Martingales, laminates and minimal Korn inequalities

Gabriele Cassese · 2025

Read Martingales, laminates and minimal Korn inequalities by Gabriele Cassese (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-2512-02784.txt.

Introduction

Gabriele Cassese works on Korn inequalities, which bound how much a material can distort overall from how it stretches. In elasticity, the full gradient of a displacement is controlled by its symmetric part alone. That symmetric part has fewer components, so the bound is far from obvious.

Cassese connects the problem to martingales from probability and to laminates from materials science. The aim is minimal constants: the smallest factors for which the inequality still holds. Sharp constants matter when engineers and mathematicians need guarantees, not estimates, for how elastic bodies deform under load. Best marks in such bounds are the ones plant and bone work cite when they need a wall they can lean on.

Catalog record

CREATORGabriele Cassese
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE77 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2512-02784

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Martingales, laminates and minimal Korn inequalities Gabriele Cassese Abstract Korn’s inequalities show that the L 2 L^{2} -norm of ∇ u \nabla u can be controlled by the L 2 L^{2} -norm of Sym ⁡ ( ∇ u ) \mathrm{Sym}(\nabla u) , which only has d ⁡ ( d + 1 ) / 2 d(d+1)/2 components. In [J. Math. Pures Appl. 148 (2021), pp. 199-220] Chipot posed the question of how many scalar measurements are needed to have a Korn-type control on ∇ u \nabla u when u u is in H 0 1 ​ ( Ω ) H_{0}^{1}(\Omega) and H 1 ​ ( Ω ) H^{1}(\Omega) , introducing the minimal numbers N ⁡ ( d , Ω ) N(d,\Omega) and N ′ ​ ( d , Ω ) N^{\prime}(d,\Omega) respectively. He proved general bounds and calculated several low-dimensional values of N , N ′ N,N^{\prime} . We reframe Chipot’s problem in the language of rank-one convexity and quasiconvexity and obtain a purely algebraic characterisation of when such inequalities hold, which yields the sharp bounds N ⁡ ( d , Ω ) \displaystyle N(d,\Omega) = 2 ​ d ​ ( 1 − o ⁡ ( 1 ) ) \displaystyle=2d(1-o(1)) N ′ ​ ( d , Ω ) \displaystyle N^{\prime}(d,\Omega) = 2 ​ d − 1 . \displaystyle=2d-1. As a consequence, we recover and streamline several of Chipot’s results, we obtain a dimension-optimal Korn inequality and several sharp estimates for the best constant for various Korn-type inequalities. Generalisations to the rectangular case and to general L p L^{p} estimates are also considered. The central new ingredient of our approach is a systematic connection between laminates and martingales which produces explicit families of laminates realising these bounds. This method is of independent interest in the calculus of variations: for instance, we use it to obtain a new quick and quantitative proof of Ornstein’s non-inequality, valid for all first order homogeneous operators in ℝ 2 × 2 \mathbb{R}^{2\times 2} and for a large class of operators in general dimensions (including Korn’s ∇ u + ∇ u t 2 \frac{\nabla u+\nabla u^{t}}{2} and ∇ u + ∇ u t 2 − div ⁡ ( u ) ​ Id d \frac{\nabla u+\nabla u^{t}}{2}-\mathrm{div}(u)\frac{\mathrm{Id}}{d} ). 1 Introduction Korn’s first inequality, in its arguably most well known form, is the statement that for all u\in C^{\infty}_{c}(\mathbb{R}^{d},\mathbb{R}^{d}), and all p\in(1,\infty) we have \|\nabla u\|_{L^{p}(\mathbb{R}^{d})}\lesssim_{p,d}\|\mathrm{Sym}(\nabla u)\|_{L^{p}(\mathbb{R}^{d})}, where \mathrm{Sym}(A):=\frac{A+A^{t}}{2} is the symmetric part of A and the norm considered is the Frobenius norm for matrices, i.e. \|A\|^{2}=\mathrm{Tr}(AA^{t}) . Korn’s second inequality is more general: for p\in(1,\infty) given a (Lipschitz) domain \Omega in \mathbb{R}^{d} , for all u\in W^{1,p}(\Omega,\mathbb{R}^{d}) we have \|u\|_{W^{1,p}(\Omega)}\lesssim_{p,\Omega}\|u\|_{L^{p}(\Omega)}+\|\mathrm{Sym}(\nabla u)\|_{L^{p}(\Omega)}. These two inequalities have many applications (see [ 28 ] and references therein), but these are not the inequalities in their strongest form. For example, if we take \mathrm{Sym}_{0}(A):=\mathrm{Sym}\left(A-\mathrm{Tr}(A)\frac{\mathrm{Id}}{d}\right) (i.e., the projection of A onto its trace-free symmetric part), we still have 1 1 1 Note that the second form of this inequality does not hold for d=2 \|\nabla u\|_{L^{p}(\Omega)}\lesssim_{p,\Omega}\|\mathrm{Sym}_{0}(\nabla u)\|_{L^{p}(\Omega)}. Chipot analysed this problem more precisely in [ 22 ] : namely, he asked what is the smallest number of coordinates that one needs on the right-hand side for an inequality such as Korn’s to hold (at least for p=2 )? More formally, Chipot studied the quantity N(d), defined as Definition 1 . Let N(d) be the smallest integer k for which there exist \ell_{1},\dots,\ell_{k} linear functionals on the space of d\times d matrices M_{d}(\mathbb{R}) such that (1) \|\nabla u\|_{L^{2}(\mathbb{R}^{d})}^{2}\lesssim\sum_{i=1}^{k}\|\ell_{i}(\nabla u)\|_{L^{2}(\mathbb{R}^{d})}^{2} holds for all u\in C^{\infty}_{c}(\mathbb{R}^{d},\mathbb{R}^{d}) 2 2 2 In [ 22 ] , N is defined for u\in C^{\infty}_{c}(\Omega,\mathbb{R}^{d}) , so a priori N depends on the domain. However, one easily sees that N is independent of the domain, so we skip that here. . Similarly, he defined N^{\prime}(d) as Definition 2 . Let \Omega be an open subset of \mathbb{R}^{d} . Let N^{\prime}(d,\Omega) be the smallest integer k for which there exist \ell_{1},\dots,\ell_{k} linear functionals on M_{d}(\mathbb{R}) such that (2) \|u\|_{H^{1}(\Omega)}^{2}\lesssim\|u\|_{L^{2}(\Omega)}^{2}+\sum_{i=1}^{k}\|\ell_{i}(\nabla u)\|_{L^{2}(\Omega)}^{2} holds for all u\in C^{\infty}(\Omega,\mathbb{R}^{d}) 3 3 3 More care is required in handling \Omega in this case. Indeed, the regularity of \Omega seems to play a more important role, as it is well known that there exist open sets \Omega on which Korn’s second inequality fails (see [ 47 ] ). . As we will see, N^{\prime}(d,\Omega) does not depend on the domain provided the domain is regular enough (say, Lipschitz), so we will continue using N^{\prime}(d) to denote N^{\prime}(d,\Omega) for any bounded Lipschitz domain. Henceforth, we will also assume all domains to be bounded and Lipschitz. Chipot proved many important properties of N(d) and N^{\prime}(d), which we briefly remind the reader: Theorem 1.1 ( [ 22 ] ) . For any d,k\in\mathbb{N} we have 1. (Proposition 2.1) N(d)\leq\frac{d(d-1)}{2}+1 2. (Theorem 2.3) N(d)\geq d and the inequality is strict if d is odd. 3. (Theorem 2.4, Theorem 4.1) N(2)=2,N(3)=4,N(4)=4,N(8)\leq 16 4. (Theorem 3.3) N(kd)\leq k^{2}N(d) 5. (Theorem 5.1) For any bounded Lipschitz domain \Omega\subset\mathbb{R}^{2} , N^{\prime}(2,\Omega)=3. This in particular implies that N^{\prime}(2) is well defined. 6. (Theorem 5.2) For any bounded Lipschitz domain \Omega , N^{\prime}(d,\Omega)\geq N(d). In particular, the natural question of the growth of N(d) remained open: does N(d) grow quadratically, or linearly, or at a different rate? It is not hard to see that one can without loss of generality assume the functionals to be orthogonal to each other, so the question is essentially how many coordinates of \nabla u are needed to define the norm on W^{1,p} . In particular, if N(d)\approx d holds, this implies that, as d grows, only a vanishingly small number of coordinates of \nabla u are really needed to determine a norm, i.e. the standard gradient norm is quite redundant. In this regard, N^{\prime}(2)\neq N(2) is rather surprising, as there seems to be no intrinsic difference between the two spaces. Our results hinge on the following observation (various forms of this claim have been known in the community of calculus of variations for a long time. See remark 3 for background and history of this result) Theorem 1.2 . Given a field \mathbb{K} , let n_{\mathbb{K}}(d,k) denote the maximal dimension a subspace Y of M_{d}(\mathbb{K}) can have without intersecting the set \mathcal{R}_{k} of non-zero matrices of rank at most k . Then (3) N(d)=d^{2}-n_{\mathbb{R}}(d,1) N^{\prime}(d)=d^{2}-n_{\mathbb{C}}(d,1) This connection with algebra allows us to attack the problem from a very different angle and this will lead us to prove several interesting properties of N(d),N^{\prime}(d), by taking advantage of what is known about n_{\mathbb{R}}(d,1) . We mention here the most important ones. The first, and perhaps most surprising, results we obtain are Corollary 1.2.1 . For each d\in\mathbb{N} larger than one we have N(d)\leq 2d-2. This bound is attained if d=2^{n}+1 for some n\in\mathbb{N} . Moreover, asymptotically we have N(d)\sim 2d , so the bound is asymptotically sharp. Corollary 1.2.2 . For each d\in\mathbb{N} we have N^{\prime}(d)=2d-1. In particular, the inequality N^{\prime}(d)\geq N(d) is always strict when d>1 . We also manage to improve on the best previously known lower bound: Corollary 1.2.3 . Equality N(d)=d is attained if and only if d\in\{1,2,4,8\}. Moreover, we have N(d)\geq 2^{\lceil\log_{2}(d)\rceil}. These corollaries by themselves are enough to calculate

Read the complete text