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 />
for electron temperatures up to 100 keV. Refinement <strong>of</strong> <strong>the</strong> Lee and More [392] transport<br />
model for strongly coupled plasmas has been carried out by Dejarlais [340] while for partially<br />
degenerate, magnetized plasmas a complete set <strong>of</strong> transport coefficients has been derived by<br />
Brown and Haines [393, 394].<br />
Pereira and Davis [51] have <strong>review</strong>ed <strong>the</strong> experimental data <strong>of</strong> x-rays from <strong>dense</strong> Z-<strong>pinch</strong>es<br />
and compared <strong>the</strong> results with <strong>the</strong>oretical modelling. The experimental I 4 scaling <strong>of</strong> x-ray<br />
yield for neon K shell and krypton L shell radiation is explained. Also <strong>the</strong>re is a marked radial<br />
distribution <strong>of</strong> radiation emission, <strong>the</strong> XUV emission coming broadly from <strong>the</strong> <strong>pinch</strong>ed region<br />
but <strong>the</strong> harder radiation, i.e. <strong>the</strong> s<strong>of</strong>t x-rays, being emitted from many hot spots along <strong>the</strong> axis.<br />
This is particularly true for higher Z materials such as tungsten. Even harder radiation (inner<br />
shell transitions) from 10 to 100 keV electrons is also detected at stagnation. The origin <strong>of</strong><br />
this electron acceleration process is still not firmly established, and in [51] is usually assumed<br />
to be caused by a resistive or inductive axial electric field. However it should be noted that<br />
runaway electrons have singular orbits and can only occur within one electron Larmor radius<br />
<strong>of</strong> <strong>the</strong> axis. Such orbits are discussed in section 2.2; fur<strong>the</strong>rmore as shown in [113] <strong>of</strong>f-axis<br />
energetic relatively collisionless electrons drift towards <strong>the</strong> axis with a E r /B ϑ drift—a sort<br />
<strong>of</strong> high energy filter—where <strong>the</strong>y are free to accelerate in <strong>the</strong> applied electric field. In linear<br />
transport <strong>the</strong>ory <strong>the</strong> velocity dependence <strong>of</strong> <strong>the</strong> collision frequency leads to <strong>the</strong> Ettingshausen<br />
heat flows (see section 2.6) in <strong>the</strong> direction <strong>of</strong> J × B, <strong>of</strong> which <strong>the</strong> runaway electron drift is an<br />
extreme example. Thus on <strong>the</strong> axis when <strong>the</strong> axial electric field has convected or diffused in<br />
(perhaps anomalously quickly by <strong>the</strong> vortices <strong>of</strong> <strong>the</strong> fine scale m = 0 instabilities discussed in<br />
section 5.8) <strong>the</strong> electron distribution function could be very non-Maxwellian. In <strong>the</strong> presence <strong>of</strong><br />
fine-scale instabilities and viscous heating, <strong>the</strong> rapid conversion <strong>of</strong> magnetic energy to <strong>the</strong>rmal<br />
energy will be accompanied by a greatly enhanced E z field, and hence enhanced Ettingshausen<br />
advection <strong>of</strong> hotter particles to <strong>the</strong> axis, leading to hot spots and runaway electrons [395].<br />
Due to <strong>the</strong> limitations <strong>of</strong> resolution in x-ray images <strong>of</strong> <strong>the</strong> <strong>pinch</strong> at stagnation, <strong>the</strong> fastest<br />
growing but small amplitude m = 0 MHD instabilities that are most effective for viscous<br />
heating <strong>of</strong> <strong>the</strong> ions accompanied by rapid conversion <strong>of</strong> magnetic energy, have not yet been<br />
detected. However <strong>the</strong> more slowly growing and larger wavelength modes will overtake <strong>the</strong>se<br />
in a time <strong>of</strong> order <strong>of</strong> <strong>the</strong> radial Alfvén time, and probably are <strong>the</strong> origin <strong>of</strong> <strong>the</strong> large amplitude<br />
hot spots which radiate many <strong>of</strong> <strong>the</strong> s<strong>of</strong>t x-rays. Axial flows away from <strong>the</strong> high pressure hot<br />
spots will occur, and <strong>the</strong> line density at each spot could drop below <strong>the</strong> critical value leading<br />
to anomalous resistivity and electron and ion beam formation as discussed in section 3.14.<br />
Such a disruption would occur several Alfvén transit times after <strong>the</strong> main s<strong>of</strong>t x-ray pulse has<br />
peaked.<br />
This tendency for runaways is greatly enhanced in vacuum spark experiments, and quite<br />
hard x-rays are generated here. Bright spots in vacuum sparks have been modelled by Vikhrev<br />
et al [396] and for a Z-<strong>pinch</strong> [328] and also Maxon et al [397]. Sotnikov et al [398] have<br />
employed a hybrid 3D MHD code including <strong>the</strong> Hall term, modelling both m = 0 and<br />
m = 1 modes. The development <strong>of</strong> an electromagnetic flute mode in a precursor plasma<br />
carrying a current [399] showed <strong>the</strong> evolution <strong>of</strong> large scale structures and <strong>the</strong> generation <strong>of</strong><br />
shorter wavelength turbulence. But simulations would require 2 µm resolution in order to<br />
model <strong>the</strong> fast-growing short wavelength modes and <strong>the</strong>ir associated viscous heating <strong>of</strong> [130].<br />
There is clearly a need to understand <strong>the</strong> energy conversion process to x-rays. Whilst it is<br />
difficult to calculate <strong>the</strong> ion kinetic energy <strong>of</strong> <strong>the</strong> implosion accurately, <strong>the</strong>re is a consensus<br />
that <strong>the</strong> energy radiated can be much greater than <strong>the</strong> kinetic energy [400]. Candidates for<br />
this are reversible pdV compressional heating by <strong>the</strong> J × B forces [321]; magnetic bubbles<br />
from m = 0 instabilities coupled to resistive dissipation [331, 332]; m = 1 instabilities<br />
leading to helical structures and enhanced resistive losses [324] associated with <strong>the</strong> current<br />
95