Student Research

Currently I'm working on simulations of fluid flow between co-rotating cylinders (Couette flow). My research focuses on the effects caused by having short cylinders. The top and bottom boundaries of the cylinders can drastically alter the fluid flow by causing an imbalance between the pressure gradient and the centrifugal force resulting in an additional fluid circulation. The purpose of my research is to investigate non-steady state fluid flow as a function of Reynolds number (rotation rate of the cylinders).

Advisor: Hantao Ji

Turbulence is considered to be one of the major sources of anomalous particle transport in fusion devices. A turbulent plasma does not confine particles as well as a laminar plasma. As such, the understanding and mitigation of tokamak turbulence is crucial for magnetic confinement fusion. One possible method of suppressing turbulence is through sheared flows. If the plasma spins rapidly, the localized shearing in the flow prevents turbulent eddies from growing tolarge scales. This is particularly the case with spherical tokamaks such as NSTX. A combination of a low aspect-ratio and powerful neutral beam injection cause the NSTX plasma to rotate quickly and set up sheared flows. Thus, it is highly probable that certain turbulent modes are suppressed in NSTX. I am currently using the GYRO gyrokinetic code to run non-linear simulations of NSTX discharges to study the effects of shearing (particularly from ExB forces) on turbulent transport. GYRO is a massively-parallel non-linear tokamak microturbulence simulation tool developed by Drs. Jeff Candy and Ron Waltz at General Atomics in San Diego, California.

Advisor:Greg Hammett

My thesis topic is "Power Balance in Neutral-beam-heated Discharges on the National Spherical Torus Experiment (NSTX)." I am investigating the effect of various phenomena on the power balance in NSTX. In particular, I am investigating three main effects.

First is the effect of neutral atoms on central power balance. Ordinarily, fusion plasmas are fully ionized (there are few or no neutral particles). However, in the edge region, neutral atoms can be plentiful. In a spherical torus, these neutral particles can interact with fast ions, whose gyroradius is a very large fraction of the minor radius of the machine. So edge neutrals might have a significant effect on central power.

Second, I am investigating the effect of magnetic ripple in NSTX. Magnetic ripple is the variation in field strength due to the finite number of coils. Preliminary data shows this to be a minor effect.

The third topic is the shift in electron density due to plasma rotation. As the plasma rotates, it feels a centrifugal force, causing the density to shift outward and violate the 'flux surface function' assumption. This could be important in determining how and where the neutral beams deposit their energy.

Advisors: David Gates and Roscoe White

The Lithium Tokamak eXperiment (LTX) is a spherical tokamak with R0 = 0.4m, a = 0.26m, BTF ~ 3.4kG, IP ~ 400kA, and pulse length ~ 0.25s. The focus of LTX is to investigate the novel, low-recycling Lithium Wall operating regime for magnetically confined plasmas. This is accomplished through placement of a heated, conformal (to the plasma last closed flux surface) shell coated with liquid lithium inside the vacuum vessel. My research focuses on the magnetic diagnostics set and associated measurements to collect data about the plasma, and by using the Equilibrium and Stability Code (ESC), perform magnetic reconstructions of the LTX equilibrium. An extensive array of magnetic diagnostics is required to characterize the experiment, including 64 Mirnov coils (single and double-axis, internal and external to the shell), 27 flux loops, 2 Rogowskii coils, and a diamagnetic loop. Diagnostics must be specifically located to account for the presence of a secondary conducting surface and engineered to withstand high temperatures and any incidental contact with liquid lithium. The diagnostic set is therefore fabricated from robust materials with heat and lithium resistance, as well as electrical isolation from the shell, and is designed to provide the most relevant data for equilibrium reconstructions using the ESC code. This is the first code that is capable of equilibrium reconstructions of very short time-scale plasmas, in which the magnetic signals are dominated by eddy current contributions from the surrounding conducting walls. Of particular significance will be changes in the plasma profiles as recycling is lowered, and the effect on confinement time, including its scaling with plasma current, toroidal field, density, and temperature. In addition to enhancing plasma physics understanding, this research will provide further practical knowledge of liquid metal walls for application to chamber technology in both inertial and magnetic fusion. Princeton Plasma Physics Laboratory

Advisors: Bob Kaita and Dick Majeski

Gas puffing, while simple and inexpensive, is an inefficient method of fueling a plasma. On NSTX, only a small portion of the puffed gas penetrates past the magnetic separatrix and contributes to the plasma density. The rest is wasted, and resides in the scrape-off layer (SOL). This cools the plasma edge and contributes to turbulence. To improve the fueling efficiency of a gas puff, a large directed velocity can be added. The Supersonic Gas Injector (SGI) on NSTX produces a Mach 4 flow of gas to test this principle. My second year project involved modeling the SGI on NSTX with DEGAS 2, a Monte Carlo neutral particle transport code. These computational studies indicate that the directed velocity alone does not produce a fueling efficiency improvement on NSTX. The SGI does produce a tightly focused gas jet, but for the conventional gas puff, the temperatures in the SOL are sufficient to produce a large number of 2-3eV dissociation product atoms. Enough of these can penetrate the magnetic separatrix to make the fueling efficiencies comparable between the two cases. These results indicate that the fueling efficiency improvements seen experimentally with the SGI are not due to geometric effects, but rather to nonlinear physics such as shielding and local cooling, which are not included in the DEGAS 2 model. The figure below gives the atomic Deuterium densities for the two cases.

Advisor: Daren Stotler

Ponderomotive barriers are regions of localized electromagnetic field oscillating at a high frequency. Depending on the type of particles (electrons, ions, clusters, molecules, or atoms) and the parameters of the field, such barriers can either attract or repel particles, acting essentially like effective potentials. Outside the parameter domain where ponderomotive barriers behave in this simple ("adiabatic") fashion, the particle behavior is generally considered hard to control, and, as we show, resembles the motion of a quantum object in a conservative field.

We have found that even these, "nonadiabatic" ponderomotive barriers can produce robust and easily controllable operations on plasma particles. In fact, employing the nonadiabatic regime allows additional flexibility for manipulating particles by means of electromagnetic fields, as compared to conventional attraction and repulsion. For example, the new techniques include a possibility of selective separation and cooling of plasma species (somewhat similar to that in atomic physics) and even formation of one-way walls, which can repel particles from one side but transmit those from another side, hence providing a novel and highly efficient current drive mechanism. All of these effects can be practiced on particles of virtually any type, from electrons to molecules or even atoms for they originate from the fundamental properties of ponderomotive interactions.

The purpose of our ongoing, mainly analytical research is to develop a general understanding of single-particle nonadiabatic dynamics in intense high-frequency fields and apply this knowledge for suggesting new advanced applications of ponderomotive barriers for plasma science and technology.

Advisor: Nathaniel J. Fisch