Examples of using Langevin equation to solve FP equation

Examples of using Langevin equation to solve FP equation

ASCOTRacetrack for tokamak particles

Ab initio particle loading– Fusion alphas (thermonuclear, beam-target, beam-beam)– NBI-generated ions

Realistic orbit tracing– Guiding center (4th order RK, w/ 5th order error check, fast)– gyro orbit (Leap Frog, accurate, slow)

Fully 3D – 3D magnetic field– 3D Wall

Comprehensive interactions– Coulomb collisions– (Turbulent transport)– Models for relevant MHD:

• NTM-type magnetic islands• Alfven Eigenmodes

ASCOT4Technical details

Parallelization– MPI parallel testruns (HELIOS) with 16384 cores– Production runs with 4096

File I/O– Single process does all file I/O– HDF5 format, partly ASCII– Processing of data w/ MATLAB

Magnetic field interpolation– PSPLINEs

Programming approach– ASCOT completely rewritten in 2012-13– Modular F90– Only 2 libraries reguired: HDF5 & MPI

FUSION ALPHAS & ITER 1ST WALL

Axisymmetry vs ripple vs FIFusion alphas in 15MA scenario

Axisymmetric B field With bare rippleRipple w/ FI’s

Also wall shape matters...

Original wall w/ 2 limiters Present wall design w/ poloidally extended ’continuous’ limiters

TBM MOCK-UP EXP’S AT DIII-D

Ripple map w/ and w/o TBM field bump

TBM field bump

Max field perturbation along separatrix close to 5% with TBM mock-up

)()(

)()(),(

minmax

minmax

BB

BBzR

Variation of the toroidal field• with the TBM coils• in the absence of TBM coils

BT

φ

TF ripple

NBI-generated deuterons in DIII-D discharges w/ TBM mock-up

limiters

TBM mock-up coils

Beam-target DD reaction T(1 MeV)

Guiding-center following clearly not applicable follow full gyro motion (FO integration)

DD DT n (14 MeV)

Experimental neutron flux in the TBM mock-up experiment

Fraction of confined tritium in the plasma as calculated by ASCOT

M. Schaffer & al, Nucl. Fusion 51 (2011) 103028

IMPURITY INJECTION EXP’S @AUG

The case of missing 13C in AUG

Besides fast ions, tritium retention is a hot issue for ITER Global impurity migration studies in AUG by injecting 13CH4 post mortem analysis of samples from selected wall tiles ->

90% missing?? Common assumption: axisymmetric 13C deposition ASCOT simulation w/ a 3D wall and magnetic field: notable

non-axisymmetry in 13C deposition (limiters, ICRH antennas, …)

[J. Miettunen et al., NF 52 (2012) 032001]