative hemodynamics in order
to predict the treatment outcome. Additionally, they can be used to test
several endovascular alternatives, thus providing useful information for
clinicians before the real intervention takes place.|
|19/10/2012(16:15) [Lieu : Stévin, b-145]|
"A brief review on Demand-Response in Electricity"
|Power systems are evolving. Firstly, liberalization era has divided responsibilities. Secondly, increased interest in variable and decentralized generation creates every day new challenges. In this framework, large efforts are being spent on finding solutions to perceived future challenges in what regards power system stability, control and adequacy. Demand response (DR) is potentially one of the multiple solutions. DR is usually defined as a change in consumption pattern, in response to economic trigger (high energy price) or for system security. This review tries to capture what drives DR development, and to which extent it could actually deliver its promises, both at system and end-users advantage.|
|11/10/2012(13h00) [Lieu : Reaumur D382]|
"Wet coating of Geldart-C particles in a rotating fluidized bed in
a static geometry"
|05/10/2012(16:15) [Lieu : Stévin, b-145]|
"Variable Stiffness Actuator Applied to an Active Ankle Prosthesis: Principle, Energy-Efficiency, and Control"
|Series elastic actuators are very popular in rehabilitation robotics. Among other advantages, elastic elements between the actuator and the load permit to store and release energy during the task completion, such that the energy balance is improved and the motor power peak is decreased. In rhythmic tasks like walking, this reduces to design the spring stiffness such that it works at resonance. To comply with different gaits and cadences, it is therefore necessary to design Variable Stiffness Actuators (VSA). This paper proposes three contributions: (i) we apply a particular concept of VSA to an active ankle prosthesis; (ii) we discuss the relevance of using VSA to change the stiffness also within the gait cycle; and (iii) we elaborate some control strategies for this device. Our guideline is to track a mechanical design and a controller maximizing energy efficiency. We establish that a promising approach is simply to control the amount of energy stored in the elastic element.|
|21/09/2012(14:00) [Lieu : Stévin, b.044]|
George F. Wittenberg(Univ. of Maryland, USA)
"Engineering Brain Changes after Stroke Neural Basis of Functional Recovery with Repetitive Task Practice"
|Many neurological diseases result in a severe inability to reach for which there is no proven therapy. Promising new interventions to address reaching rehabilitation using robotic training devices are currently under investigation in clinical trials but the neural mechanisms that underlie these interventions are not understood. Transcranial magnetic stimulation (TMS) may be used to probe such mechanisms quickly and non-invasively, by mapping muscle and movement representations in the primary motor cortex (M1). We have investigated movement maps in healthy young subjects at rest using TMS in the robotic environment, with the goal of determining the range of TMS accessible movements, as a starting point for the study of cortical plasticity in combination with robotic therapy. Maps were created by applying 10 stimuli at each of nine locations (3x3 cm2 grid) centered on the M1 movement hotspot for normal subjects. TMS-evoked movement kinematics were measured by the robotic encoders and ranged in magnitude up to 3 cm. Maps varied by subject and by location within a subject and were affected by practice of movements in the opposite direction from the mean evoked movement. This practice consisted of about 400 movements against a spring field. We also investigated the effects of TMS over M1 on movement preparation and are starting to study the effects of stimulation strength and arm position on movement and plasticity. Thus the combination of TMS and rehabilitation robotics provides a real-time, non-invasive platform for neurophysiology based evaluation and therapy in motor rehabilitation settings.|
|26/06/2012(11h) [Lieu : Salles des Séminaires, Bâtiment Stevin, B0.044]|
Stefano Gavioli(Politecnico di Torino, Italy)
Dynamic Simulation of an ejector-based cooling system for residential
|A smart and sustainable way to produce cooling power during
the hot season is through thermal activated compression.
One of the most promising technology is the ejector cycle. An experimental
machine based on this technology has been designed
and built at Universit´e Catholique de Louvain (UCL). This study
presents a dynamic simulation model of a full house incorporating
solar collectors to produce domestic hot water and to feed an
ejector-based solar air-conditioning machine. The model takes
into account geographic location, transient thermal house behaviour,
and also integrates a performance map of the ejector
cycle based on real measurement campaign. Different types of
collectors and storage sizes have been studied as well. In this
paper, the developed model is used to analyze and optimize operation
and strategy of hot water production and air conditioning.
Results show that the full solar solution may cover low energy
|15/06/2012(13h00) [Lieu : Reaumur D382]|
"Elasto-viscoplasticity and fracture behaviour
of RTM6 epoxy resin"
|15/06/2012(14h00) [Lieu : Stevin b-145]|
"Thermoelectric generation: an effective way for heat energy harvesting?"
|Energy Harvesting is one of the main concerns of our modern society. In
heavy industry, a huge amount of energy is wasted as heat and could be
harvested by several ways. One of these promising ways is the
thermoelectricity that has the advantage of making direct heat to
electricity conversion without any moving part. This involves a really
good reliability. Despite the low conversion efficiency, thermoelectric
generators ("TEG's") can be a good way to harvest waste heat. Indeed, it
has been shown that when the heat source is free it is more interesting
to improve the power density than the conversion efficiency. In the
first part of this talk, the basics of thermoelectricity and the
applications of classic TEG will be presented. In the second part, the
major challenges to use thermoelectricity in large-scale energy wasting
application will be discussed and one way to solve them will be presented.|
|14/06/2012(16h00) [Lieu : BARB94]|
"Numerical investigation of the tip leakage flow
in different high pressure turbine blade configurations"
|In high pressure gas turbine stages, a source of significant losses is the
leakage flow in the gap between the tip of the unshrouded rotor and the
casing, called the tip clearance. In addition to that, the inlet temperature -
which in modern high pressure turbines exceeds the melting point of the
blade material can cause serious damage, especially on the tip of the
rotor blade, where cooling is rather difficult.
An adequately designed tip geometry is supposed to reduce the mass
flow through the tip gap, and thus the losses and the heat transfer in the
discussed region. The presented numerical study concerns a thorough
aero-thermal investigation of tip leakage flows, applying the finite volume
method on complete three dimensional models of a rotor blade, with four
distinct tip configurations and film cooling. Three out of the four
investigated geometries represent different tip design solutions; the fourth
one portrays a typical eroded turbine blade tip geometry, after many
hours of service.
The crucial point of the work was the grid generation, which is one of the
most time-consuming phases of such projects. For this purpose, a scriptbased
semi-automatized method was developed, with which the meshing
session was significantly speeded up. The study also provides a
comparison between the different geometries from the aerothermodynamic
point of view, including the analysis of the losses and of
the cooling effectiveness.|
|12/06/2012(16h15) [Lieu : BARB92]|
"Multimesh Iterative Schemes and Model Reduction for
Linear and Nonlinear Advection-Diffusion Equations"
|The spatial discretization of conservation laws typically yields
large-scale dynamical systems. As a result, simulating, analyzing
or controlling such models requires a high computational cost. The
goal of Model Order Reduction (MOR) consists in building a
system of much lower complexity that approximates the dynamic of
the problem accurately.
The reduced model can be interpreted as a Petrov-Galerkin
projection of the unsteady partial differential equations with
nonlocal shape and test functions. For linear time-invariant
dynamical systems, we analyze the efficiency of an iterative
method that minimizes a measure of the approximation error.
Using several discretizations on coarser meshes of the same
problem, we propose a robust choice of initial condition for such a
scheme. Furthermore, we describe extensions of this MOR
scheme for linear time-varying and nonlinear models.
Sparse large-scale linear systems of equations must be solved at
each iteration of this algorithm. Unlike direct solvers, the
computational complexity of iterative solvers can be linear with an
appropriate preconditioner. For purely elliptic problems, the
convergence of the Conjugate Gradient method with a multigrid
preconditioner does not depend on the mesh size. The extensions
to the Stokes and Navier--Stokes equations are investigated
|08/06/2012(13h00) [Lieu : Reaumur D382]|
Camille Van der Rest(iMMC)
"Manufacture and characterisation of a real
|01/06/2012(13h00) [Lieu : Reaumur D382]|
"Combined in-situ stress and thickness monitoring during electrochemical
oxidation of silicon"
|01/06/2012(14h00) [Lieu : Stevin b-145]|
"Multibody Model of a Grand Piano Action Aimed at Understanding and
Demystifying the Escapement Principle"
|The present project has been achieved in close collaboration with the
MIM, the Musical Instruments Museum of Brussels. Thanks to our multibody
modeling and analysis capabilities, we have carried out a virtual
demonstrator of a grand piano action in order to understand, explain and
demystify the functioning of the "from key-to-string" transmission, and
especially the single and double escapement principle.
The developed demonstrator takes advantage of multibody dynamics as well
as of an experimental set up to analyze this highly dynamic system. Our
work is based on previous mathematical models available in the
literature but also on the large expertise of our collaborators at the
MIM, for piano action tuning and for the historical aspects relating to
In this talk, we highlight the main issues and outcomes of the project
-- namely the multibody modeling of a grand piano action together with
the experimental validation -- and some key results related to the
tuning of the aftertouch and the double escapement principle.|
|25/05/2012(13h00) [Lieu : Reaumur D382]|
gestes pour un monde plus vert"
|25/05/2012(16h00) [Lieu : Barb94]|
"Development of a Baroclinic Discontinuous Galerkin Finite Element Model for Estuarine and Coastal Flows"
|Numerical marine models have become indispensable in ocean sciences. Despite many developments over the past decades, modelling of the coastal ocean is still an area of active research. In recent years, unstructured mesh models have gained attention, as they can better represent the complex topography of coastal domains compared to standard models. The inherent flexibility of unstructured meshes is particularly useful in multi-scale applications, where a wide spectrum of spatial scales must be captured.
This thesis deals with the development of discontinuous Galerkin finite element shallow water models, with focus on shelf sea-estuary-river network systems. Two key issues in coastal marine modelling are addressed: wetting-drying and threedimensional modelling of buoyancy driven flows.
Representing the periodic exposure and submerging of tidal flats is a complicated task for Eulerian models. Most wetting-drying techniques are model specific, and require explicit time integration, which can significantly increase the computational cost. A generic wetting-drying method for depth-averaged shallow water equations is presented, which is compatible with implicit time marching, thus improving the computational efficiency.
The latter part of the thesis is devoted for the development of a three-dimensional baroclinic model. A discontinuous Galerkin finite element discretisation is presented, combined with an explicit time integration method. Slope limiters are used to ensure stability in strongly baroclinic flows. To account for vertical mixing, the model is coupled to an established turbulence closure model library. The model is validated with standard benchmarks and a Rhine river plume simulation in an idealised geometry.|
|23/05/2012(14h00) [Lieu : Barb 00]|
Prof. A. A. Benzerga(Texas A&M University)
"High-Temperature Discrete Dislocation Dynamics"
|At low homologous temperatures, the plastic deformation
is controlled by the glide of dislocations and a host of
athermal interactions with other dislocations, precipitates and
grain boundaries. In discrete dislocation dynamics simulations
of such deformation processes, temperature effects may enter through
dislocation mobility and lattice friction as well as cross-slip.
At temperatures greater than about 1/3 of the melting point,
the climb of dislocations becomes increasingly important
leading to phenomena such as creep and dynamic recovery.
The modeling of climb as a nonconservative motion generally
requires the concurrent modeling of dislocation motion and the
diffusion of point defects into the cores of the dislocations.
In this paper we report on a self-consistent formulation of
high-temperature discrete dislocation plasticity in finite bodies,
which couples dislocation dynamics with vacancy diffusion theory.
To address the issue of disparate time scales related
to glide and climb mechanisms, an adaptive multi-time stepping
algorithm is used in the numerical implementation of the theory.
We then present a series of deformation analyses at constant
applied stress in single crystals. We show that two regimes
of power-law creep naturally emerge in the simulations,
as affected by the applied stress and test temperature.
We also systematically quantify the power law exponent in either
regime and the part of the strain rate that results from
mass transport through the diffusive flow of vacancies
due to pressure gradients.|
|21/05/2012(16h00) [Lieu : Barb94]|
"Characterisation of the relationship between microstructure evolution and work hardening of Fe-Mn-C TWIP steels"
|High-manganese austenitic TWIP (TWinning Induced Plasticity) steels are currently under development for automotive applications where their excellent mechanical properties provide the potential for significant vehicle weight saving. However, the origin of their outstanding properties is still unclear due to multiple effects occurring in these steels at different scale levels. The main hypotheses available in the literature include deformation twinning and dynamic strain ageing. Twins are usually thought to have a huge impact on the outstanding properties of the materials, either by bringing about a dynamic Hall-Petch effect and/or a composite effect.
The aim of this study is to reach a better understanding of the work hardening mechanisms of TWIP steels, especially the Fe-20%Mn-1.2%C alloy. In order to do so, a reliable method to estimate the twinning rate based on point counting analysis on EBSD (Electron Back-Scattered Diffraction) micrographs is proposed. It is then shown that there exists a first order relationship between this twinning rate and the work hardening rate in tension. Depending on the loading mode (uniaxial tension, simple shear and rolling), differences in work hardening rate are observed. These discrepancies can be related to differences in texture, twinning rate, number of activated twinning systems in each grain, twin thickness and transmission of twins across grain boundaries. The influence of complex strain paths is also investigated, by means of Bauschinger tests. Internal stresses are found to increase with strain, without saturation, due to the formation of twins. In terms of texture, in cold rolling, the <110>//ND fibre appears during straining. In tension, the <111>//TA, favourably oriented for slip of partial dislocations and the <100>//TA fibre, favourably oriented for slip of perfect dislocations, are seen to develop. Finally, tensile tests at different strain rates with the help of digital image correlation were also performed to study the dynamic strain ageing effect occurring in this type of steel. It is suggested that supplementary hardening could come from reorientation of Mn-C pairs in the cores of the dislocations. |
|15/05/2012(11h00) [Lieu : Stevin b-145]|
"Comment inventer aujourdhui en chirurgie ? Observation dune synergie entre chirurgiens, roboticiens et industriels"
|Trois anthropologues (Caroline Moricot, Marina Maestrutti, Université de Paris 1 Panthéon-Sorbonne et Marie-Christine Pouchelle, Centre National de la Recherche Scientifique) se penchent depuis 2010 sur la manière dont cinq partenaires institutionnels (2 hôpitaux, 2 entreprises, 1 laboratoire universitaire de robotique) travaillent en synergie autour dun projet commun, largement financé par OSEO (1.600 000 euros). Il ressort de leurs premières observations (terrain en cours jusquen 2014) que les aspects purement techniques du projet (projet qui consiste à inventer un dispositif robotisé permettant de réaliser des biopsies optiques à lintérieur de la cavité abdominale, sans prélèvement de tissu et à léchelle cellulaire) ne sont pas seuls en cause pour la réussite de la démarche. Interviennent aussi des enjeux commerciaux, la culture professionnelle des partenaires, le mode de management adopté, les clivages socio-professionnels ente hospitaliers (chirurgiens/anatomopathologistes) et le rapport que les patients entretiennent avec les biopsies (bien que ce dernier point ne soit traité par aucun des partenaires en présence).
Est aussi en question la fiabilité de la recomposition (mosaïcing) des images obtenues à léchelle cellulaire. Nous faisons lhypothèse que dans ce cas le regard des experts est co-construit à la fois par le territoire exploré, par loutil robotisé qui permet de recueillir les images et par le sens que leur donne celui ou celle qui les interprète. Comment représenter ce que lil ne peut voir ? Comment seffectue lélaboration des dispositifs intermédiaires quil faut placer entre lil et la matière regardée ? Quels choix président à cette invention ? Que décide-t-on de montrer ? Quels sont les référents ? Enfin comment la confiance se construit-elle autour de tels dispositifs ?|
|11/05/2012(14h00) [Lieu : Stevin b-145]|
"Robotran: a powerfull symbolic software for multibody systems analysis"
|This seminar will present the multbody software Robotran which has been
developed at the CEREM for more than twenty years. It will start with a
brief overview of multibody concepts and will describe the modelling
approach developed at the CEREM. In particular, the symbolic generation
of the equations of motion will be introduced.
Then, a particular attention will be paid to explain the Robotran
software principles and to demonstrate its user-friendliness. The
various kinds of analysis (equilibrium, dynamic simulation, actuator
force estimation, control, etc.) and application domains (vehicle
dynamics, robotics, biomechanics, etc.) will be highlighted using
research and industrial application examples. This will emphasize the
reliability of Robotran for industrial problems and its flexibility and
its portability for research applications of any kinds, which always
require challenging improvements.|
|11/05/2012(13h00) [Lieu : Reaumur D382]|
of new beta-metastable titanium alloys with high ductility induced by
simultaneous TRIP and TWIP effects"
|04/05/2012(13h00) [Lieu : Ba 01]|
"Effect of internal stress and microstructure
on the hydriding behaviour of Pd thin films"
|27/04/2012(16f15) [Lieu : Bâtiment Stevin, Salles de Séminaires, b.044]|
Ehsan Tavakoli(Amirkabir University of Technology (Tehran Polytechnic)
"Large Eddy Simulation of Turbulent Swirling Flows"
|Turbulent Swirling flows play an important role in many flows that are encountered in nature and in industrial applications. Swirl extremely affects the physics of the flow and leads to various types of vortical structures which can be practical in stabilizing the flow. Combustion at high Reynolds numbers usually may lead to lifted flames or even blow off, but this instability can be prevented by locally reducing the flow speed with the aid of self-induced vortical structures. This is the main idea of swirl burners in industrial applications during the past decades.
The present talk is about numerical simulation of turbulent swirling flows with respect to bluff-body geometry of Sydney swirl burner. The simplified geometry is modeled in cylindrical coordinates with azimuthal direction of homogeneity. This is very useful in deriving averaged turbulent characteristics and turbulent viscosity in methods such as the dynamic Smagorinsky model. In spite of all these benefits, azimuthal grid refinement near the axis is inevitable and yields to numerical instability, in addition to the mathematical singularity on the axis which has to be removed. The numerical procedure and statistical turbulent results of swirl flows are presented and discussed in this talk. |
|27/04/2012(16h00) [Lieu : Stevin b-145]|
Francois Baudart and Francis Labrique(iMMC)
"Optimal torque control in normal and fault tolerant operation modes of
SMPM synchronous motors using a tooth coil technology"
|Surface mounted permanent magnet (SMPM) synchronous motors using a tooth
coil technology are well suited for building high performance fault
tolerant electromechanical actuation systems. In the first part of the
presentation we summarize the main features of polyphase permanent
magnet synchronous motors using a tooth coil technology. In the second
part, considering the most probable fault which is the loss of feeding
of one phase, we show how to build a efficient controller of the torque
developed by the motor which allows to maintain the torque unchanged
after a fault occurrence.|
|20/04/2012(13h00) [Lieu : Reaumur D382]|
"Microstructure and composition influence on the plastic flow and
damage properties of Duals-Phase steels: Experiments (and modelling)"
|02/04/2012(11h00) [Lieu : Euler a.207]|
Ludovic Noels(LTAS, ULg)
"Development of non linear mechanic solvers based on Gmsh library"
|The finite element method requires a discretization of structures with a
mesh. This one can be obtained with Gmsh, which is a 3D finite element
grid generator with a build-in CAD engine and postprocessor.
Its design goal is to provide a fast, light and user-friendly meshing
tool with parametric input and advanced visualization capabilities.
The developed solvers use Gmsh as a C++ library. Thus, the solvers can
use all C++ classes defined in Gmsh to describe the geometry and the
mesh, facilitating the management of the mesh and the post processing,
as all results are stored directly via Gmsh.
Based on these exiting tools, a non-linear solver for mechanics has been
implemented. This solver includes the general libraries as material
laws, time integration, force archiving ...
It is implemented in parallel with the use of Continuous and/or
Discontinuous Galerkin methods and takes advantage of mesh partitioning
obtained via Gmsh.
From this solver the developer can derive particularized solvers for
well defined problem, while still using the common libraries (as
The versatility of this approach is demonstrated by the resolution of
different kinds of mechanical problems as periodic boundary conditions,
non local damage, fracture mechanics for 3D problems and thin
structures, and computational homogenization schemes.|
|30/03/2012(16h00) [Lieu : Euler a.002]|
Laurence Brassart(Harvard University)
"Diffusion-induced plasticity in silicon electrodes during cyclic
charge and discharge of Li-ion batteries"
|Rechargeable lithium-ion batteries are energy storage systems of choice
for a variety of applications ranging from portable electronics to
electric vehicles. Their principle of operation is based on the reversible
insertion reaction of lithium atoms within the electrode material, driven
by electrochemical forces. Silicon is a promising anode material for the
next generation of lithium-ion batteries as it has the highest capacity.
However, the insertion of lithium atoms into silicon electrodes is
accompanied by huge volume changes (up to 300%), inducing large stresses
and eventually mechanical degradation of the host when the expansion is
constrained. Recent experiments indicated that these deformations can be
accommodated by plastic flow for electrodes of small size. In this talk I
will present a new model that couples plasticity and diffusion in a
consistent thermodynamic framework. The impact of plastic flow of silicon
on the cycle life and performance of lithium-ion battery will be
|30/03/2012(13h00) [Lieu : Reaumur D382]|
"Fracture toughness measurement of thin films"
|23/03/2012(14h00) [Lieu : Stevin b-145]|
"Introduction to the Orocos software framework and best practices"
|Orocos (www.orocos.org) is a Free and Open Source Software project
started in 2001 with funding from three machine tool builders, to help
them support their most advanced realtime control challenges. Since
then, the software is being used in dozens of critical realtime parts of
applications supporting millions of euros in added value. The "secrets"
of this success are (i) focus on being the best in a clear niche of the
overall control software domain, (ii) a motivated resistance against
"featuritis", and (iii) a very long-term vision to Model Driven
Engineering support via toolchains. This presentation gives an overview
of the essential concepts and primitives in the Orocos eco-system, as
well as some examples of "best practice" designs|
|23/03/2012(13h00) [Lieu : Reaumur D382]|
"Damage in 6xxx series Al alloys"
|16/03/2012(16h15) [Lieu : Stevin b-145]|
"Robotran: a powerfull symbolic software for multibody systems analysis"
|This seminar will present the multbody software Robotran which has been
developped at the CEREM for more than twenty years. It will start with a
brief overview of multibody concepts and will describe the modelling
approach developped at the CEREM. In particular, the symbolic generation
of the equations of motion will be introduced.
Then, a particular attention will be paid to explain the Robotran
software principles and to demonstrate its user-friendliness. The
various kinds of analysis (equilibrium, dynamic simulation, actuator
force estimation, control, etc.) and application domains (vehicle
dynamics, robotics, biomechanics, etc.) will be highlighted using
research and industrial application examples. This will emphasize the
reliability of Robotran for industrial problems and its flexibility and
its portability for research applications of any kinds, which always
require challenging improvements.|
|09/03/2012(13h00) [Lieu : Reaumur D382]|
"On damage nucleation in dual phase materials"
|09/03/2012(16h15) [Lieu : Stevin b-145]|
"Analytical Sky Modeling"
|In order to obtain a realistic view of the behavior of a photovoltaic
system, it is necessary to achieve computer simulations. For that
purpose, the most important data is the light irradiance of the
photovoltaic array at a small time scale (some minutes) during a
significant duration (one year or more). Unfortunately, complete
experimental data (irradiance for all module inclination and
orientation) are never available. For example, the only available
measurement result is often hourly or daily global irradiance on a
horizontal plane. Sky modeling is thus necessary in order to deduce from
the available partial data a realistic estimation of the module
irradiance. We hope also extract using an atmospheric model some
spectral characteristics of that irradiance.|
|09/03/2012(16h15) [Lieu : Bâtiment Stevin, Salle des Seminaires b0.44]|
"Numerical simulation of a flow over and through porous layers of large porosity"
|In this talk we present a continuum model and numerical results for transient flows over and through porous media of large porosity. Our targeted area of application is the simulation of forest fire spread. According to our modeling approach, the vegetation is treated as a rigid porous layer. In the first part of this talk, we briefly discuss the derivation of the continuum model and present an algorithm for its numerical treatment. In the second part of the talk, we present and analyze results from direct numerical simulations for developing flows over porous layers of large porosity. Due to the steep velocity gradients that are developed across the interface between the porous layer and the clear fluid, a Kelvin-Helmholtz instability is onset that leads to the formation of a mixing layer. Important characteristics of this layer such as growth rate, self-similarity and statistics of the fluctuating velocities are presented and discussed.|
|06/03/2012(16h15) [Lieu : SUD09]|
"Nanomechanical lab on-chip for testing thin film materials and application to Al and Al(Si)"
|The exceptional properties of thin films have encouraged the use in a wide variety of applications ranging from microelectronics and Micro Electro Mechanical System (MEMS) technologies to the field of surface functionalization of bulk metallic, ceramic, glass, and polymeric materials. The reliability of these devices and coatings heavily depends on the thermo-mechanical behaviour of the films. The key issue is that thin films exhibit thermo-mechanical properties different from the bulk counterpart. The development of testing methods to characterise the mechanical properties of thin films is thus essential. A direct adaptation of classical techniques used to test bulk materials is not straightforward due to the challenges in generating small loads and adequate gripping between actuator and test specimen.
In this thesis, a novel concept of on-chip mechanical testing based on MEMS fabrication methods is developed. The original idea is to use the internal stress generated in one layer acting as a small spring to impose a load to a test specimen from which the mechanical properties can be extracted. The on-chip technique is presented with all the details about the design, fabrication and data reduction. Two thin film material systems are studied for demonstration of the potential of the method: pure evaporated Al and sputtered AlSi(1%) both in the 100 to 1000 nm thickness, revealing thickness dependent strength and a statistical imperfection sensitive ductility. The technique is applied for performing stress relaxation experiments over long period of time which enables to extract the strain rate sensitivity parameters. |
|02/03/2012(13h00) [Lieu : Reaumur D382]|
"Réacteur à déposition plasma pour le traitement des surfaces des panneaux solaires"
|27/02/2012(14h00) [Lieu : SCES02]|
"Secondary SCRAM System for Liquid Metal Cooled nuclear Reactors: Simulation using Particle-Based Methods and Experimental Validation"
|The ability to shut down the nuclear chain reaction is of course one of the most important safety features of any nuclear reactor. One common way of shutting down the reactor is by inserting a neutron absorbing rod into the reactor core. This rod will absorb a part of the neutrons in the reactor and stop the chain reaction. However in some, very unlikely, accidents it is possible that these rods cannot be inserted into the reactor, for instance due to deformation of the core.
The purpose of this PhD was developing an additional safety system, complementing the system based on the control rods that can also work in a liquid metal coolant. As a first step, a concept using spherical absorber particles and a metal melt seal was proposed. During normal operation the absorbing particles are kept above the core by means of an Aluminum melt seal. When something goes wrong, the temperatures in the reactor will increase till the Aluminum seal will melt. When the Aluminum seal is molten, the neutron absorbing particles are no longer supported and fall down into the active core region. In this way they will shut down the reactor.
As a next step, this concept was evaluated both through experiments as well as Discrete Element Method (DEM) and Smoothed Particle Hydrodynamics (SPH) particle-based simulations.
As will be demonstrated during my public defense, both the particle flow as well as the Aluminum seal showed unexpected results which I would be happy to share with you the 27th of February 2012. |
|17/02/2012(16h15) [Lieu : BARB94]|
"Blood Flow Modeling for Patient-Specific Bypass Surgery in Lower-Limb Arteries"
|Every day in Belgium, at least one or two people will undergo a bypass surgery in
the lower-limb arteries. This medical procedure consists of replacing an occluded
section of the leg arteries with an artificial vessel, in order to allow blood to flow
downwards of this blockage. Even though this intervention is very common, failure
of this treatment within five years reaches up to 60%. In order to improve our
understanding of the causes of bypass failure, one approach is to study the local
hemodynamics (i.e. the blood flow circulation) in these vessels, as this factor
strongly influences the initiation and progression of arterial diseases.
This thesis addresses this objective by modeling the patient-specific arterial
hemodynamics. As each patient is characterized by his own physiology (e.g.
morphology, severity of the disease) and is treated by a bypass graft of fixed
characteristics (e.g. material, length, location of the anastomoses), each clinical
case needs to be considered individually.
In the first part of our work, blood pressure, velocity and flow rate in the network of
arteries are modeled with a one-dimensional model of the arteries and lumped
windkessel models of the secondary and distal vessels. The 18 parameters of this
coupled model are then computed in order to reflect the pathological condition of
the patient. Thanks to the set-up of an experimental protocol of data acquisition,
we have been able to acquire in-vivo measurements of hemodynamics in 25
patients operated with bypass surgery. The comparison of these in-vivo signals
with numerical simulations enabled us to improve and validate our patient-specific
In the second part of this thesis, the dynamic of blood is considered as impulses
of waves which are generated by the contractions of the heart, and which
propagate throughout the arterial network. By using the wave intensity analysis
method, we analyzed and characterized these waves in the pathological arteries
of the leg, and emphasized the effects of the bypass.|
|16/02/2012(14h00 -> 18h30) [Lieu : BARB91 + Hall VINCI]|
à l'intention des étudiants en master à l'EPL Organisée par les chercheurs de l'iMMC(institute of Mechanics,Materials and Civil
"Journée d'information sur la
recherche en mécanique,
matériaux et génie civil"
14h00-14h30 Auditoire Barb 91:
Information sur la carrière de chercheur.
14h45-16h00 Auditoire Barb 91:
Présentation de sujets de recherche
abordés dans les cinq pôles de l'iMMC.
16h15-18h15 Hall Vinci:
Open space (rencontre avec les professeurs,
chercheurs, assistants et mémorants) + Drink
|16/02/2012(13h00) [Lieu : Reaumur D382]|
"Development of new mechanical tests for composite materials"
|03/02/2012(13h00) [Lieu : Reaumur D382]|
"Residual stress estimation in damascene
copper interconnects using embedded rotating sensors"
|03/02/2012(16h15) [Lieu : Stevin b-145]|
"Introduction to stochastic optimization algorithms"
|This seminar aims to be an introduction to stochastic optimization
algorithms, a part of non-deterministic algorithms, and splits in two
steps. First, I present the basis of non-deterministic algorithms by
explaining the three most used : genetic algorithms, particle swarms
optimization, and simulated annealing. During the presentation, I will
explain the simplicity to use these algorithms for the optimization of a
problem when no conventional method can be used. Then I present an
implementation of these algorithms in the field of topology optimization.
|26/01/2012(13H30) [Lieu : Salle Peppermans]|
"Midis de la recherche GCE"
|Dans ce midi de la recherche, nous présenterons les grandes lignes du
projet de recherche concernant les efforts damarrage en vue du
dimensionnement des écluses de navigation. En effet, le transport de
marchandises par voie fluviale propose des avantages économiques et
environnementaux non négligeables. Ainsi les pouvoirs publics supportent
de nombreux projets de modernisation des voies existantes, voire la
création de nouvelles voies.
La question du mouvement des bateaux durant le sassement dune écluse est
un sujet sensible et la détermination des efforts dans les amarres nest
pas aisée. Actuellement, plusieurs méthodes de détermination existent
allant des plus simples aux plus complexes.
La présentation saxera autour des méthodes visées pour parvenir au terme
de ce projet de recherche. Elle sera également loccasion de présenter les
recherches actuelles prestées dans le cadre du contrat passé avec la
Direction des Recherches Hydrauliques de Châtelet.|
|20/01/2012(14h00) [Lieu : Stevin b-145]|
"Assistance for tumor surgery : From benchmark to market"
|Tumor surgery is among the most challenging surgeries. Clinical results
show a high complication rate which can reach 90% for a tumor located
within the pelvis. However, no assistance has been developed while the
technology is available for simpler surgeries such as knee arthroplasty.
Seven years ago, a project has been launched to help the surgeon during
this difficult task. The conventional method has been analyzed, and
assistance before and during the surgery has been developed and
validated. Several clinical cases have successfully benefited from these
solutions. A first spin-off project is currently ongoing to evaluate
their commercialization. This seminar presents the techniques that are
now available for the surgeons and the steps toward the commercialization|
|19/12/2011(16h15) [Lieu : Stevin B044]|
Dagmar Sternad(Department of Biology, Electrical and Computer Engineering and Physics, Northeastern University)
"Human motor performance: Exploiting stability, channeling, variability, and optimizing safety margins"
|Rehabilitation and augmentation of human movement using
technology has made significant strides in recent years. While
technology is being developed, it remains an open question what
determines the desired "skilled" movement. Research in my lab aims to
reveal the fundamental principles that underlie motor skill and learning
in the healthy nervous system -- as a necessary basis to understand
neurological dysfunction and to develop intervention. Research of my lab
has examined three interactive model skills: the discrete task of
throwing a projectile to hit a target, the rhythmic task of bouncing a
ball, and the continuous task of carrying a cup filled with coffee.
Characteristic for our approach is to start with a mechanical model of
the task and render it in a virtual environment. As such, the human
interacts with a known task environment. Based on stability analyses of
a dynamical model of the task, we study how the neuro-mechanical system
develops robust solutions to meet the task demands. Using the three
skills as model examples, we show that developing motor skill means 1)
exploiting solutions with dynamical stability, 2) finding the most
error-tolerant strategy and channeling sensorimotor noise into
task-irrelevant dimensions, and 3) optimizing safety margins and taking
advantage of small perturbations. Based on these insights into healthy
function, new intervention techniques can be developed that facilitate
learning and relearning of motor tasks.|
|16/12/2011(13h00) [Lieu : Reaumur D382]|
"A study of ductile fracture by in-situ microtomography"
|16/12/2011(14h00) [Lieu : Stevin b-145]|
Bohumil Horak(Prototype laboratory, Technical University of OSTRAVA, Czech Republic)
"Electromobility and Electromobiles"
|Within the framework of the research and development on the alternate
and renewable energy sources, the project of an ultra light car,
hydrogen - electric, was led by a team of students and technicians
supervised by Bohumil Horák, associate professor in the Faculty of
Electrotechnics and Computer Science of the VSB -- Technical University
This team has also realized the electrification of roadsters KAIPAN with
various power units. It implied complex realizations of the power
electronic such as electric drives (two separate drives with gearboxes
for two front wheels, electronic differencial) and control electronic,
inteligent charging systems, control of regenerative braking system,
etc. They also develop and produce public charging stations for
|09/12/2011(16h00) [Lieu : BARB94]|
"Finite element modeling of sediment dynamics in the Scheldt"
|A densely populated watershed and numerous industrial activities, especially around the port of Antwerp, are responsible for the Scheldt Estuary and River to be highly polluted. Water and sediment circulation are major processes contributing to the global dynamics of the various pollutions. The objective of this thesis is to develop a numerical tool in order to make possible simulations of those environmental issues.
The finite element technique enables the use of unstructured meshes, so that the spatial resolution can vary widely over the domain. In our implementation, in the scope of the SLIM (www.climate.be/slim) and TIMOTHY (www.climate.be/timothy) projects, we combine one-dimensional equations for rivers and two-dimensional equations for estuaries and seas. Nevertheless, the tidal river network, the estuary and the adjacent coastal zone are simulated in the same framework.
The Scheldt Estuary features large shallow areas that are periodically emerging at low tide. This phenomenon is a numerical challenge in estuarine modeling because the governing equations are intrinsically unable to deal with areas where the water thickness may become zero. A flux-limiting method has been developed, which modifies the discrete form of the governing equations, in order to prevent the water surface to go down where it is already very low.
The last contribution is the development of the sediment transport module. Its calibration pointed out the influence of suspended sediment concentration, salinity and biological activity on the flocculation processes, as the influence of the biological activity on the erodibility of bottom sediments. Our 1D-2D model, with a very competitive computer cost, appears to provide results as accurate as those from more complex, three-dimensional tools, traditionally deemed indispensable in sediment transport modeling. Our approach appears therefore to be very promising for long-term environmental simulations of the Scheldt.|
|02/12/2011(13h00) [Lieu : Reaumur D382]|
"Hybrids for electromagnetic absorption: electromagnetic and mechanical point of view"
|01/12/2011(13h00) [Lieu : salle Passelecq]|
"Etude de linteraction sol-pieu sous sollicitations harmoniques latérales "
|Létude de linteraction sol-pieu sous sollicitations harmoniques latérales est indispensable dans la conception des ouvrages. Ce travail présentera, dans un premier temps, une approche analytique simplifiée décrivant les réactions du sol considéré viscoélastique linéaire pour un mouvement vibratoire du pieu. Puis, une approche numérique sur la méthode des différences finies sera conçue afin de prendre en compte le comportement non-linéaire du sol.
Mots clés : Interaction sol-pieu, viscoélastique, dégradation, loi hyperbolique, comportement linéaire, comportement non-linéaire.|
|29/11/2011(13h00) [Lieu : Stevin B-145]|
"Active gravity compensation in upper limb rehabilitation robots -
motivations, strategy and first results"
|Strokes are the major source of physical disability, leading
generally to hemiplegy. Due to many factors, the number of stroke
patients increases yearly and there is a significant growth in needs for
rehabilitation medicine that robotic devices can help to meet. These
robotic devices assist the patient in the mobilization of their limbs
through the measurment of the interaction forces between the robot and
the patient. However, as it is not possible to discern muscular forces
produced by the patient from those due to the gravity action in these
measurements, the patient has to overcome their own arm weight to
perform rehabilitation exercises. Consequently, weak patients are not
able to initiate any mouvement and stay completely passive facing a
robot that actively mobilize their limb. This goes against clinical
studies demonstrating that rehabilitation is much more efficient when
the patient initiates the movement itself.So it seems interesting to
compensate the patient arm weight in order to permit an active
rehabilitation as soon as possible after stroke. This seminar proposes
to explain the strategy to achieve this compensation in upper limb
rehabilitation robots and the first results of the first essential step
: the measurement of the arm orientation.|
|24/11/2011(17h00) [Lieu : BARB94]|
"Topology optimization methods for the design of electromagnetic actuators"
|Topology optimization methods aim to design devices in an automated way. The designer defines a design space which is divided in cells, a library of materials and an objective function. The method goal is then to maximize or minimize this objective function by distributing optimally the materials in the cells.
This thesis aims to develop topology optimization methods based on deterministic algorithms for the design of electromagnetic devices. Deterministic algorithms are indeed characterized by a fast convergence compared to heuristic algorithms, but are usually faced with two main challenges: the presence of local minimizers and the handling of intermediate materials, i.e. mixes of different materials in the same cell, at the end of the optimization. The former prevents the algorithm from finding the optimal solution, whereas intermediate materials must be removed in the final design in order to obtain a manufacturable topology.
Through the consideration of study cases, this thesis introduces two groups of methods. The isotropic methods are based on materials characterized by isotropic properties. Each material is then described by a unique design variable per cell giving its proportion in the cell. These methods are usually simple to apply, but the range of problems on which they are effective is limited. The second group is called the anisotropic methods, in which an anisotropic permeability is considered for the iron material. These methods can be applied to a wider range of problems, but involve a larger number of design variables in order to represent the iron magnetic anisotropy, which increases the computation time.
The thesis besides studies a mesh refinement algorithm aiming to increase the resolution of the produced topology while limiting the computation time.
Eventually, the proposed methods are applied to the design of a switched reluctance motor, a permanent magnet synchronous motor and magnetic thrusts, in order to highlight their performances.|
|18/11/2011(13h00) [Lieu : Reaumur D382]|
Juray De Wilde(iMMC)
"Process Intensification, some
examples studied in IMAP"
|14/11/2011(16h00) [Lieu : SUD11]|
Benjamin de Brye(iMMC)
"Multiscale Finite-Element Modelling of River-Sea Continua"
|Coastal zones and rivers are most of the time modelled using different tools and approaches. Therefore, coupling two such models is a task that is far from trivial. With the finite-element method, unstructured meshes can be used and the spatial resolution can vary widely over the domain, making it possible to simulate river networks and the adjacent coastal zone in the same model. Investigating the feasibility thereof is one of the objectives of this thesis, in which the model utilized is SLIM, i.e. the Second generation Louvain-la-Neuve Ice-ocean Model (www.climate.be/slim).
Two applications have been developed. The first one is the Scheldt continuum (Belgium/The Netherlands) for which the grid covers the freshwater tidal river network around Ghent, the Scheldt Estuary and the whole continental shelf. The second is the Mahakam continuum (Indonesia). In this application, the domain is composed of three lakes connected to a river network that leads to a complex delta before reaching the sea.
The hydrodynamics was calibrated in the two domains and the model was used to compute the renewing timescales of the Scheldt Estuary and the Mahakam Delta. The renewing timescales such as the residence time, the exposure time and the age are measures of the rate at which the water initially in the estuary or the delta is renewed. Such diagnoses have direct links with many environmental variables and can be used to improve the understanding of the ecological dynamics inside the continuum.|
|10/11/2011(16h00) [Lieu : BARB94]|
"Micromechanical modeling of the progressive failure in short glass-fiber reinforced thermoplastics"
|Thermoplastics reinforced with short glass fibers are increasingly used in many industrial applications due to their attractive mechanical properties, rapid processing and relatively low manufacturing cost. Injection molding is a widely used process for the production of reinforced parts with complex shapes. However, the concentration and the orientation of the fibers vary from one point to the other. This induces a strong heterogeneity throughout the material, making the prediction of its behavior and rupture a nontrivial task. Multi-scale or micro-macro approaches have been proposed in the literature in order to take into consideration the influence of the microstructure on the macroscopic properties. Among them, Mean-Field Homogenization predicts with satisfying accuracy the macroscopic response and to a lesser extent the per-phase responses.
In this work, a new approach for modeling the behavior of thermoplastics reinforced with short glass-fibers has been developed. The constitutive law is based on an elastic-plastic formulation. The micro-macro transition is achieved in two steps using a Mori-Tanaka (MT) scheme and a Voigt scheme (or also a self- consistent scheme) while integrating phenomenological failure laws. One essential concept in this modeling approach is the Pseudograin (PG) unit. A PG is defined as a fictitious subregion of the microsctructure where all fibers are aligned and share the same aspect ratio. Failure models are applied at the PG level. An incremental MT model is then used to predict the phase averages of the stresses and strains in each PG. Once average responses are computed, a failure procedure is applied at each PG to predict its state of degradation. Finally, the overall average response is computed using a second homogenization rule.
Two models have been developed based with different treatments of damage. The first one is called the "First Pseudograin Failure" model in which failure criteria are used to predict the PG failure. The second one is the "First Pseudograin Damage" model in which continuum damage mechanics is adopted to predict the progressive degradation of a PG. Thus, the implemented methods account for the damage within the PGs. It can also be used for non-proportional loadings cases and can be extended to rate-dependent behavior.
An extensive validation of the predictions against experimental data is conducted for short glass-fiber reinforced polyamide 6,6.|
|08/11/2011(14h00) [Lieu : BARB94]|
"Coupled viscoelastic-viscoplastic modeling of homogeneous and reinforced thermoplastic polymers"
|Glass fiber reinforced thermoplastics are widely used in structural applications because of their high stiffness and failure stress. However, controlling their complex mechanical behavior must rely on numerical simulations. The main outcome of this thesis is an original mean-field approach predicting the mechanical response of fibre reinforced thermoplastics under arbitrary loading paths. The model accounts for the volume fraction and the orientations of fibers as well as the viscoelastic- viscoplastic (VE-VP) character of the surrounding matrix.
The matrix response is described in terms of a coupled VE-VP model suitable for multi-axial strains. The total strain is the sum of viscoelastic and viscoplastic parts, and the Cauchy stress is computed by assuming linear viscoelasticity and relying on a Boltzmann integral of the history of viscoelastic strains. An efficient and robust time-integration algorithm is developed, implemented in a numerical code and studied.
In order to predict the composite behavior, an affine linearization procedure is proposed for the coupled VE-VP model. This formulation leads to an incremental constitutive relation in the time domain which is form-similar to linear thermo-elasticity. Hence, existing homogenization models for linear thermoelastic composites can be applied. We also conducted direct finite element (FE) analysis of representative volume elements (RVEs) to verify the predictions of the proposed mean field homogenization (MFH) method.|
|04/11/2011(13h00) [Lieu : Reaumur D382]|
"New insight into microstructure observations with ACOM-TEM orientation
|28/10/2011(13h00) [Lieu : Reaumur D382]|
"Estimation of kinetic parameters in
a plasma torch for production of SiC nanoparticles"
|21/10/2011(13h00) [Lieu : Reaumur D382]|
"Fluage et relaxation de films minces en
|21/10/2011(16h15) [Lieu : BARB94]|
"Finite element methods for sea ice modeling"
|In order to study and understand the behavior of sea ice, numerical sea ice models have been developed since the early seventies and have traditionally been based on structured grids and finite difference schemes. This doctoral research is part of the Second-generation Louvain-la-Neuve Ice-ocean Model (SLIM) project whose objective is to bring to oceanography modern numerical techniques. The motivation for this thesis is therefore to investigate the potential of finite element methods and unstructured meshes for sea ice modeling.
The Canadian Arctic Archipelago (CAA) is a complex area formed by numerous islands and coastlines and constitutes a nice application for unstructured meshes. Our model is the first to investigate the effects of resolving the CAA on the ice cover features and the importance of the CAA in terms of mass balance is highlighted.
We further develop a Lagrangian and adaptive version of the model allowing the computational grid to move with the ice. We take advantage of the locality of the mesh adaptation procedure to update the discontinuous fields thanks to a local Galerkin projection.
Sea ice age patterns and how they change in time provide an integrated view of the recent evolution of sea ice growth, melt and circulation. We first study the vertical age profile in sea ice and analyze the age-thickness relationship in a stand-alone thermodynamic sea ice model of the Arctic. We then take advantage of the Lagrangian model to reproduce the algorithm used to compute satellite retrievals of ice age and compare with different ice age definitions. Several characteristics consistent with satellite observations are deduced from our numerical simulations.|
|14/10/2011(13h00) [Lieu : Reaumur D382]|
"Séminaire de sécurité dans les labos"
|10/10/2011(10h45) [Lieu : Barb 02]|
Jérôme Szewczyk(Instituts des Systèmes Intelligents et de Robotique (UPMC - Paris 6))
"Conception et optimisation d'un robot pour la micro-chirurgie de l'oreille moyenne"
|07/10/2011(14h00) [Lieu : Stevin b-145]|
Omar Scaglione(EPFL, Switzerland)
"Sensorless Brushless DC motor position detection"
|Brushless DC (BLDC) motors are more and more used in small
power industrial application (<500[W]). Because of their mechanical
construction, only a power source in not enough for driving them: a more
smart power electronic is needed.The aim of the presentation is to
introduce the audience to BLDC motors and specially to the electronics
able to drive a BLDC motor without the need of position sensors.After a
small introduction on the mechanical construction differences between
Brushless DC (BLDC) and brushed DC motors, the advantages and drawbacks
of these two drives will be highlighted. A special focus will then be
placed on the need of knowing the rotor position in order to correctly
drive a BLDC motor. Thereafter, the advantages of sensing this position
without the need of extra sensors will be discussed. For illustrating
different techniques of sensorless position detection, some practical
examples will be shown.|
|07/10/2011(13h00) [Lieu : Reaumur D382]|
"Porous anodising of silicon for
|03/10/2011(13h00) [Lieu : Reaumur D382]|
"From experimental measurements to finite elements simulations. How simulations can help to understand the behaviour of a
|30/09/2011(13h00) [Lieu : Ba00]|
"Processing of C fibres reinforces Mg matrix composites via pre-infiltration with Al"
|Mg matrix composites appear as suitable competitors to Al alloys for a wide number of applications, as they allow for a significant weight reduction while exhibiting potentially comparable mechanical properties. And a large variety of processing routes have already been considered for their production, including 'liquid state' processes such as squeeze casting. These techniques necessitate a very careful control of the wetting behaviour and of the possible interfacial reactions between the reinforcements and the molten metal, along with the solidification process. And this is more particularly true in the case of Mg matrix composites with carbon fibers reinforcements, as Mg alloys are known to exhibit a poor wetting behaviour on carbon substrates.
Various processes aiming at enhancing the wettability of carbon fibers have thus already been investigated, and among those, the deposition of a metallic layer on the fibers appear particularly promising since the wetting of a metal on a metallic substrate is known to be usually good. However, in the present work, an alternative 'liquid' processing route has been investigated. By making use of the well-known favourable effect of pre-treating the carbon fibers in an aqueous solution of K2ZrF6 an effect that is generally attributed to the K2ZrF6 bringing about the dissolution of the thin but dense alumina layer covering the free surface of the molten aluminium -, it has indeed been possible to impregnate carbon fibers by dipping them in a liquid aluminium bath, and the feasibility of further using these pre-impregnated carbon fibers as reinforcements in Mg matrix composites has been assessed.|
|26/09/2011(15h00) [Lieu : BARB94]|
"Spark Plasma Sintering: Evaluation critique dune méthode de frittage rapide et de son applicabilité aux céramiques techniques et composites"
|La technique SPS (Spark Plasma Sintering), également connue sous lacronyme FAST (Field Assisted Sintering Technique) est une méthode de frittage rapide dont le principe consiste en lapplication simultanée dune pression uniaxiale et dun courant pulsé de grande intensité. Ce courant traverse le moule et les pistons en matériau conducteur et la poudre à densifier, pour autant quelle soit conductrice, et chauffe le système par effet Joule jusquà des vitesses pouvant atteindre mille degrés par minute. Ces conditions de frittage particulières permettent de densifier rapidement bon nombre de matériaux en un temps relativement court et à des températures généralement plus faibles quen frittage conventionnel.
Cette technique de frittage est originale et relativement nouvelle. Cependant, les mécanismes sous-jacents ne sont pas encore bien compris/connus à lheure actuelle. Ainsi, lexistence, dans la poudre à densifier, détincelles ou dun plasma au cours du frittage, est toujours sujet à débat, raison pour laquelle lacronyme FAST devrait, en principe, être préféré à lacronyme SPS.
Dans ce travail, nous tentons dévaluer quels sont les réels avantages et limitations de la technique SPS en la comparant à des méthodes de frittage plus conventionnelles (Hot Pressing). La première partie du document concerne donc une étude paramétrique de la technique sur base dessais réalisés sur lalumine. Ce matériau bien connu est souvent utilisé comme référence dans létude du frittage à létat solide.
Nous évaluons aussi linfluence des propriétés électriques et thermiques des matériaux à fritter en étudiant la densification de composites. Les systèmes choisis sont Al2O3-TiC et AlN-TiC, tenant compte de la conductivité électrique élevée du TiC et de la grande conductivité thermique de lAlN. En outre, dans les conditions expérimentales choisies, aucune réaction ne se produit entre les différentes phases des composites.
Nous tentons de dégager des pistes concernant les mécanismes en jeu et les effets potentiels du courant et/ou du champ électriques en utilisant une méthode danalyse isotherme mais aussi la théorie non isotherme des courbes maitresses de frittage, qui conduit à la détermination dénergies dactivation apparentes|
|23/09/2011(14h00) [Lieu : Stevin b-145]|
"ShouldeRO, an alignment-free two-DOF rehabilitation robot for the shoulder complex"
|ShouldeRo is a rehabilitation robot aimed to assist the
shoulder movements of stroke patients during their rehabilitation
process. This robot has the general form of an exoskeleton, but is
characterized by an action principle on the patient no longer requiring
a tedious and accurate alignment of the robot and patient's joints. It
is constituted of a poly-articulated structure whose actuation is
deported and transmission is ensured by Bowden cables. It manages two of
the three rotational degrees of freedom (DOFs) of the shoulder, i.e.
flexion-extension and adduction-abduction. Quite light and compact, its
proximal end can be rigidly fixed to the patient's back on a rucksack
structure. As for its distal end, it is connected to the arm through
passive joints and a splint guaranteeing the robot action principle,
i.e. exert a force perpendicular to the patient's arm, whatever its
configuration. A first prototype of this robot was built, and
experimental tests were performed with this one on a healthy person.
They have shown the limits and the weaknesses of the solution but also
its potential, enough to motivate the development of a new version of
|16/09/2011(13h00) [Lieu : Reaumur D382]|
"A simplified approach to the modelling of sintering"
|05/09/2011(15h00) [Lieu : SUD11]|
"Experimental and numerical study of ammonia combustion"
|The problems of oil resources and CO2 emissions becoming increasingly alarming, the search for alternatives to fossil fuels is an important concern of our society. Even though hydrogen has been recognized as a promising fuel, implementing a global hydrogen-based economy is at present a non-feasible approach unless a suitable storage medium could be found. To bypass such difficulties, the use of ammonia in a modified spark ignition engine has been suggested. Since hydrogen must still be produced to obtain ammonia in large amount, it can be seen as a H2 vector. The successful application of ammonia as an alternative transportation fuel should be based on a detailed understanding of its combustion characteristics. The main objective of this thesis is thus the study of the ammonia combustion reaction in several conditions of pressure and temperature.
First of all, a low pressure ammonia flames structure study allows a more fundamental investigation of the ammonia consumption and products formation pathways. The elaboration of a kinetic mechanism able to reproduce the experimentally observed phenomena is the main objective of this first part. Then, the kinetic of ammonia under atmospheric pressure is investigated thanks to an ammonia laminar burning velocity study. Because of its low value, the laminar flame speed of ammonia in air is to date still unknown and in that way this part makes a significant contribution in the field of ammonia combustion research. This part allows also to validate the kinetic mechanism at atmospheric pressure. Finally, the combustion of ammonia in a spark ignition engine is experimentally investigated. The effects of the compression ratio, the spark timing and the equivalence ratio on the exhaust gases composition and the engine performances are analyzed. Such a study has never been performed for ammonia to the best of our knowledge.|
|02/09/2011(13h00) [Lieu : Reaumur D 382]|
"Soudage hétérogène par
friction-malaxage de l'aluminium: lien entre paramètres, température,
duretés et résistance en traction"
|02/09/2011(14h00) [Lieu : Stevin b.-145]|
Mahdy Eslamy(Université de Jena, Allemagne)
"Powered Ankle Knee Orthoprosthesis (design and control issues)"
|26/08/2011(13h00) [Lieu : Reaumur D382]|
"Evolution des contraintes
mécaniques mesurées lors de dépôts par sputtering de films transparents
et conducteurs en ZnO:Al ainsi que de films de palladium"
|16-19/08/2011() [Lieu : Vinci building]|
"Summer course: Mechanics of geomorphic fluid-granular flows"
|Flows involving mixtures of liquid and granular materials are ubiquitous
in nature as well as in industrial processes. Mixtures of water and
grains play a fundamental role in shaping rivers, lakes and sea beds.
Their study is important for many applications including hazard
mitigation associated with river flows, coastal and deep-sea currents,
landslides, and the design of jetting and dredging operations. To model
such flows, one must address both the fluid mechanics of free-surface
and stratified flows, and the complex rheology of granular materials.
This is manifested most clearly in the ability of liquid-granular
currents to form, flow over, and erode their own deposits. The proposed
series of lectures will explore the mechanics of these geomorphic flows
using a combination of empirical and mathematical approaches: tabletop
experiments, physical reasoning, derivation and solution of governing
equations, comparison with experimental data and discussion of field
examples. The series will include a series of 4 half-day lectures,
followed by lab activities and informal discussions in the afternoon.
Although not compulsory, interested participants will be encouraged to
prepare short presentations focusing on their research, which will guide
and enrich the scope of the course itself, the discussions and the lab work.|
|08/07/2011(13h00) [Lieu : Reaumur D 382]|
"Electrochemical oxydation of metals"
|01/07/2011(13h00) [Lieu : Reaumur D 382]|
"Multiscale modeling of adhesive joint fracture: energy separation and constraint
|27/06/2011(17h30) [Lieu : Passelecq Auditorium(VINCI)]|
Dov Prof. Leshchinsky (University of Delaware, USA)
"Recent developments in limit analysis and limit state design of reinforced soil"
|In this conference two powerful techniques which have recently become available to geotechnical engineers involved in the design of conventional and reinforced soil constructions will be described.
In the first presentation Prof. Dov Leshchinsky will describe how the diagnostic Safety Map (SM) tool can be used to design reinforced slopes and walls. The objective of such design is to produce economical layout and strength of reinforcement. Such problems often have multiple weak zones, complex geometry and soil strata, and are subjected to porewater pressure distributions, surcharge and seismic loadings. Through realistic example problems it is demonstrated how the SM can quickly assist in identifying an efficient reinforcement layout that meet the design requirements. It is shown that the SM does not require additional computational efforts; rather, it utilizes an existing data generated in stability analysis that otherwise is discarded.
In the second presentation Drs. Colin Smith and Matthew Gilbert will describe how the recently developed Discontinuity Layout Optimization (DLO) limit analysis procedure can bridge the gap between simple hand-based analysis methods and much more complex incremental techniques (e.g. based on non-linear finite elements), providing an easy to use general purpose geotechnical analysis tool that may be directly applied in ultimate limit state design. A DLO-based software application, LimitState:GEO, has been developed for use in industry and academia, and has been available since 2008. In the presentation the underlying technology will be briefly explained and the broad scope of the DLO procedure illustrated through the analysis of a variety of geotechnical problems including foundations, retaining walls and slopes. Finally, the modelling of soil reinforcement will be examined through a study of reinforced earth wall stability, with comparisons made with the tie-back method often used in practice.
|27/06/2011(16h15) [Lieu : Coubertin11]|
"Biomechanics of running in weightlessness on a treadmill equipped with a subject loading system."
|One countermeasure used during long duration spaceflight to maintain bone and muscle mass is a treadmill equipped with a subject loading system (SLS) that simulates gravity. To date, little is known about the biomechanics of running in weightlessness on such a treadmill-SLS system. We have designed an instrumented treadmill/force-plate to compare the biomechanics of running in weightlessness to running on Earth. Gravity was simulated by two pneumatic pistons pulling downwards on a subjects harness, with a force approximately equal to body weight on Earth. Four transducers, mounted under the treadmill, measured the three components of the reaction force exerted by the tread-belt under the foot. A high-speed video camera recorded the movements of limb segments while electromyography of four lower limb muscles was registered. Experiments in weightlessness where conducted during European Space Agency parabolic flight campaigns. Control experiments were performed on the same subjects on Earth. When running on the treadmill with a SLS, the bouncing mechanism of running is preserved. Depending on the speed of progression, the ground reaction forces, contact and aerial times, muscular work and bone stress differed by a maximum of ± 5-15% during running on the treadmill with a SLS, as compared to on Earth. The movements of the lower limb segments and the EMG patterns of the lower limb muscles were also comparable. Thus the biomechanics of running on Earth can reasonably be duplicated in weightlessness using a treadmill with a SLS that generates a pull-down force close to body weight on Earth.|
|17/06/2011(14h00) [Lieu : Auditoire Euler]|
Stéphane Berbenni(Univ. P. Verlaine, Metz)
"Approches Micromécaniques et Longueurs Internes en Plasticité des Matériaux Cristallins"
|Mes activités de recherche trouvent comme thème fédérateur le développement d'approches micromécaniques pour plusieurs classes de comportement (élasto-plastique et élasto-visco-plastique) et plusieurs topologies de microstructures principalement dans les matériaux cristallins (métaux), ainsi qu'une extension de ces méthodes vers la prise en compte des longueurs internes de la microstructure dictées par les processus discrets collectifs à l'origine de la déformation plastique, et des effets de taille observés expérimentalement. La démarche scientifique part du développement d'approches microphysiques aux échelles fines pour aller jusqu'aux échelles macroscopiques, par différents types de transitions d'échelle. Je souligne dans ce séminaire à l'attention de l'équipe l'insuffisance des approches dites à champs moyens, notamment parce qu'elles ne prennent pas en compte les longueurs internes inhérentes a la microstructure de déformation des matériaux polycristallins (initiales ou induites par la déformation inélastique). Les conséquences de l'existence de ces longueurs internes sur la formation et l'évolution des contraintes internes, des textures locales dans le matériau, ainsi que de son comportement global sont ainsi analysées. Tout en employant les préceptes et hypothèses de la mécanique des milieux continus, ce travail se démarque des approches dites phénoménologiques, qui ne prennent pas en compte la présence et l'évolution des hétérogénéités physiques de la microstructure. Des applications des techniques présentées pour des matériaux industriels tels qu'aciers, nickel, alliages d'aluminium seront montrées. Quelques perspectives actuelles seront discutées (brièvement).
|17/06/2011(16h30) [Lieu : MORE51]|
"Micro-macro modeling of debonding in inclusion reinforced elastic
|Mechanical behavior of inclusion-reinforced composites is affected by
the nature of the bond between the constituents. Imperfect bonding may
result either from the manufacturing process or from damage development
when the composite is loaded. For linear elastic materials, it lowers
the effective elastic modulus. For non-linear (e.g. plastic) material
models, debonding between harder and softer phases may be at the origin
of micro cracking, shear banding, or ductile damage.
The first part of the thesis proposes particular solutions allowing the
determination of the effective properties of the composite for
particular cases of loading, such as the deviatoric tension loading. The
solution is based on the theory of linear elasticity coupled with mean
field homogenization schemes.
The second and main part of the thesis is based on the generalization of
the concept of Eshelbys solution, to account for nonlinear interface
debonding between the constituents of a composite material. The solution
is incremental and it can work with any kind of cohesive law and under
any kind of external loading. Two simplifying hypotheses are adopted.
The stress field inside the spherical particle as well as the compliance
across the interface is considered to be uniform. The Mori-Tanaka method
is extended based on the generalization of Eshelbys solution to account
for the effect of nonlinear interface debonding. The extended Mori
Tanaka method is able to predict the macroscopic mechanical behavior of
a composite material with any kind of cohesive law and external loading.
|17/06/2011(13h00) [Lieu : Reaumur D 382]|
"Low temperature transformations in Ti-5553 alloy : a way of improvement"
|06/06/2011(16h00) [Lieu : Passelecq - Vinci]|
Octavio Sequeiros(Shell International Exploration and Production, Rijswijk, The Netherlands)
"Turbidity currents and the Oil Industry"
|Turbidity currents present a two-fold challenge for the oil industry.
They represent at the same time a ressource and a threat. On the one
hand many of the subsea oil and gas reservoirs are entraped in
turbidites deposits accumulated by past turbidity currents over
geological time scales off the continental shelves of major estuaries.
On the other hand, present-day turbidity currents running down submarine
canyons and channels pose an economic and environmental hazard for
submarine equipment, cables and pipelines. At both scales, the industry
is crucially interested in developing a better understanding of the flow
dynamics of turbidity currents, their interactions with the seabed or
offshore structures, and their depositional signatures and related
stratigraphy. After a general introduction to sediment-laden gravity
flows, the presentation will offer a brief glimpse on the two challenges
from both an engineering and a geological perspective|
|30/05/2011(16h15) [Lieu : Barb 91]|
"Combustion in Homogeneous Charge Compression Ignition engines:
experimental analysis using ethyl esters and development of a method to
include detailed chemistry in numerical simulations."
|In the global energy context, many solutions are suggested to change our
energy dependence and solve our CO2 emission issues. For transportation,
biofuels are mentioned as a possible alternative to fossil fuels. Yet many people
argue that they could hardly replace oil. The low efficiency of the conversion
from biomass to biofuels implies other main issues like land use and food
competition. More simple conversion routes, using less energy-consuming
steps, could be used. However, the major drawback is then the lack of
compatibility with conventional engines. Instead of adapting the fuel to the
engines, the opposite has been put forward recently with the homogeneous
charge compression ignition (HCCI). This engine does not intrinsically rely on
specific fuel properties. Given that a fairly homogeneous air-fuel mixture is
prepared, it can be operated on a large range of fuels.|
In the first part of this thesis, we analyzed the esters produced from fermentable
wastes by a biochemical process. We investigated experimentally the impact of
these esters on the HCCI.
In practice, the development of new engine concepts and the characterization of
fuels rely more and more on numerical simulations. However, the accurate and
comprehensive modelling of the highly nonlinear combustion processes of
realistic fuels is extremely demanding. It requires detailed mechanisms that
involve hundreds of chemical species and thousands of elementary reactions.
The computational cost of these simulations is generally prohibitive. This
implies, first, that fairly detailed CFD simulations can only include global
mechanisms, which only describe the main steps of the combustion; and
second, that very detailed reaction mechanisms are mainly limited to simplified
In the second part of this thesis, we developed a new method that significantly
mitigates the computational burden of detailed chemistry in complex CFD
simulations. The developed numerical method, named tabulation of dynamic
adaptive chemistry (TDAC), speeds the CFD simulations up to 900 times with a
very small simulation error compared to the direct numerical integration of the combustion mechanisms
|23/05/2011(16h15) [Lieu : Euler a.002]|
"Finite Element modelling of grain-scale heterogeneities in polycrystalline
|Macroscopic properties of crystalline solids depend inherently on their
underlying microscopic structure. Studying the mechanisms operating at the
microstructural scale during the various thermomechanical processes to which such
materials may be subjected offers a valuable insight into their final in-use
properties. The objective of this work is to investigate grain scale heterogeneities
in polycrystalline aggregates subjected to large strains using the Crystal
Plasticity Finite Element Method (CPFEM). For this purpose, highly resolved
simulations, where each grain is represented explicitly, are needed. The first part
of this work is devoted to a detailed account of the numerical framework implemented
for such simulations. A classical elastic-viscoplastic crystal plasticity model is
combined to a non-linear parallel finite element framework. The discretization of
the digital microstructures is performed using non-conforming unstructured meshes.
Most importantly, a level set approach is used to describe grain boundaries and to
guide an adaptive anisotropic meshing strategy. Automatic remeshing, with
appropriate transport of variables, is introduced in the proposed framework. In the
second part of this work, the robustness and flexibility of our approach is
demonstrated via different CPFEM applications. The deformation energy is used to
assess heterogeneities in polycrystalline aggregates, highlighting the need to
perform adaptive meshing so as to achieve a good compromise between accuracy and
computation time. These grain-scale heterogeneities are to be accurately predicted
during the deformation simulation if subsequent static recrystallization modelling
is to be performed. An example of linking between the deformation and static
recrystallization steps, using the proposed common approach, is illustrated. In
terms of global texture predictions, the CPFEM framework is validated for a highly
resolved model polycrystal subjected to more than 90 % thickness reduction in
rolling. The importance of automatic remeshing in avoiding excessive mesh
distortion, in such applications, is demonstrated. Most importantly, microtexture
analysis is performed on digital microstructures that correspond, in a discrete
sense, to an actual microstructure observed experimentally. Intragranular
misorientation predictions and virtual 2D orientation maps are compared to the
experimental ones, highlighting the difficulties pertaining to the validation of
such grain-scale predictions.|
|18/05/2011(14h30) [Lieu : Auditoire Euler]|
François Auslender(Univ. B. Pascal, Clermont Ferrand)
"Effective and apparent behaviors of random linear composites: new bounds and RVE sizes"
|Improved bounds for the effective behavior of linear elastic matrix-inclusions composites as well as a statistical analysis of their apparent properties are presented. As shown by Huet (J. Mech Phys Solids 1990; 38:813-41), the effective properties of random linear composite are bounded by ensemble averages of the apparent elastic behavior of cubic (in 3D) or square (in 2D) Volume Elements (VE), computed with either affine displacement boundary conditions (BC) or uniform traction BC. These bounds converge to the effective properties when the size of the VE tends to infinity, but often provide almost explicit predictions for domains of finite size. In the case of a large contrast of the constituents (rigid inclusions or pores), the discrepancy between the bounds remains however significant even for large values of the VE size. This occurs because the contribution to the total potential (or complementary) energy of the particles (or pores) which intersect the edges of the VE becomes unphysically very large when uniform BC are directly applied to the particles. To avoid such a limitation, we considerer non regular VEs constructed as the union of the Voronoï cells associated with the centers of the particles contained in the corresponding classical regular VE, each Voronoï cell being composed of an inclusion surrounded by the matrix, thus forbidding any direct application of BC to the particles. By making use of the classical energy bounding theorems, new bounds of the effective behavior are derived. Their application to a two-phase composite composed of an isotropic matrix and aligned identical fibers randomly distributed in the transverse plane leads to sharper bounds which converge quickly with the VE size, even for infinite contrasts. Lastly, a statistical study of the variability of the apparent behavior is performed by means of Monte Carlo simulations and various criteria to define a minimal size
of a Representative Volume Elements are discussed.|
|13/05/2011(16h10) [Lieu : Stevin, B044]|
Nikolaos Prasianakis(Paul Scherrer Institut, Switzerland)
"Development of advanced lattice Boltzmann algorithms for the modeling of processes in power generation systems"
|The study of processes occurring at small scales in various energy conversion systems, such as fuel cells (SOFCs, PEFCs) and micro-combustors, can be aided by the use of advanced lattice Boltzmann models (LB) that can handle heat transfer, mixing effects and reactive flows.
Heat transfer effects can be efficiently studied with a new thermal LB model which can simulate the compressible Navier-Stokes and Fourier equations for large temperature and density variations. For multi-component reactive flows, a LB model that allows the description of various species with arbitrary ranges of molecular weights is used.
Due to its kinetic theory origin, the lattice Boltzmann method proves to be a competitive tool for simulation of micro-flows in complex geometry domains. Simulation results of the flow through porous media, obtained via X-ray tomography, will also be presented. (Anode of a solid-oxide fuel cell; Gas diffusion layer of polymer electrolyte fuel cell).|
|12/05/2011(10h00) [Lieu : Euler a.007]|
Imad Elmahi(Université d'Oujda, Maroc)
"An adaptive finite volume solver for pollutant and sediment
transport in shallow water flows on unstructured meshes: Application
to the strait of Gibraltar and Nador lagoon"
|In this talk we will present and discuss a numerical algorithm based
on a finite volume method, to compute pollutant and sediment transport
by shallow water flows on non-flat topography. The method uses
unstructured meshes, incorporates upwinded numerical fluxes and slope
limiters, and is used in conjonction with a dynamical
refinement-unrefinement mesh procedure to provide sharp resolution of
steep bathymetric gradients that may form in the approximate solution. The
resulting scheme is non-oscillatory and possesses conservation property
that conserves the pollutant mass during the transport process. Numerical
results are presented for some test examples which demonstrate the
accuracy and robustness of the scheme and its applicability
in predicting pollutant transport by shallow water flows. We will also
apply the solver to compute pollutant and sediment transport event in
the Strait of Gibraltar and Nador lagoon.|
|11/05/2011(13h00) [Lieu : Reaumur D 382]|
"Multiscale characterisation of the work hardening mechanism in Fe-Mn based TWIP steels"
|10/05/2011(16h00) [Lieu : SCES 03]|
"Power transfer and vibrator-pile-soil interactions within the framework of vibratory pile driving"
|The vibratory driving technique is used for driving piles and sheet piles into the ground, by imparting to the profile a longitudinal vibratory motion. Performance of the technique results from interactions between the driving unit, the pile and the soil. The objective of the present thesis is to develop a better understanding of those interactions.|
The research is based on a detailed study of experimental results, consisting in full scale vibratory driving tests and experience databases. A large amount of data was available thanks to the partners of this research (BBRI, Arcelor-Mittal, Deltares, LCPC). A new full scale test campaign was also conducted for investigating specific questions such as the influence of vibratory parameters and sheet pile movements on the efficiency of the technique, in terms of both driving time and power consumption. Analysis of the experimental results has allowed some shortcomings in the current methods addressing pile vibrodrivability to be identified. Shortcomings are mainly related to the assumptions regarding the vibratory action actually applied to the pile (force and frequency), the (sheet) pile movement, and the definition of soil parameters.
The second part of the thesis is devoted to the improvement of vibrodriving prediction methods based on theoretical and experimental conclusions. In particular, suggestions are made for correcting the definition of the vibratory force and taking into account the possible lack of power from the power pack; while the definition of the soil parameters is reviewed based on the literature and experimental studies. Also the possibility of extending vibrodriving models for the prediction of power consumption is investigated, as well as the possibility to relax the rigid body assumption through use of the finite element approach.
|06/05/2011(13h00) [Lieu : Reaumur D 382]|
Sébastien Michotte (iMMC)
"L'oxyde de Zinc dopé Aluminium : évolution des contraintes mécaniques associées à son dépôt et exemples d'applications "solaires""
|05/05/2011(16h30) [Lieu : Barb 91]|
"Homogenization of elasto-(visco)plastic composites: history-dependent incremental and variational approaches"
|When subjected to complex loading conditions, elasto-plastic composite materials can demonstrate a complex mechanical behavior. Predicting the effective (or homogenized) response of such materials requires a micro-mechanical approach, in which the individual response of the constituents as well as their topological arrangement are explicitly taken into account. This thesis focuses on mean-field (MF) approaches, which aim to propose semi-analytical constitutive models based on assumed interaction relations between the constituents and making use of a partial description of mechanical fields through their statistical moments (mean and variance) in the phases.
The general objective of the research is to provide a reliable micro-macro constitutive model for elasto-(visco)plastic composites suited for general loading histories and constructed on sound theoretical foundations. Two main research orientations are successively investigated. First, we consider incremental approaches previously developed within the research group at UCL. These methods are based on a linearization of the local constitutive relations over each time interval, so that available models for linear composites can then be applied. Original strategies are proposed to capture field heterogeneities into the modeling. The most significant contribution of the present work is based on a radically different approach. It relies on a variational formulation of the homogenization problem according to which the macroscopic stress-strain relation derives from an effective incremental potential. The predictions of the different models are verified against reference, FE results obtained on representative volume elements of the microstructure of several two-phase composites.
|04/05/2011(16h15) [Lieu : Sud 11]|
"Vortex particle-mesh method with combined immersed boundary and mesh refinement techniques. Application to bluff-body and wake-vortex flows"
|The present work joins a global effort in the development of efficient and accurate numerical tools for the simulation of fluid flows. More specifically, we focused on external flows, a fluid dynamics discipline which actually pervades applied sciences and engineering. These are the flow encountered in aircraft or car aerodynamics, wind energy,... or biological locomotion. These are only a few examples of the class of problems this work intends to address. In our problems, we will assume the flow to be incompressible and the inertial forces to dominate the viscous stresses: we are dealing with moderate and high Reynolds number flows.|
Furthermore, the present work treats the flow equations in a very distinct approach. We use vortex particle-mesh method (VPM). This class of methods rather uses the vorticity formulation of the Navier-Stokes equations: the vorticity field is the primary variable rather than the velocity and pressure fields. The major advantage comes from the compactness of the vorticity field for some flows of interest (e.g., external flows). A limited number of particles are then required to discretized the entire flow. Moreover, VPM can intrinsically take into account unbounded-flow boundary conditions. First an immersed boundary technique has been developed and adapted to capture the flow past arbitrary shape bodies. The motivation is to benefit from the enhanced efficiency provided by the vortex particle-mesh method in terms of computational cost. The required Poisson equation is solved using an efficient grid-based solver together with the parallel fast multipole method.
Second, an accurate approach handling hierarchically refined meshes has been developed. We use vortex particle patches of varying resolution. The originality of this work is the combination of the handling of multiple flow resolutions together with a full 3-D Navier-Stokes solver for unbounded incompressible flows. This method also benefit from the versatility of the parallel fast multipole method in order to solve the Poisson equation and the associated straightforward domain decomposition. This mesh refinement technique has been also used in the context of Large Eddy Simulation of a turbulent vortex wake rollup.
|29/04/2011(13h00) [Lieu : Reaumur D 382]|
Sampath Kumar Yerra(iMMC)
"On damage nucleation in duplex stainless steel"
|08/04/2011(13h00) [Lieu : Reaumur D 382]|
"L'acier et la Belgique"
|01/04/2011(13h00) [Lieu : Reaumur D 382]|
"The new instrument of our lab: the nanoindenter/nanoscratch"
|23/03/2011(10h30) [Lieu : Stevin B.044]|
Nicola Vitiello(Scuola Superiore Sant'Anna, Italy)
neuro-robotics paradigm: NEURArm and NEUROExos"
|The neuro-robotics paradigm is a design approach to develop robotic
systems to be used in functional support, personal assistance and
neuro-rehabilitation. While usually the robotic device is considered as
a "tool" for neuroscientific studies, a breakthrough is obtained if the
two scientific competences and methodologies converge to develop
innovative platforms to go beyond robotics by including novel models to
design better robots. Two case-study platforms will be described:
NEURArm, an anthropomorphic robotic arm with a bio-mimetic compliant
actuation system, and NEUROExos, a variable impedance elbow exoskeleton
for motion assistance and rehabilitation|
|22/03/2011(10h30) [Lieu : Stevin B044]|
Marko Munih(University of Ljubljana, Slovenia)
"Kinematic, static and psychophysiological measurements in
|Various techniques today allow for kinematic measurements. When the
variables, e.g. joint angles are not directly available, the sensory
fusion methods may assist. The interaction forces and moments are
further required for estimation of moments and forces that move the
limbs and on time axis appear before the change of kinematic variables.
Both, kinematic and static interaction forces are required for
calculating variables that can not be measured directly, for example
Psychophysiological measurements is another layer of sensing that
enables rough insight into valence and arousal of the person. With this
the bio-cooperative control may be realized. The interaction becomes
bi-directional and the technical rehabilitation system takes into
account the user's properties, intentions and actions, as well as
environmental factors. For example, an actuated orthoses should be able
to detect the patient's effort and engagement in order to optimize
participation and support the patient only as little as needed|
|18/03/2011(16:00) [Lieu : Bâtiment Stevin, Salle des Séminaires b.044]|
"Thermodynamics of granular materials: towards a consistent continuum theory for partially ordered systems."
|Granular materials, which are defined as large conglomerates of discrete particles, are the second most employed material in industry, the first one being water. These conglomerates are typically immersed in a carrier fluid, thus forming a heterogeneous, two-phase mixture. Due to their micro-structure, these mixtures do not exhibit scale segregation and, therefore, constitute a partially ordered system. In fact, flows of such mixtures are different from flows of simple fluids in many important ways. For example, these mixtures support shear stresses at equilibrium, they compact, they tend to clog when driven through area constrictions, they can produce avalanches, etc.
In the first part of this talk we will introduce a thermodynamic framework for the study of the mixtures of interest, according to which the constituents of the mixture are treated as open and continuum subsystems that interact with each other. We will also show how this framework can be used for the systematic reduction of two-phase flows to the incompressible limit, via a generalization of the low-Mach number approximation. The second part of this lecture is devoted to the presentation of a comprehensive mathematical theory and numerical results for granular materials at equilibrium.|
|01/03/2011(16h00) [Lieu : Auditorium Sud11]|
"Finite Element Methods for Coastal Flows: Application to the Great Barrier Reef"
|Traditional ocean models are based on finite difference methods and solve the governing equations on structured grids. This research is part of the Second Louvain-la-Neuve Ice-Ocean Model (SLIM) whose objective is to bring to oceanography the geometrical flexibility of unstructured grids. This model is based on the finite element method which is widely used in engineering analysis. This thesis presents the development of SLIM for a coastal hydrodynamic application. The first step consists in generating unstructured meshes specifically adjusted to the needs of coastal or oceanic applications. We then compare five finite element pairs in term of both their stability and their accuracy for solving the shallow water equations. Finally, an optimized implementation is developed to reduce the time spent in the spatial integrals required by the finite element formulation.
As the worlds largest reef network, the Great Barrier Reef constitutes a unique ecosystem threatened by climate change, ocean acidification and agriculture. Finite elements are well designed for dealing with the Great Barrier intrinsic multi-scale features and complex topology. With a resolution up to 100 m near islands and reefs, our model is the first able to capture small scale features like tidal jet and eddies in the wake of islands in the same simulation as the global circulation on the continental shelf. Our model is used for several studies on the Great Barrier Reef with the objective to understand complex ecosystems and predict the human influence on their evolution.|
|07/02/2011(9h00) [Lieu : Stevin B.044]|
Sandra Soares Frazao(iMMC)
"Challenges in flood modelling"
|Considering that nowadays about one third of the natural disasters
in the world are due to floods, mitigation of their consequences is becoming
a major concern for public authorities. Flood management strategies are
usually based on simulation results for different scenarios, considering
both extreme meteorological events and potential failure of man-made
structures and flood-protection devices. In-depth knowledge and
understanding of the physical processes underlying flood propagation are
thus mandatory to obtain reliable modelling results but also an accurate
picture of the associated uncertainty. Indeed, the level of uncertainty
might be very high in cases of complex natural environment. Defining
adequate simplifications and appropriate modelling options thus constitutes
one of the key challenges of efficient flood modelling. This will be
illustrated by means of specific examples related to (i) the choice between
one- or two-dimensional modelling, (ii) complex topographies either natural
or man-made, and (iii) the presence of sediments and the associated
|07/02/2011(14h00) [Lieu : Stevin B.044]|
"Water and sediment in interaction: which one is moving the other?
Experiments and numerical simulations of geomorphic flows, from the
morphological changes triggered by dam-break floods to the sculpting of
submarine canyons by turbidity currents"
|Sediment-laden flows occur in a variety of contexts of interest to
the civil engineer. The fluid and solid phases in presence interact with
each other and with the sediment bed through a range of processes, to such
an extent that it is not even trivial to answer the simple question of which
one is moving the other.|
At first, one will focus on the fluvial setting, and investigate
catastrophic floods that mobilize vast amounts of sediment and trigger
extensive morphological changes. Impacts of climate changes, the alteration
of hydrological regimes of rivers, and the growing pressure of anthropogenic
activities, makes this field of non-classical floods a very active and
timely field of research. One such example is the flood resulting from the
sudden collapse of a dam in an alluvial valley, a study case now becoming a
landmark for the scientific community trying to extend sediment transport
theories, originally developed for steady uniform flow with moderate
transport rates, to very transient flow with intense bedload.
We present a series of idealized laboratory dam-break experiments, whose
observations initially guided the formulation of an innovative two-layer
shallow-water framework for sediment transport, accounting for sediment
inertia and the effects of granular dilatancy. The model was found robust,
yet still relies on empirical coefficients uneasily calibrated for transient
flow. More recent developments with a newly proposed imaging technique to
measure vertical profiles of velocity and sediment concentration, guided the
development of a surprisingly powerful, parameter-free theory that views the
bedload layer as a fluid-granular shear layer and imports contributions from
the realm of stratified density currents and dry granular flows. Initially
developed for steady conditions, it predicts the complete internal dynamics
of intense bedload layers and is now being extended and validated on the
Secondly, one will leave the alluvial setting and plunge underwater to see
that solid-liquid flows active in the submarine setting present striking
similarities with river flows, despite one essential difference: if water
drives the sediment along rivers, it is the (suspended) sediment that drives
the water along steep submarine canyons to trigger strong currents named
"turbidity currents" or "subaqueous debris flows". By sculpting submarine
canyons and deep-sea fans, producing oil-rich deep-sea deposits, or
endangering the stability of offshore structures, cables and pipelines,
these flows create a rich environment for research topics of engineering
significance. We present a series of illustrative experiments aimed at
understanding how such flows develop in the downstream direction and
stratify in the vertical direction, and move sediment in suspension or as
bedload. We develop a shallow-water description of long-standing turbidity
currents and illustrate its potential to reproduce observed submarine
|25/01/2011(13h à 14h) [Lieu : Salle Passelecq (bâtiment Vinci)]|
"Two-dimensional two-layer shallow-water model for river flows with
significant sediment transport"
|La dynamique des écoulements en rivière et le transport de sédiments sont intimement
liés. Leur interaction régit une grande variété découlements pouvant conduire à des
modifications considérables de la morphologie des rivières, des ondes consécutives à une
rupture de barrage aux écoulements dans un canal présentant une variation brusque de sa
géométrie. Afin de prédire les interactions entre lérosion du lit, les dépôts de sédiments
et le courant, le transport des sédiments doit être décrit attentivement.
Le modèle bidimensionnel (2D) dont il sera question lors de ce midi de la recherche est
basé sur lhypothèse deau peu profonde et propose une représentation de lécoulement
sous forme de deux couches. La couche supérieure est faite deau claire alors que la
couche inférieure, appelée couche de transport par charriage, est constituée dun mélange
dense deau et de grains mobiles. Ces couches reposent sur un lit statique. Chaque couche
présente une vitesse moyennée sur son épaisseur et une densité distinctes de celles de
lautre couche. La continuité et la conservation de la quantité de mouvement sont
imposées sur chaque couche. Le modèle tient compte (1) de lentrainement des grains au
travers de linterface du lit et (2) des échanges de quantité de mouvement entre les
couches grâce à la définition dun taux dérosion, qui dépend lui-même des contraintes de
cisaillement appliquées sur chaque face des interfaces entre couches. Les contraintes de
cisaillement sont écrites sous la forme dexpressions de Chézy. Lhypothèse de
concentration granulaire distincte dans la couche de transport et dans le lit rend possible
la dilatation de la couche de transport lors de lérosion du lit.
Le système de 7 équations à 7 inconnues est résolu numériquement par la méthode des
volumes finis sur un maillage triangulaire non-structuré. La résolution numérique ne sera
pas détaillée lors de cette présentation.
Le modèle numérique est confronté à différents cas de rupture de barrage et découlement
permanent sur fond mobile dans des configurations délargissement ou de rétrécissement.
Les résultats des simulations sont comparés à des mesures réalisées sur des canaux du
|22/12/2010(14h) [Lieu : Euler-002]|
Nicolas Moës(Ecole Centrale de Nantes, Institut GeM, UMR CNRS 6183.)
"Mécanique configurationelle et Modélisation par level set
des surfaces de discontinuité :
un couple théorique naturel dans le cadre numérique X-FEM
|Nous ferons tout d'abord un rappel rapide de la mécanique configurationelle qui traite des changements d'énergie dans les structures suite à des changements de configuration. Ces configurations sont généralement définies par des surfaces de discontinuités (fissures, interface matériaux, front de changement de phase,
). Grâce à l'approche X-FEM, le fait que ces surfaces ne soient pas maillées n'est pas gênant pour le calcul. Ces surfaces peuvent être localisées sur un maillage quelconque par une level set.
Cela permet de mettre en place sur maillage fixe ou de type adaptatif octree des algorithmes puissants de résolution de problèmes où l'inconnue principale est la position d'une surface de discontinuité représentée par une level set.
Les applications visées sont le délaminage des composites, la résolution de problèmes d'inéquations variationelles et l'iniation de fissures par la mécanique de l'endommagement. Des résultats numériques illustrerons la pertinence des approches proposées.
|16/12/2010(14h) [Lieu : Euler-002]|
Paul-Emile Bernard(Ecole Centrale de Nantes, Institut GeM, UMR CNRS 6183.)
"The Thick Level Set (TLS) model for damage growth modeling"
|The extended finite element method (X-FEM) is successfully used since a decade to represent cracks propagation using non-conforming meshes.
However, it is now commonly admitted that fracture mechanics alone may not model the full scenario of the degradation of solids under mechanical loading. Crack initiation for instance requires damage mechanics. Moreover, some numerical questions remain about the X-FEM method, as effective approaches to represent the cracks branching or merging processes.
In this framework, we developped the Thick Level Set (TLS) approach. The main idea is to represent the damage as a direct function of a level set. Many defects can then be represented with only one level set. A direct consequence is that, because of the properties of the signed distance function, we obtain an efficient non-local damage model, mandatory to avoid spurious localization and mesh dependance. The model is validated using both classical benchmarks and more complex test cases including crack branching and merging.
|16/12/2010(13h à 14h) [Lieu : Salle Passelecq (bâtiment Vinci)]|
Jean Shimatu(Université catholique de Louvain)
"Caractérisation des sols argileux en vue de leur valorisation comme matière première dans la fabrication des briques cuites en République Démocratique du Congo"
|Le Kasaï Oriental est une province située au centre de la République Démocratique du Congo et qui est devenu célèbre par ses ressources en diamants. Il a une superficie de près de six fois la Belgique pour huit millions dhabitants.
Depuis des décennies, cette province connait les prix les plus élevés en termes de construction. La raison majeure étant limportation de la quasi-totalité des produits de base (ciment, chaux, tuiles, tôles,
). Il a donc été nécessaire détudier les différentes possibilités de production de matériaux de construction à partir des produits locaux. Lune des pistes explorées est la brique de terre cuite, un produit à base de sols argileux.
Lors de cet exposé, nous présentons les résultats de recherche de sols argileux dans une province où 85% de sols sont sableux. Les différents échantillons de sols prélevés dans les deux grandes villes de la province ont été soumis à plusieurs analyses physico-chimiques et montrent quune partie dentre eux sont susceptibles dêtre valorisés comme produits de base de la brique cuite.|
|08/12/2010(16h30) [Lieu : Auditorium Sces 1]|
"Modélisation et Optimisation d'actionneurs pieézoeélectriques linéaires à onde progressive."
|Dans le cas des moteurs linéaires, la génération d'une onde progressive basée sur une technique similaire est beaucoup plus complexe à mettre en oeuvre et ne permet pas de créer une onde de flexion exempte d'une composante stationnaire qui, elle, est dépourvue de propriété d'entraînement.
Nous avons dès lors développé des modèles analytique et numérique qui ont permis de déterminer les conditions d'excitation (la fréquence et la position des excitateurs) qui maximisent la proportion d'onde progressive par un processus d'optimisation basé sur un critère original d'analyse de la qualité de l'onde de flexion. Nous avons, ensuite, validé expérimentalement ce critère sur un prototype de stator de moteur linéaire à onde progressive.
Nous avons, par ailleurs, utilisé les compétences du CEREM en matière de modélisation de systèmes multi-physiques par une approche multicorps pour construire un modèle complet du moteur car, outre les avantages qu'elle offre en matière de rapidité de calcul, cette approche se prête particulièrement bien à la modélisation de mécanismes d'entraînement par friction et à l'intégration des excitateurs piézoélectriques dans le modèle.|
|17/11/2010(13h00) [Lieu : Passelecq - Vinci]|
Sylvie Van Emelen et Christophe Petre(iMMC)
"La Meuse, de la modélisation des crues à un modèle de gestion des voies navigables."
|Depuis plus de dix ans, le secteur "Hydraulique" s'intéresse aux écoulements de la Meuse pour le compte du Sethy (Service d'Etudes Hydrologiques du Service Public de Wallonie). br>
Le but premier de ces conventions de recherche était la modélisation des crues de la Haute Meuse et de ses effets à court terme, tout en considérant que les barrages étaient effacés. Au fil des ans, le modèle s'est étoffé : utilisation d'un système d'information géographique (GIS) pour le calcul des profils de sections, extension du modèle à la Meuse Moyenne, calcul et visualisation sur cartes des zones inondées (GIS), intégration du comportement des barrages, prise en compte des centrales hydroélectriques, intégration des futurs algorithmes de régulation des automates de barrages,
La première partie de ce midi de la recherche sera consacrée à la description du modèle de calcul des écoulements de la Meuse : Hydroaxe. Nous présenterons ensuite les différents travaux en cours : maintenance du modèle et nouveaux développements, ou encore, prolongement du réseau jusqu'à Monsin, en aval de Liège.|
|09-10/11/2010(09:00) [Lieu : Auditoire SUD01]|
IMCN & iMMC(UCL)
"Journées microscopie électronique à transmission"
|Mardi 9 Novembre :|
09h00 - 12h00 : Les bases de lam icroscopie électronique à transmission by Dr. Laurent Legras (EDF R&D, Moret-sur-Loing)
13h30 - 15h00 : Analyse EELS (Electron Energy Loss Spectroscopy) by Pr. Karine Varlot-Masenelli (INSA, Lyon)
15h00 - 16h00 : FIB (Focused Ion Beam), faisceau d'ions focalisés by Dr. Laurent Legras (EDF R&D, Moret-sur-Loing)
16h00 - 16h30 : Pause-café
16h30 - 18h00 : Sessions posters + stands industriels
Mercredi 10 Novembre :
09h00 - 10h30 : Diffraction by Pr. Jean-Paul Morniroli (Université Lille 1)
10h30 - 10h45 : Pause-café
10h45 - 11h45 : L'ultramicrotomie en
science des matériaux. by Pr. Jean-Michel Gloaguen (Université Lille 1)
13h15 - 14h15 : Sessions posters + stands industriels
14h15 - 15h15 : Microanalyse X by Dr Jacky Ruste (EDF R&D, Moret sur Loing)
15h15 - 16h15: Cryo TEM Gatan, Inc.
|09/11/2010(14h00) [Lieu : Barb 93]|
Yannick Champion(Institut de Chimie et des Matériaux Paris-Est, CNRS-Université Paris Est)
"A comprehensive dislocations model to explain deformation and ductility evolution in metals with small grain size."
|Observation of near perfect elasto-plasticity for copper with average grain size of 100 nm (prepared by powder metallurgy), in tension and with 12% of deformation, rapidly led to the conclusion that metals become strain rate sensitive at small grain size. According to the Art's criterion, increase in strain rate sensitivity induces delay in necking and then plasticity. From this rheological parameter, one main draw a "macroscopic" activation volume which gives information in the microscopic events controlling the deformation. From transmission electron microscopy observations in high resolution mode and quantitative analysis, one have been able, with Yves Bréchet (SIMAP-CNRS INP Grenoble) to put forward that dislocations-grain boundaries interactions and grain boundary sliding are the relevant elements of a micro-mechanism for plastic deformation. Hence, a comprehensive model was proposed thanks to a modified Orowan equation and tested with respect to the variation of the activation volume as a function of stress.
Experimental measurements of the activation volume was performed between room temperature and 393K, at 243K and 77K for a copper nanostructure with a grain size of 100 nm. A constant activation volume is observed at low stress (or high temperature) followed by a linear increase of activation volume with stress (inverse Cottrell-Stokes behavior) and ends with a inverse stress dependence (Cottrell-Stokes behavior) at large stress (or low temperatures). The model is able to fit the whole experimental behaviors.|
|02/11/2010(16:00) [Lieu : Barb 94]|
"Dynamic Behavior of Vibrated Dry Sand : Sphere Penetration Experiments and Discrete Element Modeling of Vibro-fluidization"
|The present research focuses on the characterization of the dynamic properties of vibrated dry sand. The interest for this engineering issue comes from the study of vibrodriving and vibrocompaction processes applied to granular soils.Original experiments able to characterize the behavior of dry sand subjected to vertical vibration are first presented. When cohesionless soil placed in a cylindrical container is vertically vibrated under the gravitational field, three different types of dynamic behavior are distinguished, depending on the acceleration amplitude (a): the densification behavior (a/g < 1), the instability surface behavior (a/g ? 1), and the vibro-fluid behavior (a/g > 1). In the densification range, the sand simply settles. When the acceleration is increased beyond 1g, granular convection is observed and there is an instability in the sand mass leading to the emergence of an inclined free surface. If the acceleration is further increased, the sand becomes fully vibro-fluidized.Results of sphere penetration experiments (SPE's) are also presented. The sinking logs of the SPE's conducted at UCL do not follow Stokes law, as initially suggested by the 1962 Barkan experiments, but depend on the sand dynamic regime. Hence, the concept of vibro-viscosity, proposed by Barkan (1962), is to be reappraised in our opinion. Discrete Element Modeling (DEM) of vibrated dry granular assembly is performed to track down the fundamental origin of vibro-fluidization. At low acceleration amplitude (a/g = 0.5), the numerical sample behaves like an elastic solid. But, once the acceleration amplitude is set to 1.02g, cyclic vibro-fluidization is noticed and the emergence of "fluid" stresses is observed, underscoring the dual nature of vibrated dry cohesionless soil. Hence, vibrations result in a particular state of granular matter. The "vibro-fluid" state can be characterized by a phase transition. There is a change in the structure of the granular material: the transformation of a solidlike material into a complex fluid identified by an abrupt change in its properties resulting in its shear strength degradation.|
|02/11/2010(17:00) [Lieu : Barb 93]|
"Multiphysics Modelling of Multibody Systems : application to Railway Pneumatic Suspensions"
|For several decades, Computer Aided Engineering (CAE) has been widely used in many fields such as fluid mechanics, structural dynamics, control, etc. Within this context, multibody dynamics (MBS) deals with systems composed of several rigid or flexible bodies which are connected by joints and interact with each other and with their environment via internal or external forces. Multibody dynamics now covers a wide range of applications, from robotics to the human body, and is especially suitable for the analysis of road and rail vehicles. Concerning the latter, a noticeable evolution has followed the use of pneumatic suspensions which replace conventional helical springs by air chambers. Such a system allows to adapt the suspension height and stiffness to a varying payload and provides better comfort and sound insulation. However, in order to achieve these interesting properties, the air suspension involves a complete pneumatic circuit composed of auxiliary tanks, pipes, restriction orifices and several valves. These various components can be combined in many ways leading to an increasing design complexity. From an industrial point of view, resorting to a simulation tool could be very beneficial in designing more efficient suspensions.
The present thesis deals with the modelling of coupled multibody and pneumatic systems with application to railway pneumatic suspensions. The goal is to be able to take the complete pneumatic circuit into account and analyse its interactions with the other parts of the vehicle. It is first shown how the air suspension circuit can be suitably modelled using and energy balance and mass flow formulation. We emphasize how the circuit configuration can influence the choice of a model.
Then, experiments carried out on a real air spring are presented. We explain how the model parameters can be derived and the heat transfer between the suspension and the atmosphere is addressed. Dynamic tests are then analysed, showing a good match between experimental and simulation results.
Finally, a model of a metro car and its pneumatic suspension is implemented. The overall vehicle performances are thus analysed for tests such as curve passing or rail twist in which components like pneumatic valves play an important role. Finally, thanks to the developed approach, novel suspension morphologies inspired from automotive suspensions are explored and compared to well-established suspension designs.|
|29/10/2010(16:00) [Lieu : Barb 93]|
"Conception d'un robot interactif pour la rééducation des membres supérieurs de patients cérébrolésés"
|Many areas of medicine perceive in the progress of robotics, data processing and in the field of multimedia, an opportunity of developing robotic assistance tools to improve the therapist working conditions and to provide more efficient and better care to patients. Physical and rehabilitation medicine, being aware of the above scientific advancement, has developed upper limbs rehabilitation robots for hemiplegic patients. However, existing solutions show significant gaps in terms of mobility, ergonomics and safety.
Based on this observation, the present thesis aims to develop a solution that meets these high level criteria. It presents an original design approach, based on preliminary choices offering a significant advantage : the ability to decouple the electromechanical robot structure aspects from those related to its control.
In terms of structure, the proposed solution consists of five modules governing the movements of the shoulder, the elbow and the wrist, and operating on an original action principle which, contrary to classical exoskeletons, can clearly disregard the problem related to the alignment of the robot and patient degrees of freedom.
In terms of control, the proposed solution uses an admittance virtual system to reproduce a compliant behaviour, like a real therapist, through different operating modes. This control process is validated in laboratory and hospital by implementing it on a planar end-effector robot.|
|20/10/2010(13:00) [Lieu : Passeleq - Vinci]|
Adolphe Tchéhouali(Laboratoire d'Etude en Mécanique Appliquée - Ecole Polytechnique d'Abomey Calavi, Université d'Abomey Calavi, Bénin)
"Mise au point et caractérisation de micro béton contenant des granulats de verre recyclé au Bénin (Afrique de l'Ouest)"
|Le problème environnemental que posent les déchets non biodégradables tels que les bouteilles non réutilisables devient de plus en plus préoccupant au Bénin (Afrique de lOuest) au regard des quantités énormes produites dans les grandes villes. L'une des rares voies de recyclage de ces déchets est de les stocker dans les procédés de construction. Dans cette étude, les granulats de verre issus du broyage de déchets de bouteilles ont été introduits dans le micro béton en substitution au sable fin dans des proportions allant de 5% à 30%. Les résultats d'essais mécaniques ont montré qu'il y a un accroissement substantiel (de l'ordre de 50%) des résistances de flexion est de compression. Cet accroissement est proportionnel au taux de granulats de verre introduit. Les expériences en cours, visant à optimiser les paramètres clé de dosage (granulométrie, proportion de granulat de verre, rapport E/C, type de ciment, type d'adjuvant) permettront d'accroître la performance du matériau et déboucheront sur des applications de plus en plus intéressantes (réalisation d'une variété d'éléments de construction à sections réduites mais très résistants tels que les tuiles, toitures en dalles minces et légères, les tuyaux, les buses, les pavés etc.)|
|07/10/2010(13:00) [Lieu : Passelecq (Vinci)]|
Mounir Bouassida(Unité de recherche ingénierie géotechnique de l'Ecole Nationale d'Ingénieurs de Tunis)
"Etude numérique pour la prédiction de l'amélioration d'une argile molle suite à la mise en place de colonnes ballastée."
|L'amélioration de sols mous par la technique des colonnes ballastées se fait par incorporation d'un matériau d'apport qui augmente la capacité portante, réduit le tassement et accélère la consolidation. En outre, le procédé de mise en place par expansion latérale du matériau d'apport engendre une consolidation du sol mou avant le chargement dû à l'ouvrage. Ainsi une amélioration des caractéristiques du sol mou a bien lieu, elle a été confirmée à partir d'essais in situ réalisés avant et après la mise des colonnes. Dans le cadre d'une action
recherche-développement une analyse numérique a été lancée depuis 2001
pour prédire l'amélioration du module de déformation d'un sol mou suite à la mise en place de colonnes ballastées. La conférence retrace les étapes et expose les résultats de l'action, intentée entre l'unité de recherche ingénierie géotechnique de l'Ecole Nationale d'Ingénieurs de Tunis et du groupe Vibroflotation (Marseille, France), concernant le thème indiqué.
|17/09/2010(14h00) [Lieu : Barb 93]|
"Discontinuous finite element methods for two - and three - dimensional marine flows "
|Numerical modeling is now, along with experiment and theory, part of the scientific method. Numerical models for marine flows exist for more than forty years. These models have evolved dramatically, with improved numerical methods, and much more accurate modeling of unresolved phenomena. However, most of the mainstream marine models still rely on the old numerical paradigm, based on finite difference methods and structured grids.
The development of new numerical models, based on state-of-the art numerical methods on unstructured grids, is now an area of active research.
Those unstructured meshes allow a faithful representation of the coastlines and the choice of the resolution following guidelines from the physics and not from the numerics.
This thesis fits within this research. It is part of the development of SLIM, the Second-generation Louvain-la-neuve Ice-ocean Model. This model uses finite element methods on unstructured meshes made up of triangles for two-dimensional modeling, and triangular prisms for three-dimensional modeling.
Numerical aspects are considered in this work. First, a comparison of several finite-element discretizations of the two-dimensional shallow water equations is performed. Second, to handle flows on the sphere, a novel algorithm is described, based on local coordinate systems, that allows a discretization free from singularities.
Finally, a prototype three-dimensional baroclinic model is presented. The model features a spatial discretization built upon discontinuous finite elements and a time integration performed with an implicit mode splitting|
|10/06/2010(11h00) [Lieu : Euler A 002]|
Eric Wolanski(James Cook University & Australian Institute of Marine Science
"The importance of the biology in estuarine fine sediment dynamics"
|An increase in human activities in river catchments is resulting in increased muddiness of estuaries in many
regions of the world. This increased muddiness has management implications, such as
-increased turbidity and decreased quality of life for the human population,
-storing pollutants (e.g. heavy metals) for decades to centuries and making them available if the mud is eroded
away or dredged,
-changing some coasts from sandy to muddy, which is a significant and usually permanent environmental
Over the last few decades, much of the knowledge of fine sediment (mud) dynamics, and its modelling, came
from the engineering community. Modelling mud dynamics by engineers seemed so 'simple', the belief was
that mud was just a messy fluid and that its behaviour could be modelled by adding a few equations for erosion
and deposition, with the parameters derived from laboratory experiments and the models could then easily be
'fine-tuned' against some field data for the collection of which ingenious probes were designed.
To everyone's surprise, these engineering models were unable to reproduce much of the field observations for
muddy estuaries. A reason for this failure is that these models largely neglect the biology, which has a major
influence in controlling (1)the settling of mud flocs (2)the resuspension of settled mud (3)the dewatering (consolidation) (4)the patchiness (5) the nutrient dynamics
This leads now to view mud and muddy waters as a living body, not just a messy fluid as originally seen by
engineers. This enables us to highlight research priorities to advance the knowledge of mud dynamics by
quantifying the physics-biology links.|
|09/06/2010(11h00) [Lieu : Euler A 002]|
Eric Wolanski(James Cook University & Australian Institute of Marine Science
"The LOICZ biochemical model for turbid estuaries"
|The LOICZ biochemical model is a useful tool to understand where the nutrients go when reaching an estuary,
are they flushed to sea or are they causing eutrophication of the estuary or coastal waters. The model is simple
to use, yet realistic, and has been used widely worldwide. To use the model, some field data are needed on the
inflows and outflows of water and nutrients from the estuary. Such data are relatively easy and inexpensive to
obtain. The model has problems however in turbid estuaries, i.e. where the suspended fine sediment (mud)
concentration (SSC) is high. The problem is not with the reduction of light available for photosynthesis, since
this is already incorporated in the model by assimilating field data on the status of nutrients and chlorophyll-a,
instead the problem is due to the fine sediment absorbing dissolved nutrients in particulate forms in the
turbidity maximum zone of tidal estuaries (thus behaving like a sponge inhibiting eutrophication) or desorbing
nutrients in quasi tideless coastal lagoons (thus facilitating eutrophication). The new LOICZ model for turbid
estuaries, produced by Eric Wolanski and Dennis Swaney, incorporates this process through a partitioning
coefficient Kd, which can also be readily calculated from a few additional field data. The model has been
applied to four turbid Indian estuaries. The results will be discussed.|
|08/06/2010(11h00) [Lieu : Euler A 002]|
Eric Wolanski(James Cook University & Australian Institute of Marine Science
"Box modelling in estuarine ecohydrology"
|Box models are used by estuarine ecosystem modellers who simplify the ecosystem by amalgamating species at
the same trophic level or habitats in order to assess the throughflows of material or energy in the ecosystem.
Two examples are given. The first example refers to the marine environments in Tokyo, Shanghai, Pearl
Harbor, the Pearl Estuary, Hong Kong, Manilla, Bangkok, Ho Chi Minh City, Klang, Singapore, Jakarta, and
Darwin. Not one of these harbours was developed in an ecologically sustainable way, with environmental
impacts ranging from degradation to ecosystem collapse. In some cases development continues to occur at the
cost of environmental degradation; in others rehabilitation of the marine ecosystem is vigorously addressed
following increased wealth, though the ecosystems remain severely depleted. The key mechanisms preventing
ecosystem recovery in wealthy countries' harbours appear to be habitat loss, habitat degradation and habitat
modification. Currently there is no successful world's best practice for environmental management of harbours,
as these 12 case studies demonstrate. Box models can be used to evaluate which ecosystems could be recovered
by realistic remediation measures and which one cannot be recovered. The second example refers to the
Guadiana Estuary in Portugal, where the estuary is impacted by flow regulation at the Alqueva dam. An
estuarine ecosystem model is used to assess the effectiveness of various flow regulation regimes in recovering
or maintaining the health of the estuary.|
|01-03-08-10/06/2010(9h-12h) [Lieu : Reaumur d382]|
Jean-Michel Missiaen(Institut Polytechnique de Grenoble)
"Analyse Quantitative des Microstructures (cours de formation doctorale)"
| Apprentissage des méthodes d'analyse quantitative de la
microstructure d'un matériau à partir d'images 2D (microscopies optique,
électronique ou à champ proche) ou 3D (tomographie X). Le cours s'adresse
principalement à des ingénieurs, doctorants ou chercheurs dans le domaine des
matériaux mais pourra être d'intérêt également pour tout chercheur intéressé
par l'analyse d'images.
|27/05/2010(16:00) [Lieu : Euler - A002]|
"Mechanical behavior of Advanced High Strength Steels resistance spot welds in Cross Tension and Tensile Shear"
|Advanced High Strength Steels (AHSS) are key materials in the conception of car body structures, permitting to reduce their weight while improving their behavior in crash conditions. Nevertheless, the weldability of AHSS presents some particular aspects, complex failure types involving partial or full interfacial failure being encountered more often than with conventional mild steels during destructive testing.
Different approaches have been investigated in this study in an attempt to understand the behavior of AHSS spot welds and model their load bearing capacity.
First, a statistical analysis of experimental spot welding data available at ArcelorMittal was conducted in order to get a global picture of spot welds mechanical behavior. An empirical model was proposed for the prediction of spot welds strength. Next, an experimental protocol was developed to overcome the limits of such a model and obtain a better understanding of failure mechanisms. Interrupted Cross Tension and Tensile Shear tests were performed and spot welds failure was investigated with optical micrographs, SEM fractography and 3D-tomography in order to follow the three-dimensional crack paths due to the complex loading modes. A limited number of failure zones and damage mechanisms could be distinguished. A Finite Elements model of spot welds behavior was then developed, which requires an appropriate description of local Heat Affected Zones mechanical properties. This has been obtained by experimental simulations of the spot welding thermal cycles with a Gleeble machine, followed by classical mechanical tests. Spot welds strength and failure type could be predicted as a function of the geometrical features of the weld assembly and the loading mode.|
|26/05/2010(13h00) [Lieu : LAVO51 - Bât. Lavoisier]|
"Cristallisation de substances d'intérêt pharmaceutique présentant un polymorphisme cristallin"
|Description d'une étude de cristallisation présentant une transition polymorphique. Les phénomènes sous-jacents de ce procédé ont été étudiés, en analysant d'un côté
l'aspect chimique du composé et l'interaction avec le solvant, et d'un autre côté les aspects d'ingénierie. Cette étude a été menée en combinant diverses techniques expérimentales (sondes in-situ, ...) ainsi que des outils de modélisation (FLUENT, ...) dans le but d'obtenir une vue d'ensemble du procédé.|
|19/05/2010(Barb 93) [Lieu : 14h30]|
"Phase transformations and mechanical properties of the
Ti-5553 beta-metastable titanium alloy"
|The microstructure investigation carried out in this work deals with the
formation of the alpha phase during isothermal heat- and thermomechanical
treatments. Various kinds of defects are controlling the heterogeneous
nucleation of alpha particles. Some treatments were specially designed to
highlight a particular nucleation mechanism, like ageing on pre-bent
specimens for the influence of deformation bands, or the use of fast
quenching for the trapping of excess vacancies. However, the isothermal
ageing treatments can only represent a part of the picture: As the heating
and cooling rates cannot be infinite, they influence the phase
transformations. In situ techniques are thus needed for monitoring the
phase transformation during continuous heating and cooling. The free
vibrational method, measuring the change of elastic modulus and the wave
damping during heat treatment, is compared to other more conventional
techniques like DSC, dilatometry, or post mortem metallography and
The relationship between microstructure and mechanical properties was
addressed at several levels. At the local scale, the nano-indentation
technique was coupled with SPM imaging and EBSD to measure the hardness
and elastic modulus of each phase, taking into account phase boundaries
and crystal orientation. At the macro scale, tensile tests on notched
specimens were used to link the microstructural features, the stress state
and the fracture strain via a micromechanical model based on the growth of
cavities nucleated on weaker points of the microstructure. The Ti-5553
alloy presents a very specific behavior compared to other titanium alloys.
High temperature tensile tests were also performed in order to extract
flow stresses, strain-rate sensitivity coefficients, activation energies
for plastic deformation, for strain-rates ranging from creep to dynamic
|27/04/2010(13h00) [Lieu : Ba91]|
Yves Bréchet(INP Grenoble)
"From materials selection to materials design"
|organized in the context of the course "Selection of Materials" given by T. Pardoen.|
|22/04/2010(16:30) [Lieu : Euler A002]|
"Multi-scale modeling of elasto-plastic composites"
|Homogenization techniques aim at predicting the overall behavior of heterogeneous media taking into account the geometrical and mechanical properties of their constituents. Accurate predictions may be obtained from full-field calculation (e.g. by the FE method) of the local stresses and strains throughout a statistically representative volume element (RVE) of the microstructure. However, such approach requires high computational costs, and even becomes unrealistic when composites are involved in simulations of real structures. Mean-field homogenization schemes provide a cost-effective alternative. Interactions between the constitutive phases are then accounted for in a simplified way. For composites having linear elastic constituents, reliable models have been available for quite a long time. However, extending these schemes to the nonlinear regime is still challenging, and strongly depends on the constitutive laws of the constituents.
This talk is devoted to composites having elasto-(visco)plastic phases. Classical strategies for dealing with such composites are briefly reviewed. Next, a new approach based on a thermodynamic formulation of the equations of elasto-plasticity is presented. The new method is shown to compete with other existing schemes, while providing a rigorous theoretical framework.
|29/03/2010(13h - 14h) [Lieu : Salle Passeleq (Vinci)]|
"Modelling the flow of a flood through a urban area with a porosity approach"
|The consequences of floods can be disastrous, especially when they occur in an inhabited area. Floods can be produced by heavy rains making the water level of a river to rise, by a tsunami or by the break of a structure of retention like a dam or dikes. It is important to be able to predict the water level and the velocity of the water in a urban area if such a catastrophe occurs, for example to draw emergency plans. In this presentation a model to predict the flow of water during a flood event in a city will be exposed. The model consists in assimilating a city to a large-scale porous medium in which the pores are the streets, and the obstacles, the buildings. Governing equations for the flow are derived based on this assumption. Experiments have been done in laboratory on a simplified city, where the streets are all perpendicular to each other. The comparison between numerical results and experimental measures present a good agreement but show that the treatment of the boundaries of the city and the evaluation of the drag force of the buildings are two delicate key issues. |
|24/03/2010(16h30) [Lieu : Auditorium Sud 11]|
Olivier Cartiaux(iMMC, CEREM)
"Evaluation d'outils d'assistance naviguée et robotisée à la découpe osseuse pour la chirurgie des tumeurs du pelvis
|Cette thèse s'inscrit dans le projet /Sarcome/ que nous avons mené en
collaboration avec le Laboratoire d'orthopédie de l'UCL et les Cliniques
universitaires Saint-Luc de Bruxelles. Ce projet porte sur le traitement
chirurgical des tumeurs osseuses du pelvis. Il s'agit en effet d'un
problème complexe pour lequel la procédure conventionnelle de découpe
osseuse réalisée sur le patient souffre encore aujourd'hui d'une
imprécision importante. Les travaux présentés dans cette thèse
s'attachent particulièrement à évaluer des outils d'assistance naviguée
et robotisée intégrés dans la procédure en vue d'améliorer les
performances des chirurgiens lors des découpes osseuses.
Tout d'abord, nous proposons une approche systématique de définition et
de mesure de la qualité d'une découpe conformément à la norme
internationale ISO1101. Notre approche se base sur un simulateur de
découpes d'une géométrie osseuse simplifiée et a permis de valider la
/localisation/ comme nouvel indicateur de qualité d'une découpe.
Ensuite, nous proposons une étude quantitative permettant la comparaison
de pistes d'amélioration des découpes sur base de performances réelles.
En particulier, nous investiguons les pistes /navigation optique/ et/
robot actif/ en les mettant en oeuvre sur un simulateur de découpes d'une
géométrie pelvienne. La précision des découpes réalisées, en termes de
leur /localisation/, est améliorée quand les chirurgiens utilisent la
navigation optique et davantage quand un robot industriel est intégré
dans la procédure.
Finalement, la procédure assistée par la navigation a été appliquée avec
succès sur un patient réel. Il s'agit du premier cas clinique pour
lequel les gestes de résection tumorale et de reconstruction par greffe
ont été chacun navigués au cours d'une même intervention.
|24/03/2010(15h) [Lieu : Ba20]|
Xavier Sauvage(Univ. de Rouen)
"Nanoscale 3D characterization by Atom Probe Tomography"
|Atom Probe Tomography (APT) is a projection microscope combined with a time of flight mass spectrometer. During the last twenty years this technique has been widely used to measure the composition of nanoscaled particles in various metallic materials and to display their distribution and morphology. It is also a powerful instrument to reveal chemical gradients or to highlight segregation along structural defects like grain boundaries or dislocations. This talk will start with a brief description of the technique followed by some examples of application with a special emphasis on the relationship between nanostructures and mechanical properties.|
|22/03/2010(16:30) [Lieu : A207, Bâtiment Euler]|
Prof. Sigfried Schmauder(IMWF. University of Stuttgart)
"Multiscale simulation of the embrittlement of Cu containing steels"
|22/03/2010(14:00) [Lieu : A207, Bâtiment Euler]|
"Activités à l'UCL dans le domaine de la micromécanique des métaux et composites"
|12/03/2010(08h30) [Lieu : Stevin B 044]|
Benoit Herman(CEREM, UCL, et ISIR (Institut des Systèmes Intelligents et de
Robotique), Université Pierre et Marie Curie, Paris)
"Human-robot interactions in medical robotics: an integrated approach to improve
safety and ergonomics"
|Ergonomics and safety are two major issues in modern robotics ? especially for
medical applications ? where human beings and machines interact and collaborate with
each other to carry out their duties. Thirty years ago, a few industrial
manipulators were customized to fulfil minimal safety constraints for milling bones
in orthopedic surgery. The rapid broadening to various specialties (e.g.
neurosurgery, urology, gynecology, and cardiac surgery) required the development of
novel devices that would meet surgeons' needs in specific applications.
In this context, the electro-mechanical structure and the control algorithm and
interface must be designed at once, since all these aspects interact and cannot be
optimized separately. Through the example of EVOLAP, a robotic camera positioner for
minimally invasive laparoscopic surgery, this presentation introduces a global
design method for medical devices. First, a weighted objectives tree was constructed
with clinicians to clarify their needs and to lay down specifications. Subsequently,
an original mechanical architecture was proposed so as to meet these requirements. A
preliminary passive prototype was built in order to introduce the solution to
surgeons and gather their first impressions. Actuators were then dimensioned by
multibody modelling, which was also used to assess kinematic and dynamical
performances of the device. Careful attention was paid to the optimal selection of
laparoscope kinematics, so as to induce accurate image motions. A clinical trial was
finally performed to estimate the overall ergonomics and usability of the system and
to identify further avenues of improvement.
We conclude this presentation by expounding current research aiming at the
improvement of robot guidance implementing advanced control such as visual servoing,
and at reducing invasiveness in the present trend of miniaturization.
|12/03/2010(10h45) [Lieu : Stevin B 044]|
Andreas Muller(Mechanics and Robotics, University Duisburg-Essen, Germany)
"Model-based control of asynchronous linear direct drives"
|Novel asynchronous linear direct drive concepts have been proposed for special
applications, such as traction systems of rail shuttles. An attempt to modeling the
dynamics of such electromechanical systems (EMS) reveals a significant difference of
the equations of motion when compared to those of classical multibody systems (MBS).
The fundamental difference is that the kinetic state functions depend
non-quadratically on the generalized velocities - in contrast to classical MBS where
the kinetic energy is quadratic. Moreover, this property impedes the use of modern
model-based control concepts.
Passivity has evolved as a design concept for model-based control of Euler-Lagrange
systems, i.e. systems whose governing motion equations arise from energetic state
functions via a variational principle. The passivity property gives rise to the
stability of the controller. Vital for the passivity is a certain skew symmetry
property of the motion equations. The theory developed so far has exclusively
focused on energetic state functions that are quadratic in the generalized
velocities. For such systems the required skew symmetry condition is fulfilled.
Electromechanical systems, in general, and asynchronous direct drives in particular,
may possess energetic state functions that are not quadratic in the generalized
velocities, e.g. systems that are described by the Maxwell equations for moving
media (Minkowski equations). Hence, such systems cannot be treated with the
classical passivity-based control techniques.
In this presentation an augmented PD control scheme for such systems is proposed. In
the quadratic case, the skew symmetry property results from the Riemanian geometry
of the configuration space, which is lost for non-quadratic state functions.
Therefore, firstly, the system dynamics is represented in the event space,
considered as a Finsler space, and the skew symmetry property is shown for this
Finsler geometry. This admits, to develop a passivity-based augmented PD controller
in the event space. Secondly, this control law is restricted to the configuration
space giving rise to a passivity-base control law. The approach is demonstrated on
an asynchronous direct drive.
|11/03/2010(08h30) [Lieu : Stevin B 044]|
Renaud Ronsse(Biorobotics Laboratory, EPFL, Lausanne)
"Flexible human augmentation using adaptive oscillators"
|Recovering proper upper- and lower-limb functionalities after a stroke or a spinal
cord accident necessitates intensive rehabilitation exercises. These are however
time- and energy consuming for the physiotherapists, who are in charge of moving the
patient to elicit coordinated behaviors and to restore normal sensory-motor
functionalities. In order to take over part of the physiotherapists' burden,
rehabilitation robotics is nowadays an intensive field of investigation. The
classical approach consists of moving the patient with a sophisticated robot along a
pre-specified trajectory, corresponding to an average "normal" profile. However,
this approach suffers from several drawbacks: (1) the patient will be moved even if
she/he remains totally passive, such that the method does not promote the patient's
efforts; (2) the "normal" profile might not correspond to the patient's own "intact"
profile; (3) it has been shown that variability and even errors are necessary for
efficient learning, but a stiff position-controlled method trains only along the
average profile. The most recent controllers of rehabilitation robots deal with
these issues, giving rise to the concept of "assistance as needed". In this talk, we
will present a new way of providing assistance as needed for rhythmic movements
(e.g. walking) which is based on adaptive oscillators. These are dynamical systems
which are capable of learning the high-level parameters (frequency, phase,
amplitude, offset) of any quasi-sinusoidal movement. By deriving a simple dynamical
model of the forearm, we were able to augment rhythmical movements about the elbow,
while leaving the control of all the high-level parameters up to the participant.
Successful augmentation was demonstrated as a marked decrease in the biceps and
triceps EMG profile, while the flexibility was illustrated by instructing the
participant to change the movement frequency. We expect that this method will be of
high relevance for the field of rehabilitation robotics, exploiting both the
intrinsic robustness of dynamical systems, and the flexibility to adapt the
high-level parameters according to the patient's wish.
|26/02/2010(16h15) [Lieu : Euler A002]|
Jean-François Remacle(Université catholique de Louvain)
"High Quality Surface Remeshing Using Harmonic Maps.
Part II: Surfaces with High Genus and of Large Aspect Ratio.
|High Quality Surface Remeshing Using Harmonic Maps.
Part II: Surfaces with High Genus and of Large Aspect Ratio.
|22/02/2010(10h30) [Lieu : Euler a.002]|
Francois Henrotte(RWTH Aachen, Germany)
"Finite element techniques for large electromechanical models"
|Finite element techniques for large electromagnetic models with motion will be discussed. Two main aspects will be covered, which correspond basically to the Eulerian and Lagrangian approaches. In particular situations, electromagnetic braking applications for instance, an Eulerian approach is possible. We discuss the physical definition of the problem in terms of a convective (Lie) derivative. A 3D algorithm is proposed, based on the extrusion operator of Bossavit and the work by Heumann, and the implications on the finite element data structure are pointed out. On the other hand, finite element models of electromechanical devices adopt in general a Lagrangian point of view. Special techniques to account for the relative motion of different parts of the system are then required. Re-meshing and mortar techniques will be compared in terms of their suitability for inclusion in large 3D computations. Finally, the definition of
electromagnetic forces in a continuous medium will be briefly discussed in order to show how the Lie derivative also plays a central part in this context.|
|22/02/2010(14h00) [Lieu : Euler a.002]|
Sébastien Mariéthoz(Automatic Control Lab., ETH Zürich, Switzerland)
"Fast model predictive control, hybrid multilevel inverters: two ways to improve the performance of power electronics systems
|Power electronics is the key technology that enables some important changes
visible at different levels of our technology based society. The integration
of renewable energy and storage in power grids, the development of new
reliable power systems, of more efficient transportations systems and of
more complex lighter portable electronics devices rely on efficient power
In this context, the presentation highlights some past and recent research
that has aimed at improving dynamic performance and/or energy efficiency of
several power electronics systems. It is motivated and illustrated by
several application examples where the investigated techniques have been
The first part of the presentation focuses on the constrained optimal
control of power electronics systems. Most approaches employed for
controlling these systems have been based on classical continuous time
approaches (or their discrete equivalent) that do not handle constraints.
This is not suitable for new applications with high performance
requirements. In contrast, model predictive control handles constraints,
which generally allows obtaining better performance. The gain is
particularly pronounced when the system dynamics are fast and when
constraints become active. Over the past few years, several fast model
predictive control techniques have been introduced that enable reducing the
sampling times by several order of magnitudes down to a few microseconds on
today's low cost DSPs or fractions of microseconds on today's low cost FPGA
devices. These fast model predictive control techniques are very attractive
and suitable to systematically and reliably implement high performance
controllers for power electronics systems.
Building-up on one of the examples of the first part, the second part of the
presentation moves the focus on the design of medium voltage power
electronics systems. With today's semiconductor devices, classical power
conversion structures feature high switching losses, such that it is
preferable to employ multilevel power conversion structures. The bottlenecks
of some power conversion structures are shown as well as the expected
benefits of using hybrid multilevel inverter structures. The presented
structures will have potential applications in traction, renewable energy
production and storage, with the objective of reducing system losses and
improving power quality, which will become of paramount importance for the
development of tomorrow's transportation and power systems.
|05/02/2010(11.00) [Lieu : salle TERM, STEVIN b.044]|
"Some control issues in assisted MIS and cardiac surgery"
|Minimally invasive surgery offers important benefits for the
patient that result in less operative trauma and reduce the incidence of
post-surgical complications. But it also imposes several challenges for
the surgeons such as the lack of contact feeling, the difficulty in
compensating for the physiological motion especially within the context
of beating heart cardiac surgery where the physiological motions present
high dynamics, ... Robotic assistance has been proposed to overcome many
difficulties inherent to the minimally invasive procedures but so far
most of these difficulties are still open issues in terms of control.
During the talk, we will then present some recent works on an adaptive
force feedback controller designed for medical robotics teleoperated
systems. This controller will be illustrated through experimental
results consisting in needle insertion. We will also show how this
approach has been extended recently to motion compensation for robotic
assisted surgery with force feedback. We will develop subsequently some
new results on 3D motion tracking for beating heart surgery using
thin-plate spline deformable model. It constitutes a first step for
active cancellation of cardiac motion. Evaluation on in-vivo images
acquired from a DaVinci surgical robotic platform will be proposed.
Finally some perspectives will be given within the context of
endoluminal or transluminal surgery.
|17/12/2009(16.30) [Lieu : euler a.002]|
Thierry Coupez(CEMEF - Ecoles des Mines - ParisTech)
"Adaptation de maillage anisotrope et méthode level set convectée :
capture de surfaces et d'interfaces immergées
Anisotropic mesh adaptation and convected Level Set: capture of
immersed surfaces and interfaces"
|A partir d'une méthode locale de génération de maillage basée sur
l'amélioration de topologie et un principe de volume minimale, il est
relativement facile d'obtenir une technologie d'adaptation de maillage
anisotrope. Il faut pour cela changer la façon de mesurer en prenant en
compte un champ de métrique. Des résultats tout à fait convaincants ont été obtenus ces dernières années en construisant des métriques à partir
d'une analyse a posteriori de l'erreur d'interpolation et du hessien
d'un champ scalaire. On propose ici de construire le champ de métrique
directement aux noeuds du maillage à partir d'un tenseur de distribution
de longueurs, une analyse de l'erreur par arètes et un assemblage
simple. On montre les possibilités de cette technique sur des problèmes
d'interpolations géométriques. On se place ensuite dans le contexte de
méthode éléments finis mixtes stables et stabilisés pour des problèmes
de mécanique de fluides incompressible et de déformations de matériaux.
On associe la technique d'adaptation anisotrope à une méthode Level Set
dite convectée qui ne nécessite plus d'étape séparée de
réinitialisation. Level Set et adaptation anisotrope permettent
d'obtenir des résultats particulièrement prometteurs qui seront
illustrés sur le problème particulier de fluide buckling et d'autres
applications dans le domaine des matériaux.
The local mesh generation method based on mesh topology improvement and
a minimal volume principle, eases the design of an anisotropic mesh
adaptation tool. Indeed one only needs to account for a metric field
when length and volume measurement is required. Very convincing results
have been already obtained in the past years by deriving the metric
field from a posteriori interpolation error analysis and hessian
reconstruction. We propose here a different route to make the metric
field directly at the node of the mesh. We introduce the length
distribution tensor and show that only a 1d error analysis along the
edges is sufficient. We will show the possibilities of this technique on
geometrical interpolation problems by combination with a Level Set
representation. We address then moving free surfaces and interfaces. We
combine the convected level method with anisotropic mesh adaptation, the
flow solver being based on a stable/stabilised finite element for
incompressible fluid as well as for solid material. We will propose
examples of fluid buckling simulation showing that anisotropic
adaptation enhances the solution quality. Other examples will be
proposed in material, forming and others
|11/12/2009(17.00) [Lieu : TBA]|
"Modelling of texture and hardening of TWIP steel - Advanced finite
element representation of polycrystalline aggregates"
|Metallic alloys with enhanced mechanical properties are required
by the industry for example in order to produce cars that are safer,
lighter and more ecological. The development of micromechanical models and
their use in numerical simulations of forming processes (e.g. rolling or
deep drawing) as well as mechanical tests (e.g. uniaxial tension) helps
drawing the link between, on the one hand, physical and microstructural
properties and, on the other hand, the macroscopic mechanical response.
The first part of the thesis proposes a new crystal plasticity model
adapted to TWinning Induced Plasticity (TWIP) steel. The latter steel
grade combines high strength and ductility because of the formation
microscopic lenticular regions (called Â« twins Â» ) which are very hard and
which act as obstacles to dislocation glide. The mathematical model
accounts for this as well as the induced anisotropy and the rotation of
the crystal lattice. It is demonstrated that, by taking proper account of
the interaction of adjacent grains at the micron level, one obtains
improved predictions of the macroscopic texture development and hardening
in tension and in plane strain compression.
In the second part of the thesis, an advanced numerical representation of
single and multi-phase polycrystalline aggregates is presented. The
proposed modelling approach relies, first, on a novel algorithm for
texture discretisation valid when grains have non uniform sizes and
shapes; second, on the generation of finite element meshes selectively
refined along grain boundaries. Based on a systematic sensitivity study,
guidelines are formulated about the design of such crystal plasticity
based finite element modelling (CPFEM), Predictions obtained with the new
model microstructures offer an improved compromise of accuracy and
|06/11/2009(16.15) [Lieu : Barb92]|
"Development of a finite element marine model
|Numerical models are very helpful to understand the behaviour of the
marine system. Ocean models have been developed for more than forty
years, and their design is an area of active research. If the
representation of the physics has been highly improved, the fundamental
numerical technique has not evolved: they still use the finite
difference method on structured grids. Recent efforts focus on
developing the new generation of ocean models, taking advantage of the
potential of modern numerical methods. Based on unstructured grids, such
models allow to faithfully represent complex topographic features such
as the coastlines, narrow straits and islands. The mesh resolution can
be refined locally in regions of interest or where the dynamics is more
This PhD dissertation focuses on the development of a three-dimensional
baroclinic marine model using the Discontinuous Galerkin finite element
method. The model is described, with some results of baroclinic
simulations. The rest of the thesis is devoted to different types of
unresolved physics. Two different boundary layers are introduced: the
bottom velocity boundary layer and the boundary layer of the residence
time. Both parameterisation and representation using the extended finite
element method are discussed. A part of the thesis is dedicated to the
treatment of the horizontal density gradient in baroclinic column
models. Attention is paid to the stability of the method under
different configurations. Then, the turbulence modelling in
three-dimensional models is studied by comparing the effect of different
turbulence closure models on a simulation of the flow around a
|25/09/2009(16.15) [Lieu : Auditorium Euler]|
"Growth versus spreading in biological systems: from Malthus' model to epidemics"
|The invasion of muskrats in Europe and the propagation of the Black Death (or other horrible diseases) appear to have nothing in common. And, yet, they can be thought to result from the competition between two radically different processes, growth and spread. Basic mathematical models thereof will be introduced. It will be seen that some of these models admit solutions that can be viewed as travelling waves -- though none of the equations dealt with are of a hyperbolic nature. Qualitative and, occasionally, quantitative comparison with field data will point to the well-foundedness of the simple models of growth and spread that have been designed since the late 18th century|