POWER AND PARTICLE BALANCE

• The ion Jr(r) was incorrectly divided by the subcycling factor.
  
  
• The other problem was that e- Jr(r) is inherently noisy. The way we compute flux diagnostics currently does not take into account that some particles will be crossing cell boundaries.
  
  For the ion Jr(r) this is no problem, since for a typical timestep, the ions barely move. Hence, our ion flux diagnostics are very accurate. However, for the e- J(r), a large number of electrons cross cell boundaries for a typical timestep.
  
  Fortunately, Jr(r) is only calculated if you run with Xgrafix (diagnostics turned on), and NOT when you are doing the actual physics run to get to equilibrium. Also, in our discharges, most of the current is in the axial direction. Net current in the r direction is typically small. This implies that ion J(r) and e- J(r) are very close.
  
  To get accurate electron Jr(r) and JdotE(r) diagnostics, reduce the time step so that barely any electrons cross a cell boundary in a timestep. However, for all the other diagnostics and to do the actual run to equilibrium, it is NOT necessary to reduce the timestep. The following link shows the effect of varying the timestep dt on th e- Jr(r): Jr(r) for diff. dt
  
  

• PowInput = CollLoss + WallLoss
  
• pcreated = pilost = pelost