Aachen PK 2003 – scientific programme

Parts | Days | Selection | Search | Downloads | Help

P: Plasmaphysik

P III: HV III

P III.1: Invited Talk

Tuesday, March 18, 2003, 11:00–11:30, FO2

Avoidance, prediction and mitigation of disruptions in tokamak. — •Gabriella Pautasso¹, Andrei Savtchov², and Christiane Tichmann¹ — ¹Max Planck Institut fuer Plasmaphysik, Garching — ²Institut fuer Plasmaphysik, Forschungszentrum Jeulich

Tokamak plasmas may undergo a major instability called disruption. Disruptions are known to cause large thermal and mechanical stresses on the machine components and to be occasionally responsible for damages. Therefore several methods of avoiding, detecting and mitigating disruptions are under investigation; they are aimed to prolong the lifetime of a future reactor and their state of development is encouraging. The disruptive boundaries, which are relevant for a steady state reactor, are relatively well known. Methods to avoid specific disruptions rely on the control of density, radiation and stored energy and on the detection plasma detachment, of rotating and locked modes. Disruptions may exhibit a very short precursor phase and their avoidance becomes then difficult or impossible. Therefore parallel to the avoidance methods, predictive tools based on a statistical modelling of the operational boundaries just started being developed. Neural networks (NNs) have mainly been used and shown to be relatively successful ifor this purpose but their use is not without disadvantages. The injection of low or high Z impurity gas or solid pellets to mitigate the three major effects of disruptions on the machine, namely the mechanical forces, the heat load and the runaways, have been investigate on several machine.