ADAS Terminology

The terminology used within ADAS and OPEN-ADAS can sometimes be unfamiliar and lead to confusion. In this page, attention is drawn to a number of terms which may be ambiguous and require clarification.

Acronyms and mnemonics

ADAS uses a number of acronyms and mnemonics for certain data formats and sub-classes. There are also some simple abbreviations. These are quite widely used in the fusion community but may not be generally familiar. They occur in text both as upper and lower case forms. They include:

• ADF: ADAS data format
• CXS: charge exchange spectroscopy, normally beam driven
• MSE: motional Stark effect, affecting beam atom emission
• RR: radiative recombination - ADF08
• DR: dielectronic recombination - ADF09
• CR: collisional-radiative coefficients - ADF11. Subclasses:
  • ACD: effective recombination
  • SCD: effective ionisation
  • CCD: effective charge exchange recombination
  • PLT: effective low-level line power
  • PRB: effective recom + Brems power
  • PRC: effective charge exchange power
• GCR: generalised collisional-radiative coefficients - ADF11. Subclasses as CR +:
  • QCD: effective metastable cross-coupling
  • XCD: effective parent metastable cross-coupling
• CD: collisional-dielectronic (synonym for collisional--radiative)
• QEF: effective emission coefficient (from CXS normally) - ADF12
• SXB: ionisation per photon coefficient - ADF13
• PEC: photon emissivity coefficient - ADF15
• BMS: beam stopping coefficient - ADF21
• BME: beam emission coefficient - ADF22

Bundling and resolution

Depending on circumstances, it may not be helpful or possible to deal with the populations of excited states of ions in a plasma at the detail of individual levels. Rather, one deals with the populations of groups of levels, such as of all the levels of a term, of all the levels of an nl-shell or of all the levels of an n-shell. The collisional-radiative calculations are then referred to as being in a "bundle-nl" or "bundle-n" approximation. This bundling is sometimes described as the "resolution level" of the calculations (and of the consequential derived data. Other relevant resolution levels of ADAS data are "term" (or "LS" or "ls") and "level" (or "LSJ" or "ic"). "bundle-nl" is also called "ca". "ic" and "ca" are short for "intermediate coupling" and "configuration average" respectively. Resolution in this context is related to but a bit different from "metastable resolution".

Condensation and projection

In GCR studies, it is efficient and physically appropriate to couple a lower resolution population model, such as "bundle-n" for very many highly excited levels, to a high resolution model, such as "ic" for the set of lowest levels. Then one says that the influence of the bundle-n results for the high levels are condensed or projected onto the low levels. Projection matrix files (ADF17) and mapping files (ADF18) enable this in ADAS.

Data formats and ADF numbers

The ADAS "database" is organised using a collection of ASCII files sitting on a conventional UNIX filesystem. Files are grouped together in directories. At the top level of this database are a set of file-formats called ADAS Data Formats (or ADFs). These formats are all distinct and refer to different types of data (e.g. ADF15 are effective emission coefficients). The ADAS data format name, such as ADF15, is used to refer both to the formatting specification and the name of the top level data set directories.

Date conventions

ADAS filenames often have a two-digit year number embedded in them. These numbers should be treated as a token and do not necessarily reflect when a particular file was introduced. Typically it indicates when a particular approximation or quality level was introduced. For example, for GCR data such as ADF11 and ADF15, the year "96" was the year of its substantive introduction into the ADAS database. Information on the actual year a file was produced is available in the comments section of each file.

Sub-directory naming conventions

Sub-directories at one level below the ADAS data format commonly contain files of a particular element (an iso-nuclear set) or files of ions with the same number of electrons (an iso-electronic set). For the former "#<nuclear charge number>" appears as a post-fix of the sub-directory name (such as copmm#18 for argon). For the latter "\#<chemical symbol of first member>" appears as a post-fix of the sub-directory name (such as pec96#be for beryllium-like). Although this is present practice, some older ADAS directories are inconsistent.

File naming conventions

Actual data files, as distinct from directories, all are of form ".dat". The whole file name, if one of an automatic or semi-automatic production process are typically of the form "#<resolution>#<element><ion charge>.dat" such as ic#ar10.dat. If hand produced by a specialist, typically author initials appear in the file name. Again there are many older ADAS datasets which do not conform to the present conventions.

Derived data

Derived data are data which have been processed through a collisional-radiative model and have dependency on plasma parameters. These are data like {\em effective} ionisation/recombination rates (i.e. the driving terms for transport models) and {\em effective} emission coefficients.

Fundamental data

Fundamental data are data which are specific to an ion in isolation or composed of individual pure reactions. These are things like energy levels, spontaneous radiative rates, state resolved recombination/ionisation rates and collision cross-sections. Maxwell averaged rate coefficients (or reduced forms such as effective collision strength) between specific levels are viewed as fundamental data.

J values and statistical weights

The quantum number assignment of energy levels in ADAS data files such as ADF04 appear as (2S+1)L(J). There is a long-standing convention in ADAS that the actual meaning of the J field is "(statistical weight-1)/2" independent of the coupling scheme. This is true for level resolved data. For term resolved data we also follow the convention which means that a ³P level would have an apparent J-value of 4 but it actually gives the correct statistical weight of the term as 9. Consequently ADF04 files of varying bundling can be handled transparently in collisional-radiative modelling.

Metastable resolved

In a stage-to-stage, metastable unresolved picture, There is often confusion as to whether metastables are included or not in the population structure. They are, but they are treated as thought they were ordinary excited states and therefore assumed in quasi-static balance with the ground state. Also their contribution to the total population, summed over all levels, of the stage (which is usually assumed to be that of the ground population alone to good precision) is omitted, unless special re-normalising steps have been taken in the collisional-radiative calculations. Such information is contained in ADAS data set comments and should be looked out for to avoid misunderstanding.

Quasi-static approximation

Collisional-radiative theory is concerned with the relative relaxation times of different state populations of ions in plasmas. Appropriate separation of those with short (ordinary excited states) and those with long (ground and metastable states) relaxation times takes place according to plasma (mainly density) regime. The valid and simplifying approximation is made that those with short lifetimes can be assumed in statistical equilibrium with the instantaneous values of the long-lived populations. Then the solution of the collisional-radiative equations is easier to obtain, delivers universal derived data which are functions of local plasma conditions only and separates cleanly into the data formats required for establishing ionisation state (in dynamic plasma models) and then subsequent emission.