SPC MTs and SPs

Navigation

Ce wiki

Cette page

The table below gives an overview of Master theses and Specialization projects available at SPC.

These projects are available to students of the Physics section.

Students from sections other than Physics should contact the secretary of their section beforehand to check whether it is possible to undertake their Masters thesis at SPC.

In case of questions, please contact Marcelo Baquero or Ivo Furno.

Instructions

Take a look at the available projects in the table below.
- Rows marked No correspond to projects that have already been assigned and are no longer available. Please refrain from contacting the supervisors of those projects.
If an available project calls your attention, send an email to the Contact/Supervisor to inquire or apply.
If/when an agreement is reached with the Supervisor, ask them to let Marcelo Baquero know of the decision by email.

Resources

Info on Physics Masters theses: https://www.epfl.ch/schools/sb/sph/en/master/master-in-physics/master-project-physics/
Info on Plasma Physics Specialization projects : https://www.epfl.ch/schools/sb/sph/en/master/master-in-physics/specialization-semester/

Lists for Spring 2026

Masters theses

Available?	Contact/Supervisor	Type of project	Device/Lab	Project title	Project description	Assigned student
No	Davide Mancini (davide.mancini@epfl.ch), Paolo Ricci (paolo.ricci@epfl.ch)	Theory	TCV	Turbulence simulations of Doublets with the GBS code	Modelling and analysis of the plasma turbulence in the SOL of a tokamak device in different Doublets magnetic configurations, based on recent experiments performed in TCV.	Armand Laville (armand.laville@epfl.ch)
No	Davide Mancini (davide.mancini@epfl.ch), Paolo Ricci (paolo.ricci@epfl.ch)	Theory	TCV	Time dependent neutral solver in KINDNES	Analysis and implementation of a neutral solver that doesn't use the adiabatic approximation in the framework of KINDENS, a deterministic neutral dynamics solver for plasma turbulence simulations.	Matyas Rodriguez Szonyi (matyas.rodriguezszonyi@epfl.ch)
Yes	Francesco Carpanese (francesco.carpanese@epfl.ch)	Data analysis, modelling, machine learning	TCV	Exploration of the FGS free-boundary solution space and development of an ML surrogate	The free-boundary solver FGS provides the initial equilibrium used in time-dependent free-boundary simulations for testing magnetic controllers on TCV. Because the underlying nonlinear problem admits multiple solutions, FGS convergence depends strongly on the proximity between these solutions and on the quality of the initial guess. This often complicates obtaining the desired starting plasma for evolutive simulations. This project will systematically investigate how many alternative solutions exist near a given equilibrium, and how sensitive they are to input parameters, across a broad range of plasma conditions. The resulting large-scale database will enable training a machine-learning surrogate model for FGS. This surrogate is expected to support Bayesian equilibrium reconstruction and accelerate the evolutive free-boundary solver FGE.
Yes	Francesco Carpanese (francesco.carpanese@epfl.ch)	Data analysis, modelling, machine learning	TCV	Bayesian magnetic and kinetic equilibrium reconstruction for TCV	Equilibrium reconstruction infers the plasma shape from magnetic measurements, typically via the nonlinear least-squares solver LIUQE, which uses a frequentist uncertainty model. A Bayesian formulation, although more computationally demanding, provides principled uncertainty of quantification and naturally incorporates heterogeneous diagnostic measurements with proper error propagation. This approach has gained traction in recent years and was previously prototyped for TCV. With the tools now available at SPC, the Bayesian analysis could use a forward model constrained by physical assumptions on the internal current-density profile, enabling a more realistic reconstruction of the safety-factor profile q, which is crucial for many aspects of tokamak operation. This workflow hasn't been developed yet and will benefit several analysis for tokamak operation.
Yes	Duccio Testa (duccio.testa@epfl.ch)	Hardware, experimental	DEMO, DTT	Design and prototyping of a high-frequency magnetic sensor for current and future tokamaks, combining photolithography and Hall technologies	The project involves prototyping activities using different technologies, mostly available in-house at the EPFL, and testing of the as-built prototypes. This in turn will be used to define the final design and the manufacturing and testing specifications for the actual magnetic sensors for DEMO and DTT.
Yes	Duccio Testa (duccio.testa@epfl.ch)	Data analysis, experimental	TCV	EM and ES fluctuation measurements	Advanced statistical analyses of EM and ES fluctuation measurements on TCV, focussing on the role of the plasma shape and magnetic helicity in affecting the stability of ES and EM turbulence.
Yes	Duccio Testa (duccio.testa@epfl.ch)	Data analysis, experimental	TCV	Data analysis for the TCV fast ion experiments: gap structure vs. onset of fast ion drives modes vs. plasma shape, positive & negative triangularity	Analysis of the role of the plasma shape in affecting the stability of electro-magnetic and electro-static macroscopic instabilities driven by fast ions
Yes	Duccio Testa (duccio.testa@epfl.ch)	Theory, data analysis	TCV, ITER	Analysis of the phase-space stability of transport simulations in TCV	Use the Phase-Space-Reconstruction approach, and the ensuing evolution of the Lyapunov exponents, as a tool to determine whether a pre-programmed discharge is evolving towards or diverging from a stable attractor in phase space, and on what time scales this would happen. This will be done using ASTRA and/or RAPTOR transport simulations of TCV tokamak discharges, and if time allows then using comparison with electro-magnetic and electro-static fluctuation measurements in such pre-programmed discharges as then actually run on TCV.
Yes	Duccio Testa (duccio.testa@epfl.ch)	Numerical, data analysis	TCV, ITER, DEMO, DTT	Development of a "universal" GUI for MHD analyses	All magnetic fusion experiments require the analysis of the magnetohydrodynamic perturbation to the plasma equilibrium, and there is a set of mathematical and computational tools that can be universally adapted to all such experiments, existing and futures. Starting form existing codes, we wish to prepare a user-friendly software package, such as a Graphical User Interface (GUI), that could be generally adapted and thus deployed to multiple experiments.
Yes	Duccio Testa (duccio.testa@epfl.ch)	Numerical, data analysis	DEMO, DTT	Assessment of the long-term performance of the magnetic diagnostic systems in DEMO and DTT	Develop an assessment of a general concept for the overall magnetic diagnostic system satisfying the intended measurement requirements for DEMO and DTT, specifically designed considering the long-term but cost-effective resilience of the measurement performance to the inevitable loss of sensors.
Yes	Benoit Labit (benoit.labit@epfl.ch)	Theory, data analysis	TCV, possibly JET	Role of MHD in pedestal structure for various H-mode regime	The details of the project will be discussed with the candidate but the main idea is to use numerical tools (MHD codes) to evaluate the pedestal stability and MHD induced transport for either TCV or JET cases.
Yes	Reinart Coosemans (reinart.coosemans@epfl.ch), Benoit Labit (benoit.labit@epfl.ch)	Data analysis, modeling	TCV	Neutron rate estimates for TCV plasmas with RABBIT code	In plasmas heated with deuterium beams a deficit of the expected fusion neutron rate is an indicator of the deterioration of the fast-ion confinement, caused, for instance, by magnetohydrodynamic instabilities. The goal of the project is to estimate the neutron rate for TCV plasmas using the code RABBIT (rt-capable) and to compare the results with experimental data and with other codes like NUMEAM (which is more slower). If the results are encouraging, a coupling with transport codes like ASTRA could be envisaged.
Yes	Anton Jansen Van Vuuren (anton.jansenvanvuuren@epfl.ch), Benoit Labit (benoit.labit@epfl.ch)	Data analysis, modeling	TCV	Refined charge-exchange losses estimates for fast ions for TCV plasmas	In fusion plasmas, a precise knowing of the loss power is crucial to assess the quality of the confinement. One source of losses is charge-exchange (CX) between the fast ions and the background “cold” neutrals. The goal of the project is to evaluate CX losses for TCV plasmas (using ASCOT code) assuming a more realistic neutral density spatial distribution (from SOLPS simulations). At TCV, a large SOLPS database exists, in particular for plasma detachment where plasma density is increased by large injection of deuterium neutral molecules. Results will be compared to estimates from ASTRA simulations, the used standard tool at TCV.
Yes	Benjamin Vincent (benjamin.vincent@epfl.ch)	Experimental, data analysis	TCV	Experimental characterization and commissioning of a new system based spectral apodization filter for the Incoherent Thomson Scattering (ITS) diagnostic	Previous simulation work has demonstrated the feasibility and benefits of using a single-filter/two-spectral-channels system for analyzing scattered light from ITS. When analytically optimized to spectrally apodize the ITS spectrum in specific ways, simulation for such a setup has suggested it would offer performance similar to or even better than a polychromator. It has also been demonstrated that, in some cases, such a tool could be a promising candidate for assessing velocity profiles. The student would be involved in the characterization of such a filter design and its implementation at TCV inside an existing polychromator system, in order to validate the sensitivity improvements predicted by simulation.
Yes	Guillaume Van Parys (guillaume.vanparys@epfl.ch)	Theory	TCV	MHD stability of shaped, rotating, pressure anisotropic plasmas	Coupling of ideal MHD stability code VENUS-MHD to new equilibrium solver with rotation pressure anisotropy.

Specialization projects

Available?	Contact/Supervisor	Type of project	Device/Lab	Project title	Project description
Yes	Stefano Coda (stefano.coda@epfl.ch)	Theory	TCV	Exact characterization of the transmission properties of a TCV microwave launcher	This project involves using combined full-wave/ray-tracing codes to calculate precisely the propagation of ECRH beams through launchers, going beyond the usual paraxial approximation. COMSOL Multiphysics as well as the 2D code REFMULF can also be explored as tools for this project.
Yes	Guillaume Van Parys (guillaume.vanparys@epfl.ch)	Theory	TCV	MHD stability of shaped, rotating, pressure anisotropic plasmas	Similar to project of same title proposed for a Thesis (see list above) but with reduced scope (to be discussed with contact person).
Yes	Chizhou Wang (chizhou.wang@epfl.ch), Mengdi Kong (mengdi.kong@epfl.ch)	Theory	TCV	3D MHD modelling of ohmic and runaway electron dominant plasmas with JOREK	This project explores the MHD stability of Ohmic plasmas and plasmas that are mainly held by runaway electrons (REs) via 3D non-linear MHD modelling with the JOREK code. The high-energy REs can be generated during tokamak plasma disruptions and carry most of the plasma current. They need to be mitigated to avoid damaging the machine. A possible solution is to disperse REs over a larger ‘wetted area’ on the wall to reduce the local heat load, i.e. reaching benign termination. This is realized by deconfining REs via MHD instabilities. This project could thus contribute to the study of RE benign termination.
Yes	Francesco Carpanese (francesco.carpanese@epfl.ch)	Data analysis, modelling, machine learning	TCV	Exploration of the FGS free-boundary solution space and development of an ML surrogate	Similar to project of same title proposed for a Thesis (see list above) but with reduced scope (to be discussed with contact person).
Yes	Francesco Carpanese (francesco.carpanese@epfl.ch)	Data analysis, modelling, machine learning	TCV	Bayesian magnetic and kinetic equilibrium reconstruction for TCV	Similar to project of same title proposed for a Thesis (see list above) but with reduced scope (to be discussed with contact person).
Yes	Reinart Coosemans (reinart.coosemans@epfl.ch), Benoit Labit (benoit.labit@epfl.ch)	Data analysis, modeling	TCV	Neutron rate estimates for TCV plasmas with RABBIT code	Similar to project of same title proposed for a Thesis (see list above) but with reduced scope (to be discussed with contact person).
Yes	Anton Jansen Van Vuuren (anton.jansenvanvuuren@epfl.ch), Benoit Labit (benoit.labit@epfl.ch)	Data analysis, modeling	TCV	Refined charge-exchange losses estimates for fast ions for TCV plasmas	Similar to project of same title proposed for a Thesis (see list above) but with reduced scope (to be discussed with contact person).