Relative Tchebychev hypersurfaces which are also translation hypersurfaces (Q1026994)

Relative hypersurface theory studies an immersion \( x: M \to \mathbb{R}^{n+1} \) where \( M \) is a manifold of dimension \( n \geq 2 \), together with a transversal map \( y : M \to \mathbb{R}^{n+1} \) parallel to \( x \), i.e. \( y(p) \notin \text{im}(dx)_{p} \) but \( \text{im}(dy)_{p} \subseteq \text{im}(dx)_{p} \) for all \( p \in M \). So \( y(p) \) serves as an affine analogue of the Euclidean normal vector at \( p \). The usual regularity assumption for \( x \) is that it should be non-degenerate, meaning that all its height functions have only non-degenerate critical values. The pair \( (x,y) \) is then called a relative hypersurface. Associated to this setting are two connections \( \nabla \), \( \widehat{\nabla} \) and several tensor fields, including a pseudo Riemannian metric \( h \), the relative metric, the shape operator \( S \), a cubic form \( C \), and its trace contraction \( T \), the Chebyshev vector field. \( (x,y) \) is called a Chebyshev hypersurface if the Chebyshev operator \( L := \frac{1}{2}\,S-\frac{n}{n+2}\widehat{\nabla} T \) is a multiple of the identity. For further details [see \textit{A.-M. Li, H. L. Liu, A. Schwenk-Schellschmidt, U. Simom, C. P. Wang}, Geom. Dedicata 66, No.~2, 203--221 (1997; Zbl 0878.53010)], where it is made clear, that the class of Chebyshev hypersurfaces is very ample and contains in particular the affine spheres of the Blaschke/Berwald normalizations. The present author first gives a new characterization of Chebyshev hypersurfaces in terms of the Blaschke/Berwald geometry. In view of the richness of these surfaces, he then studies the subclass of hypersurfaces which are at the same time translation hypersurfaces, finally ending in a full local classification of these hypersurfaces. In the special case of centroaffine geometry (\( y = x \)) only paraboloids are left in the list.