On the motion by singular interfacial energy (Q5944154)

If the energy and the interfacial surface are smooth, anisotropic curvature flow equations are well understood: the speed of the interface is equal to the gradient of the energy. The case of singular interfacial energy is more complex. If the interface, is a planar curve, then there exists a unique gradient characterization of the velocity. However, the surface evolution in three dimensions is more complicated and less is known about its qualitative behavior. In this paper, the authors propose a notion of solutions to surface evolution in \({\mathbb{R}}^3\) by means of geometric subdifferentials. They prove that the speed must be equal to the minimal element, in \(L^2\) norm, of geometric subdifferentials. One should note that this notion of solutions is different from the one introduced earlier by Bellettini and Novaga. While the speed is expected to be the negative divergence of a suitable Cahn-Hoffman vector field, Bellettini and Novaga assume the existence of a bounded Cahn-Hoffman vector field with bounded divergence, while in the present paper it is required that the divergence is in \(L^2\). The paper provides also a counterexample to a conjecture formulated by Cahn and Hoffman. They have conjectured that a reasonable Cahn-Hoffman vector field whose minus divergence equals the speed can be written as a gradient of a function on a facet.