Richard William Harvey

Nondimensional transport scaling in DIII‐D: Bohm versus gyro‐Bohm resolved

The scaling of cross‐field heat transport with relative gyroradius ρ* was measured in low (L) and high (H) mode tokamak plasmas using the technique of dimensionally similar discharges. The relative gyroradius scalings of the electron and ion thermal diffusivities were determined separately using a two‐fluid transport analysis. For L‐mode plasmas, the electron diffusivity scaled as χe∝χBρ1.1±0.3* (gyro‐Bohm‐like) while the ion diffusivity scaled as χi∝χBρ−0.5±0.3* (worse than Bohm‐like). The results were independent of the method of auxiliary heating (radio frequency or neutral beam). Since the electron and ion fluids had different gyroradius scalings, the effective diffusivity and global confinement time scalings were found to vary from gyro‐Bohm‐like to Bohm‐like depending upon whether the electron or ion channel dominated the heat flux. This last property can explain the previously disparate results with dimensionally similar discharges on different fusion experiments that have been published. Experimen...

Fokker-Planck simulations mylb of knock-on electron runaway avalanche and bursts in tokamaks

The avalanche of runaway electrons in an ohmic tokamak plasma triggered by knock-on collisions of traces of energetic electrons with the bulk electrons is simulated by the bounce averaged Fokker-Planck code, CQL3D. It is shown that even when the electric field is small for the production of Dreicer runaways, the knock-on collisions can produce significant runaway electrons in a fraction of a second at typical reactor parameters. The energy spectrum of these knock-on runaways has a characteristic temperature. The growth rate and temperature of the runaway distribution are determined and compared with theory. In simulations of pellet injection into high temperature plasmas, it is shown that a burst of Dreicer runaways may also occur depending on the cooling rate due to the pellet injection. Implications of these phenomena on disruption control in reactor plasmas are discussed.

Fast wave and electron cyclotron current drive in the DIII-D tokamak

The non-inductive current drive from directional fast Alfven and electron cyclotron waves was measured in the DIII-D tokamak in order to demonstrate these forms of radiofrequency (RF) current drive and to compare the measured efficiencies with theoretical expectations. The fast wave frequency was 8 times the deuterium cyclotron frequency at the plasma centre, while the electron cyclotron wave was at twice the electron cyclotron frequency. Complete non-inductive current drive was achieved using a combination of fast wave current drive (FWCD) and electron cyclotron current drive (ECCD) in discharges for which the total plasma current was inductively ramped down from 400 to 170 kA. For steady current discharges, an analysis of the loop voltage revealed up to 195 kA of non-inductive current (out of 310 kA) during combined electron cyclotron and fast wave injection, with a maximum of 110 kA of FWCD and 80 kA of ECCD achieved (not simultaneously). The peakedness of the current profile increased with RF current drive, indicating that the driven current was centrally localized. The FWCD efficiency increased linearly with the central electron temperature as expected; however, the FWCD was severely degraded in low current discharges owing to incomplete fast wave absorption. The measured FWCD agreed with the predictions of a ray tracing code only when a parasitic loss of 4% per pass was included in the modelling along with multiple pass absorption. Enhancement of the second harmonic ECCD efficiency by the toroidal electric field was observed experimentally. The measured ECCD was in good agreement with Fokker-Planck code predictions

Electron cyclotron heating and current drive in ITER

Electron cyclotron (EC) power has technological and physics advantages for heating and current drive (CD) in a tokamak reactor, and advances in source development make it credible for applications in the International Thermonuclear Experimental Reactor (ITER). Strong single pass absorption makes heating to ignition in ITER particularly simple. At densities to 3.6*1020 m-3, with ohmic temperatures and wave frequency 170 GHz, heating in the plasma core is readily obtained. For outside launch of an ordinary mode (O mode) near the fundamental EC frequency, the optimized EC current drive (ECCD) efficiency ((n)IR/P) shows a linear temperature scaling at temperatures up to ~15 keV. For temperatures above 30 keV, the efficiency saturates at approximately 0.3*1020 A/W.m2 for a frequency of 220 GHz in an ITER target plasma with a toroidal field of 6 T, due primarily to harmonic overlap and to a lesser extent due to limitations arising from relativistic effects. The same efficiency can also be obtained at 170 GHz for the same plasma equilibrium and q profile except that the magnetic field is reduced to (170/220)*6 T=4.6 T. The ECCD efficiencies are obtained with the comprehensive 3-D, bounce averaged Fokker-Planck codes CQL3D and BANDIT3D

Second harmonic electron cyclotron current drive experiments on T-10

Results of the electron cyclotron current drive experiment at the second harmonic resonance on the T-10 tokamak are presented. High frequency (HF) power up to 1.2 MW was launched from the low field side. A maximum driven current of 35 kA and current drive efficiency ηCD = 0.05 A/W at an electron temperature Tc(O) = 4 keV and a density nc(0) = 1 × 1013 cm-3 were obtained. For low HF power, the current drive efficiency was less than predicted by the linear theory unless the effect of the elliptical polarization from non-perpendicular injection is considered, in which case the efficiency is close to the theoretical value. The experimental dependence of HF on the absorbed HF power indicated a strong increase of ηCD with power. Suppression of sawtooth oscillations and improvement of confinement during electron cyclotron heating has also been demonstrated

Investigation of the second harmonic electron cyclotron current drive efficiency on the T‐10 tokamak*

Experiments on second harmonic electron cyclotron current drive were done on the T‐10 tokamak using four gyrotrons. Total powers up to 1.2 MW at a frequency of 140 GHz were injected. Current generation by electron cyclotron (EC) waves was demonstrated in the experiments. The efficiency η of current generation and its dependence on plasma parameters were measured and it was shown that the efficiency is a nonlinear function of input power, more closely predicted by Fokker–Planck calculations than by linear theory. The interaction of EC waves with the tail of the electron distribution was shown to be important. It was also found that current density profile redistribution played an important role in the plasma behavior.

Modeling of Fast Wave Current Drive Experiments on DIII‐D

Modeling of fast wave current drive experiments for DIII‐D has been improved to include calculation of target temperature profiles consistent with the DIII‐D database and more accurate modeling of the launched spectrum. The calculations indicate that a measurable current will be driven by fast waves in the near‐term (30–200 kA). Modeling of the long‐range goal of 2 MA non‐inductive at high β indicates the proposed 18 MW of rf power will be adequate. The optimum frequency for the intermediate scenario is 120 MHz; this frequency selection is also adequate for the long‐term goals.

60 MHz fast wave current drive experiment for DIII-D

The DIII‐D facility provides an opportunity to test fast wave current drive appoach. Efficient FWCD is achieved by direct electron absorption due to Landa damping and transit time magnetic pumping. To avoid competing damping mechamisms we seek to maximize the single‐pass asorption of the fast waves by electrons. (AIP)

Inductive effects during second harmonic current drive experiments on T-10

Current drive during second harmonic, electron cyclotron heating experiments performed on the T‐10 tokamak have been simulated with the onetwo transport code to determine the effects of induction on the time evolution of the loop voltage and current density profile. Ray tracing shows the well focused rf power can generate centrally peaked current densities which exceed the Ohmic current densities by a factor of five, causing very peaked current profiles which will be unstable to sawteeth. A Kadomtsev model of the sawtooth shows that a limit cycle is quickly reached which maintains a broad current profile and requires generation of a negative dc component of the loop voltage localized near the magnetic axis. This negative electric field effectively reduces the measured current drive efficiency. A broader profile of driven current, as in the fundamental current drive experiment on T‐10, would not suffer this effect.

Interaction of fast waves with ions

To fully utilize the available power sources in DIII–D (FW, NBI, ECH), understanding of the synergism between the heating mechanisms is important. In this paper the ion distribution, under simultaneous application of NBI and FW, is calculated from Fokker‐Planck code CQL3D coupled to ray‐tracing code CURRAY. It is found that interaction between energetic ions and FW can be minimized or maximized by adjusting various parameters such as magnetic field, density, beam energy, and FW frequency. Specifically, in DIII–D, we find negligible interactions above 1.8 T and above 80 MHz, while the interaction increases at lower fields and frequencies. The results are compared with experiments in DIII–D including the calculated neutron rate. Energetic ion orbit losses may play an important role in the ion distribution, and this effect is being investigated.