Thumbnail
Access Restriction
Open

Author Ramsey, A. T. ♦ Diesso, M.
Source United States Department of Energy Office of Scientific and Technical Information
Content type Text
Language English
Subject Keyword PLASMA PHYSICS AND FUSION ♦ ERRORS ♦ PLASMA DIAGNOSTICS ♦ INVERSE SCATTERING PROBLEM ♦ PLASMA DENSITY ♦ TFTR TOKAMAK
Abstract Inversion of chordally integrated data to infer the profile of a plasma parameter (plasma density, radiated power, or, in our case, the visible continuum from bremsstrahlung) propagates errors from outer shells inward as the inversion progresses. If the inversion is done by a matrix technique, error propagation can likewise be determined by a matrix technique. Where other considerations make the matrix method undesirable, there is no clearly defined analytical method to test the reliability of the inversion. We have solved this problem by taking real data from the Tokamak Fusion Test Reactor (TFTR) visible bremsstrahlung (VB) diagnostic, applying normally distributed random noise of a known mean value, and inverting the signal. When we have done this hundreds of times at all points in the profile, we can clearly track the error propagation. At the plasma center, for example, the error is several times the average error in the relative (chord to chord) calibration of the system. This error multiplier is a powerful generalization; it depends little on the details of the inversion, and almost entirely on the profile shape. For the TFTR VB profile it ranges from 2 to 4 for most types of plasma. We have also looked at the damping of the inversion error from an error in a single chord; this damping can be rather fast in some cases. {copyright} {ital 1999 American Institute of Physics.}
ISSN 00346748
Educational Use Research
Learning Resource Type Article
Publisher Date 1999-01-01
Publisher Place United States
Journal Review of Scientific Instruments
Volume Number 70
Issue Number 1
Technical Publication No. CONF-980605-


Open content in new tab

   Open content in new tab