You also want an ePaper? Increase the reach of your titles
YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.
Plasma Phys. Control. Fusion 53 (2011) 093001<br />
Topical Review<br />
force. Perhaps as a result <strong>the</strong> growth rate did not follow <strong>the</strong> usual k 1/2 ideal MHD dependence<br />
for a Z-<strong>pinch</strong>, but decreased with k.<br />
Introduction <strong>of</strong> lower hybrid anomalous resistivity will also change <strong>the</strong> formula for <strong>the</strong><br />
Pease–Braginskii current, a point followed up by Robson [125].<br />
In section 3.14 anomalous resistivity will be introduced in a model <strong>of</strong> disruption when<br />
gaps <strong>of</strong> low density plasma exist between density islands at <strong>the</strong> time <strong>of</strong> a m = 0 disruption<br />
and <strong>the</strong>reafter.<br />
It is to be hoped that with <strong>the</strong> new generation <strong>of</strong> computers fur<strong>the</strong>r work, perhaps in 3D,<br />
full electromagnetic will be undertaken on lower hybrid turbulence, and <strong>the</strong> related modified<br />
two-stream instability.<br />
Meanwhile attention could be drawn also to <strong>the</strong> need for better diagnosed and specially<br />
designed experiments. A paper by Takeda and Imezuka [237] presents measurements <strong>of</strong><br />
anomalous resistivity associated with large amplitude noise at <strong>the</strong> lower hybrid frequency.<br />
Here <strong>the</strong> ratio v d /v i was large (∼10 2 ) and <strong>the</strong> anomalous collision frequency was essentially<br />
e . This leads to Bohm diffusion. It is perhaps permissible to speculate that <strong>the</strong> formula for<br />
ν anom given by equation (3.54) or(3.55) is only valid up to a value <strong>of</strong> e , when it becomes<br />
independent <strong>of</strong> (v d /v i ). For a hydrogen plasma this will occur at v d /v i equal to 4.2. A suitable<br />
form for <strong>the</strong> effective collision frequency νanom ∗ as a function <strong>of</strong> (v d/v i ) 2 through equation<br />
(3.54) or(3.55) for e τ e > 1isgivenby<br />
νanom ∗ = ν anom/(1+ν anom / e ). (3.56)<br />
3.14. Disruptions; ion beams and neutrons<br />
During <strong>the</strong> nonlinear evolution <strong>of</strong> <strong>the</strong> m = 0 instability <strong>of</strong> a stagnated compressional Z-<strong>pinch</strong> or<br />
plasma focus employing deuterium as <strong>the</strong> working gas, a significant pulse <strong>of</strong> neutrons appears,<br />
typically a yield <strong>of</strong> up to 10 12 neutrons [238, 239]. The origin <strong>of</strong> this has been a great source <strong>of</strong><br />
debate and controversy which has still not been resolved. Some <strong>of</strong> <strong>the</strong> proposed mechanisms<br />
are fluid-like and o<strong>the</strong>rs are kinetic in nature. An important test <strong>of</strong> any hypo<strong>the</strong>tical <strong>the</strong>ory is<br />
whe<strong>the</strong>r total axial momentum is conserved during <strong>the</strong> ion beam formation.<br />
In fact <strong>the</strong> normally principal pulse <strong>of</strong> neutrons at <strong>the</strong> time <strong>of</strong> m = 0 disruption is <strong>of</strong>ten<br />
<strong>the</strong> second pulse [240]. The likely explanation <strong>of</strong> <strong>the</strong> first pulse is essentially kinetic in origin.<br />
It was first hinted at in Trubnikov’s paper [241] and developed more realistically by Deutch<br />
and Kies [242]. Here it is considered that <strong>the</strong>re is a sheet current piston which compresses <strong>the</strong><br />
plasma as <strong>the</strong> first <strong>pinch</strong> is formed. Unlike <strong>the</strong> snowplough model, it is postulated that <strong>the</strong> ions<br />
on encountering this moving piston are reflected by it (in fact by <strong>the</strong> radial electric field in <strong>the</strong><br />
current sheet) with twice <strong>the</strong> velocity <strong>of</strong> <strong>the</strong> piston. During this process <strong>the</strong>y are essentially<br />
collisionless and also conserve <strong>the</strong>ir angular momentum (a point made by Trubnikov [241] and<br />
earlier by Haines [243]). On encountering <strong>the</strong> moving piston a second time, <strong>the</strong>re is ano<strong>the</strong>r<br />
gain <strong>of</strong> energy; indeed <strong>the</strong> faster ions have more reflections and gain even more energy, a feature<br />
<strong>of</strong> this Fermi acceleration mechanism. There will be, as a result, a high energy ion tail, and <strong>the</strong><br />
ion–ion collisions which will grow in number when <strong>the</strong> piston reaches close to <strong>the</strong> axis and <strong>the</strong><br />
density rises will also yield DD nuclear reactions, but <strong>of</strong> a supra<strong>the</strong>rmal nature. If <strong>the</strong> incoming<br />
piston is conical in shape, it will also lead to a growth in axial motion. It should be noted<br />
that multiple collisions <strong>of</strong> essentially collisionless ions are needed for this, a condition which<br />
might also pertain in certain conditions in <strong>the</strong> precursor plasma <strong>of</strong> an imploding wire array.<br />
Turning to <strong>the</strong> principal pulse <strong>of</strong> neutrons, <strong>the</strong>re are important features; (i) <strong>the</strong> neutrons<br />
occur at <strong>the</strong> time <strong>of</strong> an m = 0 instability, a point confirmed experimentally by Bernard<br />
et al [238], (ii) <strong>the</strong> neutrons have an anisotropic distribution consistent with originating from<br />
an ion beam in <strong>the</strong> z-direction: This was also confirmed by Bernard who employed a deuterated<br />
54