Pith. sign in

REVIEW 3 major objections 3 minor 54 references

The null condition in elastodynamics leads to non-uniqueness

T0 review · 3 major / 3 minor · reviewed 2026-08-08 · deepseek-v4-flash

Pith's one-line read Under a weak null condition, 2D elastodynamics has nontrivial C^1 weak solutions from zero initial data, so weak solutions are not unique.

desk verdict Novel first convex integration for Lagrangian elastodynamics, but the C^1 claim appears to contradict the paper's own admitted error estimates, so the main theorem is not supported as written. read the letter →

arxiv 2502.07521 v2 pith:5TJL2FEH submitted 2025-02-11 math.AP

classification math.AP MSC 35A0235D3035L0535L1535L72
keywords convexintegrationnon-uniquenesselastodynamicsweaknullconditionC^1solutionsCauchyproblemhyperbolicsystems
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

The paper targets the two-dimensional elastodynamic system for hyperelastic materials whose stored energy satisfies a weak null condition. It claims that this system admits nontrivial weak solutions of class $C^1([0,T] imes\mathbb{T}^2)$ starting from zero initial data. If that is right, the Cauchy problem from the zero datum has at least two weak solutions—the zero solution and the constructed one—so weak solutions are not unique. The proof is a convex integration construction: a sequence of approximate solutions is corrected by adding high-frequency waves, with the leftover error pushed to zero. The construction uses the double wave speeds of elasticity to design building blocks made of a longitudinal wave plus a small transverse correction.

What carries the argument

The load-bearing mechanism is a convex integration iteration for the approximate system $\partial_{tt}u-\operatorname{div}((\mathrm{Id}+G)\Sigma_G)=\operatorname{div}((\mathrm{Id}+G)(c+R))$. At each sub-step the approximate solution is perturbed by high-frequency waves built from Lemma 7.1 building blocks: a longitudinal wave of order one along a direction $f_i$ plus a smaller transverse wave, which are approximate solutions of the linearized elastodynamic wave operator precisely because the two wave speeds $\lambda+2\mu$ and $\lambda+\mu$ are distinct. A geometric lemma (Lemma 2.2) represents any symmetric matrix close to the identity as a sum of three projection matrices along the directions $(4,3)$, $(4,-3)$, $(0,5)$, which lets the quadratic part of the perturbation cancel the current Reynolds error. The remaining non-symmetric error terms are absorbed through a new local right-inverse of the asymmetric divergence equation $\operatorname{div}((\mathrm{Id}+\nabla u)\tilde R)=U$, which is solvable when $U$ has zero Lagrangian momentum and angular momentum; that inverse is what keeps the new Reynolds error symmetric and gives the estimate (8.16) that closes the induction.

What would settle it

Compute the first Reynolds-error residual predicted by (8.16) with the paper's schedule $\lambda_q=\lceil \varepsilon^{-b(q+3)^{2/3}}\rceil^3$, $\delta_q=\varepsilon^{2(q+3)/3}$, $b>5$: the induction can start only if the residual after the first correction is at most the claimed multiple of $\delta_2$. If that bound, or its uniform version for all $q$, fails, the approximate solutions need not converge in $C^1$ and the theorem's conclusion is not established.

Watch

Extended reading notes

Core claim

The central claim is Theorem 1.7: for any stored-energy function of the form (1.2) satisfying (1.13) (positive Lam\'e constants, convexity near the origin, and the weak null condition $3\sigma_{11}+2\sigma_{111}=0$, $\sigma_*\neq 0$), one can construct a non-zero weak solution $u\in C^1([0,T]\times\mathbb{T}^2)$ to (1.9) that emanates from zero initial data. The proof runs through Proposition 2.6, an inductive statement that, starting from the zero tuple, produces approximate solutions $(u_q,c_q,R_q)$ with Reynolds error tending to zero, and then shows the sequence converges to a genuine weak solution. The solution is not an artifact of low regularity: it is continuously differentiable, even though the construction is of convex-integration type. The paper reads this as showing that a null condition—usually a mechanism for global existence and uniqueness of small classical solutions—does not prevent non-uniqueness once one works with weak solutions.

Load-bearing premise

The whole induction closes only if the leftover error after each correction is smaller than the next error level for the chosen exponentially growing frequencies and decaying amplitudes, at every step; the authors flag this absorption step as the one that blocks positive-H\"older regularity, so the $C^1$ claim rests entirely on that estimate holding uniformly.

Editorial extensions

If this is right

  • The Cauchy problem for the elastic wave system (1.9) has at least two weak solutions from the same zero data, so weak solutions are not unique under the assumptions (1.13).
  • The non-uniqueness occurs in $C^1$, which is smoother than the $L^\infty$ or H\"older weak solutions typical of convex integration; it shows the non-uniqueness is not an artifact of very low regularity alone.
  • The characteristic double wave speeds $\lambda+2\mu>0$ and $\lambda+\mu\neq 0$ are used to build perturbations as a longitudinal wave plus a small transverse wave; materials violating $\lambda+\mu\neq 0$ fall outside this construction.
  • The iteration produces approximate solutions whose stored energy increases by a controlled amount at each step, so the limiting solution has positive energy despite zero initial data.

Reading between the lines

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

  • If Theorem 1.7 is right, then a null condition—usually a small-data global-existence device—does not restrict the weak-solution set: the weak formulation is far from well-posed even when the nonlinearity has good null structure.
  • The admitted obstruction to positive-H\"older regularity suggests that a different absorption of the non-principal quadratic terms might raise regularity; a natural next step is to try replacing the exponential decay of $\delta_q$ with a polynomial one.
  • The Lagrangian momentum/angular-momentum inverse divergence is transportable: any hyperelastic system whose stress depends symmetrically on the deformation gradient, and for which a geometric lemma of the type (2.2) holds, should admit the same non-uniqueness construction.
  • A testable robustness check is to add a fourth-order term to the stored-energy function (1.2); the paper indicates third-order terms are handled directly, so this would probe whether the error absorption survives higher-order nonlinearities.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 3 minor

Summary. The paper claims to construct nontrivial weak solutions u ∈ C^1([0, T] × T^2) to the 2D elastodynamic system (1.9) emanating from zero initial data, under the weak null condition (1.13). The proof is a 66-page convex integration scheme: an induction produces approximate solutions (u_q, c_q, R_q) solving (2.1), with longitudinal-transverse building blocks (Lemma 7.1) exploiting the double wave speeds, a Lagrangian formulation of momentum and angular momentum constraints (1.17), a local inverse of the asymmetric divergence operator (Prop. 5.10), and an inductive Proposition 2.6 whose key closing estimate is (8.16). Theorem 1.7 asserts convergence of the scheme to a nonzero C^1 weak solution from zero data.

Significance. If the main theorem were correct, this would be a striking and important result: the first convex-integration-based non-uniqueness theorem for elastodynamics in Lagrangian coordinates, extending the method from fluid systems to quasilinear hyperbolic systems with a null condition. The paper contains genuinely new ideas and a large amount of careful technical work: the conserved quantities (1.17), the solution of the asymmetric divergence equation with time-derivative bounds in Proposition 5.10, the building block Lemma 7.1, and the energy-increment control (2.13). These components are creative and internally plausible in isolation. However, the significance of the paper hinges entirely on the quantitative closing of the induction, and that closing is not supported by the stated parameter schedule; the quantitative failure identified below is not a presentation issue but affects the central claim. Credit is due for the construction framework, but the main theorem is not established as written.

major comments (3)
  1. [§8.3, Eq. (8.16); §2, Eqs. (2.3)–(2.4)] The closing estimate (8.16) of Proposition 8.3 is incompatible with the frequency-amplitude schedule (2.3)–(2.4). From (2.4), μ_{q,i} ≈ (λ_{q,i}λ_{q,i+1})^{−1/2}δ_q^{−1/4}, and for each i = 0, 1, 2 one obtains λ_{q,i+1}μ_{q,i} ≈ (λ_{q+1}/λ_q)^{1/6}δ_q^{−1/4}. The second summand in (8.16) therefore behaves as (λ_{q,i+1}μ_{q,i})^{−1}δ_{q+1}^{1/2} ≈ (λ_{q+1}/λ_q)^{−1/6}δ_q^{1/4}δ_{q+1}^{1/2} = ε^{(3q+11)/6}(λ_{q+1}/λ_q)^{−1/6}, while δ_{q+2} = ε^{(4q+20)/6}. Since λ_{q+1}/λ_q ≈ ε^{−2b(q+3)^{−1/3}} → 1 for any fixed b, the quotient of the linear-error term by δ_{q+2} grows like ε^{−(q+9)/6}, which diverges as q → ∞ independently of C_{b,σ,M} and c_0. Hence the asserted bound ≤ (1/128)c_0^4 λ_{q,i+1}^{N+r}δ_{q+2} cannot hold for all large q. This is not a missing computation but an internal inconsistency: the estimate the induction requires is false for the stated schedule. The paper's own admission in §1.3 that the linear wave error 'limits our choice of δ_q' and 'prevents us from achieving the non-uniqueness of weak solutions in C^α spaces with α>0' is exactly this failure. The same calibration problem invalidates (7.77) in the proof of Proposition 7.8, since Mμ_{q,i}λ_{q,i} ≈ M(λ_{q,i}/λ_{q,i+1})^{1/2}δ_q^{−1/4} ≈ Mε^{−(q+3)/6} cannot be bounded for large q by the claimed right-hand side ε^2δ_{q+1}δ_{q,i}^{−1/2} ≈ ε^{(2q+20)/6}. Thus the induction over q cannot be closed as written.
  2. [§3, Eq. (3.21) and Proposition 3.3] The mollification stage also fails under the stated schedule. Proposition 3.3 relies on (3.21), in particular on the inequality 120Mλ_{q,i} ≤ ℓ_{q,i}^{−1}. Using (2.3)–(2.4), λ_{q,i}ℓ_{q,i} = (λ_{q,i}/λ_{q,i+1})^{1/2}δ_q^{−1/4} = (λ_q/λ_{q+1})^{1/6}δ_q^{−1/4}, and since (λ_q/λ_{q+1})^{1/6} ≈ ε^{(b/3)(q+3)^{−1/3}}, we get λ_{q,i}ℓ_{q,i} ≈ ε^{(b/3)(q+3)^{−1/3} − (q+3)/6}. The exponent is negative and grows in magnitude like (q+3)/6 for large q, so λ_{q,i}ℓ_{q,i} ≫ 1 for all q beyond a fixed threshold (e.g., q ≥ 3 when b = 6). Consequently 120Mλ_{q,i} ≤ ℓ_{q,i}^{−1} is violated for large q uniformly in ε < ε_1^*(b,M). Since the estimates (3.11)–(3.17) are used at every step of the induction, the mollification step is not justified by the stated parameters. This is a second, independent calibration failure in the same core mechanism.
  3. [§1.3 and Theorem 1.7] Theorem 1.7 claims u ∈ C^1([0,T] × T^2), and C^1 regularity implies C^α regularity for every 0 < α < 1. Section 1.3 states that the treatment of second-order and higher-order nonlinearities 'limits our choice of δ_q' and, in particular, that 'this limitation prevents us from achieving the non-uniqueness of weak solutions in C^α spaces with α>0.' Both statements cannot hold for the same construction. If the limitation in §1.3 is genuine, then the C^1 conclusion of Theorem 1.7 and the corresponding claim in Proposition 2.6 are false as stated; if the limitation is not genuine, the discussion in §1.3 misrepresents the scope of the method. Either way, the manuscript contains a direct self-contradiction on the regularity of the constructed solutions, and this contradiction is load-bearing because the final convergence argument in §9 rests on the C^1 bounds supplied by the induction.
minor comments (3)
  1. [Abstract and §1.3] The abstract contains spacing/typo artifacts such as 'NON-UNIQUENE SS' and 'elastodyna mic', and §1.3 contains 'Rlinaer' instead of 'R_linear'; these should be corrected in a final version.
  2. [§5–§7 notation] The notational load is very heavy (superscripts (1),m,c, subscripts q,i+1, velocities A^υ_I, etc.). A summary table of the main symbols introduced in Sections 5–7 would substantially improve readability and reduce the risk of misreading the estimates.
  3. [§8.3.1, §8.4, §9] The version of the manuscript supplied for review breaks off inside §8.3.1 and does not display §8.4, §9, or the appendices. The final version must contain the complete proof of Proposition 2.6 and Theorem 1.7, because the closing of the induction depends on the estimates asserted in §8.3.

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity: the convex-integration construction is self-contained with in-paper proofs, the single self-citation [44] is not load-bearing, and the flagged tension around estimate (8.16) is a soundness risk rather than a circular reduction.

full rationale

The central claim, Theorem 1.7, asserts the existence of a nontrivial weak solution u ∈ C^1([0,T] × T^2) emanating from zero data, and it is derived through the inductive Proposition 2.6 rather than assumed. The load-bearing machinery is proven inside the paper: the compactly supported solutions to the symmetric and asymmetric divergence equations (Lemma 5.2 and Proposition 5.10, with the Isett–Oh construction re-derived in Section 5.1), the building blocks of Lemma 7.1, and the Reynolds-error estimates of Proposition 8.3. The external citations that carry method (Isett–Oh [28], De Lellis–Székelyhidi [24], De Lellis–Kwon [21], Giri–Kwon [26], Luo–Xin [42]) are independent prior works, and none of their statements is replaced by an unproved claim of the present authors. The only self-citation appears in Section 1.3: 'As shown in another of our works [44], if higher-order nonlinearities (third-order and above) are absent, it is possible to achieve non-uniqueness for Cα(α>0) continuous solutions.' That passage is a contrast explaining a difficulty; no proposition, estimate, or step in the proof of Theorem 1.7 imports a theorem from [44], so the self-citation is not load-bearing. The weak null condition (1.13) is an independent restriction on the stored-energy coefficients, and Remark 1.5 shows it is strictly weaker than Sideris's condition, so the theorem is a genuine conditional statement and not a definitional restatement of its hypothesis. No parameter is fitted to any data, and no predicted quantity reduces to an input by construction. Per the flagging instruction, I record the limitation admitted in Section 1.3: 'This absorption provides only a ε-level benefit, which limits our choice of δq in (2.3) that decays exponentially as q → ∞' and 'This limitation prevents us from achieving the non-uniqueness of weak solutions in Cα spaces with α>0.' Since C^1 regularity implies C^α for every α<1, this admitted limitation is in direct tension with Theorem 1.7, and the skeptic's arithmetic indicates that the linear error term (λ_{q,i+1}μ_{q,i})^{-1}δ_{q+1}^{1/2} in closing estimate (8.16) is asymptotically larger than δ_{q+2} under the schedule (2.3)–(2.4) for b>5. I weigh this as an internal-consistency or soundness risk that the authors and referees should resolve; it is not an instance of circularity as defined in this pass, because it concerns whether an estimate closes, not a reduction of the theorem to its own assumptions or to a self-citation chain.

Assumptions & free parameters 6 free parameters · 5 assumptions · 3 invented entities

The central claim rests on a large induction with many hand-chosen parameters and structure assumptions on the stored energy function. No physical constants are fitted to data; the parameters listed above are construction parameters. The axioms are mostly domain assumptions about the material and standard convex integration tools, while the inductive estimates are the internal burden of the proof.

free parameters (6)
  • epsilon (amplitude scale) = 0 < epsilon << 1, below epsilon_1(b,M,sigma)
    Small parameter controlling the size of all perturbations and the decay rates. The proof requires epsilon sufficiently small, depending on b, M, and the stored-energy coefficients.
  • b (frequency growth exponent) = b > 5
    Appears in lambda_q = ceil(epsilon^{-b(q+3)^{2/3}})^3. The proof requires b > 5 to close the estimates in Proposition 7.8 and 8.3.
  • M (induction constant) = M > 1 depending on coefficients of sigma
    Used in inductive estimates (2.10) and (2.16) for derivatives of the approximate solution. It is chosen after the stored-energy function is fixed.
  • delta_q (Reynolds amplitude) = epsilon^{2(q+3)/3}
    Decay sequence for the Reynolds error. Chosen by hand to be summable and to interact with lambda_q in the estimates.
  • lambda_q (base frequency) = ceil(epsilon^{-b(q+3)^{2/3}})^3
    Frequency scale of the high-frequency oscillations. The subexponential growth is central to the regularity of the limit solution and to the closing estimates.
  • derived scales lambda_{q,i}, ell_{q,i}, tau_{q,i}, mu_{q,i} = formulas (2.3) and (2.4)
    Frequency, mollification, time cutoff, and spatial cutoff scales derived from epsilon, b, and delta_q. All are construction parameters chosen to make the estimates close.
assumptions (5)
  • domain assumption Stored energy function is the cubic polynomial (1.2).
    The theorem is stated for this restricted class. Remark 1.1 asserts the general case is workable but does not provide a proof in this paper.
  • domain assumption Structural conditions (1.13), including the weak null condition 3 sigma_11 + 2 sigma_111 = 0, sigma* nonzero, lambda, mu > 0, and convexity sigma_11 sigma_22 - sigma_12^2 > 0.
    These assumptions define the class of materials. The weak null condition is used in the geometric lemma and the wave building blocks.
  • ad hoc to paper The approximate solutions satisfy the inductive estimates (2.8)-(2.11) and (2.14)-(2.19).
    The proof constructs such tuples, but the existence of the full family with all required constants is the core burden of Proposition 2.6, not an externally established fact.
  • standard math The Isett-Oh compactly supported solution operator for the symmetric divergence equation, Lemma 5.2.
    The paper provides a proof of the needed version, but the original technique is imported from Isett-Oh. It is a standard tool in the convex integration literature.
  • ad hoc to paper The building block Lemma 7.1 requires smallness of the coefficients A and lambda* + mu* nonzero.
    The existence and estimates for the longitudinal-transverse wave building blocks are proved inside the paper, but the smallness and non-degeneracy conditions are tailored to the overall scheme.
invented entities (3)
  • Longitudinal-transverse wave building blocks w_{A,f} = (f + a_A(t) f^perp) exp(i xi_{A,f})
    purpose: Serve as the principal oscillations added to the approximate solution at each step of the convex integration iteration.
    These building blocks are internal mathematical constructs with no independent physical observable. Their validity is established by the estimates in Lemma 7.1.
  • Angular momentum correction perturbations ~u_ac = ~u_L + ~u_M with functions g_L and g_M solving the ODE (7.48)
    purpose: Enforce the Lagrangian angular momentum conservation condition (5.57) needed by the local inverse divergence operator.
    These corrections are introduced purely to make the error terms fall into the class where the asymmetric divergence equation is solvable. They have no external falsifiable handle.
  • Lagrangian momentum and angular momentum constraints (1.17)
    purpose: Characterize the vector fields for which the local inverse divergence operator produces a symmetric stress matrix.
    These are mathematical structural hypotheses introduced for the construction, not independently measurable physical conservation laws.

how reviews work

0 comments
Cite this review

Pith. "Pith review of The null condition in elastodynamics leads to non-uniqueness." pith.science (2026). https://pith.science/paper/5TJL2FEH

@misc{pith2026250207521,
  author       = {Pith},
  title        = {Pith review of: The null condition in elastodynamics leads to non-uniqueness},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/5TJL2FEH}},
  note         = {Machine review of arXiv:2502.07521}
}
abstract

We consider the Cauchy problem for the system of elastodynamic equations in two dimensions. Specifically, we focus on materials characterized by a null condition imposed on the quadratic part of the nonlinearity. We can construct non-zero weak solutions $u \in C^1([0, T] \times \mathbb{T}^2)$ that emanate from zero initial data. The proof relies on the convex integration scheme. By exploiting the characteristic double wave speeds of the equations, we construct a new class of building blocks. This work extends the application of convex integration techniques to hyperbolic systems with a null condition and reveals the rich solution structure in nonlinear elastodynamics.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

54 extracted references · 54 canonical work pages

  1. [1]

    Al Baba, C

    H. Al Baba, C. Klingenberg, O. Kreml, V . M´ acha, and S. Markfelder. Nonuniqueness of admissible weak solution to the R iemann problem for the full Euler system in two dimensions. SIAM J. Math. Anal. , 52(2):1729–1760, 2020

  2. [2]

    S. Alinhac. The null condition for quasilinear wave equa tions in two space dimensions. I. Invent. Math., 145(3):597–618, 2001

  3. [3]

    X. An, H. Chen, and S. Yin. Low regularity ill-posedness f or elastic waves driven by shock formation. Amer . J. Math., 145(4):1111–1181, 2023

  4. [4]

    X. An, H. Chen, and S. Y u. Low-regularity local well-pose dness for the elastic wave system. arXiv:2411.15886, 2024

  5. [5]

    Beekie, T

    R. Beekie, T. Buckmaster, and V . Vicol. Weak solutions of ideal MHD which do not conserve magnetic helicity. Ann. PDE, 6(1):Paper No. 1, 40pp, 2020

  6. [6]

    Bˇ rezina, O

    J. Bˇ rezina, O. Kreml, and V . M´ acha. Non-uniqueness of delta shocks and contact discontinuities in the multi-dimen sional model of Chaplygin gas. NoDEA Nonlinear Differential Equations Appl., 28(2):Paper No. 13, 24, 2021

  7. [7]

    Buckmaster

    T. Buckmaster. Onsager’s conjecture almost everywhere in time. Comm. Math. Phys. , 333(3):1175–1198, 2015

  8. [8]

    Buckmaster, C

    T. Buckmaster, C. De Lellis, L. Sz´ ekelyhidi, Jr., and V . Vicol. Onsager’s conjecture for admissible weak solutions . Comm. Pure Appl. Math. , 72(2):229–274, 2019

Show all 54 references
  1. [9]

    Buckmaster, C

    T. Buckmaster, C. De Lellis, P . Isett, and L. Sz´ ekelyhidi, Jr.. Anomalous dissipation for 1/5-H¨ older Euler flows.Ann. of Math. (2), 182(1):127– 172, 2015

  2. [10]

    Buckmaster, C

    T. Buckmaster, C. De Lellis, and L. Sz´ ekelyhidi, Jr. Dissipative Euler flows with Onsager-critical spatial regula rity. Comm. Pure Appl. Math. , 69(9):1613–1670, 2016

  3. [11]

    Buckmaster and V

    T. Buckmaster and V . Vicol. Nonuniqueness of weak solut ions to the Navier–Stokes equation. Ann. of Math. (2) , 189(1):101–144, 2019

  4. [12]

    Chen, and J

    K. Chen, and J. Liu. Weak solutions of the three-dimensi onal hypoviscous elastodynamics with finite kinetic energy J. Differential Equations, 339:637–665, 2022

  5. [13]

    R. M. Chen, A. F. V asseur, and C. Y u. Non-uniqueness for continuous solutions to 1D hyperbolic systems. arXiv:2407. 02927v1, 2024

  6. [14]

    Cheskidov and X

    A. Cheskidov and X. Luo. Sharp nonuniqueness for the Nav ier–Stokes equations. Invent. Math., 229(3):987–1054, 2022

  7. [15]

    Chiodaroli, C

    E. Chiodaroli, C. De Lellis, and O. Kreml. Global ill-po sedness of the isentropic system of gas dynamics. Comm. Pure Appl. Math. , 68(7):1157–1190, 2015

  8. [16]

    Chiodaroli and O

    E. Chiodaroli and O. Kreml. Non-uniqueness of admissib le weak solutions to the Riemann problem for isentropic Eule r equations. Nonlinear- ity, 31(4):1441–1460, 2018

  9. [17]

    Chiodaroli, O

    E. Chiodaroli, O. Kreml, V . M´ acha, and S. Schwarzacher. Non-uniqueness of admissible weak solutions to the compre ssible Euler equations with smooth initial data. Trans. Amer . Math. Soc., 374(4):2269–2295, 2021

  10. [18]

    Christodoulou

    D. Christodoulou. Global solutions of nonlinear hyper bolic equations for small initial data. Comm. Pure Appl. Math. , 39(2):267–282, 1986

  11. [19]

    C. M. Dafermos. Hyperbolic Conservation Laws in Continuum Physics . Springer-V erlag, 2016

  12. [20]

    Daneri and L

    S. Daneri and L. Sz´ ekelyhidi, Jr. Non-uniqueness and h -principle for H¨ older-continuous weak solutions of the Euler equations. Arch. Ration. Mech. Anal., 224(2):471–514, 2017

  13. [21]

    De Lellis and H

    C. De Lellis and H. Kwon. On nonuniqueness of H¨ older con tinuous globally dissipative Euler flows. Anal. PDE, 15(8):2003–2059, 2022

  14. [22]

    De Lellis and L

    C. De Lellis and L. Sz´ ekelyhidi, Jr.. The Euler equatio ns as a di fferential inclusion. Ann. of Math. (2) , 170(3):1417–1436, 2009

  15. [23]

    De Lellis and L

    C. De Lellis and L. Sz´ ekelyhidi, Jr.. On admissibility criteria for weak solutions of the Euler equation. Arch. Ration. Mech. Anal., 195(1):225– 260, 2010

  16. [24]

    De Lellis and L

    C. De Lellis and L. Sz´ ekelyhidi, Jr.. Dissipative continuous Euler flows. Invent. Math., 193(2):377–407, 2013

  17. [25]

    De Lellis and L

    C. De Lellis and L. Sz´ ekelyhidi, Jr.. Dissipative Euler flows and Onsager’s conjecture. J. Eur . Math. Soc., 16(7):1467–1505, 2014

  18. [26]

    Giri and H

    V . Giri and H. Kwon. On non-uniqueness of continuous ent ropy solutions to the isentropic compressible Euler equati ons. Arch. Ration. Mech. Anal., 245(2):1213–1283, 2022

  19. [27]

    V . Giri, H. Kwon, and M. Novack. A wavelet-inspired L3-based convex integration framework for the Euler equations. Ann. PDE, 10(2):Paper No. 19, 271pp., 2024

  20. [28]

    Isett and S

    P . Isett and S. Oh. On nonperiodic Euler flows with H¨ olde r regularity. Arch. Ration. Mech. Anal., 221(2):725–804, 2016

  21. [29]

    P . Isett. H¨ older Continuous Euler Flows in Three Dimensions with Compact Support in Time. Princeton University Press, Princeton, NJ, 2017

  22. [30]

    P . Isett. A proof of Onsager’s conjecture. Ann. of Math. (2) , 188(3):871–963, 2018

  23. [31]

    F. John. Almost global existence of elastic waves of fini te amplitude arising from small initial disturbances. Comm. Pure Appl. Math. , 41(5):615–666, 1988

  24. [32]

    Klainerman

    S. Klainerman. The Null Condition and Global Existence to Nonlinear W ave Eq uations. Amer. Math. Soc., Providence, RI, 1986. 66

  25. [33]

    Klingenberg, O

    C. Klingenberg, O. Kreml, and V . M´ acha, and S. Markfeld er. Shocks make the Riemann problem for the full Euler system in multiple space dimensions ill-posed. Nonlinearity, 33(12):6517–6540, 2020

  26. [34]

    P . D. Lax. Hyperbolic systems of conservation laws. II. Comm. Pure Appl. Math. , 10:537–566, 1957

  27. [35]

    Lei and Y

    Z. Lei and Y . Zhou. Global low regularity solutions of qu asi-linear wave equations. Adv. Differential Equations, 13(1-2):55–104, 2008

  28. [36]

    Z. Lei, C. T. Sideris, and Y . Zhou. Almost global existen ce for 2-D incompressible isotropic elastodynamics. Trans. Amer . Math. Soc. , 367(11):8175–8197, 2015

  29. [37]

    Z. Lei. Global well-posedness of incompressible elast odynamics in two dimensions. Comm. Pure Appl. Math., 69(11):2072–2106, 2016

  30. [38]

    Y . Li, P . Qu, Z. Zeng, and D. Zhang. Sharp non-uniqueness for the 3D hyperdissipative Navier–Stokes equations: abov e the Lions exponent. J. Math. Pures Appl. (9) , 190:Paper No. 103602, 64 pp, 2024

  31. [39]

    Y . Li, Z. Zeng, and D. Zhang. Non-uniqueness of weak solutions to 3D magnetohydrodynamic equations. J. Math. Pures Appl. (9), 165(9):232- 285, 2022

  32. [40]

    Lindblad

    H. Lindblad. Counterexamples to local existence for qu asilinear wave equations. Math. Res. Lett., 5(5):605–622, 1998

  33. [41]

    Luo and P

    T. Luo and P . Qu. Non-uniqueness of weak solutions to 2D hypoviscous Navier–Stokes equations. J. Differential Equations, 269(4):2896–2919, 2020

  34. [42]

    Luo and Z

    T. Luo and Z. Xin. H¨ older continuous solutions to the th ree-dimensional Prandtl system. arXiv: 1804.04285, 2018

  35. [43]

    Mao and P

    S. Mao and P . Qu. Non-uniqueness for the compressible Eu ler-Maxwell equations. Calc. V ar . Partial Differential Equations. 63(7), Paper No. 186, 84 pp., 2024

  36. [44]

    Mao and P

    S. Mao and P . Qu. Non-uniqueness for the nonlinear dynam ical Lam´ e system, preprint .2024

  37. [45]

    Miao and W

    C. Miao and W. Y e. On the weak solutions for the MHD system s with controllable total energy and cross helicity. J. Math. Pures Appl. (9) , 181:190–227, 2024

  38. [46]

    Novack and V

    M. Novack and V . Vicol. An intermittent Onsager theorem . Invent. Math., 233(1):223–323, 2023

  39. [47]

    Smith and D

    H. Smith and D. Tataru. Sharp local well-posedness resu lts for the nonlinear wave equation. Ann. of Math. (2) , 162(1):291–366, 2005

  40. [48]

    T. C. Sideris. The null condition and global existence o f nonlinear elastic waves. Invent. Math., 123(2):323–342, 1996

  41. [49]

    T. C. Sideris. Nonresonance and global existence of pre stressed nonlinear elastic waves. Ann. of Math. (2) , 151(2):849–874, 2000

  42. [50]

    T. C. Sideris and B. Thomases. Global existence for thre e-dimensional incompressible isotropic elastodynamics v ia the incompressible limit. Comm. Pure Appl. Math., 58(6):750–788, 2005

  43. [51]

    Q. Wang. A geometric approach for sharp local well-pose dness of quasilinear wave equations. Ann. PDE, 3(1):Paper No. 12, 108 pp., 2017

  44. [52]

    D. Zha. Global and almost global existence for general q uasilinear wave equations in two space dimensions. J. Math. Pures Appl. (9) , 123(9):270–299, 2019

  45. [53]

    Zha and K

    D. Zha and K. Hidano. Global solutions to systems of quas ilinear wave equations with low regularity data and applica tions. J. Math. Pures Appl. (9), 142(9):146–183, 2020

  46. [54]

    H. Zhang. Local well-posedness for incompressible neo -Hookean elastic equations in almost critical Sobolev spac es. Calc. V ar . Partial Differ- ential Equations, 63(3):Paper No. 66, 20, 2024. School of MathematicalSciences, Fudan University, China. Email address, Shunkai Mao:...

Pith tools

Reviewed August 8, 2026 · model on record in the stance chip above.