Contacts: Valerie Izzo
Calculates drift orbits for runaway electrons on 3D fields
This module uses the runaway orbit capability in NIMROD, but you do not need any nimrod files or know-how to use it. You do need an equilbrium file (EFIT gfile) and a 3D fields file (probe-g output file). If you want to calculate RE orbits on 2D fields, you can skip Step 2 in the GUI (BUT you must set ‘Number of Fourier components to retain’=0) and just use an EFIT file.
The module runs NIMROD at NERSC, so you will need a NERSC account. Instructions for obtaining a new NERSC account are here: http://www.nersc.gov/users/accounts/user-accounts/get-a-nersc-account/ Then you will need to contact the PI of an existing repo, or apply for a new repo to get a run time allocation.
The equations are only valid for relativistic, passing, runaway electrons–with an assumed small pitch angle. You can’t set the electron energy to something thermal and expect to get valid orbits.
The reOrbits GUI guides you through the process
The module gives you 2 types of output:
1) For a large number (~1000’s) of REs, with random starting locations covering the region inside the LCFS and specified energy, you get spatial locations for all electrons at specified time intervals. These are all plotted on the same poloidal plane by default, but toroidal location is also stored. This can give you information about RE loss rate and loss region when you have stochastic fields.
2) For a small number (=4) of REs, you get the complete 3D orbit information. By default, a Poincare plot will be generated showing the intersection of these orbits with phi=0, but you could also plot the complete 3D path. Note that if you set the total integration time >> 1e-5 (the default) these files can become so large as to be unwieldy. You can specify the starting locations for these 4 electrons.
List of contributors sorted by number of lines authored:
333 Valerie Izzo 140 Fusion Bot 32 Orso Meneghini