Model Overview

BTR code screen with the beamline and tokamak: horizontal (top) and vertical (bottom) plane views. Standard (focused) BTR beam model, the beamlets axes are shown in violet. NBI elements and virtual surfaces are shown in black.

BTR-code (“Beam Transmission with Reionization”, [3]) was developed in 1995-2005 at RRC “Kurchatov Institute” for the purposes of NBI design and scenarios studies. Since then it is widely used by several parties as a neutral beamline simulator. BTR model (see Fig. 2) performs a 3D direct tracing of the beam particles (PIC) from the ion source and delivers power map on each solid surface (the amount of beamline surfaces reaches 500), resulting from beam direct interception and re-ionized fluxes. The background fields and gas conditions are applied to imitate particles conversions while tracing. With a great amount of particles in a model (up to 10⁹ ) BTR provides the most detailed injected beam geometry and optics, while the running speed is kept high due to permanent optimization: at present running 10⁵ particles takes less than 1 minute on 2-core Win-system. BTR executable is a small Win32/64 application (~6MB, MS Visual C++). Here are the main reasons for BTR extension to plasma.

BTR runs the most detailed spatial neutral beam geometry, which includes the array of beamlets (elementary beams, >1000 total); every single beamlet is represented by 10²–10⁵ rays (or big particles) according to the experimental bi-gauss angular distribution. Each of the elementary beamlets is supposed to carry individual parameters of angular profile and axis focusing in horizontal and vertical planes.

Big particles are 3D traced with account of all the background electro-magnetic fields, thus the atoms directions scattering evolved along the source ions neutralization path is also calculated.

Due to the beam direct interception by the injector channels walls and solid elements, and with the neutral beam re-ionization in the beam-duct, the final shape of the injected beam and its velocity distributions differ from the shapes obtained from a direct extrapolation of source beam to plasma entrance. It means that BTR is capable to calculate the injected beam spatial power penetration profile in plasma with higher precision, as compared with traditional simplified beam models. More accurate approach, used by BTR, can be useful in particular for the calculations of the fast ions losses and plasma current generation, where the accuracy of pitch angle calculations (the angle between the ion velocity and magnetic field vector) is critical, as even small deviations lead to erroneous losses of fast ions and the resulting power loads on FW.

BTR model of beam tracing is simple, straightforward, self-reproduced (no random parameters), and manually tuned. It imposes no limits on particle statistics or power maps resolution, so all the beam power footprints, the shine-through maps, the beam integral power profile, and ionization points distribution in the plasma bulk can be easily obtained for any (even sparse) plasmas.

Standard beam model (focused or realistic beam) in BTR reproduces the detailed accelerated beam structure: apertures positions on the grounded grid (IS exit electrode), the source beamlets (elementary beams) individual angular parameters (divergence, aiming, steering). The input source beam geometry in BTR can be set as a regular or irregular array of beamlets. Basic beam structure and beamlets focusing in DEMO-FNS are outlined in Fig. 2 - a regular beam array of 1280 beamlets. The beamlets’ source positions are prescribed by IS apertures arrangement, while their grouping to segments and focusing are optimized for minimal beam losses for a 2-channel structure [8].