Kinetic modelling for neoclassical transport of high-Z impurity particles using a binary collision method

Y. Homma, S. Yamoto, Y. Sawada, H. Inoue, A. Hatayama

Research output: Contribution to journalArticlepeer-review

9 Citations (Scopus)


A new kinetic model for neoclassical impurity particle transport simulation has been developed. Our model is able to simulate the following two effects, which have been theoretically predicted, but neglected in most of the existing kinetic impurity transport simulations in the SOL (scrape-off layer)/Divertor plasmas of tokamak; (1) the neoclassical inward pinch (NC IWP) due to the density gradient of background plasmas and (2) the neoclassical temperature screening effect (NC TSE, outward transport) caused by the plasma temperature gradient. The IWP and TSE, both proportional to the impurity charge number Z, become especially important for higher-Z impurities such as tungsten. In this paper we focus on the case where background plasmas are in the Pfirsch-Schlüter regime. The velocity distribution of background plasma ions is modelled by a distorted Maxwellian distribution, which includes the Pfirsch-Schlüter flow velocity and the Pfirsch-Schlüter heat flux density, in order to reproduce the NC IWP and NC TSE. A series of test simulations have been performed for a toroidal magnetic field geometry. Characteristics of the neoclassical transport, such as dependencies on the safety factor and on the impurity charge number, have been confirmed.

Original languageEnglish
Article number036009
JournalNuclear Fusion
Issue number3
Publication statusPublished - 2016 Feb 5


  • distorted Maxwellian
  • fusion plasma
  • impurity
  • inward pinch
  • kinetic simulation
  • neoclassical transport
  • temperature screening effect

ASJC Scopus subject areas

  • Nuclear and High Energy Physics
  • Condensed Matter Physics


Dive into the research topics of 'Kinetic modelling for neoclassical transport of high-Z impurity particles using a binary collision method'. Together they form a unique fingerprint.

Cite this