Constant rank operators in Korn-Maxwell-Sobolev inequalities

Paul Stephan; Peter Lewintan

REVIEW 2 major objections 4 minor 29 references

Constant rank operators in Korn-Maxwell-Sobolev inequalities

T0 review · 2 major / 4 minor · reviewed 2026-08-11 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read For operators of reduced constant rank, the paper proves a Korn-Maxwell-Sobolev inequality whose left-hand side subtracts the projection Π_B Π_{ker A} P, and establishes the borderline L^1 case under a reduced cancellation condition.

desk verdict The reduced-constant-rank idea is genuinely new and the p=1 argument is neat, but Theorem 2.5 is false as stated because Π_B is not defined when B is only reduced constant rank; the stress-test counterexample holds. read the letter →

arxiv 2412.14866 v1 pith:IDSOXDCI submitted 2024-12-19 math.AP

Peter Lewintan , Paul Stephan This is my paper · ORCID

classification math.AP MSC 35A2326D1035Q7435Q7546E35

keywords Korn-Maxwell-SobolevinequalitiesconstantrankoperatorsreducedcancellinglimitingSobolevL1estimatesincompatibletensorfieldsprojection

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper asks which operators can appear on the right-hand side of a Korn-Maxwell-Sobolev inequality when the classical elliptic condition fails. It answers that a weaker algebraic condition — reduced constant rank relative to the part map A — is sufficient, provided the left-hand side is corrected by subtracting the projection Π_B applied to the kernel-of-A projection. The authors prove the inequality for all 1

What carries the argument

The machinery consists of three pieces: (1) the Fourier projection operator Π_A defined by projecting the Fourier transform onto ker A[ξ], which is bounded iff A has constant rank (Lemma 2.1); (2) the algebraic reduced constant rank condition on the restricted symbols B[ξ]|_{ker A}, which is shown in Lemma 3.1 to be equivalent to the projected inequality with a negative-Sobolev norm on BP; (3) for p=1, the existence of an ellipticity complex for constant rank operators — a companion operator L with ker L[ξ] = B[ξ](ker A) — and the strong Bourgain-Brezis estimate for cocancelling operators, which bridge from $L^{1}$ to the negative-Sobolev norm.

What would settle it

Take a candidate pair such as (A,B)=(tr,Curl) and compute the symbol-level multiplier bound: if there exists ξ0 where the symbol of P − Π_B Π_{ker A} P cannot be controlled by |A[ξ]| + |B[ξ]|, a smoothed plane-wave sequence will violate the claimed inequality; this finite-dimensional linear-algebra computation settles the question at a stroke.

Watch

Extended reading notes

Core claim

The central discovery is that the reduced constant rank condition — B[ξ] restricted to ker A having constant rank for all nonzero ξ — is sufficient for the projected Korn-Maxwell-Sobolev inequality ‖P − Π_B Π_{ker A} P‖_{$W^{{k-1,p*}}$} ≤ c (‖A[P]‖_{$W^{{k-1,p*}}$} + ‖BP‖_{L^p}) for 1<p<n, and in the limiting case p=1 if additionally the intersection of the image symbols ∩_ξ B[ξ](ker A) is trivial. The proof works by splitting P into its ker A and (ker A)⊥ components, bounding the latter pointwise by A[P], and applying the constant-rank projection estimate of Lemma 2.1 to the former. For p=1, a companion ellipticity complex is used to transfer to a cocancelling operator and apply the strong limiting Sobolev estimate.

Load-bearing premise

The borderline case p=1 rests on a cited theorem that every constant-rank operator admits an ellipticity complex (a companion operator L whose symbol kernels match B[ξ](ker A)); if that theorem were false or inapplicable for some pair A,B, the step bounding BP in the negative Sobolev space would collapse.

Editorial extensions

If this is right

For any linear map A and constant rank operator B with reduced constant rank, the projected inequality holds; when B is elliptic relative to ker A the projection vanishes and the classical reduced-elliptic KMS inequalities are recovered.
The combination (A,B)=(tr,Curl) is admissible: the paper derives an explicit inequality controlling P up to the Helmholtz-type projection of the deviatoric part, with the L^1 norm of Curl P on the right.
Lemma 3.1 gives an equivalence: the reduced constant rank condition is necessary and sufficient for the L^q inequality with negative-Sobolev BP norm, so any future improved inequality must be tight against this condition.
The p=1 borderline works under reduced cancellation, matching the structure known from the elliptic case and identifying the extra algebraic condition that must be checked in applications.
The constant rank framework subsumes the classical elliptic classification as a special case, showing that constant rank is a natural level of generality for projection-based coercivity estimates.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

If the open necessity question in Section 4 is resolved affirmatively, reduced constant rank would be the exact threshold for the projected inequality, making the projection correction the canonical quantifier of non-ellipticity in this setting.
The projection correction Π_B Π_{ker A} may serve as the correct right-hand-side object for error estimates in mixed finite element methods for non-elliptic incompatibilities, where the trace/deviation structure appears in plasticity models.
The reliance on an ellipticity complex for the p=1 step suggests that extending the theorem to operators with nonconstant rank, or to variable coefficients, would need a new mechanism; a constructive alternative to the potential-theoretic existence result would sharpen the argument.
The explicit (tr,Curl) inequality gives a ready test case for numerical experiments in relaxed micromorphic elasticity: one can check whether the Helmholtz-type projection of the deviatoric distortion is the exact null-space correction observed in simulations.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

Quick take: the core idea here is real and worth keeping — swapping reduced ellipticity for reduced constant rank and paying for it with a projection correction is a sensible extension, and the p=1 route through Van Schaftingen's cocancelling estimate is clean. But the main theorem is not correct as written, and the stress-test note is right to call it out.

The problem is that Π_B is never defined for an operator that only has reduced constant rank. The paper defines Π_A only for constant rank operators. Under the reduced constant rank hypothesis, the full symbol B[ξ] may jump rank, and the orthogonal projection onto its kernel is then a discontinuous function of ξ, not a bounded Fourier multiplier. The proof of Lemma 3.1 applies Lemma 2.1 with domain ker A, which yields the projection inside ker A onto ker(B[ξ]|_{ker A}). That projection is not the same as Π_B applied to Π_ker A P. The proof silently identifies the two, and they typically differ whenever ker B[ξ] is not contained in ker A.

The stress-test counterexample is valid. Take n=3, ker A = span{e1,e2}, and B[ξ] = [[0, |ξ|^2, ξ_2^2],[0,0,ξ_1^2]]. Restricted to ker A, this has rank 1 for every nonzero ξ, so the reduced constant rank hypothesis holds. But B itself has rank 2 for ξ_1≠0 and rank 1 for ξ_1=0, so it is not globally constant rank, and Π_B is not a bounded multiplier. For P=(0,φ,0), A[P]=0 and BP is smooth, so the right side is finite. The left side, P − Π_B Π_ker A P, has a Fourier contribution supported on the hyperplane ξ_1=0; its inverse Fourier transform has a component constant in x_1, so it is not in W^{1,p*}(R^3). The inequality fails.

This is a load-bearing flaw, not a cosmetic one. The intended applications, particularly (tr, Curl), do have B with global constant rank, and in that class the theorem is likely true. The fix is either to add the assumption that B has global constant rank, or to change the correction term to the projection inside ker A onto ker(B[ξ]|_{ker A}) and restate the theorem accordingly. The authors also need to say explicitly where Π_B is defined when B lacks global constant rank.

Credit where it is due: the paper is short, the algebra in Lemma 3.1 is clean, and the use of Raita's potential theorem for the p=1 case is a sensible route. The necessity discussion in Section 4 is honest about what remains open. The minor issues are real but small: reference [24] is never cited, and the abstract seems to blur the distinction between [9] and [11].

Verdict: this deserves a serious referee, but not acceptance as is. It needs a major revision to make the theorem true, either by restricting to global constant rank or by redefining the projection. I would not cite it in its current form, but I would look again after a fix.

Referee Report

2 major / 4 minor

Summary. The paper studies Korn-Maxwell-Sobolev (KMS) inequalities for operators B that are of reduced constant rank relative to a linear map A. The main result, Theorem 2.5, asserts that if the restricted symbols B[ξ]|_{ker A} have constant rank, then for 1<p<n the homogeneous Sobolev norm of P - Π_B Π_{ker A} P is controlled by ‖A[P]‖_{W^{k-1,p*}} + ‖BP‖_{L^p}, with a variant for p=1 under an additional reduced cancellation condition. The proof proceeds through Lemma 3.1, which states an equivalence between the reduced constant rank condition and an L^q estimate with a negative Sobolev norm of BP, followed by Sobolev embedding and, for p=1, Van Schaftingen's Bourgain-Brezis estimates.

Significance. If the main theorem were correct, it would extend the KMS inequality framework from the reduced elliptic setting to the reduced constant rank setting, covering the motivating example (A,B)=(tr,Curl). The paper is short, clearly written, and its proof strategy relies on established external results (constant rank estimates, potentials, limiting Sobolev inequalities), all of which are appropriate. However, the central claim is false as stated: the reduced constant rank hypothesis does not make the Fourier projection Π_B a bounded or even well-defined operator on the relevant function spaces, and a concrete counterexample satisfies the hypotheses while violating the conclusion. The advertised examples, such as B=Curl, do have global constant rank, so a repaired statement is plausible, but the current theorem overreaches.

major comments (2)

[Theorem 2.5(i), Lemma 3.1] The main theorem is false as stated because the hypothesis that B[ξ]|_{ker A} has constant rank does not imply that the full symbol family B[ξ] has constant rank, and the projection Π_B is only a well-defined bounded Fourier multiplier in the latter case. Concretely, take n=3, V=R^3, ker A = span{e1,e2}, and let B be the second-order operator with symbol B[ξ] = [[0, |ξ|^2, ξ_2^2],[0, 0, ξ_1^2]]. Then B[ξ]|_{ker A} has rank 1 for all ξ≠0, but B[ξ] has rank 2 when ξ_1≠0 and rank 1 when ξ_1=0. For P=(0,φ,0) with φ∈C_c^∞(R^3), we have A[P]=0 and BP=(-Δφ,0)∈L^p for every p>1, so the right-hand side of the claimed inequality is finite. On the Fourier side, the e3-component of P - Π_B P equals (φ̂(ξ)/2) 1_{ξ_1=0}, whose inverse Fourier transform is δ(x_1) times a Schwartz function of (x_2,x_3); this distribution is not in W^{1,p*}(R^3). Hence the left-hand side is infinite while the right-hand side is finite, contradicting Theorem 2.5(i).
[Lemma 3.1] The proof applies Lemma 2.1 with E=ker A and u=Π_{ker A}P. In that application, the projection produced by Lemma 2.1 is the orthogonal projection onto ker(B[ξ]|_{ker A}) inside ker A (equivalently, onto ker A∩ker B[ξ] within V), not the projection Π_B onto ker B[ξ] in V. These two projections differ whenever Π_B does not map ker A into itself. In the counterexample above, at frequencies with ξ_1=0 the projection Π_B sends vectors in ker A = span{e1,e2} to vectors with a nonzero e3 component, so it does not preserve ker A. The proof silently identifies the two projections and therefore does not establish the stated estimate for the left-hand side with Π_B. The same identification issue affects the necessity direction (a)⇒(b), where testing with P∈C_c^∞(R^n;ker A) gives an estimate involving Π_B rather than the restricted projection.

minor comments (4)

[Proof of Theorem 2.5(i)] The upgrade from the L^q estimate in Lemma 3.1 to the W^{k-1,p*} estimate is only stated, not shown; it requires applying Lemma 3.1 to derivatives ∂^α P for |α|=k-1 and using that A and B commute with constant-coefficient derivatives. This should be spelled out.
[Theorem 2.5] The phrase 'has constant rank for all ξ∈R^n\{0}' is ambiguous; it should say that there exists r such that rank B[ξ]|_{ker A}=r for every ξ≠0.
[Example 2.6] In the displayed formula for Π_BΠ_{ker A}P with B=Curl, a brief derivation of the row-wise Biot-Savart formula would help the reader; as written, the passage from the abstract projection to the convolution kernel is implicit.
[References] Reference [24] is a Master's thesis; if journal policy allows, this is acceptable, but the authors may wish to cite a published version or a more standard reference for the same material.

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity: the derivation reduces to external constant-rank and Bourgain–Brezis theorems, with only minor self-citations for technique.

full rationale

The main inequality in Theorem 2.5 is derived from Lemma 3.1, whose sufficiency half applies Lemma 2.1 (cited to Fonseca–Müller [7], Schulenberger–Wilcox [23], and Guerra–Raită [14]) to the restricted operator B|_{\ker A}. Lemma 2.1 is a published external characterization of constant rank operators and does not itself contain the reduced-constant-rank inequality or the correction projection Π_BΠ_{\ker A}; it is an input, not the target. The p>1 case then uses only standard Sobolev embedding. The borderline p=1 case invokes Raită's potential theorem [22, Thm 1] to obtain L with ker L[ξ] = B[ξ](ker A) and Van Schaftingen's strong Bourgain–Brezis estimate [25, Thm 9.2]; both are published external results whose hypotheses are checked from the reduced constant rank and reduced cancellation assumptions. The paper's own earlier works [9,10,11,15] are cited for terminology, background, and proof strategy; the phrase 'we can proceed like in [11]' indicates imitation of a method, not a black-box citation of the theorem being proved. No fitted parameter is relabeled as a prediction, and no algebraic condition is defined in terms of the inequality it is used to prove. The paper even flags in Section 4 that the necessity counterpart of (4.1) remains open, which is an honest incompleteness rather than a circularity. Any concern that Π_B is not well-defined under mere reduced constant rank is a mathematical correctness risk, not a circular dependency; for that reason the score is low.

Assumptions & free parameters 0 free parameters · 4 assumptions · 0 invented entities

The paper introduces no free parameters and no invented entities. Its central claim rests on three external theorems: the constant rank characterization, Ratiu's potential theorem, and Van Schaftingen's Bourgain-Brezis estimate, all standard published results. The algebraic hypotheses, reduced constant rank and reduced cancelling, are stated as conditions rather than fitted quantities. The correction operator Pi_B Pi_{ker A} P is a defined Fourier multiplier, not a hand-chosen entity; it arises naturally from the constant rank structure.

assumptions (4)

standard math Constant rank characterization (Lemma 2.1): a k-th order homogeneous constant coefficient operator A has constant rank if and only if the L^q norm of u minus its kernel projection is controlled by the negative Sobolev norm of Au.
This is the backbone of Lemma 3.1 and Theorem 2.5, cited to [7,14,23]. The paper does not reprove it.
standard math Existence of an ellipticity complex / potential: for a constant rank operator B on ker A there exists a homogeneous constant coefficient operator L such that ker L[xi] = B[xi](ker A) for all nonzero xi.
Invoked in the proof of Theorem 2.5(ii) via [22, Thm 1]; it is a deep published result needed for the p=1 case.
standard math Van Schaftingen's strong Bourgain-Brezis estimate for cocancelling operators: the negative Sobolev norm of f is controlled by the negative Sobolev norm of Lf plus the L^1 norm of f.
Used in Theorem 2.5(ii) and cited to [25, Thm 9.2]; essential for the limiting p=1 argument.
standard math Sobolev embedding L^p embeds into the negative Sobolev space W^{-1,p*} for 1 < p < n.
Used to pass from Lemma 3.1 to Theorem 2.5(i); a classical embedding.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Constant rank operators in Korn-Maxwell-Sobolev inequalities." pith.science (2026). https://pith.science/paper/IDSOXDCI

@misc{pith2026241214866,
  author       = {Pith},
  title        = {Pith review of: Constant rank operators in Korn-Maxwell-Sobolev inequalities},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/IDSOXDCI}},
  note         = {Machine review of arXiv:2412.14866}
}

abstract

We focus on Korn-Maxwell-Sobolev inequalities for operators of reduced constant rank. These inequalities take the form \[ \|P - \Pi_{\mathbb{B}} \Pi_{\ker\mathscr{A}} P\|_{\dot{\mathrm{W}}^{k-1, p^*}(\mathbb{R}^n)} \le c \, (\|\mathscr{A}[P]\|_{\dot{\mathrm{W}}^{k-1, p^*}(\mathbb{R}^n)} + \|\mathbb{B} P\|_{\mathrm{L}^p(\mathbb{R}^n)}) \] for all $ P \in \mathrm{C}_c^\infty(\mathbb{R}^n; V) $, where $ V $ is a finite-dimensional vector space, $ \mathscr{A} $ is a linear mapping, and $ \mathbb{B} $ is a constant coefficient homogeneous differential operator of order $ k $. In particular, we can treat the combination $(p,\mathscr{A},\mathbb{B},k)=(1,\operatorname{tr},\operatorname{Curl},1)$. Our results generalize the techniques from Gmeineder et al. (Math.Mod.Met.Appl.Sci,34:03,2024; arXiv:2405.10349), which exclusively dealt with reduced elliptic operators. In contrast to the reduced ellipticity case, however, the reduced constant rank case necessitates to introduce a correction, namely the projection $\Pi_\mathbb{B}$ on the left-hand side of the inequality.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

29 extracted references · 28 canonical work pages

[24]

Neﬀ, P.; Pauly, D.; Witsch, K.-J.: Poincar´e meets Korn via Maxwell: Extending Korn’s ﬁrst inequality t o incompatible tensor ﬁelds. J. Diﬀerential Equations 258(4) (2015), pp. 1267–1302

work page 2015
[9]

Conti, S.; Garroni, A.: Sharp rigidity estimates for incompatible ﬁelds as consequ ence of the Bourgain Brezis div-curl result. C. R. Math. Acad. Sci. Paris 359(2) (2021), pp. 155–160

work page 2021
[11]

Fonseca, I.; M ¨uller, S.: A-quasiconvexity, lower semicontinuity, and Young measures. SIAM J. Math. Anal. 30 (6) (1999), pp. 1355–1390

work page 1999
[1]

(1.1) For further relaxation of the models, one might ask for a be/t_ter un derstanding of the possible part maps that may appear on the right-hand side of such inequalitie s

I/n.sc/t.sc/r.sc/o.sc/d.sc/u.sc/c.sc/t.sc/i.sc/o.sc/n.sc Korn’s inequalities provide the coercivity of the bilinear for ms appearing in linear elasticity or ﬂuid mechanics and in their most prominent version it reads (f or n ≥ 2): ‖Du‖L2(Rn) ≤ c ‖sym Du‖L2 (Rn) ∀ u ∈ C∞ c (Rn; Rn). (1.1) For further relaxation of the models, one might ask for a be/t_ter u...

work page 2024
[2]

/T_his brings us to more general diﬀerential operators, namely constant rank operators

C/o.sc/n.sc/s.sc/t.sc/a.sc/n.sc/t.sc /r.sc/a.sc/n.sc/k.sc /o.sc/p.sc/e.sc/r.sc/a.sc/t.sc/o.sc/r.sc/s.sc /T_he motivation of the present note is to understand in which sense the la/t_ter combination(A , B) = (tr, Curl) can appear on the right-hand side of ( KMS-I), see Example 2.6 below. /T_his brings us to more general diﬀerential operators, namely consta...

work page
[3]

(b) ⇒ (a)

P/r.sc/o.sc/o.sc/f.sc /o.sc/f.sc /t.sc/h.sc/e.sc /m.sc/a.sc/i.sc/n.sc /t.sc/h.sc/e.sc/o.sc/r.sc/e.sc/m.sc /T_he proof of the main /T_heorem 2.5 follows from the next lemma: Lemma 3.1. Let n ≥ 2, 1 < q < ∞, k ∈ N, V ,~V and W be ﬁnite dimensional spaces, A : V → ~V be linear and B be a k-th order linear homogeneous constant coeﬃcients diﬀerent ial operator...

work page
[4]

However, for a sharp res ult, the necessity is also obligatory

N/e.sc/c.sc/e.sc/s.sc/s.sc/i.sc/t.sc/y.sc In terms of applications, the suﬃciency of our results is already avaluable tool to check the validity of the desired coercivity estimates. However, for a sharp res ult, the necessity is also obligatory. /T_hus, in Lemma3.1 we establish the equivalence between the reduced constant rank condition and the validity o...

work page
[5]

Arnold, D.N.; Douglas, J.; Gupta, C.P.: A family of higher order mixed ﬁnite element methods for plane elasticity. Numer. Math. 45 (1984), pp. 1–22

work page 1984

Show all 29 references

[6]

ESAIM Control Optim

Bauer, S., Neﬀ, P., Pauly, D., Starke, G.: Dev-Div- and DevSym-DevCurl-inequalities for incompatible square tensor ﬁelds with mixed boundary conditions. ESAIM Control Optim. Calc. Var. 22(1) (2016), pp. 112–133

2016
[7]

Calder ´on, A.P.; Zygmund, A.: On singular integrals. Amer. J. Math., 78:289–309, 1956

1956
[8]

Conti, S.; Faraco, D.; Maggi, F.: A new approach to counterexamples to L1 estimates: Korn’s inequality, geometric rigidity, and regularity for gradients of separately conve x functions. Arch. Ration. Mech. Anal. 175 (2005), no. 2, pp. 287–300

2005
[10]

Sharp trace and Korn inequalities for diﬀerential operator s

Diening, L.; Gmeineder, F. Sharp trace and Korn inequalities for diﬀerential operator s. Potential Analysis (2024), pp. 1–54

2024
[12]

Garroni, A.; Leoni, G.; Ponsiglione, M.: Gradient theory for plasticity via homogenization of discr ete dislocations. J. Eur. Math. Soc. (JEMS) 12(5) (2010), pp. 1231–1266

2010
[13]

Gmeineder, F.; Lewintan, P.; Neﬀ, P.: Korn-Maxwell-Sobolev inequalities for general incompatibilities. Math. Mod. Meth. Appl. Sci. (M 3AS) 34(03) (2024), pp. 523–570

2024
[14]

Gmeineder, F.; Lewintan, P.; Neﬀ, P.: Optimal incompatible Korn-Maxwell-Sobolev inequalities in all dimensions. Calc. Var. PDE 62 (2023), 182

2023
[15]

submi/t_ted, arXiv Preprint 2405.10349

Gmeineder, F.; Lewintan, P.; Van Scha/f_tingen, J.: Limiting Korn-Maxwell-Sobolev inequalities for general i ncompatibil- ities. submi/t_ted, arXiv Preprint 2405.10349

arXiv
[16]

Indiana Univ

Gmeineder, F.; Rait ¸˘a, B.; Van Scha/f_tingen, J.: On limiting trace inequalities for vectorial diﬀerential o perators. Indiana Univ. Math. J. 70(5) (2021), pp. 2133–2176

2021
[17]

Gmeineder, F.; Spector, D.: On Korn-Maxwell-Sobolev inequalities. J. Math. Anal. Appl. 502(1) (2021),125226

2021
[18]

Guerra, A.; Rait ¸˘a, B.: On the necessity of the constant rank condition for Lp estimates. C. R. Math ´ematique 358 (9-10) (2020), pp. 1091–1095

2020
[19]

Lewintan, P.; M ¨uller, S.; Neﬀ, P.: Korn inequalities for incompatible tensor ﬁelds in three sp ace dimensions with confor- mally invariant dislocation energy. Calc. Var. PDE 60 (2021), 150

2021
[20]

Murat, F.: Compacit´e par compensation. Ann. Sc. Norm. Super. Pisa, Cl. Sci., IV. 5 (1978), pp. 489–507

1978
[21]

Neﬀ, P., Ghiba, I.-D., Lazar, M., Madeo, A.: /T_he relaxed linear micromorphic continuum: well-posedness of the static problem and relations to the gauge theory of dislocations. /Q_uart. J. Mech. Appl. Math. 68(1) (2015), pp. 53–84

2015
[22]

Neﬀ, P., Ghiba, I.-D., Madeo, A., Placidi, L., Rosi, G.: A unifying perspective: the relaxed linear micromorphic continuum. Contin. Mech. /T_hermodyn.26(5) (2014), pp. 639–681

2014
[23]

Neﬀ, P.; Pauly, D.; Witsch, K.-J.: On a canonical extension of Korn’s ﬁrst and Poincar ´e’s inequality to H(Curl). J. Math. Sci. (N. Y.) 185(5) (2012), pp. 721–727

2012
[25]

D. S. Ornstein: A non-equality for diﬀerential operators in the L1 norm. Arch. Rational Mech. Anal. 11 (1962), 40–49

1962
[26]

Rait ¸˘a, B.: Potentials for A-quasiconvexity. Calc. Var. PDE 58 (2019), 105

2019
[27]

R.; Wilcox, C

Schulenberger, J. R.; Wilcox, C. H.: Coerciveness inequalities for nonelliptic systems of part ial diﬀerential equations. Ann. Mat. Pura Appl., IV. Ser. 88 (1971), pp. 229–305

1971
[28]

Master /T_hesis, Konstanz, 2023

Stephan, P.: A Unifying KMS-/T_heorem: From Ellipticity to the Constant Ran k Condition. Master /T_hesis, Konstanz, 2023

2023
[29]

Van Scha/f_tingen, J.: Limiting Sobolev inequalities for vector ﬁelds and canceli ng linear diﬀerential operators. J. Eur. Math. Soc. (JEMS) 15(3) (2013), pp. 877–921. (Peter Lewintan) I/n.sc/s.sc/t.sc/i.sc/t.sc/u.sc/t.sc/e.sc /f.sc/o.sc/r.sc A/n.sc/a.sc/l.sc/y.sc/s.sc/i.sc/s....

2013

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.

[1] [24]

Neﬀ, P.; Pauly, D.; Witsch, K.-J.: Poincar´e meets Korn via Maxwell: Extending Korn’s ﬁrst inequality t o incompatible tensor ﬁelds. J. Diﬀerential Equations 258(4) (2015), pp. 1267–1302

work page 2015

[2] [9]

Conti, S.; Garroni, A.: Sharp rigidity estimates for incompatible ﬁelds as consequ ence of the Bourgain Brezis div-curl result. C. R. Math. Acad. Sci. Paris 359(2) (2021), pp. 155–160

work page 2021

[3] [11]

Fonseca, I.; M ¨uller, S.: A-quasiconvexity, lower semicontinuity, and Young measures. SIAM J. Math. Anal. 30 (6) (1999), pp. 1355–1390

work page 1999

[4] [1]

(1.1) For further relaxation of the models, one might ask for a be/t_ter un derstanding of the possible part maps that may appear on the right-hand side of such inequalitie s

I/n.sc/t.sc/r.sc/o.sc/d.sc/u.sc/c.sc/t.sc/i.sc/o.sc/n.sc Korn’s inequalities provide the coercivity of the bilinear for ms appearing in linear elasticity or ﬂuid mechanics and in their most prominent version it reads (f or n ≥ 2): ‖Du‖L2(Rn) ≤ c ‖sym Du‖L2 (Rn) ∀ u ∈ C∞ c (Rn; Rn). (1.1) For further relaxation of the models, one might ask for a be/t_ter u...

work page 2024

[5] [2]

/T_his brings us to more general diﬀerential operators, namely constant rank operators

C/o.sc/n.sc/s.sc/t.sc/a.sc/n.sc/t.sc /r.sc/a.sc/n.sc/k.sc /o.sc/p.sc/e.sc/r.sc/a.sc/t.sc/o.sc/r.sc/s.sc /T_he motivation of the present note is to understand in which sense the la/t_ter combination(A , B) = (tr, Curl) can appear on the right-hand side of ( KMS-I), see Example 2.6 below. /T_his brings us to more general diﬀerential operators, namely consta...

work page

[6] [3]

(b) ⇒ (a)

P/r.sc/o.sc/o.sc/f.sc /o.sc/f.sc /t.sc/h.sc/e.sc /m.sc/a.sc/i.sc/n.sc /t.sc/h.sc/e.sc/o.sc/r.sc/e.sc/m.sc /T_he proof of the main /T_heorem 2.5 follows from the next lemma: Lemma 3.1. Let n ≥ 2, 1 < q < ∞, k ∈ N, V ,~V and W be ﬁnite dimensional spaces, A : V → ~V be linear and B be a k-th order linear homogeneous constant coeﬃcients diﬀerent ial operator...

work page

[7] [4]

However, for a sharp res ult, the necessity is also obligatory

N/e.sc/c.sc/e.sc/s.sc/s.sc/i.sc/t.sc/y.sc In terms of applications, the suﬃciency of our results is already avaluable tool to check the validity of the desired coercivity estimates. However, for a sharp res ult, the necessity is also obligatory. /T_hus, in Lemma3.1 we establish the equivalence between the reduced constant rank condition and the validity o...

work page

[8] [5]

Arnold, D.N.; Douglas, J.; Gupta, C.P.: A family of higher order mixed ﬁnite element methods for plane elasticity. Numer. Math. 45 (1984), pp. 1–22

work page 1984