Sharp \(L^1\)-Poincaré inequalities correspond to optimal hypersurface cuts (Q494597)

Let \(\Omega \subset \mathbb{R}^n\) be a convex set. If \(u : \Omega \to \mathbb{R}\) has zero mean, then the Poincaré inequality states that \(\|u\|_{L^p} \leq c_p \mathrm{diam} (\Omega) \| \nabla u\|_{L^p}\), where \(c_p\) is a constant depending on \(p\). A result of \textit{G. Acosta} and \textit{R. G. Duran} [Proc. Am. Math. Soc. 132, No. 1, 195--202 (2004; Zbl 1057.26010)] states that the best constant for the \(L^1\) norm is \(c_1 = \frac{1}{2}\). The main result of the paper is to prove the inequality \(\|u\|_{L^1(\Omega)} \leq \frac{2}{\log 2} M(\Omega) \| \nabla u\|_{L^1(\Omega)}\), where \(M(\Omega)\) is the average distance between a point of \(\Omega\) and the center of gravity of \(\Omega\). The author applies this result to prove inequalities for the surface area of hyperplane cuts of \(\Omega\).