On Polyconvexity and Almgren Uniform Ellipticity With Respect to Polyhedral Test Pairs
Maciej Lesniak · 2025
Read On Polyconvexity and Almgren Uniform Ellipticity With Respect to Polyhedral Test Pairs by Maciej Lesniak (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2501-12472.txt.
Introduction
Maciej Lesniak, a solo author, proves a shape energy result in a 2025 math paper. The work concerns energies on surfaces inside flat space, where the energy can favor some directions over others. A classical test from Almgren checks whether such an energy is well behaved. Lesniak shows that passing that test uniformly against flat faced test chains forces a strong convexity condition on the energy density, extending a recent result by other authors.
Catalog record
| CREATOR | Maciej Lesniak |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 41 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2501-12472
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
On Polyconvexity and Almgren Uniform Ellipticity With Respect to Polyhedral Test Pairs Maciej Lesniak Abstract We study anisotropic geometric energy functionals defined on a class of k k -dimensional surfaces in a Euclidean space. The classical notion of ellipticity, coming from Almgren, for such functionals is investigated. We prove a variant of a recent result of De Rosa, Lei, and Young and show that uniform ellipticity of an anisotropic energy functional with respect to real polyhedral chains implies uniform polyconvexity of the integrand. 1 Introduction Let \mathbf{G}_{0}(n,k) denote the Grassmannian of oriented k -planes in \mathbf{R}^{n} . Given a continuous integrand F:\mathbf{G}_{0}(n,k)\to(0,\infty) , one defines anisotropic energy functional \Phi_{F}:\mathbf{P}_{k}(\mathbf{R}^{n})\to\mathbf{R} , acting on k -dimensional polyhedral chains (see § 4 for relevant definitions) in \mathbf{R}^{n} by \Phi_{F}(T)=\int F(\vec{T}(x))\,\mathrm{d}\|T\|(x)\quad\text{for }T\in\mathbf{P}_{k}(\mathbf{R}^{n}), where \|T\| is a Radon measure over \mathbf{R}^{n} and \vec{T} is a \|T\| -measurable function with values in \mathbf{G}_{0}(n,k) . Assuming F is even (i.e. F(-\xi)=F(\xi) for \xi\in\mathbf{G}_{0}(n,k) ) we may consider F to be defined on the unoriented Grassmannian and then the functional \Phi_{F} can be extended to act on varifolds by setting \Phi_{F}(V)=\int F(T)\,\mathrm{d}V(x,T) for a k -dimensional varifold V in \mathbf{R}^{n} ; see [ All72 ] . Almgren studied the regularity of varifolds which are local minimisers of \Phi_{F} and proved that they are almost everywhere regular, provided F is uniformly elliptic (abbreviated AUE) [ Alm68 , Theorems 1.4 and 1.7] . If F is elliptic, then a flat k -disc (or k -cube) is the unique minimiser of \Phi_{F} among all competitors with the same ( k-1 )-dimensional boundary as the disc (or cube). Uniform ellipticity gives additionally quantitative control on \Phi_{F} ; see § 5 . This notion heavily depends on the choice of competitors. When competitors are rectifiable currents, any uniformly convex norm on {\textstyle\bigwedge}_{k}\mathbf{R}^{n} gives rise to a uniformly elliptic integrand; cf. [ Fed69 , 5.1.2] . Burago and Ivanov [ BI04 , Theorem 3] showed that if competitors are rectifiable currents (with real coefficients), then semi-ellipticity is equivalent to being extendibly convex , i.e., there exists an extension of the integrand to a norm on the whole {\textstyle\bigwedge}_{k}\mathbf{R}^{n} . Their proof relies on showing, in non-constructive way, the existence of a polyhedral surface with prescribed tangent planes; see [ BI04 , Theorem 1, Theorem 2] . However, if competitors are allowed only to be integral currents, then there exists a continuous integrand defined on \mathbf{G}_{0}(4,2) for which the associated energy functional is elliptic, but the integrand cannot be extended to a norm on the whole {\textstyle\bigwedge}_{2}\mathbf{R}^{4} ; see [ BI04 , Theorem 5] . In [ DLY24 , Theorem 3.2] , De Rosa, Lei, and Young provided a constructive proof of [ BI04 , Theorem 2] and established an analogous theorem to [ BI04 , Theorem 3] for Lipschitz multivalued functions, namely, that ellipticity for multigraphs is equivalent to polyconvexity ; see [ DLY24 , Theorem 1.7] . The classical notion of polyconvexity , coming from calculus of variations, is defined for functionals acting on functions; cf. [ Giu03 , Chapter 5] . In this setting integrands are real-valued functions defined on the space of linear maps \Hom(\mathbf{R}^{k},\mathbf{R}^{n}) – for the purposes of the following discussion let us call them classical integrands as opposed to geometric integrands of the type \mathbf{G}_{0}(n,k)\to\mathbf{R} ; cf. [ Fed69 , 5.1.1 and 5.1.9] . The authors of [ DLY24 ] use polyconvexity for classical integrands and briefly describe its connection with being extendibly convex for related geometric integrands; see [ DLY24 , Remark 1.4, Definition 1.6, Definition 1.7] . Later they (as we do as well) call a geometric integrand F polyconvex in case it is extendibly convex. As a service to the community, we present a detailed formal discussion of the relation between extendibly convex geometric integrands and polyconvex classical integrands in section 3 . A notion stronger than polyconvexity is uniform polyconvexity (abbreviated UPC) which involves certain constant c>0 ; see 3.21 . In this article, we show that uniform ellipticity for polyhedral k -chains with constant c implies uniform polyconvexity with any constant smaller than c . This can be seen as a refinement of some results contained in [ BI04 ] and [ DLY24 ] . Theorem (see 5.10 for a precise statement) . Assume F is a geometric integrand satisfying \Lip F<\infty . If F is Almgren uniformly elliptic with respect to polyhedral k -chains with real coefficients and with ellipticity constant 0<c<\infty , then F is uniformly polyconvex with any constant 0<\tilde{c}<c . Our proof is based on [ DLY24 , Theorem 3.2] , which allows to construct polyhedral k -chains with Gaussian measures close to a given measure and provides control of the geometry of the constructed surface. Note also that the reverse implication (with \tilde{c}=c ) holds by a simple argument involving the Stokes theorem; cf. [ Fed69 , 5.1.2] . This paper is organised as follows. In Section 2 , we recall the basic terminology that will be used throughout the paper. In Section 3 , we present a formal discussion on the relationship between the notions of polyconvex and extendibly convex integrands, demonstrating their equivalence. Moreover, in the same section, we recall the definition of uniform polyconvexity for integrands. Section 4 is devoted to the classical definitions of polyhedral chains and Gaussian measures. Finally, in Section 5 , we define ellipticity for polyhedral test pairs and prove the main result. 2 Preliminaries In principle we shall follow the notation of Federer; see [ Fed69 , pp. 669–671] . In particular, A\!\sim\!B is the set-theoretic difference and \mathscr{P} denotes the set of positive integers. Diverting from Federer, if -\infty<s<r<\infty , we occasionally write (s,r) , [s,r] , (s,r] , and [s,r) to denote appropriate intervals in \mathbf{R} . We shall use letters k and n to denote two integers satisfying 0\leq k<n . We say that a function f defined on some open subset of a Banach space with values in another Banach space is of class \mathscr{C}^{l} if it is l -times differentiable and \mathrm{D}^{l}f is continuous; cf [ Fed69 , 3.1.11] . 2.1 Definition (cf. [ Fed69 , 1.3.2] ) . By \Lambda(n,k) we denote set of all increasing maps of \{1,\ldots,k\} into \{1,\ldots,n\} . Remark . Note that any \lambda\in\Lambda(n,k) is uniquely determined by \im\lambda which is a k -element subset of \{1,2,\ldots,n\} ; thus, choosing \lambda\in\Lambda(n,k) corresponds to choosing k -elements out of \{1,2,\ldots,n\} . 2.2 Definition . Let \xi\in{\textstyle\bigwedge}_{k}\mathbf{R}^{n} . We define the vector space associated with \xi by \asssp\xi=\mathbf{R}^{n}\cap\bigl\{v:v\wedge\xi=0\bigr\}\,. 2.3 Remark ( [ Fed69 , 1.6.1] ) . An element \xi\in{\textstyle\bigwedge}_{k}\mathbf{R}^{n} is simple if and only if \dim\asssp\xi=k . 2.4 Definition ( [ Fed69 , 3.2.28(b)] ) . We define the oriented Grassmannian \mathbf{G}_{0}(n,k) as \mathbf{G}_{0}(n,k)={\textstyle\bigwedge}_{k}\mathbf{R}^{n}\cap\bigl\{\xi:\xi\text{ is simple},\,|\xi|=1\bigr\}\,. 2.5 Remark . Let \xi\in{\textstyle\bigwedge}_{k}\mathbf{R}^{n} . Recall from [ Fed69 , 1.8.1] that \|\xi\|=|\xi| if and only if \xi is simple. 2.6 Definition . Given a vectorspace X we write \mathrm{End}({X}) for the space of all linear maps mapping X to X , i.e., \mathrm{End}({X})=\Hom(X,X) . 2.7 Definition ( [ Fed69 , 1.7.4] ) . A linear map j\in\Hom(\mathbf{R}^{k},\mathbf{R}^{n}) is called an orthogonal injection if j^{*}\circ j=\mathds{1}_{\mathbf{R}^{k}} . If p\in\Hom(\mathbf{R}^{n},\mathbf{R}^{k}) and j=p^{*} is an orthogonal