Suppose that the rigid wall surrounding the plasma is perfectly conducting (i.e.,
, where
is the time constant of the wall—see Section 3.10), and that there is no current
flowing through the external magnetic field-coil [i.e.,
, where
is the normalized coil current—see Equation (3.194)]. In this case, Equations (3.187),
(3.188), and (7.5), yield
 |
(9.2) |
where
is the (real dimensionless) perfect-wall tearing stability index,
the normalized reconnected helical magnetic flux at the rational surface [see Equation (3.184)],
and
the normalized helical sheet current density at the rational surface [see Equation (3.183)]. It is clear from the previous equation that
 |
(9.3) |
which implies that zero electromagnetic torque is exerted at the rational surface. (See Section 3.13.)
Equation (9.2) yields
 |
(9.4) |
which can be combined with Equations (8.1), (8.108), and (9.3) to give the so-called Rutherford island width evolution equation [13]:
 |
(9.5) |
Here,
is the full radial width of the magnetic island chain that develops at the rational surface,
the
minor radius of the surface, and
[see Equation (8.124)].
We conclude that, in the nonlinear regime, the width of the magnetic
island chain grows algebraically in time on the resistive diffusion time,
. Obviously, this is a much slower time evolution than that predicted in the linear regime.