10/2020 - Present, COMSOL AB
Specializing in the development of the "computational backbone" of multiphysics simulations, my work focuses on bridging the gap between complex CAD geometries and robust numerical solvers.
Boundary Layer Meshing Engineering advanced algorithms to automatically generate high-aspect-ratio elements near solid boundaries. This is critical for capturing steep velocity and thermal gradients in turbulence modeling and heat transfer applications.
Mesh-Based Geometry & Copy Technologies Developing scalable C++ routines for "Copy Mesh" operations and mesh-to-geometry associations. These technologies allow for the efficient transfer of discretized data across symmetric or repetitive assemblies, significantly reducing pre-processing time for large-scale industrial models.
Mesh Defeaturing & Boolean Operations Supporting robust mesh-level operations including Union and Intersection. These tools enable efficient mesh defeaturing—removing small or irrelevant geometric details from the mesh—to streamline the transition from complex CAD to simulation-ready models.
Finite Voids for Acoustics Developing specialized support for generating meshes in finite voids. This is essential for acoustics and other applications where the fluid or air domain enclosed within a structure must be accurately discretized for wave propagation analysis.
Performance & Reliability Utilizing Visual Studio’s advanced profiling and diagnostic tools to optimize meshing kernels. My focus is on ensuring that mesh generation is not only geometrically accurate but also memory-efficient and robust enough to handle "dirty" CAD data without solver failure.
08/2019 - 10/2020, COMSOL AB
Advanced Fluid Models Specialized in the implementation and validation of inelastic non-Newtonian fluid models (see Polymer Flow Module - COMSOL® 5.6 Release Highlights). This involved developing solvers that account for shear-dependent viscosity, essential for industrial applications in chemical processing and polymer rheology.
Turbulence & Interface Problems Expert in modeling high-Reynolds-number flows using advanced turbulence models. My work includes solving complex interface problems where fluid behavior must be coupled with structural or thermal constraints, ensuring numerical stability and physical accuracy.
Nov 2017 – 2019
As an external collaborator of this project, I developed an HPC software to simulate diffusion MRI in the neurons published on Neuromorpho database.
Jul 2017 – 2019
I was one of the main developers of the online course on high-performance finite element method: Part I, Part II.
CFD studies of VAWTs (Swedish Energy Agency, 2015-2018)
Aug 2015 – 2019
I had been involving in the project as a doctoral researcher in computer science. My main tasks consisted of (1) implementing a framework to simulate turbulence of a rotating vertical axis wind turbine whose CAD was given by the Uppsala University by using the adaptive stabilized finite element method developed within the group recent years, (2) validating numerical results against experimental data and (3) developing a slip velocity model for internal interfaces in a fluid-structure interaction framework.
Aug 2015 – 2019
I was one of the developers of the software with two main contributions: (1) computational diffusion MRI and (2) simulating turbulence of a vertical axis wind turbine.
MAPIE
Feb 2014 – May 2015
In this project, I worked as a postdoctoral researcher with the main focus on the ADER-DG method for the elastic wave equations. The method was implemented in Matlab using the nodal basis functions for the 1D problem with some numerical verifications and comparisons against other methods. This work facilitated a reference for the implementation of higher dimensional problems on the structure code OOFE developed at MSSMAT, Ecole Central Paris.
Nov 2010 – Jan 2014
I proposed a finite element method to solve the Bloch-Torrey equation applied to diffusion MRI. This was the first use of this technique in diffusion MRI. Some applications of the method for studying the diffusion MRI signal inside multi-compartment models were considered. I also proposed an efficient one-dimensional model for accurately computing the dMRI signal inside neurites trees to test the validity of a semi-analytical expression for the dMRI signal arising from neurites trees. I ended the project with 7 peer-reviewed journals, one conference proceedings and one finite element code developed in the FEniCS.
2015-2020, KTH Royal Institute of Technology, Stockholm, Sweden.
DD2325 Applied Programming and Computer Science.
DD2363 Methods in Scientific Computing.
DD2365 Advanced Computation in Fluid Mechanics.
DD2437 Artificial Neural Networks and Deep Architectures.
DD1331 Fundamentals of Programming,
DD1388 Program System Construction Using C++.
DD1354 Models och Simulation.
DA2210 Philosophy of Science
Online course MOOC-HPFEM.
18-09-2025, Linear Systems & Applications, DIS Study Abroad in Scandinavia
2007-2009, High-school teacher of maths and computer science, Vietnam.
2015-2020, KTH Royal Institute of Technology, Stockholm, Sweden.
Title: High Performance Finite Element Methods with Application to Simulation of Vertical Axis Wind Turbines and Diffusion MRI.
Supervisors: Prof. Johan HOFFMAN and Assoc. Prof. Johan JANSSON.
2010-2014, INRIA Saclay-Equipe DEFI, CMAP, Ecole Polytechnique, Palaiseau, France.
Title: A finite elements method to solve the Bloch-Torrey PDE applied to diffusion magnetic resonance imaging of biological tissues.
Supervisors: Dr. Jing-Rebecca LI and Dr. Denis GREBENKOV.
2009-2010, Pôle Universitaire Français Ho Chi Minh, Vietnam,
M.S. internship in Applied Mathematics, LPMC, Ecole Polytechnique, Palaiseau, France.
Topic: Pulsed-gradient spin-echo monitoring of restricted diffusion in multilayered structures.
Supervisor: Dr. Denis GREBENKOV.
2003-2007, CanTho University, CanTho, Vietnam.
Thesis: Genetic algorithm for integral problems.
Supervisor: Dr. Bao Quoc Truong.
We generate volume meshes for a population of 36 pyramidal and 29 spindle neurons. They are distributed in the anterior frontal insula (aFI) and the anterior cingulate cortex (ACC) of the neocortex of the human brain. They share some morphological similarities such as having a single soma and dendrites branching on opposite sides. This population consists of 20 neurons for each type in aFI, and 9 spindles, 16 pyramidals in ACC.
In this project, we developed two MPI-based solvers to solve the Poisson equation on Cartesian grids: the Jacobi and the conjugate gradient (CG) methods. The finite difference method will be used for space discretization. Thanks to the funtionalities of the MPI virtual topology, the computational domain is decomposed into subdomains and then each subdomain is assigned to a MPI process. The performance analysis will also be taken into account in this project.
FEM tools COMSOL Multiphysics, FEniCS, FreeFem, ANSA, Salome, GMSH.
Machine Learning tools PyTorch, TensorFlow.
Languages & Tools C++ (Expert/C++11/14/17), Python, MATLAB, Shell Scripting.
Libraries PETSc, MPI, OpenMP, FEniCS.
Workflow Agile/Scrum, Git, Visual Studio (Profiling/Debugging), Unit Testing.
Hoang-Trong-An Tran, Master 2 Internship Ecole Polytechnique, 2015.
With main supervisor Dr. Jing-Rebecca Li.
Gökce Tuba Masur, KTH Master thesis, 2017.
With main supervisor Prof. Johan Hoffman.
Henning Spett, KTH Master thesis, 2019
With main supervisor Assoc. Prof. Johan Jansson.