NPRE Events Calendar

View Full Calendar

Final Exam (Dissertation Defense) Md Fazlul Huq

Event Type
Other
Sponsor
Department of Nuclear, Plasma & Radiological Engineering
Location
101A Talbot Laboratory
Date
Jun 14, 2024   10:00 am - 12:00 pm  
Speaker
Md Fazlul Huq, Ph.D. Candidate
Contact
Nuclear, Plasma & Radiological Engineering
E-Mail
nuclear@illinois.edu
Phone
217-300-5517
Views
3
Originating Calendar
NPRE Events

Md Fazlul Huq, Ph.D. Candidate

Dr. Davide Curreli, Director of Research

June 14, 2024 | 10:00am - 12:00pm CST 

This final examination will be held in 101A Talbot Laboratory.

Zoom: Meeting ID: 833 6744 0660, Password:  906311

Hybrid Meshing Techniques for Full-Orbit Particle-in-Cells Simulation of Full Scrape-Off Layers including Plasma-Surface Interaction and Monte Carlo Collision Effects

ABSTRACT:  Some of the most significant challenges that fusion research is dealing with today are directly related to how the plasma behaves in the Near and Far Scrape-off Layer (SOL) regions, and their interaction with the Plasma Facing Components (PFC). The best first-principle approach to model the plasma behavior in the SOL is to perform full-orbit kinetic simulations of the SOL by solving the multi-species electrostatic Boltzmann equation using a Particle-in-Cell numerical approach. However, the computational requirements of this type of simulations pose substantial challenges, particularly in terms of time and resources needed. Even with hybrid parallel Particle-in Cell (PIC) codes, such as hPIC2, trying to simulate the entire poloidal cross-section of the SOL of a tokamak soon becomes impractical due to the multiscale and complex nature of the plasma and to the requirement of uniform meshes typical of PIC approaches.

 In this work, we explore how to reduce the computational demands of full-scale tokamak PIC simulations by relaxing the mesh requirements. We propose and analyze an anisotropic structured-unstructured mesh, hereon referred as ``hybrid mesh'', specifically tailored for full-scale PIC simulations of the SOL of a tokamak. An important feature of our meshing methodology is that it allows to spatially resolve the plasma sheath forming at the plasma-material interface, a feature not tackled in previous approaches. In the near SOL region, where large density and temperature gradients and large fluxes are present, we used an anisotropic, multi-block, boundary-layer-refined, structured mesh. This structured mesh ensures to capture steeper gradients in this region. For the far SOL region, which experiences weaker ionization and gradients and has a more complex geometry, we used an unstructured mesh. This unstructured mesh is well-suited to accommodate the complex geometry of the far SOL region. A highly resolved mesh is employed in the high potential gradient sheath region at the vicinity of the first wall throughout the device. This ensures proper refinement in this critical region, capturing the sharp electric potential gradient normally occurring within the sheath. This new mesh reduces the number of degrees of freedom significantly, thus enabling sheath-resolved full-orbit PIC simulations of the entire SOL.

 We tested our hybrid mesh algorithm by successfully running a full-orbit hPIC2 simulation of the full Scrape-Off Layer of the WEST tokamak. The simulation includes a dynamic wall model via coupling to the RustBCA code across the full extension of the first wall of the device. The tungsten sputtered as a consequence of ion irradiation enters into the SOL plasma and is further ionized into multi-step ionization (ZW = 0,+1,+2,...,+18 included). The simulation produces distributions of W impurity charge in the near-wall layer of the SOL, along with a prediction of the gross erosion fluxes over the full extension of the tokamak wall. Our model allows to estimate from first principles, at a reasonable computational cost, which wall regions are eroded the most by plasma exposure.

link for robots only