Joubine Aghili

Research Engineer

Avatar

Laboratoire Jean Dieudonné
Univ. de Nice UMR 7351
Parc Valrose, 28 avenue Valrose
06108 Nice CEDEX 2
Office: 3rd floor, room 701.
joubineaghili.fr

Vitae

  • Ph.D in Applied Maths supervised by D. Di Pietro
  • MSc & BSc in Applied Maths

Interests

Numerical analysis of PDEs, Nonconforming methods, discretization of fluid flow in porous medium, model order reduction, Automatic learning from PDEs, Computational Science & HPC techniques.


Languages

French, English, Farsi, C++



Current Position Postdoctoral research fellow at INRIA (Team COFEE) & Univ. Côte d'Azur, collaboration with ANDRA (Agence nationale pour la gestion des déchets radioactifs), since oct. 2017.

Simulation des écoulements liquide gaz dans les milieux de très faible perméabilité avec réseaux de fractures discrètes

Research activity

Darcy flows in fractured porous media

Part of the COFFEE team (COmplex Flows for Energy and Environnement), I am currently working on Darcy flows in a low permeability porous medium having a discrete fractured network. This work is supervised by R. Masson and K. Brenner with the collaboration of Laurent Trenty (ANDRA), find below two teasers

computations made on cartesian and triangular meshes (derived from FVCA5 meshes) with different fracture networks.


Non-conforming Methods


Reduced order modeling
  • In progress... Solve in real time, within the web browser, a parametrized Oseen PDE, discretized by a HHO(0) method on a triangular mesh using Deep Learning : HHOLIVE

Communications

Publications
  1. JA & D. Di Pietro. An advection-robust Hybrid High-Order method for the Oseen problem, to appear in Journal of Scientific Computing, 2018. arXiv | HAL
  2. JA, B. Ruffini & D. Di Pietro. A hp-Hybrid High-Order method for variable diffusion on general meshes, Comput. Meth. Appl. Math., 17(3):359-376, January 2017. DOI | HAL
  3. JA. Numerical resolution of partial differential equations with variable coefficients, Ph.D thesis, December 2016. TEL
  4. JA, S. Boyaval & D. Di Pietro. Hybridization of Mixed High-Order methods on general meshes and application to the Stokes equations, Comput. Meth. Appl. Math., 15(2):111-134, January 2015. DOI | HAL

Talks/Poster
  • (upcoming) June 2018 — CMWR2018, St-Malo — Hybrid-dimensional two-phase flow in fractured porous media with interface solver.
  • (upcoming) March 2018 — Jeunes EDPistes français, Univ. Lorraine — TBA.
  • 2017 — Séminaire des doctorants, IMAG — Numerical reduction of the parametrized diffusion equation.
  • 2016 — ACSIOM Workgroup, IMAG — Numerical reduction of the parametrized diffusion equation.
  • 2015 — Poster presentation at 3rd Intl. Workshop Reduced Basis, POD and PGD Model Reduction Technique, ENS Cachan, Numerical reduction of the parametrized diffusion equation. pics
  • 2015 — Laboratoire d'Hydraulique St-Venant, Ecole des Ponts & EDF Chatou, Numerical reduction of the parametrized diffusion equation , November 2015.
  • 2015 — 13th European Finite Element Fair, Hybridization of mixed high-order methods on general meshes and applications to the Stokes equations, Prague. slides & pics
  • 2014 — 10th SEME, Impedance matrix computation for the numerical simulation of multi-conductor transmission lines, Strasbourg
  • 2014 — I3M Ph.D students seminar — EDP paramétriques & Bases réduites.

Computer science

Feel free to check my Github repo
CS
  • Languages:
    • Daily usage: C++, Matlab/Octave, Bash, LaTeX
    • Frequent usage: HTML/CSS, Fortran, PHP
    • Notions: Python, Perl, CUDA, MySQL, OpenMP
  • Libraries:
    • Daily usage: Eigen, STL
    • Frequent usage: ViennaCL, SFML
    • Notions: TensorFlow, MathJS, p5js, GNU Sci. Library, PETSc, boost, GLUT
  • Softwares:
    • Daily usage: GNU/Make, SVN, FreeFem++(-MPI), Emacs
    • Frequent usage: Paraview, CMake, Git, Scilab
    • Notions: Mathematica, Maple, Photoshop, GIMP
  • OS:
    • GNU/Linux: ArchLinux, Ubuntu, Fedora, Debian
    • Windows 98
Personal projects


Misc.

Teaching

  • 2017 — 16.5h of training classes: Numerical Analysis (HLMA405) at Université de Montpellier.
  • 2017 — 27h of training classes: Mathematics for Chemistry II (HLMA204) at Université de Montpellier.
  • 2017 — 78h of training classes: Linear Algebra II (HLMA201) at Université de Montpellier.
  • 2016 — 36h of training classes: Biomaths (HLMA103) at Université de Montpellier.
  • 2016 — 39h of training classes: Mathematics for Chemistry (HLMA102) at Université de Montpellier.
  • 2015 — 39h of training classes: Mathematics for Chemistry (HLMA102) at Polytech'Montpellier.
  • 2015 — 9h of training classes: Introduction to MatLab, L2 Géosciences, UM.
  • 2015 — 15h of training classes: Introduction to MatLab, L2 Maths, UM.
  • 2014 — 48h of training classes: Analysis & Algebra I (HLMA101) at Polytech'Montpellier.
  • 2014 — 9h of training classes: Introduction to MatLab, L2 Géosciences, UM.
  • 2014 — 15h of training classes: Introduction to MatLab, L2 Maths, UM.

Hobbies

(...)