Liesbeth  Vanherpe

News

›› 01/01/2018
Open sourcing of BluePyMM, Blue Brain Python Cell Model Management.

›› 01/01/2018
Started working at ASML, Veldhoven, the Netherlands.

›› 05/12/2017 at 17:00
OSPF Concert, Eglise de Puplinge, Puplinge, Switzerland.

›› 05/11/2017 at 17:00
OSPF Concert, Eglise Notre Dame des Grâces, Lancy, Switzerland.

›› 03/11/2017 at 20:00
OSPF Concert, Salle Frank Martin, Geneva, Switzerland.

›› 11/2017
Release of the Simulation Neuroscience MOOC on the edX platform; I contributed to the development of the exercises.

›› 12-13/09/2017
HBP Young Researchers Event, Campus Biotech, Geneva, Switzerland.

›› 23/06/2017 at 19:00
OSPF concert, Fête de la Musique, Salle de spectacle de l'Institut International de Lancy, Grand-Lancy, Switzerland.

›› 06-09/06/2017
Human Brain Project Brain Simulation Platform Hackathon, Campus Biotech, Geneva, Switzerland.

›› 26/04/2017
Member of the 9th edition of the Réseau romand de mentoring pour femmes.

›› 09/04/2017 at 17:00
OSPF Concert, Eglise Notre Dame des Grâces, Lancy, Switzerland.

›› 07/04/2017 at 20:00
OSPF Concert, Temple de Coppet, Coppet, Switzerland.

›› 06-10/03/2017
Workshop on Big Data Management Systems in Business and Industrial Applications, Stuttgart, Germany; I'm a member of the program committee.

 

News Archive

Conference Archive

 

E-mail

›› See Liesbeth Vanherpe @ EPFL.

Disclaimer

›› This is a personal page.

Most recent update

›› 09 July 2018 20:42:11

 

Layout

›› The layout of this page is based on the theme purgatory shuffle, to be found on Open Source Webdesign.

›› The header image of this page is made with Brayns, a visualiser for ray-traced rendering of neuronal morphologies, mainly developed by Cyrille Favreau at the Blue Brain Project.

 

Valid XHTML 1.0 Strict   Valid CSS!

The goal of my PhD research was to develop efficient methods to perform realistic, three-dimensional simulation of grain growth with phase field models. This work involved partial differential equations, parallel computing, sparse algorithms, simulation and finding new materials science results, in collaboration with Nele Moelans, Stefan Vandewalle and Bart Blanpain.

PhD Defence

Practical information

  • Title: Acceleration strategies for phase field simulation of grain growth in polycrystalline materials (Dutch: Versnellingsstrategieën voor faseveldsimulatie van korrelgroei in polykristallijne materialen)
  • Date and time: 09/09/2010, 17:00
  • Location: Auditorium “De Tweede Hoofdwet”, Thermotechnisch Instituut, Kasteelpark Arenberg 41, 3001 Heverlee
  • Text: PhD text available through KU Leuven

Abstract

The microstructure of many materials consists of multiple grains with different crystallographic orientations. Under certain circumstances, such as increased temperature, the smaller grains will shrink and disappear under the influence of surface tension. One of the modelling techniques that is explored for the simulation of this phenomenon, called grain growth, is phase field modelling. Two phase field models in particular are the main interest of this thesis, namely the continuum field model and the multi-phase field model. Both models represent a polycrystalline microstructure with a large set of phase field variables, where each variable corresponds to a single crystallographic orientation. However, realistic three-dimensional grain growth simulations with these models can demand significant amounts of computation power.

In this thesis, we present a sparse bounding box algorithm designed to perform efficient phase field simulations of grain growth. The algorithm shows significant improvements over existing techniques as its computational requirements scale with the grid size instead of with the number of crystallographic orientations involved. Furthermore, a nonlinear multigrid solver, based on the Full Approximation Scheme, is constructed to solve the multi-phase field model for multiple phase field variables. Experiments with this solver show that its convergence rates are independent of the grid size.

The applicability of the bounding box algorithm is illustrated by three-dimensional simulations of grain growth in the presence of spheroid second-phase particles and of grain growth in a microstructure with anisotropic boundary energy. From the former simulations, it is found that the pinning effect of a particle distribution is stronger for increasing volume fraction, and for increasing aspect ratio of the particles. Furthermore, a generalised Zener type relation is proposed. The second type of simulations is performed in a microstructure whose boundary energy is described by a Read-Shockley type dependence. Simulation results show that the low-angle boundaries are clearly preferred during grain growth. The anisotropic formulation of the boundary energy is furthermore observed to change the individual growth rates of the grains as a function of the number of grain faces.