Simulation of collisional effects on divertor pumping in JT-60SA

C. Gleason-González, S. Varoutis, X. Luo, K. Shimizu, T. Nakano, K. Hoshino, H. Kawashima, N. Asakura, Chr Day, S. Sakurai

Research output: Contribution to journalArticlepeer-review

8 Citations (Scopus)


In this work, the exhausted neutral gas flow is modeled for two cases of a simplified JT-60SA sub-divertor geometry and compared via three different approaches, namely (i) a collisionless approach based on the ProVac3D code, (ii) the DSMC approach based on the DIVGAS code that can be run with and without consideration of particle collisions, and (iii) the NEUT2D approach which has been extensively used in the past for the JT-60 design. In a first case study, the transmission probability was calculated by the 3 approaches and very good agreement is found between NEUT2D-ProVac3D whereas discrepancies between DIVGAS and NEUT2D are found and further analyzed. In the second case, the assessment of collisions is done by means of DIVGAS. Simulations showed that the flow is found in the transitional regime with Kn numbers between 0.1 and 0.4. The DIVGAS collisionless case yielded lower values of temperature than the collisional one by factors of 0.5–0.8 in regions near the inlets of the sub-divertor whereas in regions near the pump and the chevron, the temperature difference is marginal. Moreover, a relative percentage difference of 16–40% in pressure values was found between collisionless and collisional approaches.

Original languageEnglish
Pages (from-to)693-699
Number of pages7
JournalFusion Engineering and Design
Publication statusPublished - 2016 Nov 1
Externally publishedYes


  • DSMC
  • Divertor pumping
  • Sub-divertor gas dynamics
  • Tokamak particle exhaust
  • Vacuum

ASJC Scopus subject areas

  • Civil and Structural Engineering
  • Nuclear Energy and Engineering
  • General Materials Science
  • Mechanical Engineering


Dive into the research topics of 'Simulation of collisional effects on divertor pumping in JT-60SA'. Together they form a unique fingerprint.

Cite this