URL study guide
https://tue.osiris-student.nl/onderwijscatalogus/extern/cursus?cursuscode=3MF502&collegejaar=2025&taal=enDescription
In this Master class the students will be introduced to computational approaches to study Nuclear Fusion plasmas. Computational approaches provide a scalable alternative to physically studying these plasmas in a lab setting or in an actual fusion device. They allow us to study plasma phenomena through numerical studies based on mathematical approaches, iterate more quickly on nuclear fusion designs, or even help us control plasmas by predicting plasma behavior ahead of time as part of a control system. Therefore, computational approaches have become an indispensable part of nuclear fusion research.This course is a combination of hands-on practical topics where you learn how to leverage computers to do computation, and a series of lectures where you learn about how these topics are relevant in the field. After an introduction to the topic of leveraging computers for scientific purposes we will start from scratch, first introducing you to the basic techniques any computational physicst needs to know, such as the Command Line Interface (CLI) in Linux, source control with git, and scientific programming with Fortran.
After that we will discuss various numerical techniques, their strengths and their limitations, and how to leverage them. These include
- Discretization: The process of converting our real-world continuous experience into the grid-based discrete world of computers.
- Finite differences: The application of discretization to differential equations.
- Finite elements: An alternative to finite differences where we divide a domain into smaller elements to solve differential equations.
Finally, we will have special attention for High Performance Computing (HPC) techniques that open the door to leveraging much more powerful computers or clusters of computers. We will briefly introduce you to
- Alternative languages: Ranging from low-level languages such as CUDA, to newer high level languages that aim to blur the difference between programming locally on your laptop to programming on a supercomputer.
- Parallelization of algorithms: Leveraging more compute power to accomplish a task.
- HPC clusters: When and how to use them.
Objectives
- Obtaining in-depth insight into a selected topical issue in the field of fusion,
- Being able to apply the fusion knowledge in a practical application
- Getting the latest update of the state-of-the-art techniques, and challenges in fusion
- Train in effective group work
- Improve on presenting/reporting
- Familiarize in studying and assessing literature – recent papers