PSFC Student Seminars

All Seminars are on Wednesday at 5:15pm, unless otherwise noted.
NW17-218, 175 Albany Street, Cambridge
For further information: info@psfc.mit.edu

Nov 15, 2022

Investigating the robustness of machine learning-based disruption prediction with synthetic training data

Andrew Maris

MIT PSFC

Data-driven algorithms have shown great promise in magnetic fusion applications, but these techniques can be unreliable when there are few examples for the algorithms to learn from. This is will be especially true for predicting disruptions on ITER, where the device will only be able to experience a small number of disruptions at full power over its lifetime. In this study, we investigate whether synthetically generated time series can augment small training sets of real data to achieve robust data-driven predictions of rare events. We perform numerical experiments by varying the amount of real and synthetic data in the training set of machine learning (ML) models tasked with predicting MHD-driven disruptions at DIII-D. We find that synthetic data appears to reduce underfitting and leads to higher prediction accuracies for standard neural networks.

4:00pm

Nov 8, 2022

Designing defect-tolerant high-temperature superconductor magnets

Richard Ibekwe

MIT PSFC

While the physical and operational characteristics of HTS offer high tolerance to defects, a rigorous physical understanding and quantification at relevant scale has not been established. To begin addressing this shortcoming, we present results from experiments with a copper cable containing a soldered stack of five HTS tapes and a physical defect.

4:00pm  |  NW17-218

Nov 1, 2022

Using the Stix Finite Element RF Code to Investigate the Power-Phasing Scan on Alcator C-Mod

Christina Migliore

MIT PSFC

​Ion cyclotron range of frequency (ICRF) power plays an important role in heating and current drive in fusion devices. However, experiments show that in the ICRF regime there is a formation of a radio frequency (RF) sheath at the material and antenna boundaries that influences sputtering and power dissipation. Methods to mitigate the formation of the near-field RF sheath have been studied through the means of optimizing the ICRF antenna. 

4:00pm  |  NW17-218

Oct 25, 2022

A review of acoustic techniques in cryogenic refrigeration and opportunities for new developments

Jacob Adams

MIT PSFC

Cryogenic refrigeration is vital in many areas of nuclear science and engineering including superconducting magnets for nuclear fusion and quantum computing. This presentation will begin by describing the working principles behind different types of cryogenic refrigerators and giving historical background to their development. Then, we will see how the pulse-tube cryogenic refrigerator successfully replaced mechanical components with acoustic waves to reduce cost and improve reliability. Finally, innovative acoustic refrigeration concepts inspired by musical instruments will be presented

4:00pm  |  NW17-218

Sep 20, 2022

Edge turbulence studies across confinement regime transitions at ASDEX Upgrade

Rachel Bielajew

MIT PSFC

Edge turbulence is thought to play an important role in tokamaks in the transition between the low confinement operating regime, L-mode, and high confinement operating regimes such as H-mode. High confinement regimes that are free of damaging Edge-Localized Modes, such as the “improved confinement regime” I-mode, are promising for future reactor operation.  However, open questions remain about the role of turbulence in regulating transport in I-mode, and changes in edge turbulence leading up to the I-mode and H-mode transitions. 

4:00pm  |  NW17-218

Dec 10, 2019

Non-relativistic pair plasma turbulence

Lucio Milanese

MIT

Achieving a more comprehensive understanding of electron-positron (pair) plasmas is important to interpret observations and explain the dynamics of astrophysical systems and phenomena such as the magnetosphere of pulsars, the accretion disks of black holes, gamma-ray bursts and astrophysical jets.  Magnetized electron-positron plasmas in the contexts of astrophysical jets and pulsar wind nebulae are thought to be in a turbulent state, as the large separation between the energy injection scale and the dissipation scale generates an extended turbulent inertial range. We present results from theoretical and numerical efforts aimed at elucidating the turbulent dynamics of strongly magnetized, low beta, sub-relativistic electron-positron plasmas. The key concepts of Kolmogorov energy cascade and critical balance will be introduced and discussed.   

5:00pm  |  NW17-218

Dec 3, 2019

Feasibility study of electron-scale electron temperature fluctuation diagnostics

Xiang Chen

MIT

Measurements of turbulent fluctuations of physical quantities such as density, temperature, electric field, play an important role in the study of transport. The electron-scale electron temperature fluctuations (denoted as T̃­­ee)are predicted to exist in fusion plasmas and they're also predicted to contribute significantly to heat loss in a fusion reactor. Diagnostics for T̃­­ee  is crucial for better understanding plasma transport and predicting the performance of plasmas in next-generation tokamaks like ITER and DEMO. However, to date, diagnostics for T̃­­ee  are still missing. We aim to use simulations to explore the feasibility of making measurements of T̃­­ee  in the core plasma of tokamaks and stellarators, which paves the way for further hardware installation of this diagnostic on a fusion device. The simulations are carried out using CGYRO, a gyrokinetic code for collisional plasmas developed by General Atomics, and the results are compared with the ones obtained with an old code GYRO. The good agreements between the two codes lay a solid foundation for further study.

5:00pm  |  NW17-218

Dec 3, 2019

Inverse transfer of magnetic energy through magnetic reconnection

Muni Zhou

MIT

A wide range of space and astrophysical systems, such as the solar corona, heliosheath and Weibel-produced magnetic field in supernova shocks, of which the dynamics are governed by turbulence and reconnection, can be conceptualized as an ensemble of interacting flux ropes. We investigate magnetic field dynamics in a system of parallel flux ropes as well as more generic magnetically-dominated turbulent systems, focusing on the inverse magnetic energy transfer.  An analytical model is introduced and shown to capture the evolution of the main quantities of interest, as borne out by our 2D and 3D reduced magnetohydrodynamics (RMHD) and 2D particle-in-cell simulations. Magnetic reconnection is identified as the key mechanism enabling the inverse transfer and setting its properties: magnetic energy decays as T̃­­-1, where  T̃­­ is time normalized to the reconnection timescale; and the field correlation length grows as T̃­­1/2.  Critical balance is shown (by magnetic structure functions) to govern the aspect ratio of the flux ropes in 3D RMHD simulations. This quantitative description of inverse energy transfer could improve our understanding of longstanding problems such as coronal heating, galactic magnetogenesis, and high-energy emission in gamma-ray bursts.

5:00pm  |  NW17-218

Pages

© 2024 Plasma Science and Fusion Center