betaFeedback

WorksMathematics

Energies with values in networks and geometric quasiconvexity

Abstract

We study the asymptotic behaviour, as ε→0+\varepsilon\to 0^+, of the pp-Dirichlet energy of maps u ⁣:Ω⊂R2→R2u\colonΩ\subset\mathbb{R}^2 \to \mathbb{R}^2 constrained to take values in the rescaled periodic network εZ\varepsilon\mathcal Z, where Z\mathcal Z is a standard square network in R2\mathbb{R}^2. The oscillations are carried by the constraint in the target, not by the energy density, and their effect depends on pp. For 1<p<21<p<2 the constraint disappears in the limit: the ΓΓ-limit is an integral functional on the whole of W1,p(Ω,R2)\rm{W}^{1,p}(Ω, \mathbb{R}^2 ), whose density is given by an asymptotic cell formula on tilted cells. This density is equal to (∣a1∣+∣a2∣)p(|a_1|+|a_2|)^p on rank-one matrices a⊗νa\otimesν, ∣ν∣=1|ν|=1, and exceeds its convex envelope by a quantity comparable to ∣det⁡A∣ ∣A∣p−2|\det A|\,|A|^{p-2}, with constants of order (2−p)−1(2-p)^{-1}. Consequently, although the original energy density is convex, the limit density is quasiconvex but neither convex nor polyconvex: quasiconvexity is generated by the geometry of the constraint. For p≥2p\ge2 the constraint survives in the limit as det⁡∇u=0\det\nabla u=0, and the ΓΓ-limit is the degenerate energy ∫Ω(∣∇u1∣+∣∇u2∣)p dx\int_Ω(|\nabla u_1|+|\nabla u_2|)^p\, \rm{d} x, that is, the pp-energy of uu considered as a map into the normed plane (R2,∣⋅∣1)(\mathbb{R}^2,|\cdot|_1), the metric limit of the networks.

Dedicated to David Kinderlehrer on the occasion of his 85th birthday

Introduction and main results

In this paper, we study the asymptotic behavior of a prototypical family of variational integrals defined on vector-valued Sobolev functions, and whose values are constrained to lie on a periodic network, which is seen at a scale ε\varepsilon that is small compared to the scale of observation. Given a bounded Lipschitz domain Ω⊂R2\Omega\subset\mathbb{R}^{2} and an exponent p>1p > 1, we consider

Fεp(u)={∫Ω∣∇u(x)∣p dxif u(x)∈εZ for almost every x∈Ω,+∞otherwise,(1)F_{\varepsilon}^{p}(u) = \begin{cases} \int_{\Omega} \lvert\nabla u(x) \rvert^{p}\,\mathrm{d}x & \text{if } u(x) \in\varepsilon\mathcal{Z} \text{ for almost every } x \in\Omega,\\ +\infty& \text{otherwise,} \end{cases} \tag*{(1)}

where Z\mathcal{Z} is a periodic network in R2\mathbb{R}^{2} and ∣⋅∣\lvert\cdot\rvert denotes the Frobenius norm on R2×2\mathbb{R}^{2 \times2}. The prototypical example, that is our focus here, is the union of the coordinate lines

Z={(a,b)∈R2:a∈12+Z or b∈12+Z}=⋃i∈Z2(i+∂Q),Q=(−12,12)2.(2)\mathcal{Z} = \left\{(a,b) \in\mathbb{R}^{2} : a \in\frac{1}{2} + \mathbb{Z} \text{ or } b \in\frac{1}{2} + \mathbb{Z}\right\} = \bigcup_{i \in\mathbb{Z}^{2}} (i + \partial Q), \qquad Q = \left(-\frac{1}{2},\frac{1}{2}\right)^{2}. \tag*{(2)}

The effective domain of FεpF_{\varepsilon}^{p} is the class of constrained Sobolev maps W1,p(Ω,εZ)={u∈W1,p(Ω,R2):u(x)∈εZ for almost all x∈Ω}W^{1,p}(\Omega,\varepsilon\mathcal{Z}) = \{u \in W^{1,p}(\Omega,\mathbb{R}^{2}) : u(x) \in\varepsilon\mathcal{Z} \text{ for almost all } x \in\Omega\}. Constraints of this type arise naturally in models of media with a fine internal architecture, where admissible deformations, order parameters or internal variables are forced onto a network of preferred states: lattice or network-like microstructures, strongly anisotropic media, and discrete systems in which the order parameter is quantized. Moreover, this can be regarded as a prototype for energies with values on an oscillating manifold, with the heuristic motivation that low-energy functions will be located approximately on a network with small oscillations within the manifold (which is the case for curves [18]). In all such models, one is led to ask what effective continuum theory survives the limit ε→0+\varepsilon\to0^{+}.

We will refer to this limit as a homogenization process, although it is not a usual one. Two features distinguish (1) from the energies usually encountered in homogenization theory. First, the integrand is the pp-Dirichlet integrand: all oscillations are carried by the constraint, not by the energy density. Second, and more importantly, the oscillations take place in the target space rather than in the spatial xx-variable. The classical homogenization results deal with integrands f(x/ε,∇u)f(x/\varepsilon,\nabla u) oscillating in xx, and the effective energy is obtained through a cell formula on the domain (see Braides [15], Müller [55] and the monograph [19]); here, instead, the cell problem lives in the target, and the relevant cell is the pre-image QA=A−1(Q)Q_A=A^{-1}(Q) of the target cell under the affine datum with gradient AA. This is why, unlike in the classical setting, the homogenization formula depends on the rank of the matrix AA. More importantly, although the original integrand is convex, the homogenized density is not: it is quasiconvex but neither convex nor polyconvex, and the distance from its convex envelope is controlled by the determinant. This lack of convexity is produced by the geometry of the constraint alone and it is an important aspect of the present work.

The main goal of the paper is to characterize the Γ\Gamma-limit of FεpF_\varepsilon^p as ε→0+\varepsilon\to0^+ and to understand how the interplay between pp and the geometry of Z\mathcal{Z} affects the limit model. A dichotomy appears at p=2p=2: for p<2p<2 the constraint disappears in the limit, but it leaves its trace in a genuinely quasiconvex energy density; for p>2p>2 it survives as the differential constraint det⁡∇u=0\det\nabla u=0. For p<2p<2 the result is the following.

Theorem A (subcritical regime: geometric quasiconvexity). Let 1<p<21<p<2. Then FεpF_\varepsilon^p Γ\Gamma-converges, with respect to the strong LpL^p-topology, to

Fhomp(u)=∫Ωfhomp(∇u(x)) dx,u∈W1,p(Ω,R2).F_{\mathrm{hom}}^p(u)=\int_{\Omega}f_{\mathrm{hom}}^p(\nabla u(x))\,\mathrm{d}x,\qquad u\in W^{1,p}(\Omega,\mathbb{R}^2).

where fhomp:R2×2→[0,+∞)f_{\mathrm{hom}}^p:\mathbb{R}^{2\times2}\to[0,+\infty) is a continuous, pp-homogeneous, quasiconvex integrand with the following properties:

  1. (asymptotic homogenization formula) on matrices of rank two

fhomp(A)=lim inf⁡T→+∞∣det⁡A∣T2inf⁡{∫TQA∣∇v∣p dx:u∈W1,p(TQA,Z), u=uA in a neighbourhood of ∂TQA}(3)f_{\mathrm{hom}}^p(A)=\liminf_{T\to+\infty}\frac{|\det A|}{T^2}\inf\left\{\int_{TQ_A}|\nabla v|^p\,\mathrm{d}x:u\in W^{1,p}(TQ_A,\mathcal{Z}),\ u=u_A\text{ in a neighbourhood of }\partial TQ_A\right\} \tag*{(3)}

where uAu_A is the projection of the linear function AxAx on Z\mathcal{Z} (see Definition 2.3);

  1. (qualified two-sided estimates) there exist c‾,C‾>0\underline{c},\overline{C}>0 independent of pp and explicitly computable, such that

N(A)p+c‾(p−1)12(1−p)∣det⁡A∣(2−p)∣A∣2−p≤fhomp(A)≤N(A)p+C‾∣det⁡A∣(2−p)∣A∣2−p(4)N(A)^p+\underline{c}(p-1)^{\frac{1}{2(1-p)}}\frac{|\det A|}{(2-p)|A|^{2-p}}\leq f_{\mathrm{hom}}^p(A)\leq N(A)^p+\overline{C}\frac{|\det A|}{(2-p)|A|^{2-p}} \tag*{(4)}

for every A≠0A\ne0, where, if A1A_1 and A2A_2 denote the rows of AA,

N(A):=∣A1∣+∣A2∣.(5)N(A):=|A_1|+|A_2|. \tag*{(5)}

In particular, fhomp(A)=N(A)pf_{\mathrm{hom}}^p(A)=N(A)^p on rank-one matrices, and the function fhompf_{\mathrm{hom}}^p is neither convex nor polyconvex.

The density fhompf_{\mathrm{hom}}^p is given by an asymptotic cell formula on the tilted cells TQA=TA−1(Q)TQ_A=TA^{-1}(Q), with the projected linear map uA=P0(A⋅)u_A=P_0(A\mathbin{\cdot}) as boundary datum (Definitions 2.3 and 2.7); by (4) it has standard pp-growth, ∣A∣p≤fhomp(A)≤C∣A∣p|A|^p\leq f_{\mathrm{hom}}^p(A)\leq C|A|^p. The restriction p<2p<2 is exactly what makes the radial projection onto the boundary of a cell an admissible operation in W1,pW^{1,p}, the singularity ∣x∣−p|x|^{-p} integrable in the plane precisely for p<2p<2 (Lemma 2.2): every Sobolev map can be projected onto the network at a uniformly controlled energy cost, which is why the limit domain is the whole of W1,p(Ω,R2)W^{1,p}(\Omega,\mathbb{R}^2).

The lower bound in (4) shows that fhompf_{\mathrm{hom}}^p is not convex. Indeed, writing I=12(2e1⊗e1)+12(2e2⊗e2)I=\frac{1}{2}(2e_1\otimes e_1)+\frac{1}{2}(2e_2\otimes e_2), by (i) and (ii) we have

12fhomp(2e1⊗e1)+12fhomp(2e2⊗e2)=2p<2p+cp−12−p812(1−p)2p−22≤fhomp(I).(6)\frac{1}{2}f_{\mathrm{hom}}^p(2e_1\otimes e_1)+\frac{1}{2}f_{\mathrm{hom}}^p(2e_2\otimes e_2)=2^p<2^p+c\frac{p-1}{2-p}8^{\frac{1}{2(1-p)}}2^{\frac{p-2}{2}}\leq f_{\mathrm{hom}}^p(I). \tag*{(6)}

As a consequence, fhompf_{\mathrm{hom}}^p is not polyconvex either: a polyconvex function with growth of order p<2p<2 must be convex, since an affine function of (A,det⁡A)(A,\det A) lying below a function with subquadratic growth cannot depend on det⁡A\det A.

Geometric quasiconvexity. We consider fhompf_{\mathrm{hom}}^p as a quasiconvex function generated by the geometry of the constraint, from which the title of this paper comes. It is worth commenting on its form. The two terms in (4) have a transparent meaning. The first is the convex envelope of fhompf_{\mathrm{hom}}^p. Indeed, NpN^p is convex and lies below fhompf_{\mathrm{hom}}^p; conversely, if A1,A2≠0A_1,A_2\ne0 and λ=∣A1∣/N(A)\lambda=|A_1|/N(A), then

A=λ(1λe1⊗A1)+(1−λ)(11−λe2⊗A2)A=\lambda\left(\frac{1}{\lambda}e_1\otimes A_1\right)+(1-\lambda)\left(\frac{1}{1-\lambda}e_2\otimes A_2\right)

References

  1. [1]E. Acerbi and N. Fusco, Semicontinuity problems in the calculus of variations, Arch. Rational Mech. Anal. 86 (1984), 125–145.DOI
  2. [2]E. Acerbi and N. Fusco, An approximation lemma for W^{1,p} functions, in: Material Instabilities in Continuum Mechanics (Edinburgh, 1985–1986), Oxford Univ. Press, New York, 1988, 1–5.
  3. [3]A. Agazzi et al., Sharp higher integrability theory, Part I: The L^p-norm of the Beurling–Ahlfors transform, preprint, arXiv:2609.28311, 2026.arxiv.org/abs/2609.28311
  4. [4]J.-J. Alibert and B. Dacorogna, An example of a quasiconvex function that is not polyconvex in two dimensions, Arch. Rational Mech. Anal. 117 (1992), 155–166.DOI
  5. [5]L. Ambrosio, Metric space valued functions of bounded variation, Ann. Scuola Norm. Sup. Pisa Cl. Sci. (4) 17 (1990), 439–478.
  6. [6]L. Ambrosio, N. Fusco and D. Pallara: Functions of bounded variation and free discontinuity problems. Oxford Mathematical Monographs. The Clarendon Press, Oxford University Press, New York, 2000. xviii+434 pp.
  7. [8]J.-F. Babadjian and V. Millot, Homogenization of variational problems in manifold valued Sobolev spaces, ESAIM Control Optim. Calc. Var. 16 (2010), 833–855.
  8. [9]E. J. Balder, A general approach to lower semicontinuity and lower closure in optimal control theory, SIAM J. Control Optim., 22 (1984), pp. 570–598.DOI
  9. [10]J. M. Ball, Convexity conditions and existence theorems in nonlinear elasticity, Arch. Rational Mech. Anal. 63 (1977), 337–403.DOI
  10. [11]J. M. Ball and R. D. James, Fine phase mixtures as minimizers of energy, Arch. Rational Mech. Anal. 100 (1987), 13–52.
  11. [12]J. M. Ball and F. Murat, W^{1,p}-quasiconvexity and variational problems for multiple integrals, J. Funct. Anal. 58 (1984), 225–253.DOI
  12. [13]F. Bethuel, H. Brezis and F. Hélein, Ginzburg–Landau Vortices, Progress in Nonlinear Differential Equations and their Applications 13, Birkhäuser, Boston, 1994.
  13. [14]J. Bourgain, H. Brezis and P. Mironescu, Lifting in Sobolev spaces, J. Anal. Math. 80 (2000), 37–86.DOI
  14. [15]A. Braides, Homogenization of some almost periodic coercive functional, Rend. Accad. Naz. Sci. XL Mem. Mat. (5) 9 (1985), no. 1, 313–321.
  15. [16]A. Braides, Loss of polyconvexity by homogenization, Arch. Rational Mech. Anal. 127 (1994), 183–190.DOI
  16. [17]A. Braides, G. Buttazzo and I. Fragalà, Riemannian approximation of Finsler metrics, Asymptot. Anal. 31 (2002), 177–187.DOI
  17. [18]A. Braides, A. Cancedda and V. Chiadò Piat, Homogenization of metrics in oscillating manifolds, ESAIM Control Optim. Calc. Var. 23 (2017).arxiv.org/abs/1511.04250
  18. [19]A. Braides and A. Defranceschi, Homogenization of Multiple Integrals, Oxford Lecture Series in Mathematics and its Applications, 12, Oxford Univ. Press, New York, 1998.DOI
  19. [20]A. Braides, M. Maslennikov and L. Sigalotti, Homogenization by blow-up, Appl. Anal. 87 (2008), no. 12, 1341–1356.DOI
  20. [21]H. Brezis and L. Nirenberg, Degree theory and BMO. I. Compact manifolds without boundaries, Selecta Math. (N.S.) 1 (1995), 197–263.
  21. [22]M. R. Bridson and A. Haefliger, Metric Spaces of Non-Positive Curvature, Grundlehren der Mathematischen Wissenschaften 319, Springer-Verlag, Berlin, 1999.DOI
  22. [23]D. Burago, Periodic metrics, in Representation Theory and Dynamical Systems, Adv. Soviet Math. 9, Amer. Math. Soc., Providence, 1992, 205–210.
  23. [24]G. Cassese, A solution to Morrey’s problem in ℝ^{2×m}, preprint, arXiv:2608.03488, 2026.arxiv.org/abs/2608.03488
  24. [25]D. Chiron, On the definitions of Sobolev and BV spaces into singular spaces and the trace problem, Commun. Contemp. Math. 9 (2007), 473–513.DOI
  25. [26]B. Dacorogna, Direct Methods in the Calculus of Variations, 2nd edition, Applied Mathematical Sciences 78, Springer, New York, 2008.
  26. [27]B. Dacorogna, I. Fonseca, J. Malý and K. Trivisa, Manifold constrained variational problems, Calc. Var. Partial Differential Equations 9 (1999), 185–206.DOI
  27. [28]B. Dacorogna and J.-P. Haeberly, On convexity properties of homogeneous functions of degree one, Proc. Roy. Soc. Edinburgh Sect. A 126 (1996), 947–956.DOI
  28. [29]G. Daskalopoulos and C. Mese, Harmonic maps between singular spaces I, Comm. Anal. Geom. 18 (2010), 257–337.DOI
  29. [30]J. Eells and B. Fuglede, Harmonic Maps Between Riemannian Polyhedra, Cambridge Tracts in Mathematics 142, Cambridge University Press, Cambridge, 2001.
  30. [31]H. Federer and W. H. Fleming, Normal and integral currents, Ann. of Math. (2) 72 (1960), 458–520.DOI
  31. [32]I. Fonseca and G. Leoni, Modern Methods in the Calculus of Variations: Lᵖ Spaces, Springer Monographs in Mathematics, Springer, New York, 2007.DOI
  32. [33]I. Fonseca and S. Müller, Quasi-convex integrands and lower semicontinuity in L¹, SIAM J. Math. Anal. 23 (1992), no. 5, 1081–1098.DOI
  33. [34]L. Gennaioli, Lipschitz continuity of harmonic maps between RCD(K,N) spaces and CAT(κ) spaces, preprint, arXiv:2607.27010, 2026.
  34. [35]L. Gennaioli, N. Gigli, H.-C. Zhang and X.-P. Zhu, Comments on the regularity of harmonic maps between singular spaces, J. Geom. Anal. 36 (2026), Paper No. 208.arxiv.org/abs/2408.00479
  35. [36]D. Gilbarg and N. S. Trudinger, Elliptic Partial Differential Equations of Second Order, Springer-Verlag, Berlin, 2001.
  36. [37]M. Gromov, Partial Differential Relations, Springer-Verlag, Berlin, 1986.DOI
  37. [38]M. Gromov and R. Schoen, Harmonic maps into singular spaces and p-adic superrigidity for lattices in groups of rank one, Publ. Math. Inst. Hautes Études Sci. 76 (1992), 165–246.
  38. [39]A. Guerra and J. Kristensen, Automatic quasiconvexity of homogeneous isotropic rank-one convex integrands, Arch. Ration. Mech. Anal. 245 (2022), 479–500.DOI
  39. [40]R. Hardt, D. Kinderlehrer and F.-H. Lin, Existence and partial regularity of static liquid crystal configurations, Comm. Math. Phys. 105 (1986), 547–570.DOI
  40. [41]R. Hardt and F.-H. Lin, Singularities for p-energy minimizing unit vectorfields on planar domains, Calc. Var. Partial Differential Equations 3 (1995), 311–341.
  41. [42]S. Hildebrandt, H. Kaul and K.-O. Widman, An existence theorem for harmonic mappings of Riemannian manifolds, Acta Math. 138 (1977), 1–16.DOI
  42. [43]W. Hurewicz and H. Wallman, Dimension Theory, Princeton University Press, Princeton, 1941.
  43. [44]M. B. Karmanova, Area and co-area formulas for mappings of the Sobolev classes with values in a metric space, Sibirsк. Mat. Zh. 48 (2007), 778–788.DOI
  44. [45]D. Kinderlehrer and P. Pedregal, Gradient Young measures generated by sequences in Sobolev spaces, J. Geom. Anal. 4 (1994), 59–90.DOI
  45. [46]B. Kirchheim, Rectifiable metric spaces: local structure and regularity of the Hausdorff measure, Proc. Amer. Math. Soc. 121 (1994), 113–123.
  46. [47]B. Kirchheim, Rigidity and Geometry of Microstructures, MPI-MIS Lecture Notes 16, Leipzig, 2003.
  47. [48]B. Kirchheim and J. Kristensen, On rank one convex functions that are homogeneous of degree one, Arch. Ration. Mech. Anal. 221 (2016), 527–558.DOI
  48. [49]N. J. Korevaar and R. M. Schoen, Sobolev spaces and harmonic maps for metric space targets, Comm. Anal. Geom. 1 (1993), 561–659.DOI
  49. [50]P. Logaritsch and E. Spadaro, A representation formula for the p-energy of metric space-valued Sobolev maps, Commun. Contemp. Math. 14 (2012), 1250043.arxiv.org/abs/1112.0122
  50. [51]A. Lytchak and S. Wenger, Area minimizing discs in metric spaces, Arch. Ration. Mech. Anal. 223 (2017), 1123–1182.
  51. [52]A. Lytchak and S. Wenger, Energy and area minimizers in metric spaces, Adv. Calc. Var. 10 (2017), 407–421.arxiv.org/abs/1507.02670
  52. [53]M. Marcus and V. J. Mizel, Transformations by functions in Sobolev spaces and lower semicontinuity for parametric variational problems, Bull. Amer. Math. Soc. 79 (1973), 790–795.DOI
  53. [54]C. B. Morrey, Jr., Quasi-convexity and the lower semicontinuity of multiple integrals, Pacific J. Math. 2 (1952), 25–53.DOI
  54. [55]S. Müller, Homogenization of nonconvex integral functionals and cellular elastic materials, Arch. Rational Mech. Anal. 99 (1987), no. 3, 189–212.DOI
  55. [56]S. Müller, On quasiconvex functions which are homogeneous of degree 1, Indian Univ. Math. J. 41 (1992), 295–301.
  56. [57]S. Müller and V. Šverák, Convex integration for Lipschitz mappings and counterexamples to regularity, Ann. of Math. (2) 157 (2003), 715–742.
  57. [58]S. B. Nadler, Continuum Theory: An Introduction, Marcel Dekker, New York, 1992.
  58. [59]P. Pansu, Croissance des boules et des géodésiques fermées dans les nilvariétés, Ergodic Theory Dynam. Systems 3 (1983), 415–445.DOI
  59. [60]Yu. G. Reshetnyak, Sobolev classes of functions with values in a metric space, Siberian Math. J. 38 (1997), 567–583.DOI
  60. [61]F. Rindler, Calculus of Variations, Springer, 2018.
  61. [62]T. Serbinowski, Harmonic maps into metric spaces with curvature bounded above, Ph.D. thesis, University of Utah, 1995.
  62. [63]E. M. Stein, Singular Integrals and Differentiability Properties of Functions, Princeton Mathematical Series 30, Princeton University Press, Princeton, NJ, 1970.
  63. [64]V. Šverák, Quasiconvex functions with subquadratic growth, Proc. Roy. Soc. London Ser. A 433 (1991), 723–725.DOI
  64. [65]V. Šverák, New examples of quasiconvex functions, Arch. Rational Mech. Anal. 119 (1992), 293–300.DOI
  65. [66]L. Székelyhidi Jr., On quasiconvex hulls in symmetric 2 × 2 matrices, Ann. Inst. H. Poincaré Anal. Non Linéaire 23 (2006), 865–876.
  66. [67]G. T. Whyburn, Analytic Topology, American Mathematical Society Colloquium Publications 28, New York, 1942.DOI

Paper details

vlnc:W_rcfva5k5
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Mathematics

A large language model

View statement
Looking for related papers…
Contents