This paper considers a fast track to non-energy applications of nuclear fusion that is associated with the ‘fusion for neutrons’ (F4N) paradigm. Being a useful product accompanying energy, fusion neutrons are more valuable than the energy released in DT reactions and they are urgently needed for research purposes and to develop and validate modern technologies. In the near future neutron yield in fusion devices will become significantly larger than that of fission and accelerator sources. This paper describes a compact tokamak fusion neutron source based on a small spherical tokamak (FNS-ST) with a MW range of DT fusion power and considers the key physics issues of this device. The major and minor radii are ∼0.5 and ∼0.3 m with magnetic field ∼1.5 T, heating power less than 15 MW and plasma current 1–2 MA. The production rate of DT neutrons of (3–10) × 1017 n s−1 and their flux at the first wall of 0.2 MW m−2 ensure that the device is capable of fusion–fission demonstration experiments. The problems of major concern are discharge initiation, current drive, plasma—fast ion beam stability and high first wall and divertor loads. The conceptual design provides solutions to these problems and suggests the feasibility of the FNS-ST.

The targeted plasma parameters of the compact spherical tokamak (ST) Globus-M have basically been achieved. The reasons that prevent further extension of the operating space are discussed. The operational limits of Globus-M together with an understanding of the limiting reasons form the basis for defining the design requirements for the next-step, Globus-M2. The recent experimental and theoretical results achieved with Globus-M are discussed, the operational problems and the research programme are summarized and finally, the targeted Globus-M2 parameters are presented. The magnetic field and plasma current in Globus-M2 will be increased to 1 T and 0.5 MA, respectively. The plasma dimensions will remain unchanged. With auxiliary heating at a high average plasma density, the temperatures will be in the keV range and the collisionality parameter with ν * 1 will define the operational conditions. Noninductive current drive will be a major element of the programme. The engineering design issues of Globus-M2 tokamak are discussed and the technical tokamak parameters are confirmed by thermal load and stress analysis simulations. The experimental results obtained on Globus-M2 and the limits of its performance should clarify the feasibility of an ST-based super compact neutron source.

The first experiments on noninductive current drive (CD) using lower hybrid waves in a spherical tokamak are described. Waves at 2.45 GHz were launched by a 10 waveguide grill with 120° phase shift between neighbouring waveguides. The experimental results for a novel poloidal slowing-down scheme are described. The CD efficiency is found to be somewhat larger than that predicted theoretically whilst at the same time being somewhat less than that for standard tokamak lower hybrid CD. Geodesic acoustic modes (GAM) have been discovered in Globus-M. GAMs are localized 2-3 cm inside the separatrix. The GAM frequency agrees with theory. The mode structures of plasma density and magnetic field oscillation at the GAM frequency have been studied. Fast particle confinement during neutral beam injection has been investigated and numerically simulated. Alfvén instabilities excited by fast particles were detected by a toroidal Mirnov probe array. Their excitation conditions are discussed and the dynamics of fast ion losses induced by Alfvén eigenmodes is presented. Preliminary experiments on the isotopic effect influence on global confinement in the ohmic Nuclear Fusion

In this paper we present the fusion related activities of the Plasma Physics Division at the Ioffe Institute. The first experiments on lower hybrid current drive (LHCD) in a spherical tokamak performed at the Globus-M tokamak (R = 0.36 m, a = 0.24 m, B t = 0.4 T, I p = 200 kA) with a novel poloidally oriented grill resulted in an RF driven current of up to 30 kA at (100 kW, 2.5 GHz), exceeding the modelling predictions. At the FT-2 tokamak (R = 0.56 m, a = 0.08 m, B t = 3 T, I p = 30 kA) experiments with a traditional toroidally oriented grill revealed no strong dependence of the LHCD density limit on the H/D ratio in spite of LH resonance densities differing by a factor of 3. Microwave Doppler reflectometry (DR) at the Globus-M, and DR and heavy ion beam probe measurements at the tokamak TUMAN-3M (R = 0.53 m, a = 0.24 m, B t = 1.0 T, I p = 190 kA) demonstrated geodesic acoustic mode (GAM) suppression at the L to H transition. Observations at FT-2 using Doppler Enhanced Scattering showed that the GAM amplitude is anti-correlated both spatially and temporally to the drift turbulence level and electron thermal diffusivity. For the first time turbulence amplitude modulation at the GAM frequency was found both experimentally and in global gyrokinetic modelling. A model of the L-H transition is proposed based on this effect. The loss mechanisms of energetic ions' (EI) were investigated in the neutral beam injection (NBI) experiments on Globus-M and TUMAN-3M. Empirical scaling of the 2.45 MeV DD neutron rate for the two devices shows a strong dependence on toroidal field B 1.29 t and plasma current I 1.34 p justifying the B t and I p increase by a factor of 2.5 for the proposed upgrade of Globus-M. Bursts of ∼1 MHz Alfvenic type oscillations correlating with sawtooth crashes were observed in ohmic TUMAN-3M discharges. The possibility of low threshold parametric excitation of Bernstein and upper hybrid waves trapped in drift-wave eddies resulting in anomalous absorption in electron cyclotron resonance heating (ECRH) experiments in toroidal plasmas was identified theoretically. A novel method of radial correlation Doppler reflectometry is shown to be capable of measuring the turbulence wave-number spectrum in realistic 2D geometry. On the progress in design and fabrication of three diagnostics for ITER developed in the Ioffe institute is reported: neutral particle analysis, divertor Thomson scattering and gamma spectroscopy.

Key issues of design of the divertor and the first wall of DEMO-FNS are presented. A double null closed magnetic configuration was chosen with long external legs and V-shaped corners. The divertor employs a cassette design similar to that of ITER. Water-cooled first wall of the tokamak is made of Be tiles and CuCrZr-stainless steel shells. Lithium injection and circulation technologies are foreseen for protection of plasma facing components. Simulations of thermal loads onto the first wall and divertor plates suggest a possibility to distribute heat loads making them less than 10 MW m −2 . Evaluations of sputtering and evaporation of plasma-facing materials suggest that lithium may protect the first wall. To prevent Be erosion at the outer divertor plates either the full detached divertor operation or arrangement of the renewal lithium flow on targets should be implemented. Test bed experiments on the Tsefey-M facility with the first wall mockup coated by Ве tiles and cooled by water are presented. The temperature of the surface of tiles reached 280-300 °С at 5 MW m −2 and 600-650 °С at 10.5 MW m −2 . The mockup successfully withstood 1000 cycles with the lower thermal loading and 100 cycles with higher thermal loading.

An approach to the integrated modelling of plasma regimes in the projected neutron source DEMO-FNS based on different codes is developed. The integrated modelling allows to eliminate uncertainties in external parameters for such tasks as plasma current ramp up, steady-state plasma consistency, plasma stability and heat load onto the wall and divertor plates. The following codes are employed for the integrated modelling. ASTRA transport code is used for adjustment of the steady-state regime parameters, NUBEAM Monte Carlo code for NBI incorporated into the ASTRA code, DINA free boundary equilibrium and evolution code, SPIDER free boundary equilibrium and equilibrium reconstruction code, KINX ideal MHD stability code, TOKAMEQ free boundary equilibrium code, edge and divertor plasma B2SOLPS5.2 code and Semi-analytic Hybrid Model (SHM) code for self-consistent description of the core, edge and divertor plasmas based on the experimental scaling laws. The consistent steady state regime for the DEMO-FNS plasma and the plasma current ramp up scenario are developed using the integrated modelling approach. Passive copper coils are considered for suppression of the instability to vertical displacement.

A concept of the divertor and the technology for organizing the edge plasma in a fusion neutron source based on a spherical tokamak (FNS ST) are described. The experimental data on the characteristics of the peripheral plasma in modern tokamaks are extrapolated to the FNS ST conditions with the help of semi analytical models. The effects depending on the magnetic configuration and on the geometry and mate rials of the divertor and the first wall elements are considered. Possible designs of the FNS ST divertor and the first wall are described. Using an original model, it is shown that the maximum density of the heat flux at the divertor plates in a double null magnetic configuration does not exceed 5-6 MW/m 2 , which complies with modern engineering capabilities. Methods for further improvement of the FNS ST divertor concept are analyzed.

No abstract