Homeomorphic Sobolev extensions and integrability of hyperbolic metric

Xilin Zhou

REVIEW 4 major objections 6 minor 15 references

Homeomorphic Sobolev extensions and integrability of hyperbolic metric

T0 review · 4 major / 6 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read This paper proves that a Jordan domain's φ-integrable hyperbolic metric yields homeomorphic W^{1,p} extensions for all p<2 precisely when ∫_1^∞ 1/φ(s) ds is finite, and constructs a sharp counterexample when it diverges.

desk verdict A solid phi-scale generalization of the L^q-extension theorem, but the sharpness construction as written has a repairable geometric inconsistency. read the letter →

arxiv 2506.10216 v1 pith:PGQPTKLM submitted 2025-06-11 math.CV

Xilin Zhou This is my paper · ORCID

classification math.CV MSC 46E3530C6258E20

keywords SobolevhomeomorphismextensionhyperbolicmetricJordandomainphi-integrabilitydyadiccrosscutsinternaldiametersharpthreshold

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 when every homeomorphic parametrization of the boundary of a planar Jordan domain can be extended to a Sobolev homeomorphism of the whole disk. Building on the recent L^q-integrability criterion, the author replaces the power q by an arbitrary weight function φ and proves that the extension exists for every boundary homeomorphism exactly when φ satisfies a mild subadditivity and the tail integral of 1/φ is finite. When that tail diverges, the paper constructs a Jordan domain with φ-integrable hyperbolic metric whose boundary has a parametrization admitting no $W^{{1,1}}$ homeomorphic extension, so the threshold is sharp. The result settles the refined-scale question for functions like t(log(e+t))^α: the dividing line is α>1.

What carries the argument

The engine is the Koski–Onninen dyadic crosscut criterion [11, Theorem 3.1] (Theorem 2.2 here): a boundary homeomorphism extends to a $W^{{1,p}}$ homeomorphism if one can exhibit a dyadic family of boundary arcs and, for each arc, a crosscut Γ_{n,j} inside the domain such that the length series (2.1) converges and the crosscuts are pairwise disjoint. The key estimate (Lemma 2.3) bounds the squared Euclidean length of a single crosscut (the image under the Riemann map of a short hyperbolic geodesic) by a constant times (∫_1^∞ 1/φ(s) ds)(∫_Δ φ(h_Ω(z0,·)) dz), where Δ is the region between the crosscut and the boundary arc. The proof moves the two geodesic endpoints to ±1 by a Möbius transformation, decomposes the upper half-plane into the annular regions A_m, uses the lower bound h(ω,T(0)) ≥ |m| log 2, and applies Cauchy–Schwarz; the finiteness of the tail integral makes the series over scales convergent. The counterexample rests on a separate construction: a half-disk glued to trapezoids with heights a_n satisfying ∑ $a_n^{2}$ φ(n) < ∞ while ∑ a_n = ∞, then folded into rectangles of summable widths to produce a Jordan domain of infinite internal diameter.

What would settle it

Plot the hyperbolic geodesics joining consecutive dyadic boundary points $e^{{2π i j/2^n}}$ in the unit disk: if any two geodesics from different dyadic scales meet in the open disk, the pairwise-disjointness condition used in the proof of Theorem 1.4 would fail for the natural family, and the proof would need a different crosscut selection to stand.

Watch

Extended reading notes

Core claim

The central claim is Theorem 1.4: let Ω be a Jordan domain, z0∈Ω, and let φ be continuous, strictly increasing to infinity, with φ(s+t) ≤ M(φ(s)+φ(t)). If ∫_Ω φ(h_Ω(z0,z)) dz < ∞ and ∫_1^∞ ds/φ(s) < ∞, then every homeomorphic parametrization φ:∂D→∂Ω has a homeomorphic extension in $W^{{1,p}}$(D,C) for every p∈[1,2). Theorem 1.5 shows the condition on the tail integral is necessary in a strong sense: when it diverges, there exists a Jordan domain satisfying the φ-integrability condition (for instance with G=D when φ allows) whose boundary has a homeomorphism with no $W^{{1,1}}$ extension. The paper also proves that the hypotheses of the positive theorem force the internal diameter of Ω to be finite, and that infinite internal diameter always prevents some boundary homeomorphism from having a $W^{{1,1}}$ extension (Theorem 1.6).

Load-bearing premise

The positive extension proof assumes that the crosscuts chosen from the dyadic family are pairwise disjoint across all scales, a geometric property that the paper states without proof; if two of these crosscuts crossed in the interior, the Koski–Onninen criterion could not be applied to that family.

Editorial extensions

If this is right

The L^q-integrability theorem of [4] for q>1 is a special case, and the proof extends to critical scales like φ(t)=t(log(e+t))^α with α>1, yielding W^{1,p} homeomorphic extensions for all p<2.
The counterexample shows that for α≤1, there are Jordan domains with φ-integrable hyperbolic metric but whose boundary homeomorphisms are not all W^{1,1}-extendable; in particular, the threshold α>1 is sharp.
Under the hypotheses of Theorem 1.4, the internal diameter of Ω is finite (Lemma 2.6), so any Jordan domain with infinite internal diameter is automatically outside the extension class.
The sufficient condition is independent of the chosen base point z0, because the hyperbolic metric changes by a controlled factor under a change of base point, and the integral condition remains finite.
The theorem gives a parameter-free criterion: the extension holds for any φ satisfying the mild quasi-subadditivity, not just power functions.

Reading between the lines

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

The tail integral ∫_1^∞ 1/φ(s) ds appears as a quantitative weight: the constant in the length estimate (Lemma 2.3) is proportional to the square root of this tail, so one can expect explicit bounds on the W^{1,p} norm of the extension in terms of this tail and the total φ-mass, which the paper does not state.
The folding construction in Theorem 1.5 creates domains with infinite internal diameter; a similar folding with different scalings might produce domains where W^{1,p} extension exists exactly for p < p0 for some p0∈[1,2), filling the gap between the W^{1,1} failure and the all-p<2 theorem.
The proof's disjointness condition could be automated: for a given Jordan domain, checking pairwise disjointness of the dyadic geodesic images might be decidable from the conformal map, and if it fails for one dyadic system, one could try other dyadic schemes before concluding non-extendability.

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

Xilin Zhou has a decent paper here. Theorem 1.4 extends the L^q-integrability result of Bouchala et al. to a sharp phi-scale: if the hyperbolic metric is phi-integrable and the tail of 1/phi converges, then every boundary parametrization has a W^{1,p} homeomorphic extension for p<2. That is a real generalization, and the proof adapts the Koski-Onninen dyadic crosscut machinery competently. Lemma 2.3 is the heart, and the Cauchy-Schwarz summation is clean. The sharpness idea—using a domain of infinite internal diameter—is natural, and Theorem 1.5 is a plausible claim.

The soft spots are mostly fixable, except one that is not. In Theorem 1.6, the arcs A_n are declared to have length pi/(4n). Their total length diverges, so they cannot be pairwise disjoint on the unit circle. In fact A_4 already reaches pi and A_5 starts before pi and crosses it. Because the intervals I(theta_n, delta_n) are pairwise disjoint, injectivity of phi forces the A_n to be disjoint; they are not, so phi is not well-defined. That is a genuine flaw in the counterexample as written. It is easy to repair—take the lengths to be summable but still with |A_n| 4^n non-summable, e.g., |A_n| = c 4^{-n}—but the current text does not do it.

Two smaller gaps: the proof invokes Theorem 2.2 without checking that the crosscuts Gamma_{n,j} are pairwise disjoint across different dyadic scales. This is probably true for hyperbolic geodesics between boundary points coming from a homeomorphic image of a nested dyadic set, but it should be stated or proved. And Theorem 1.5 borrows [4, Lemma 3.2] without stating the lemma or verifying its hypotheses, so the final integrability transfer is not independently checkable.

The citation pattern is fine: the author cites the prior paper he coauthored because that is the result being generalized.

Bottom line: the main theorem is likely correct and worth publishing. The counterexample needs repair before the sharpness claim is established. I would send this to a serious referee and ask for those specific revisions.

Referee Report

4 major / 6 minor

Summary. The paper studies homeomorphic Sobolev extensions of boundary parametrizations of planar Jordan domains under a general integrability condition on the hyperbolic metric. Theorem 1.4 asserts that if ∫_Ω φ(h_Ω(z0,z)) dz < ∞ with φ satisfying Assumption 1.3 and ∫_1^∞ 1/φ(s) ds < ∞, then every boundary parametrization extends as a homeomorphism in W^{1,p}(D,C) for all p∈[1,2). Theorem 1.5 claims sharpness: when the tail integral diverges, a Jordan domain can be built satisfying the φ-integrability condition but admitting a boundary parametrization with no homeomorphic W^{1,1} extension. The sufficiency proof follows the Koski–Onninen crosscut method and generalizes a lemma from Bouchala et al. The sharpness proof constructs a domain of infinite internal diameter and invokes a new Theorem 1.6. The main technical problems are that the crosscut disjointness required by Theorem 2.2 is never verified, and the counterexample construction in Theorem 1.6 uses boundary arcs whose total length diverges, making them unable to be pairwise disjoint; this invalidates the sharpness proof as written.

Significance. If the gaps are repaired, the sufficiency theorem would be a natural and valuable generalization of the L^q-integrability result of Bouchala et al. [4] to a Orlicz-type scale, and the proposed sharpness condition (1.4) is the right critical exponent in the model case φ(t)=t(log(e+t))^α. The paper uses published tools rather than circular reasoning; the strategy of Lemma 2.3 follows the established method of [4, Lemma 2.4], and the counterexample is explicit and potentially checkable. However, in its present form the sharpness claim is not established because the central counterexample is internally inconsistent, and the sufficiency proof omits a load-bearing geometric verification. The paper is likely repairable, but the current version cannot be accepted.

major comments (4)

[Section 3, proof of Theorem 1.6] The arcs A_n = {e^{i(α+β)} : α=π−π/2^n, β∈[0,π/(4n)]} have length π/(4n), so ∑_{n≥1} |A_n| = ∞. Since each g_n maps A_n homeomorphically onto a pairwise disjoint interval I(θ_n,δ_n), injectivity of the boundary parametrization φ forces the A_n to be pairwise disjoint. This is impossible: for instance A_4 ends at angle π and A_5 begins at angle π−π/32 < π, so A_4∩A_5 is nonempty. Consequently φ is not well-defined and the lower-bound argument for the W^{1,1} energy collapses. Because Theorem 1.5 is deduced from Theorem 1.6, the sharpness claim is not proved. The construction is repairable by taking summable lengths such as π/4^n, but as written the counterexample is invalid.
[Section 2, proof of Theorem 1.4] Theorem 2.2 is applied to the family of crosscuts Γ_{n,j}=f(γ_{n,j}) for all n≥n_0 and j=1,…,2^n, but condition 2 of Theorem 2.2 — pairwise disjointness of all crosscuts apart from endpoints — is never verified. Lemma 2.6 proves disjointness only for the geodesics belonging to a single finite cycle {x_0,…,x_k}. For dyadic intervals at different dyadic scales, disjointness is not automatic from the construction and requires an argument, for example showing that a child geodesic lies in the component of D cut off by its parent geodesic. Without this verification, the series estimate (2.1) cannot be applied to the permitted family, so the proof of Theorem 1.4 is incomplete at a load-bearing point.
[Section 3, proof of Theorem 1.6, first paragraph] The proof invokes 'the internal geodesic γ⊂Ω connecting f(0) and f(eω)' in the case d_I(f(0),f(eω))=∞. Internal geodesics are defined only for finite internal distance, so this requires a limiting argument using approximate geodesics between points at finite distance. The subsequent construction of the sequence {x_n} and the inequalities (3.2)–(3.3) depend on this object. This is repairable, but as written it is a gap in the derivation of (3.1).
[Section 3, proof of Theorem 1.5] The folding construction that converts the unbounded domain R into a bounded Jordan domain Ω is described only informally: the text refers to 'twisting parts', 'pipes', and figures, and states choices such as m_d and s_d with 'It is easy to check' rather than a rigorous verification. In addition, the final conclusion ∫_Ω φ(h_Ω(z0,z)) dz < ∞ is justified by citing [4, Lemma 3.2] without stating the lemma or verifying its hypotheses after the folding. Since Theorem 1.5 depends on this construction, the sharpness example is not fully checkable as written.

minor comments (6)

[Throughout] There are several typos: 'Thenrem 1.4' in the proof of Theorem 1.4, 'Cauchy–Schwartz' should be 'Cauchy–Schwarz', 'homeomorpic' in the proof of Theorem 1.5, and 'line segement' in the proof of Theorem 1.6.
[Section 2, definition of internal distance] The internal distance d_I is defined only for points in Ω, but Lemma 2.6 and Theorem 1.6 use expressions such as d_I(f(0),f(ω)) for boundary points ω∈∂D and diam_I(Ω)=sup_{x,y∈∂Ω}d_I(x,y); the limiting definition for boundary points should be stated explicitly.
[Lemma 2.3] The sets A_m are initially defined as subsets of the unit disk with polar coordinates, but later they are used as subsets of H+ in expressions such as h_{H+}(ω,T(0)) and g(A_m); the change of variables under T should be made explicit.
[Lemma 3.3] The summation-by-parts formula in part (ii) contains index inconsistencies (k vs j) and appears to drop a factor; a standard integral-comparison proof would be simpler and clearer.
[Section 3, proof of Theorem 1.5] The definition of i_n and the sentence 'We assume, for convenience, that the sequence {i_n} is strictly increasing' need justification; the display (3.11) also has an indexing typo (the lower limit should be i_n+1 rather than i_{n+1}).
[Section 3, proof of Theorem 1.6] In the radial-energy estimate, the line 'the equivalence constant relies on ϵ' is vague, and the argument should explicitly state that Φ(ηe^{it}) lies in D(Φ(0),ϵ) by the choice of η, which is needed to compare the radial image length with the internal distance from the boundary of D(Φ(0),ϵ).

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: main theorem is derived from an external extension criterion; self-citations to [4] are independently argued tools, not definitional inputs.

full rationale

The paper's central derivation, Theorem 1.4, does not reduce to its own inputs by construction. The hypothesis (1.3) and the tail condition (1.4) enter through Lemma 2.3, which bounds the length of a single crosscut by a product of the tail integral and the phi-integral over a region. The Sobolev-extension conclusion is then obtained by summing the resulting dyadic series and applying the external Koski-Onninen criterion, Theorem 2.2. No parameter is fitted to the target class W^{1,p}; instead the series converges for every p in [1,2) because of a geometric factor, so the extension statement is not statistically or definitionally forced by the hypothesis. Lemma 2.3 explicitly generalizes [4, Lemma 2.4], and the author is a coauthor of [4], but the cited estimates are published results with independent proofs and are not the theorem being established. Theorem 1.5 uses the assumed existence of G with finite phi-integrability to obtain the auxiliary integrability (3.6) via Lemma 3.2, then constructs a Jordan domain with infinite internal diameter and uses Theorem 1.6 to obtain non-extendability; the final transfer from quasi-hyperbolic to hyperbolic integrability cites [4, Lemma 3.2], an external lemma, rather than restating the conclusion. I found no equation in the paper that is, by construction, identical to the quantity it purports to establish, nor any fitted input renamed as a prediction. The proof may have geometric gaps, such as the unverified pairwise disjointness of dyadic crosscuts when applying Theorem 2.2 and the apparent overlapping of the arcs A_n in Theorem 1.6, but these are correctness issues, not circularity, and do not raise the circularity score.

Assumptions & free parameters 0 free parameters · 6 assumptions · 0 invented entities

The central claim rests on standard conformal mapping, the Gehring-Hayman comparability, and the Koski-Onninen dyadic criterion. Assumption 1.3 and the existence of a test domain G in Theorem 1.5 are hypotheses, not hidden inputs. There are no free parameters fitted to data and no invented entities.

assumptions (6)

standard math Riemann mapping theorem and Caratheodory's theorem provide a conformal homeomorphism f:D→Ω extending to the boundary.
Used throughout Sections 2 and 3 to transfer hyperbolic geometry from D to Ω.
standard math Gehring-Hayman theorem: internal distance in a Jordan domain is comparable to length of hyperbolic geodesics.
Used in Lemma 2.6 to bound internal diameter by sums of geodesic lengths.
standard math Koski-Onninen dyadic criterion, Theorem 2.2, is taken as a black box.
The extension theorem in Theorem 1.4 is obtained by verifying its hypothesis.
domain assumption Assumption 1.3 on phi: continuous, strictly increasing, tends to infinity, quasi-subadditive.
This is the class of functions for which the main theorem is stated; all examples must lie in it.
domain assumption Theorem 1.5 assumes existence of a simply connected domain G with ∫_G phi(h_G(g0,z)) dz finite.
This is the non-vacuity condition used to build the counterexample.
standard math Quasi-hyperbolic and hyperbolic metrics are comparable in simply connected planar domains.
Used in Theorem 1.5 to estimate ∫ phi(k_R) and transfer to h_Ω.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Homeomorphic Sobolev extensions and integrability of hyperbolic metric." pith.science (2026). https://pith.science/paper/PGQPTKLM

@misc{pith2026250610216,
  author       = {Pith},
  title        = {Pith review of: Homeomorphic Sobolev extensions and integrability of hyperbolic metric},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/PGQPTKLM}},
  note         = {Machine review of arXiv:2506.10216}
}

abstract

Very recently, it was proved that if the hyperbolic metric of a planar Jordan domain is $L^q$-integrable for some $q\in (1,\infty)$, then every homeomorphic parametrization of the boundary Jordan curve via the unit circle can be extended to a Sobolev homeomorphism of the entire disk. This naturally raises the question of whether the extension holds under more general integrability conditions on the hyperbolic metric. In this work, we examine the case where the hyperbolic metric is $\phi$-integrable. Under appropriate conditions on the function $\phi$, we establish the existence of a Sobolev homeomorphic extension for every homeomorphic parametrization of the Jordan curve. Moreover, we demonstrate the sharpness of our result by providing an explicit counterexample.

Figures

Figures reproduced from arXiv: 2506.10216 by the authors.

**Figure 1.** Transformation between domains We set g := f ◦ T −1 . By [4, (2.4), (2.6), and the inequality right after (2.6)], the following results hold: For all m ∈ Z \ {0} and ω ∈ Am, hH+ (ω, T(0)) ≥ |m| log 2, (2.2) ℓ(g(Cm)) ≤ c1 2 |m| |g ′ (zm)|, (2.3) |g(Am)| ≥ c2 2 2|m| |g ′ (zm)| 2 , (2.4) where c1 and c2 are positive constants independent of m. Then, by (2.2), (2.4), along with the invariance of the hyperbolic metric un… view at source ↗

**Figure 2.** Construction of {ηn} and {I(θn, δn)} Let ω0 be defined as in (3.1). Without loss of generality, we assume that ω0 = e iπ. We construct the sequence of intervals {I(θn, δn)}n∈N+ satisfying the following properties (1) The intervals {I(θn, δn)}n∈N+ are pairwise disjoint, and ω0 ∈/ I(θn, δn); (2) For all n ∈ N +, θn = e itn where 0 < t1 < t2 < . . . < π, and limn→∞ tn = π; (3) For all n ∈ N +, it holds that inf ω∈I(θn,… view at source ↗

**Figure 3.** The domain R Set rn(s) := cM an − s(an−an+1) an for all s ∈ [0, an]. Recall that kR denotes the quasi-hyperbolic metric of the domain R. By (3.8), it holds that kR(0,(s + bn, t)) ≤ kR(0,(s + bn, 0)) + kR((s + bn, 0),(s + bn, t)) ≲ Xn k=1 ak cMak+1 + ˆ t 0 1 rn(s) − r dr ≲ n + log rn(s) rn(s) − t , where the equivalent constants do not rely on n, s and t. Then, by Lemma 3.1 and Lemma 3.2, we obtain ˆ Rn ϕ(kR(… view at source ↗

Figures from the paper (2 more)

**Figure 4.** Figure 4: The folding domain 2cMam1 2cMam2 Rm2−2 Rm2−1 · · · Rm2Rm2+1 Rm2+s1−1 Rm2+s1 Rm2+s1+1 2cMam2 [PITH_FULL_IMAGE:figures/full_fig_p015_4.png]

**Figure 5.** Figure 5: The twisting part, and the choice of s1 Since the sequence {an} is decreasing, and by (3.14) and the construction process, we have wn ≲ cM X Kn d=1 amd ≲ am1 + K Xn−1 d=1 2ln md+1 − md . By (3.14) and the Cauchy–Schwartz inequality, we conclude that (ln) 2 4 ≤ mdX+1−1 …

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

15 extracted references · 15 canonical work pages

[4]

Bouchala, J

O. Bouchala, J. J¨ a¨ askel¨ ainen, P. Koskela, H. Xu, X. Zhou, Homeomorphic Sobolev extensions of parametrizations of Jordan curves, J. Funct. Anal. 228 (4) (2025) 110721, 22pp

work page 2025
[1]

S. S. Antman, Nonlinear problems of elasticity, Applied Mathematical Sciences, vol. 107, Springer-Verlag, New York, 1995

work page 1995
[2]

Astala, T

K. Astala, T. Iwaniec, G. Martin, Elliptic Partial Differential Equations and Quasiconformal Mappings in the Plane, Princeton University Press, 2009

work page 2009
[3]

J. M. Ball, Convexity conditions and existence theorems in nonlinear elasticity, Arch. Rational Mech. Anal. 63 (1976/77), no. 4, 337-403

work page 1976
[5]

P. G. Ciarlet, Mathematical elasticity Vol. I. Three-dimensional elasticity, Studies in Mathe- matics and its Applications, vol. 20. North-Holland Publishing Co., Amsterdam, 1988

work page 1988
[6]

Hencl, P

S. Hencl, P. Koskela, Lectures on mappings of finite distortion. Lecture Notes in Mathematics,

work page
[7]

F. W. Gehring, W. K. Hayman, An inequality in the theory of conformal mapping. J. Math. Pure. Appl. 41(9) (1962), 353-361

work page 1962
[8]

Iwaniec, G

T. Iwaniec, G. Martin, Geometric Function Theory and Non-linear Analysis, Oxford Mathe- matical Monographs, Oxford University Press, 2001

work page 2001

Show all 15 references

[9]

Koskela, A

P. Koskela, A. Koski, J. Onninen, Sobolev homeomorphic extensions onto John domains, J. Funct. Anal. 279 (2020), no. 10, 108719, 17 pp

2020
[10]

Koski, J

A. Koski, J. Onninen, Sobolev homeomorphic extensions, J. Eur. Math. Soc. 23 (2021), no. 12, 4065-4089

2021
[11]

Koski, J

A. Koski, J. Onninen, The Sobolev Jordan–Sch¨ onflies problem, Adv. Math. 413 (2023) 108795

2023
[12]

Yu. G. Reshetnyak, Space mappings with bounded distortion, American Mathematical Soci- ety, Providence, RI, 1989

1989
[13]

G. C. Verchota, Harmonic homeomorphisms of the closed disc to itself need be inW 1,p, p <2, but notW 1,2, Proc. Amer. Math. Soc. (2007), vol. 135, no. 3, 891–894

2007
[14]

Y. R.-Y. Zhang, Schoenflies solutions with conformal boundary values may fail to be Sobolev, Ann. Acad. Sci. Fenn. Ser. A I Math. (2019), vol. 44, no. 2, 791-796. X. Zhou, Department of Mathematics and Statistics, University of Jyv¨askyl¨a, P.O. Box 35 (MaD), FI-40014, Finland...

2019
[2096]

Springer, Cham, 2014

2014

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.

[1] [4]

Bouchala, J

O. Bouchala, J. J¨ a¨ askel¨ ainen, P. Koskela, H. Xu, X. Zhou, Homeomorphic Sobolev extensions of parametrizations of Jordan curves, J. Funct. Anal. 228 (4) (2025) 110721, 22pp

work page 2025

[2] [1]

S. S. Antman, Nonlinear problems of elasticity, Applied Mathematical Sciences, vol. 107, Springer-Verlag, New York, 1995

work page 1995

[3] [2]

Astala, T

K. Astala, T. Iwaniec, G. Martin, Elliptic Partial Differential Equations and Quasiconformal Mappings in the Plane, Princeton University Press, 2009

work page 2009

[4] [3]

J. M. Ball, Convexity conditions and existence theorems in nonlinear elasticity, Arch. Rational Mech. Anal. 63 (1976/77), no. 4, 337-403

work page 1976

[5] [5]

P. G. Ciarlet, Mathematical elasticity Vol. I. Three-dimensional elasticity, Studies in Mathe- matics and its Applications, vol. 20. North-Holland Publishing Co., Amsterdam, 1988

work page 1988

[6] [6]

Hencl, P

S. Hencl, P. Koskela, Lectures on mappings of finite distortion. Lecture Notes in Mathematics,

work page

[7] [7]

F. W. Gehring, W. K. Hayman, An inequality in the theory of conformal mapping. J. Math. Pure. Appl. 41(9) (1962), 353-361

work page 1962

[8] [8]

Iwaniec, G

T. Iwaniec, G. Martin, Geometric Function Theory and Non-linear Analysis, Oxford Mathe- matical Monographs, Oxford University Press, 2001

work page 2001