Contraction of areas vs. topology of mappings (Q2636916)

From MaRDI portal
scientific article
Language Label Description Also known as
English
Contraction of areas vs. topology of mappings
scientific article

    Statements

    Contraction of areas vs. topology of mappings (English)
    0 references
    18 February 2014
    0 references
    From the introduction: ``The \(k\)-dilatation of a map between Riemannian manifolds measures how much the map stretches \(k\)-dimensional volumes. If \(F\) is a \(C^1\) map, we say that \(Dil_k(F)\leq \lambda\) if each \(k\)-dimensional surface \(\Sigma\) in the domain is mapped to an image with \(k\)-dimensional volume at most \(\lambda Vol_k(\Sigma)\). The 1-dilation is the same as the Lipschitz constant. We will study how the \(k\)-dilation of \(F\) is related to its homotopy class. The \(k\)-dilation describes the local geometry of \(F\), and we want to understand how the local geometry of \(F\) influences its global topological features. We focus on maps from the unit \(m\)-sphere to the unit \(n\)-sphere. We begin with the following question: if a map \(F:S^m\rightarrow S^n\) has small \(k\)-dilation, does the map \(F\) have to be contractible? If a map \(F:S^m\rightarrow S^n\) has \(Dil_1F<1\), then it is contractible. If \(Dil_1F<1\), then the diameter of the image of \(F\) is \(<\pi\), and so \(F\) is not surjective. In [Commun. Pure Appl. Math. 57, No. 8, 1110--1126 (2004; Zbl 1067.53056)], \textit{M.-P. Tsui} and \textit{M.-T. Wang} proved that maps with small 2-dilation are also contractible. Tsui-Wang inequality: If \(F:S^m\rightarrow S^n\) has \(Dil_2F<1\), (and if \(m,n\geq 2\)), then \(F\) is contractible. In contrast, we will show that maps with small 3-dilation may not be contractible. Example. There is a sequence of homotopically non-trivial maps \(F_j:S^4\rightarrow S^3\) with \(Dil_3(F_j)\rightarrow 0\). Our goal is to study this phenomenon. We study the following question. Main Question. Suppose that \(f_j:S^m\rightarrow S^n\) is a sequence of maps, all in the same homotopy class, with \(Dil_k(F_j)\rightarrow 0\). What can we conclude about the homotopy class of \(F_j\) ? Our main result describes the situation for maps from \(S^m\) to \(S^{m-1}\). Main Theorem. Fix an integer \(m\geq 3\). If \(k>(m+1)/2\), then there is a sequence of homotopically non-trivial maps \(F_j\) from \(S^m\) to \(S^{m-1}\) with \(k\)-dilation tending to zero. On the other hand if \(k\leq (m+1)/2\), then every homotopically non-trivial map from \(S^m\) to \(S^{m-1}\) has \(k\)-dilation at least \(c(m)>0\). In the first half of our theorem, we have to construct some homotopically non-trivial maps with tiny \(k\)-dilations. Our construction gives the following more general result. An \(h\)-principle for \(k\)-dilation. Suppose that \(F_0\) is a map from \(S^m\) to \(S^n\) with \(m>n\) and \(k>(m+1)/2\). Then for an \(\epsilon>0\), we can homotope \(F_0\) to a map \(F\) with \(k\)-dilation less than \(\epsilon\). \dots'' The work ends with formulating some open problems and three conjectures in the context.
    0 references
    0 references
    0 references
    \(k\)-dilation
    0 references
    thick tubes
    0 references
    Hopf invariant
    0 references
    directed volume
    0 references
    perpendicular pair
    0 references
    Federer-Fleming deformation operator
    0 references
    (local) grids
    0 references
    cubical cycles (chains)
    0 references
    geometric (quantitative) control
    0 references
    quantitative embedding
    0 references
    0 references
    0 references
    0 references