Title:	Dipolar Magnets and Frustration
DNr:	SNIC 2016/10-17
Project Type:	SNIC Large Compute
Principal Investigator:	Patrik Henelius <henelius@kth.se>
Affiliation:	Kungliga Tekniska högskolan
Duration:	2016-07-01 – 2017-07-01
Classification:	10304
Keywords:

Abstract

This proposal aims to investigate frustrated dipolar magnets. The main goals are to calculate the demagnetizing factors necessary for any quantitative study of magnetic susceptibilities in insulating magnets, to understand the maximum that appears in the magnetic susceptibility for certain frustrated magnets, and to investigate the freezing process in the spin ice family of materials. Our preliminary results indicate that the demagnetizing factors that have been tabulated, and much cited, so far are not valid for insulating, lattice-based magnets. Furthermore, we find that the maximum in the spin ice susceptibility is a magnetic analog of the Joule temperature in the theory of classic gases. The scientific methods we use in this project are tailored to the specific parts. In order to calculate the demagnetizing factors we are developing a method to solve the field equations iteratively, capable of handling several million dipolar-coupled spins. In order to determine the magnetic Joule and Boyle temperatures we are developing a magnetic analog of the classical theory of thermodynamics, and apply large-scale parallel tempering simulations to verify our theory . Finally we are writing a massively parallel Monte Carlo code to study the freezing process in the spin-ices. Our algorithm is capable of calculating the neutron structure factor for systems ten times larger than previously studied, with an equal increase in the spatial resolution. Of paramount importance in all parts of the project is a very close collaboration with our experimental colleagues. In the collaborations we develop state-of-the-art computational tools to meet the many challenges encountered when investigating frustrated magnetic systems. Having completed the development of several of the program packages a large allocation on Beskow would mean that we can could calculate neutron spectra, demagnetizing factors and critical temperatures and publish the results of our current projects.