http://arxiv.org/abs/1409.3299
Observations of several types of neutron stars indicate surface temperature inhomogeneities. In recent years magneto-thermal simulations have supported the idea that the magnetic field and anisotropic heat conduction play important roles in generating these inhomogeneities. Simulations rely on crustal microphysics input heretofore calculated at the level of a plasma model — neglecting lattice structure and electron polarizability. We focus on the low density outer envelope, treating both of these elements by a proper periodization of the magnetic Thomas-Fermi model. Our solution method involves a novel domain decomposition and we describe a scalable implementation using \textit{Hypre}. The method may be seen as a prototype for the general class of problems involving nonlinear charge screening of periodic, quasi-low-dimensionality structures, e.g. liquid crystals. Findings include low density $c'<0$ elastic instabilities for both bcc and fcc lattices, reminiscent of the situation in some light actinides, and phonon thermal conductivity three orders of magnitude larger than that derived from the plasma model. The former result suggests there is a symmetry-lowering transition to a tetragonal or orthorhombic lattice. The latter indicates transport anisotropy may be greatly reduced within $\sim10$ meters of the surface, giving the effect of a “heat-spreader cladding” which may significantly increase the size of polar hot spots and alter pulse profiles.
T. Engstrom, V. Crespi, B. Owen, et. al.
Fri, 12 Sep 14
4/61
Comments: 8 pages, 5 figures, submitted to ApJ on 9/10/14
You must be logged in to post a comment.