THEORY OF MAGNETOSTRICTION WITH APPLICATIONS TO TBXDY1-XFE2

We present a new approach to magnetostriction that is formulated to describe materials with large magnetostriction. The main idea of the theory is to derive precisely from lattice considerations the potential wells of the anisotropy energy. The theory exhibits frustration in the sense explored by the authors in the rigid case, with fine domains modelled by minimizing sequences. It is applied to the material TbxDy1-xFe2 and predicts accurately the domain structures observed by Lord in growth twinned crystals, and suggests a mechanism of magnetostriction involving a switch from a coarse domain structure to a different finer domain structure.