Optimization of tokamak disruption scenarios
Publicerad
Författare
Typ
Examensarbete för masterexamen
Program
Modellbyggare
Tidskriftstitel
ISSN
Volymtitel
Utgivare
Sammanfattning
Research in the field of fusion science has been propelled by its potential to alleviate
humanity’s reliance on fossil fuels. One of today’s most promising approaches
to generating thermonuclear fusion energy uses magnetic confinement of hydrogen
fuel in the plasma state. The tokamak concept, which has achieved the best fusion
performance so far, is used in the three reactor-scale devices (ITER, SPARC and
STEP) currently being constructed — they aim to achieve a positive energy balance,
thereby demonstrating the scientific feasibility of magnetic confinement fusion
energy.
A major open issue threatening the success of these tokamaks is plasma disruption.
In these off-normal events the plasma loses most of its thermal energy on a
millisecond timescale, exposing the device to excessive mechanical stress and heat
loads. In addition, in the high-current devices currently under construction, one
of the most important related problems is posed by currents carried by electrons
accelerated to relativistic energies, called runaway electrons. If these were to strike
the inner wall unmitigated, it may cause potentially irreversible damage to the device.
The methods proposed to mitigate these dangerous effects of disruptions, such
as massive material injection, are characterized by a large number of parameters,
such as when to inject material, in which form and composition. This poses an
optimization problem that involves a potentially high-dimensional parameter space
and a large number of disruption simulations.
In this work, we have developed an optimization framework that we apply to
numerical disruption simulations of plasmas representative of ITER, aiming to find
initial conditions for which large runaway beams and excessive wall loads can be
avoided. We assess the performance of mitigation when inducing the disruption
by massive material injection of neon and deuterium gas. The optimization metric
takes into account the maximum runaway current, the transported fraction of the
heat loss — affecting heat loads — and the temporal evolution of the ohmic plasma
current — determining the forces acting on the device.
Beskrivning
Ämne/nyckelord
fusion plasma, disruption mitigation, runaway electrons, Powell’s method, Bayesian optimization