Singular minimal surfaces. (Q4407664)

In this survey article the author discusses various results, many of them proved by himself, related to the variational integral \(E_\alpha(u)= \int u^\alpha \sqrt{1+| Du|^2 dx}\) and its extremals which solve the equation \(\text{div}(Du/\sqrt{1 +| Du|^2})= \alpha/(u \sqrt{1 + | Du|^2})\). Here \(u\) is a nonnegative function defined on some domain \(\Omega\subset \mathbb{R}^n\) and \(\alpha\) is a real parameter. For \(\alpha= 1E_\alpha\) can be interpreted as the potential energy of the graph of \(u\) considered as a heavy surface in a vertical gravitational field whereas for any \(\alpha= m\in\mathbb{N} E_m\) is the surface area of the rotational surface \(\{(x, u(x)\omega)\mid x\in\Omega, \omega\in \mathbb{S}^m\subset \mathbb{R}^{m+1}\}\subset \mathbb{R}^{n+ m+1}\). Setting \(n= 1\) the author uses the latter interpretation to discuss the \((m+ 1)\)-dimensional analogue of the family of catenoids and their significance for Plateau's problem. The author goes on to explain how \(E_\alpha\) may be defined for any \(\alpha\geq 0\) using the function spaces \(\text{BV}(\Omega)\) and he describes the pertinent existence and regularity results. The regularity of solutions depending of course on whether they do or do not have zeroes, corresponding criteria are cited. In the last section entire solutions (i. e. extremals defined on all of \(\mathbb{R}^n\)) are discussed. Though in general, as the author points out, smooth entire solutions may exist, he cites two nonexistence results of his; one holds for solutions of strictly sublinear growth, the other one for stable solutions in the range \(\alpha+ n< 5.23\dots,\alpha> 0\). For the entire collection see [Zbl 1007.00073].