TFTR Discharges Research Articles

Investigation ofTemeasurements discrepancies between ECE and Thomson diagnostics in high-performance plasmas in JET

For high-temperature JET and TFTR discharges, electron cyclotron emission (ECE) measurements of central electron temperature were systematically found to be up to 20% higher than those taken with Thomson scattering. In recent high-performance JET discharges, centralTemeasurements, performed with LIDAR Thomson scattering and the X-mode ECE interferometer, have been studied in a large database, including deuterium (DD), and deuterium-tritium plasmas (DT). Discrepancies betweenTemeasurements have been observed outside of the experimental uncertainties. ECE measurements, at highTe, have been found to be higher or lower than those of LIDAR, depending on the specific plasma scenario. In addition, discrepancies between the peaks of the second and third harmonic ranges of the ECE spectrum have been interpreted as evidence for the presence of non-Maxwellian features in the electron distribution function. These comparisons seem to suggest that such features can be found in most of the high-performance scenarios selected in this JET database.

Core fusion power gain and alpha heating in JET, TFTR, and ITER

Profiles of the ratio of fusion power and the auxiliary heating power qDT are calculated for the TFTR and JET discharges with the highest neutron emission rates, and are predicted for ITER. Core values above 1.3 for JET and 0.8 for TFTR are obtained. Values above 20 are predicted for ITER baseline plasmas.

Testing a model for triggering sawtooth oscillations in tokamaks

Simulations of JET and TFTR discharges with the BALDUR integrated modeling code are used to test a sawtooth model that consists of sawtooth triggering mechanisms [Porcelli et al., Plasma Phys. Contolled Fusion 38, 2163 (1996)] together with a modified version of the Kadomtsev sawtooth reconnection model [Kadomtsev, Sov. J. Plasma Phys. 1, 389 (1975)]. In simulations of 12 TFTR low confinement (L-mode) and 5 JET high confinement (H-mode) discharges, sawtooth triggering conditions are examined in order to determine which physical mechanisms are responsible for the crashes. It is found that most sawtooth crashes in the simulations are triggered by the m=1 resistive internal kink instability in plasmas that are in the semicollisional regime. However, in some discharges, some of the sawtooth crashes are triggered as a consequence of the driving force for the internal kink instability overcoming the fast ion stabilization. In rare instances, a sawtooth crash is triggered when the driving force for the internal kink instability overcomes the stabilization produced by the diamagnetic rotation of thermal ions. Generally, the median sawtooth period is found to increase as the magnetic reconnection fraction is increased. Optimal agreement with experimental data for the discharges considered is obtained with a magnetic reconnection fraction of approximately 37% although there is considerable scatter in the sawtooth periods observed in each discharge, both in the experimental data and in the simulations.

Observations concerning the injection of a lithium aerosol into the edge of TFTR discharges

A new method of actively modifying the plasma-wall interaction was tested on the Tokamak Fusion Test Reactor. A laser was used to introduce a directed lithium aerosol into the discharge scrape-off layer. The lithium introduced in this fashion ablated and migrated preferentially to the limiter contact points. This allowed the plasma-wall interaction to be influenced in situ and in real time by external means. Significant improvement in energy confinement and fusion neutron production rate as well as a reduction in the plasma Zeff have been documented in a neutral beam heated plasma. The introduction of a metallic aerosol into the plasma edge increased the internal inductance of the plasma column and also resulted in prompt heating of core electrons in ohmic plasmas. Preliminary evidence also suggests that the introduction of an aerosol leads to both edge poloidal velocity shear and edge electric field shear.

Rollover: confinement degradation during TFTR discharges

The time evolution is described of TFTR neutral beam heated plasmasthat often featured a decreasing energy content called `rollover'. Rolloveraffects the gross plasma parameters such as energy content and fusion powerproduced, as well as the central plasma parameters and the plasma transportcoefficients. Rollover was studied using a database of all the TFTR plasmasproduced over a two year period. The plasma behaviour at the time of peakenergy content was related to the behaviour at a later quasi-equilibrium time.The main conclusion is that limiter power handling issues were the underlyingcause of TFTR rollover. This study re-emphasizes the fact that power handlingis an important issue for plasma facing components in tokamaks.

Bursts of electron cyclotron emission during disruptions of high β discharges in TFTR

During sudden disruptions of tokamak plasma discharges at high β, short and intense bursts of electron cyclotron emission (ECE) are observed. A unique combination of two fast grating polychromator instruments located at different toroidal positions is used to measure the emission and characterize these bursts. New insights into the 3D dynamics of these disruptions and the accompanying bursts of ECE have been obtained. An explanation of the bursting is presented which is consistent both qualitatively and quantitatively with observations, predicting for example radiation enhancement factors of about 10. Bursting can be explained not in terms of enhanced excitation of emission but rather in terms of reduction of absorption of thermal emission. Bursting is consistent with a modification to the electron distribution function f e due to a rapid energy or particle exchange between hot electrons and cold electrons from the edge, momentarily reducing the velocity gradient of fe in the thermal region. Large ELM (edge-localized mode) events also exhibit bursts of ECE due to a similar sequence of events.

Distinction between DD and DT neutrons in the TFTR neutron collimator

A difficult part of the analysis of the neutron collimator data is the distinction between contributions from DD and DT neutrons. This is important for the majority of TFTR discharges, which use neutral deuterium heating beams and relatively small amounts of tritium. At present, the DT-neutron contribution is determined from a comparison of the chord-integrated neutron collimator data with measurements of the DT-neutron source strength from general wide-angle detectors, using an iterative computational method. A more satisfactory solution should be obtained with detectors that measure the contributions from DD and DT neutrons, and gamma radiation, directly. Pulse-shaping techniques are insufficient, since the detectors in the neutron collimator are operated in current mode. The presently used data analysis is described and the concept of a new multilayered plastic detector that makes use of the differences in the attenuation of the various radiation components is presented. The dimensions of the detector are based on calculations of the attenuation for the expected neutron energy spectrum.

Distinction between DD and DT neutrons in the TFTR neutron collimator (abstract)

A difficult part of the analysis of the neutron collimator data is the distinction between contributions from DD and DT neutrons. This is important for the majority of TFTR discharges, which use neutral deuterium heating beams and relatively small amounts of tritium. At present, the DT-neutron contribution is determined from a comparison of the chord-integrated neutron collimator data with measurements of the DT-neutron source strength from general wide-angle detectors, using an iterative computational method. A more satisfactory solution should be obtained with detectors that measure the contributions from DD and DT neutrons, and gamma radiation, directly. Pulse-shaping techniques are insufficient, since the detectors in the neutron collimator are operated in current mode. The presently used data analysis is described and the concept of a new multilayered plastic detector that makes use of the differences in the attenuation of the various radiation components is presented. The dimensions of the detector are based on calculations of the attenuation for the expected neutron energy spectrum.

Measurements of confined alphas and tritons in the MHD quiescent core of TFTR plasmas using the pellet charge exchange diagnostic

The energy distributions and radial density profiles of the fast confined trapped alpha particles in DT experiments on TFTR are being measured in the energy range 0.5 - 3.5 MeV using the pellet charge exchange (PCX) diagnostic. A brief description of the measurement technique which involves active neutral particle analysis using the ablation cloud surrounding an injected impurity pellet as the neutralizer is presented. This paper focuses on alpha and triton measurements in the core of MHD quiescent TFTR discharges where the expected classical slowing-down and pitch angle scattering effects are not complicated by stochastic ripple diffusion and sawtooth activity. In particular, the first measurement of the alpha slowing-down distribution up to the birth energy, obtained using boron pellet injection, is presented. The measurements are compared with predictions using either the TRANSP Monte Carlo code and/or a Fokker - Planck Post-TRANSP processor code, which assumes that the alphas and tritons are well confined and slow down classically. Both the shape of the measured alpha and triton energy distributions and their density ratios are in good agreement with the code calculations. We can conclude that the PCX measurements are consistent with classical thermalization of the fusion-generated alphas and tritons.

Fokker-Planck 3-D modelling of axisymmetric collisional losses of fusion products in TFTR

The results of a 3-D (in constants of motion space) Fokker-Planck simulation of the collisional losses of fusion products in axisymmetric DT and DD discharges in TFTR are presented. The distributions of escaped ions over poloidal angle and pitch angle, and their energy spectra, are obtained. Axisymmetric collisional losses of fusion products are found to be less than 2 to 5%. The distribution of confined fusion products is shown to be strongly anisotropic and non-uniform in the radial co-ordinate, mainly for slowed-down fusion products with small longitudinal energy. A comparison between these modelling results and experimental data is made

Studies of energetic confined alphas using the pellet charge exchange diagnostic on TFTR

Results from recent DT experiments on TFTR to measure the energy distribution and radial density profile of fast confined alphas with the use of Li pellets and neutral particle analysis are presented. When a pellet is injected into the plasma, a toroidally extended ablation cloud is formed. A small fraction of the fusion alphas incident on the cloud is converted to helium neutrals as a result of electron capture processes. The escaping energetic helium neutrals are analysed and detected by the neutral particle analyser. Radially resolved energy spectra of trapped confined alphas in 0.5-2 MeV range and radial alpha density profiles are presented in this paper. The experimental data are compared with modelling results obtained with the TRANSP Monte Carlo code and with a specially developed Fokker-Planck post-processor (FPP) that uses the alpha source distribution produced by TRANSP. Comparison of the experimental data with TRANSP and FPP shows that the alphas in the plasma core of sawtooth free discharges in TFTR are well confined and slow down classically. The energy and radial profiles distributions outside the plasma core show the influence of stochastic ripple losses on alphas. Measurements for sawtoothing plasmas show a significant outward radial transport of trapped alphas

Systematic comparison of a theory-based transport model with a multi-tokamak profile database

A theoretical model of flux-surface-averaged radial transport in tokamaks has been tested and calibrated against a well-documented set of temperature and density profiles from a pre-defined set of discharges from seven tokamaks. The transport theory includes neoclassical, drift/ηi, circulating electron mode, kinetic ballooning, neoclassical MHD, and resistive ballooning effects. Allowing for no explicitly adjustable free parameters, the nominal theory results are compared with experimental density and temperature profiles from the reference set of discharges from Alcator-C, ASDEX, ISX-B, TFTR, DIII-D, and JET. Profile results are also given for an additional set of six discharges including Alcator-C, ASDEX, DIII-D, TFTR, JET, and JFT-2M in addition to five TFTR discharges as part of a ρ* and β scan. Employed is a statistical model of calibration/modeling and measurement variances allowing detailed analysis of results and further calibration of the model along with a well defined procedure for setting the time dependent boundary conditions at an appropriate location just inside the inner-most "closed" magnetic flux surface.

Z eff profiles measured by visible and soft x-ray bremsstrahlung arrays in TFTR (abstract)a)

The Zeff profiles from the tangential array of visible bremsstrahlung (VB) emission (5235 Å) have shown several interesting features in TFTR. The observed Zeff profile was flat during the ohmic phase and hollow during the neutral-beam heating phase. In order to crosscheck the measured Zeff profiles from the VB array, a poloidal array which measures bremsstrahlung emission in the soft x-ray region has been analyzed. The analysis of array measurements in the two different spectral regions was performed on the same geometry determined from the Abel inversion of the electron density (interferometry). The two independent measurements of Zeff profiles have been consistent with each other for most of the discharges in TFTR. However, in a discharge where the two measurements are inconsistent, the Zeff profile deduced from the VB array indicates some degree of problems. In this paper, the detailed analysis of the bremsstrahlung array (both poloidal and tangential) measurements is presented. Particularly, the effects of geometry in the inversion process, such as a poloidal elongation of flux surfaces, are investigated.

Observation of a localized transition from edge to core density turbulence in the TFTR tokamak.

A localized transition zone is observed between turbulent, long-wavelength density fluctuations (${\mathit{k}}_{\mathrm{\ensuremath{\perp}}}$2 ${\mathrm{cm}}^{\mathrm{\ensuremath{-}}1}$) in the core and edge of beam-heated discharges in TFTR. The core turbulence is unimodal while the edge has two counterpropagating modes, both of which are spectrally distinct from the core fluctuations. The fluctuation amplitude in the core scales with the global energy confinement time while in the edge it does not. These observations suggest that distinct modes are responsible for turbulence in the two regions and that only the core mode is directly related to global confinement.

Analysis of erosion and transport of carbon impurity in the TFTR inner bumper limiter region

Carbon sputtering and transport on the TFTR inner graphite bumper limiter is investigated with the impurity transport code REDEP. Analysis is carried out for a series of ohmic discharges in TFTR. Predictions for Z eff in the core plasma agree well with in situ experimental measurements. Runaway self-sputtering of carbon is predicted at low densities and high edge plasma temperatures when the limiter surface was purged of deuterium. Surface erosion and deposition is analyzed. In general, redeposition reduces the peak erosion by about a factor of 5. Analysis is also carried out for a typical neutral beam heated discharge with a noncircular plasma. Spatial surface erosion and deposition profiles are compared qualitatively with beta backscattering measurements of metal deposition found on the limiter.

Observation of density fluctuations from the infrared bremsstrahlung emission on TFTR

The infrared emission of beam heated TFTR discharges is studied using the high collection efficiency TFTR Thompson scattering system. It is found that emission in the vicinity of 9800 Å can be used to study fluctuations in the bremsstrahlung emission. Low-frequency (<40 kHz) fluctuations that are associated with MHD modes and high-frequency (≳200 kHz) bursts of fluctuations are observed. Both coherent and incoherent fluctuations exist. In some cases, the observed coherent fluctuations can be interpreted as being due to density fluctuations as high as 5%. The rms value of broadband density turbulence in the near central regions of the plasma is estimated to be about 1%.

Measurement of the nonaxisymmetric heat load distribution on the first wall of TFTR due to locked modes

The first wall of TFTR is covered in large part (23%) by an inner-wall bumper limiter which is the primary power handling structure in TFTR. The limiter is comprised of more than 2000 tiles, and is instrumented with a large number (≳100) of thermocouples in a two-dimensional (2D) array, primarily for protection of the wall. While only about 5% of the tiles are monitored, this thermocouple system is nevertheless capable of mapping details in the nonaxisymmetric, as well as symmetric, heat load patterns encountered under different conditions. In particular, helical heating patterns are observed in discharges which have locked modes. The helical patterns clearly match the expected trajectories based on the m/n mode numbers obtained from Mirnov coils (m/n=2/1 and 4/1), so that the thermocouple system can and was used to identify the existence and mode number of a locked mode. While TFTR discharges rarely suffer from locked modes, locked modes always alter the heating pattern. The locked modes are found to very significantly redistribute the heat load for both ohmic and NBI heated discharges. Locked modes can make what were the coldest areas into the hottest areas, and vice versa. Locked modes also can alter the heat pattern resulting from the frequent disruptions which occur as a result of a locked mode.

Status and Plans for TFTR

AbstractRecent research on TFTR has emphasized optimization of performance in deuterium plasmas, transport studies and studies of energetic ion and fusion product physics in preparation for the D-T experiments that will commence in July of 1993. TFTR has achieved full hardware design parameters, and the best TFTR discharges in deuterium are projected to QDT of 0.3 to 0.5.The physics phenomena that will be studied during the D-T phase will include: tritium particle confinement and fueling, ICRF heating with tritium, species scaling with tritium, collective alpha-particle instabilities, alpha heating of the plasma and helium ash buildup. It is important for the fusion program that these physics issues be addressed to identify regimes of benign alpha behavior, and to develop techniques to actively stabilize or control instabilities driver by collective alpha effects.

Effect of the boundary plasma on TFTR ohmic discharges

The role of the boundary plasma in determining the power and particle balance of tokamak discharges is discussed. Detailed boundary plasma measurements of edge density, edge temperature, deuterium influx and carbon influx are reported from ohmically heated TFTR discharges over a range of plasma densities. The experimental results are compared with predictions from a simple zero-dimensional model based on power and particle balance. Reasonable agreement is obtained. More detailed impurity modelling is performed with the LIM impurity production and transport code. The complementary modelling approaches reveal, amongst other results, the important role of the sputtering yield at the limiter in determining the central effective charge of the discharge, the ability of a densified boundary plasma to screen impurities from the central plasma and the importance of cross-field particle transport to the TFTR limiter. The authors demonstrate that, under the conditions of this experiment, the behaviour of the boundary plasma and its effect on the central plasma appear to be explicable using rather simple considerations of power and particle balance

Deuterium recycling, confinement and limiter flux in TFTR

The neutrals code DEGAS and the transport code SNAP were used to model recycling during steady state phases of Ohmic and neutral beam-heated discharges in TFTR. The flux of deuterium from the inner limiter is calculated to be 15–45 times the total D α emission rate, with ≈ 2/3 of the deuterium flux resulting from D + flow in the scrape off to the limiter, and ≈ 1/3 resulting from D° scattering off the limiter. The total D + ionization rate in the core plasma is 6–24 times the total D α emission rate, and is larger than the total neutral beam fueling rate. The D α emission rate, limiter flux, and core ionization rates scale approximately as the square of the volume averaged electron density.