Initialising ...
Initialising ...
Initialising ...
Initialising ...
Initialising ...
Initialising ...
Initialising ...
Fukuda, Takeshi
Purazuma, Kaku Yugo Gakkai-Shi, 75(12), p.1377 - 1395, 1999/12
no abstracts in English
Hoshino, Katsumichi; JFT-2M Group
Proc. of 1993 Workshop on MHD Computations; Applied to the Controlled Thermonuclear Fusion Research, 0, p.111 - 122, 1994/02
no abstracts in English
Shimada, Michiya
no journal, ,
The disruption brings out large heat load and electromagnetic load on in-vessel components. The electromagnetic load is within an acceptable range, but its mitigation is highly desirable. The immense heat load will melt the surface of in-vessel components, forcing them to be replaced. Preemptive injection of massive material injection by gas or pellet to radiatively cool the core plasma is being considered. Neural network has been developed in a number of devices. However, neural network requires training, which is feasible in present devices, where heat load associated with disruption is acceptable. Neural network is very hard to implement in ITER, since disruption would make ITER inoperable. Therefore development of disruption prediction method that does not need training is an urgent issue. In the long term, development of in-vessel components which are operable after disruption is important. Use of liquid metal is a possibility.
