Search

The formation of electron beams and their transport in a mirror-type guiding magnetic field are modeled numerically for two types of an experimental electro-optical system in the accelerating diode of an electron source with a plasma low-pressure arc-discharge emitter. The observed behavior of currents is interpreted and the limiting values of currents transported in the system are obtained.

A problem of surface melting under the impact of plasmas is one of the most important in the development of future magnetic confinement fusion reactors. The expected high heat loads can lead to melting of tungsten chosen as the material of a divertor and the first wall of ITER. The tungsten melt can move under the action of forces and quickly deform a surface. This article shows that at high temperatures, a tungsten vapor can be considered as a conductive material with conductivity high enough to be taken into account during simulation of current through a sample. We describe the mechanism of current generation by the thermo-emf due to non-uniform heating of the vapor/condensed substance interface without external sources of charges. Even without direct contact with the external plasma, the density of this current is high enough to cause noticeable movement of the melt in an external magnetic field. Simulation shows that the melt moving observed at BETA may be caused not by a current of an electron beam but by non-uniformity of heating and can be reproduced by any other similar heat source in the external magnetic field.

Melting of the surface of tungsten exposed to a pulsed electron beam has been simulated numerically. Comparison of the experimentally measured at BETA facility time dependence of the radius of the molten region with the calculated data has shown that the surface cooling caused by evaporation has a significant effect on the temperature distribution and melting of the material at sufficiently high densities of the surface heating power. This result validates the created theoretical model of the tungsten melting and evaporation in exposure to a pulsed electron beam. The studied mechanism of the limitation of the surface temperature is different from the well-studied vapor shielding. The presented model is a step to correct interpretation of the erosion caused by the melt motion and splashing in exposure to the ITER-relevant pulsed heating by electron beam.

The paper reports on a comparative study of two different types of diode system for producing intense submillisecond electron beam in a source with low-pressure arc plasma emitter. The first is a planar multi-aperture diode comprising cathode and anode performed as molybdenum grids with 241 round openings of few millimeters in diameter. In the second system electrons are accelerated in a sheath between the grid-stabilized cathode plasma and the open boundary of beam-generated anode plasma. It was demonstrated in experiments that latter system should be advantageous in terms of maximally achievable beam current and total beam energy content, but there are drawbacks connected with slow anode plasma build-up that require further study.

Experimental research on behavior of rolled tungsten plates under intense transient heat loads generated by a powerful (a total power of up to 7 MW) long-pulse (0.1–0.3 ms) electron beam with full irradiation area of 2 cm2 was carried out. Imaging of the sample by the fast CCD cameras in the NIR range and with illumination by the 532 nm continuous-wave laser was applied for in-situ surface diagnostics during exposure. In these experiments tungsten plates were exposed to heat loads 0.5–1 MJ/m2 with a heat flux factor (Fhf) close to and above the melting threshold of tungsten at initial room temperature. Crack formation and crack propagation under the surface layer were observed during multiple exposures. Overheated areas with excessive temperature over surrounding surface of about 500 K were found on severely damaged samples more than 5 ms after beam ending. The application of laser illumination enables to detect areas of intense tungsten melting near crack edges and crack intersections.

A new test facility for experimental simulation of transient heat load expected in the ITER divertor during unmitigated events is developed. Application of a long-pulse (0.1–0.3 ms) wide-aperture (up to 2 cm 2 ) electron beam as a heating device provides powerful energy loads at a tungsten target with F HF > 250 MJ m −2 s −0.5 . Dynamics of tungsten particles in the ablation plume is investigated with a novel for PSI experiments small-angle laser light scattering technique. The threshold of intense droplet generation and dynamics of particles sizes are estimated.

This paper presents a model for calculating deformations and mechanical stresses around a crack normal to surface that appeared under pulsed heat load. The model was applied to calculation of stresses that may lead to formation of cracks along the surface, which are observed when tungsten is exposed to ITER-relevant heat load. It was found that such stresses might be not negligibly small in comparison with the ultimate tensile strength, and thus the appearance of cracks normal to the surface may leads to development of cracks parallel to the surface. The calculated deformation of the region around a crack is in good agreement with the experimental data. The deformations calculated can be a basis for experimental detection of formation of cracks normal and parallel to the surface. Keywords: Pulsed heat load, Tungsten, Cracking, Theoretical simulation, Elasticity, Electron beam, Profilometry

Damage of divertor tungsten tiles as a result of transient plasma processes is one of the key problems of the ITER tokamak. Even if the temperature of the divertor tiles is maintained below the melting threshold, their thermal deformation, accumulation of residual stresses and cracking will occur. Cracking reduces the resistance of the metal to subsequent loads, therefore, the study of this process is important. In materials science, non-destructive methods for measuring mechanical stresses are very limited. In this work, a new method is proposed that allows non-contact investigation of stresses in polycrystalline tungsten plates by their bending. A corresponding diagnostic system and theoretical model are implemented. A number of experiments were carried out using this diagnostic system at the BETA facility at the Budker Institute to study the dynamics of tungsten plates bending with different thickness, subjected to pulsed heating of different intensities. The experimental results showed the qualitatively expected process of the plate bending, as well as good agreement with the theoretical model. This proved that the method described in the work is promising for use in order to study the residual stresses in a pulse-heated metal sample.

Experiments on the effect of fast heat loads on the surface of tungsten were carried out on the BETA facility at the Budker Institute. Tungsten samples were uniformly heated by an electron beam with a heat flux factor below the melting threshold. During and shortly after exposure, the 2D surface temperature distribution was measured, as well as the temperature history on selected surface areas. Active diagnostics using the scattering of CW laser light on a surface exposed by the electron beam allowed us to monitor the damage dynamics. At the heating stage, an increase in the surface roughness occurred, caused by inhomogeneous elastic and plastic deformations of the heated layer. As the cooling progressed, the residual plastic deformations remained. Simultaneously with the modification of the surface, bending of samples with a thickness of 3-4 mm occurred. The bending dynamics of the sample was measured by the intensity of a converging laser beam reflected from the back surface of the sample, polished to a mirror state. The residual sag due to bending increases with the heat load similarly as residual roughness of the front surface of the sample. These data, together with simultaneously measured temperature dynamics and the spatial heating profile, can provide an experimental basis for the numerical calculation of the residual stresses in the sample. The data obtained in situ were compared with those measured outside the vacuum chamber with X-ray diffraction, optical profiler, and optical interferometer. At the stage of cooling, after a sufficient intensity of heating, the second stage of damage took place — the cracking of the surface layer. The time before the start of this relatively fast process usually exceeded the time to achieve a DBTT by 1–4 orders of magnitude.

The generation of a 250-μs-wide electron beam in a plasma-emitter diode is studied experimentally. A plasma was produced by a pulsed arc discharge in hydrogen. The electron beam is extracted from a circular emission hole 3.8 mm in diameter under open plasma boundary conditions. The beam accelerated in the diode gap enters into a drift space in the absence of an external magnetic field through a hole 4.1 mm in diameter made in the anode. The influence of electron current deposition at the edge of the anode hole on the beam’s maximum attainable current, above which the diode gap breaks down, is studied for different accelerating voltages and diode gaps. The role of processes occurring on the surface of the electrodes is shown. For an accelerating voltage of 32 kV, a mean emission current density of 130 A/cm2 is achieved. The respective mean strength of the electric field in the acceleration gap is 140 kV/cm. Using the POISSON-2 software package, the numerical simulation of the diode performance is carried out and the shape of steady plasma emission boundaries in the cathode and anode holes is calculated. The influence of the density of the ion current from the anode plasma surface on the maximum attainable current of the electron beam is demonstrated.

Development of a new linear device for confinement of fusion plasmas is under way in the Budker Institute of Nuclear Physics, Novosibirsk. The new device combines features of existing GOL-3 and GDT...

Sub-THz radiation can be generated by conversion of plasma waves into electromagnetic (EM) radiation in a plasma with strong Langmuir (LT) turbulence produced via a two-stream instability of a high current relativistic electron beam (REB). Nonlinear plasmon-plasmon merging results in the generation of photons nearby the 2nd harmonic of the plasma frequency 2ω p (“2ω p -process”). For plasma densities of 1014 − 1015 cm−3, these frequencies are in the range of sub-THz waves at 370–570 GHz. The specific power density of sub-THz-wave emission from plasmas in the multi-mirror magnetic trap GOL-3 (at BINP) during injection of a 10-μs-REB with a current density of about 1 kA/cm2 at plasma densities n e ≈ 5∙1014 cm−3, electron temperatures T e ≈ 1.5 keV and magnetic induction B ≈ 4 T was measured to be approx. 1 kW/cm3 in the frequency band around 300 GHz. In the case of a weakly relativistic 100-μs-electron beam (90 keV) with 250 A/cm2 the corresponding results are 700 W/cm3 around 90 GHz with an efficiency of 1–2 % at n e ≈ 3∙1013 cm−3 (total power ≈ 30 kW). Theoretical investigations show that at a density of n e ≈ 3∙1015 cm−3 and a turbulence level of 5 % the generated sub-THz power can reach ≈ 1 MW/cm3.

Surface melting of tungsten under exposure to a pulsed electron beam was simulated numerically, the evaporation process taken into account. The calculation is based on the experimental time dependence of the total beam power. The model of the tungsten heating process is based on solving the two-phase Stefan problem. The position of the phase boundary depends on discontinuous time- and space-nonlinear coefficients and boundary conditions. The aim of the study is to provide a detailed resolution of the heat flow deep into the material with a fine spatial grid step. As compared with the size of the tungsten plate, the heating depth is very small. The problem statement under consideration is multiscale. Further expansion of the model involves taking into account microcracks. Micro-cracks occur during the cooling process after exposure and affect the temperature of the tungsten surface during the subsequent heating process. The article presents a modeling of cracks of different geometries typical for this process. The results of the calculations correlate with the experimental data obtained on the experimental test facility BETA at BINP SB RAS.

Results of experiments on injection of a sequence of several electron beam pulses (80–90 keV, 20–75 A, ∼30 μs) into 1019–1021 m−3 deuterium plasma are presented. The first beam pulse was injected into deuterium gas, the following pulses propagated through residual plasma that survived the inter-pulse periods. A typical pulse duration was of the order of 106 ω p − 1 that is much longer than most of the interaction-related processes. However, we noticed significant differences in plasma processes between the first pulse and the rest ones. A suggested physical factor behind those differences is longitudinal plasma current that develops during the first beam pulse.

A novel BETA test facility (Beam of Electrons for materials Test Applications) was developed at the Budker Institute to study the erosion of materials directly during the impact of intense thermal shocks. A powerful (up to 7 MW) long pulse (100–300 μs) electron beam is applied for experimental simulation of fast transient heat loads with the intensity probable in the ITER divertor. The heat flux parameter on a target can be widely varied (FHF = 10–300 MW m−2 s0.5) from a value significantly below the melting threshold to a value much higher, within the area of about 1 cm2. The use of an electron beam to simulate the thermal impact on the material surface makes it possible to employ a variety of optical diagnostics for in situ observations of the dynamics of surface erosion processes during intense thermal shocks. These distinctive features make BETA a promising tool in the research of material surface erosion mechanisms and for experimental verification of various analytical and numerical models associated with these mechanisms. The first results obtained with this facility include fast (10 μs exposure) imaging of the heated target in the near-infrared range and in the reflected light of 532 nm continuous wave (CW) laser, visualization of ejected tungsten particles using fast ICCD and CCD cameras with the minimal exposure of 2 μs and 7 μs respectively. The dynamics of dust particles ejected from the heated surface is investigated using a multichannel recording of the light of 532 nm CW-laser scattered on the dust particles. The present paper describes the first results of use of two new in situ methods: continuous recording of light scattered from the tungsten surface and three-dimensional tracking of tungsten particles using three viewing angles. The first method makes it possible to observe the dynamics of development of roughness and cracking of the polished tungsten surface, which manifest themselves as two successive processes separated by a large time delay. The second method allows us to track dust particles ejected from the melt layer, and thus determine the time and place of particle ejection from the surface.

The paper reviews recent experimental results from GOL-3. Currently efforts are focused on further development of a physical database for multiple-mirror confinement systems and also on an upgrade of plasma heating systems of GOL-3 device. In general, current GOL-3 parameters demonstrate good prospects of a multiple-mirror trap as a fusion reactor.

A test facility developed at the Budker Institute is designed for experimental simulation of effects of ITER-scale transient heat loads on plasma facing materials. Employing a long-pulse electron beam for creation of transients has advantages of low direct pressures on surface of material and lesser limitation of power density by vapor shielding. In addition, relative modest attendant background light allows using a set of diagnostics for observation of erosion dynamics. The set includes fast imaging in the near infrared, scattering of light of continuous wave laser on target surface and on droplets ejected from the melt layer.

Wide-area (2 cm2) high-power (up to 7 MW) submillisecond electron beam source was applied for generation of intense pulsed heat loads on tungsten sample. Various diagnostics sets were used for in-situ research of the surface damage: IR imaging of the front view of the target, capturing of the reflection pattern of continuous wave laser radiation, recording of intensity of thermal radiation from several spots of the surface. Sample was exposed to a various heat loads near and above melting threshold. Formation of the crack network on the surface and its development with successive pulses was observed. Two-dimensional temperature distribution was obtained after heating at the cooling stage. Melting and spalling of material were shown on severe damaged target. Propagation of the cracks along the surface at a depth of 0.1 – 0.2 mm was revealed with transverse microsections. Circular motion of the molten layer was found with subsequent exposures.

The results of experiments and numerical simulation of compression of an intense electron beam in increasing field of magnetic mirror and its transmission through a linear magnetic system are presented. This system is used as a test module for a section of the GOL-3 magnetic trap. The study was carried out for the beam with a spread in initial electron pitch angles and a non-uniform radial current distribution. An analytical estimate of the vacuum beam-current limit was obtained taking into account the initial pitch angles of electrons and the beam compression in the magnetic field. The simulation results are compared with the results of experiments; in both cases similar phenomena of beam reflection and diode breakdown are observed.

The paper presents an overview of plasma-material interaction studies at the Siberian Synchrotron and Terahertz Radiation Centre. The measurements of recrystallization and surface texturing are demonstrated. The growth of grain size after the exposure by electron beam was detected using two-dimensional diffractometry. The orientation of crystal structure was detected in the tungsten irradiated by deuterium plasma. The residual stresses in irradiated tungsten were measured. The predicted structure of deformations after pulsed heat load is confirmed. First in-situ experiments confirmed the presence of crystal plane rotation effect during a pulsed heat load.

A test facility for experimental simulation of transient heat load expected for ELMs type I events in ITER is developed at BINP SB RAS. Dynamics of tungsten particles in the ablation plume is investigated by small-angle light scattering technique and using fast CCD and ICCD cameras. The threshold of intense particle generation, sizes and velocities of particles ejected from the surface are estimated.

A test facility for experimental simulation of transient heat loads in ITER divertor with the use of high power electron beam is developed at the Budker Institute of Nuclear Physics. This report presents an experimental study of the absorption of the electron beam with an incident heat power flux of 10-50 GW/m2 on the area of about 2 cm2 of a tungsten target. The electron beam has duration of 0.1-0.3 ms and electron energy of 80-95 keV, Diagnostics for measuring of the beam parameters on the target are briefly discussed. Results of measurement of the beam profile and calorimetry of beam energy deposited in tungsten sample at ELM-like heat load are presented.

A novel high-power (10 MW) sub-millisecond electron beam is developed for injection into the open (linear) plasma devices. The beam is produced by extraction of electrons from a plasma of pulsed ar...

Magnetic activity of plasma during the injection of 2-8 MW non-relativistic electron beam of sub-millisecond duration in GOL-3 is studied.The main parameters of fluctuations of the magnetic field a...

First experimental results on microwave radiation at double plasma frequency from plasma heated by ~100 keV, 2-10 MW electron beam of ~100 μs duration in GOL-3 are presented. Measurements were done...

An incoherent Thomson scattering system for observing local unsteady cavities that form in a turbulent plasma in the course of Langmuir collapse is described. Using this diagnostics, the density cavities in a laboratory plasma with developed strong Langmuir turbulence were directly observed for the first time and their spatial and temporal characteristics were determined.

Experiments on strong Langmuir turbulence (LT) driven by electron beam are reported. The technique of cold high-current relativistic electron beam (REB) permits to set up experimental conditions that are practically important but difficult for theoretical treatment of LT. These conditions include strong kinetic effects of plasma non-Maxwellian electrons, ion-acoustic oscillations which are weakly damped due to plasma non-isothermality and dispersion of Langmuir waves that are considerably modified by external magnetic field. A relatively dense plasma permits the use of the Thomson scattering method for observation of spectra of plasma fluctuations, electron distribution function, and local dynamics of plasma density. LT is studied in two operating modes that are characterized by moderate and increased current of REB. The experimental results with moderate REB current do not support the widely accepted picture when most of the Langmuir oscillations are trapped in density cavities. The energy flow through turbulence to plasma electrons is explained without major contribution of fully developed collapse, whereas, with increased REB current dynamic density cavities of spatial scale much less than the size of turbulent region are directly observed.

The regime of strong Langmuir turbulence characterized by the plasma nonisothermality and by the presence of an appreciable non-Maxwellian hot-electron component was experimentally studied. Turbulence was excited in the preliminary produced plasma by the relativistic electron beam. Thomson scattering of laser IR radiation served as the main diagnostic method. The spatial spectra of the Langmuir turbulence and of the attendant ion-sound turbulence were studied using Thomson collective scattering. Thomson incoherent scattering was used for studying the plasma electron distribution function and searching for the local dips of plasma density. Stark spectroscopy of turbulent microfields and the method of observation of plasma radiation at the double plasma frequency were also used. Based on the experimental data, the mechanism of Langmuir oscillation damping by plasma electrons was analyzed. The Langmuir wave conversion induced by the ion-sound turbulence is the most probable channel for energy transfer from the turbulence to plasma electrons, the low-frequency fluctuations being the direct consequence of the strong Langmuir turbulence.

Recently high plasma parameters were reached in experiments on plasma heating in the multi-mirror open trap GOL-3. In these experiments deuterium plasma with density 10-10 cm in the 12-m trap with corrugated magnetic field of 4.8T/3.2T was heated by a relativistic electron beam of eU ≤ 0.8 MeV, I ≤ 20 kA, j ~ 1-2 kA/cm, pulse duration ~ 9 μs and angular spread θ ≤ 0.2 rad. The electron temperature Te ~ 1-4 keV and ion one to Ti ~ 1-2 keV were reached. After the beam pulse the electron temperature of the plasma quickly (~ 20 μs) decreased to ~100 eV. In turn, this leads to an increase in the rate of cooling of ions through the ion-electron collisions, which, together with particle losses determines the energy confinement time of plasma as 0.5-1 ms. To increase the plasma parameters a prolonged heating of the plasma by the electron beam is proposed. The paper considers basic physical phenomena in the beamplasma interaction and required parameters of the longpulse beam. A choice of long pulse beam parameters is based on the obtained experimental results and scalings. Then estimates of the expected plasma parameters under the influence of the new beam have been done.

A collective CO 2 laser scattering method for experimental studies of local dynamics of strong Langmuir turbulence driven by a relativistic electron beam is described. The preliminary results of temporal behavior of Langmuir oscillations are presented.

The Thomson scattering technique has been employed for the observation of turbulent waves excited by a relativistic electron beam. The frequency and k-spectra of the Langmuir waves are measured both in the excited and damped regions. The waves interact directly with the beam electrons inside a narrow section of k-sppace amid the region occupied by the non-resonant waves, which concentrate near k ∼ ωpe/c. Both resonant and non-resonant waves hold equal (to within an order of magnitude) amounts of energy. Incoherent scattering of the second harmonic of a Nd : glass laser is used to measure the time history of the electron distribution function including the super thermal tails. The typical beam and plasma parameters are as follows: (ne ∼ 1015 cm−3, nb/ne ∼ 10−4, B = 1–4 T, tb = 100–200 ns). The peculiarity of these experiments is the presence of a strong magnetic field (ωBec/ωpe VTe ⪢ 1, but cope ⪢ ωBe) which, together with plasma non-isothermality (Te ⪢ Ti), sets conditions typical for laboratory and space plasmas. At present, the processes under these conditions have not been adequately investigated, either experimentally or theoretically, especially for our case of developed Langmuir turbulence, when the spatial and temporal scales far exceed those for a single caviton. This work is centred on the study of developed Langmuir turbulence in a magnetized plasma.

Strong Langmuir turbulence driven by a relativistic electron beam has been investigated by the laser scattering technique. The special features of the experiments are as follows: (i) in the regime under study, Langmuir turbulence is well developed, and the temporal and spatial scales of the turbulent region far exceed those of a single caviton; (ii) the dispersion of the electron plasma waves is governed by the magnetic field, although ωpe≫ωBe; and (iii) the temperature of plasma electrons is much higher than that of ions, i.e., the damping of ion‐acoustic waves is small. k‐Spectra of electron plasma waves are measured in a broad spectral region by means of CO2‐laser scattering. Criterion of modulational instability of observed spectra is estimated. The role of conversion and scattering of Langmuir waves by ion‐acoustic waves in nonlinear energy transfer is discussed.

There are described electromagnetic spectra of radiation emitted by magnetized plasma during sub-relativistic electron beam in a double plasma frequency band. Experimental studies were performed at the multiple-mirror trap GOL-3. The electron beam had the following parameters: 70–110 keV for the electron energy, 1–10 MW for the beam power and 30–300 μs for its duration. The spectrum was measured in 75–230 GHz frequency band. The frequency of the emission follows variations in electron plasma density and magnetic field strength. The specific emission power on the length of the plasma column is estimated on the level 0.75 kW/cm.

In the experiments an relativistic electron beam (REB)-plasma interaction, the foilless injection of REB from a magnetized diode is of special interest due to the low spread angle of the beam and high repetition rate of the shots. In the experiments presented, the problem of diode shortening in the presence of a dense plasma from the interaction chamber has been solved using a special drift pipe as an anode of the foilless diode. The electron beam (U/sub d//spl sim/0.7 MeV, t/sub b//spl sim/100-200 ns) produced by an axially symmetric magnetically insulated diode has been injected into a magnetized hydrogen plasma column with density ranging from 1/spl middot/10/sup 15/-3/spl middot/10/sup 16/ cm/sup -3/. It has been found that the anode pipe substantially reduces the plasma flow into the diode gap, but does not stop it completely, thus the REB generation in a plasma-filled diode has taken place. In some regimes of the beam generation it becomes possible to increase significantly the injected current and total energy deposition of the beam in comparison with the case of a vacuum magnetized diode of the same geometry. The experiments have shown effective dense plasma heating under the foilless injection.

The traditional way of relativistic electron beam (REB) injection in the experiments on REB- plasma interaction is the injection through an anode foil that separates vacuum diode from plasma chamber. The presence of separating foil leads to the following: (1) replacement of destroyed foil is required after each shot, and (2) the beam angular characteristics making worse. A beam with low angular spread can be obtained from foilless diode placed into strong guiding magnetic field; the problem is how to avoid the diode shortening in the presence of a dense plasma from the interaction chamber. In the experiments on studying of Langmuir turbulence, carrying out on GOL-M device it becomes possible to avoid a separating foil and to obtain a foilless injection of REB into a dense (

An infrared interferometer built according to the Mach–Zehnder design without a frame is described. The interferometer was used in measurements of the plasma-density dynamics inside a plasma-filled diode under conditions of intense vibrations. Such measurements can be performed under conditions of a quite short plasma lifetime, compared to the characteristic vibration period of the interferometer elements.

Recent results of the experiments at GOL-3 facility are presented. In present configuration of the device, plasma with a density of 10 14 ÷10 16 cm -3 is confined in a 12-meter-long solenoid, which comprises 55 corrugation cells with mirror ratio B max /B min =4.8/3.2 T. The plasma in the solenoid is heated up to 2-4 keV temperature by a high power relativistic electron beam (∼1 MeV, ∼30 kA, ∼8 μs, ∼120 kJ) injected through one of the ends. Mechanism of experimentally observed fast ion heating, issues of plasma stability and confinement are discussed.