Case XGC-204 crashed after 65000 time steps. It is possible that time limit for this simulation has been set incorrectly.
Case XGC-205 crashed after 50000 time steps. Same reason (?)
Profiles for plasma density, temperatures and electric field are given below.
Radial electric field remains very noisy in the plasma core. Both the radial electric field and plasma density profiles are found to be very different comparing to the case when electron dynamics has been diabled (case XGC-203).
Previous XGC-0 results with with electron dynamics disabled show significant difference in radial electric field profiles. Simulations with electron disabled has larger values for radial electric field in the plasma core. It looks that there is a charge separation in the plasma core (see Fig 1).
Fig. 1. Case XGC-203: DIII-D discharge 136674.01405; no electrons; 80 radial zones
Two new simulations are submitted on Franklin:
XGC-204 for the DIII-D discharge 096333.0333, and
XGC-205 for the DIII-D discharge 136674.01405.
Another problem that needs to be address is particle losses (see Fig. 2).
Some changes to the code includes increased number of radial zones for radial electric field:
efld_npsi is increased from 35 to 70 in modules.F90
Gunyoung suggested to disable the bounce operation from
the simulation inner boundary in the simulations that include magnetic axis. Cases XGC-200 — XGC-205 don’t include the magnetic axis and start at sml_inpsi equal to 0.01.
Fig. 2. Case XGC-203: DIII-D geometry g136674.01405; 80 radial zones; no electrons.
