Pith. sign in

REVIEW 1 major objections 4 minor 32 references

Properties of the phase boundary in the parabolic problem with hysteresis

T0 review · 1 major / 4 minor · reviewed 2026-08-12 · deepseek-v4-flash

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

Pith's one-line read Hysteresis phase boundary is Hölder-1/2, Lipschitz with smooth data

desk verdict Solid interface-regularity results for transverse data; the advertised non-existence for non-transversal data is weaker than the abstract claims. read the letter →

arxiv 2411.17512 v1 pith:CQPRNQXN submitted 2024-11-26 math.AP

Darya Apushkinskaya , Sergey Tikhomirov , Nina Uraltseva This is my paper · ORCID
classification math.AP MSC 35R3535K5535B6547J40
keywords hysteresisparabolicequationphaseboundaryfreetransversalityHöldercontinuityLipschitzregularitynon-transversaldata
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 studies the one-dimensional heat equation with a discontinuous hysteresis source term, in which the domain is divided into two phases separated by a free boundary. The central claim is that when the initial data cross the switching thresholds transversally, the free-boundary problem has a solution on a small time interval, and each branch of the phase boundary is a monotone Hölder function with exponent $1/2$. For initial data in $W^2_\infty(\Omega)$, the same construction upgrades this to a Lipschitz phase boundary, with an explicit Lipschitz constant $4N_2/m$ that is the paper's main quantitative result. For non-transversal initial data, the paper shows that a well-defined interface of the considered type cannot exist. A curious reader should care because the result converts a discontinuous, history-dependent problem into a standard linear-parabolic one with controlled interface regularity.

What carries the argument

The carrying mechanism is the transversality condition $|\partial_x\varphi(b_i)| \ge m > 0$ at each initial switching point, which yields the uniform bound $\partial_x v \ge m/4$ on a fixed $\sigma$-neighborhood of each $b_i$ for small times. This makes the level set $\{v = \alpha\}$ or $\{v = \beta\}$ a graph $x = a_i(t)$, whose Hölder modulus follows from pointwise estimates of $v$. To obtain monotone branches, the authors take the upper or lower monotone envelope $a_{i,0}(t)$ of each level curve and apply the Schauder fixed-point theorem on a closed convex set of monotone curves. The new mechanism for Lipschitz regularity is a pair of one-sided estimates on the time difference quotient $u^{(h)} = (u(x,t+h)-u(x,t))/h$: $\min_{Q_i^*} u^{(h)} \ge -N_2$ when $\varphi(b_i)=\alpha$, and $\max_{Q_i^*} u^{(h)} \le N_2$ when $\varphi(b_i)=\beta$, derived from Lemma 4.1's bound $|v(x,t)-\varphi(x)| \le N_0 t$, Lemma 4.2's bound in the intermediate rectangles, and the sign of the residual of the difference quotient. These one-sided estimates control the re-crossing speed of the threshold, and the transversality slope converts this into a bound on the interface's slope.

What would settle it

Take a transverse initial datum with $\varphi(b)=\alpha$ and $\partial_x\varphi(b)=m>0$, and measure numerically the maximal time difference quotient of the solution in the switching rectangle as $h\to 0$; if this quantity exceeds $N_2$, or if a branch $s_i(t)$ has a chord slope exceeding $4N_2/m$, then the one-sided estimate (4.8) and the Lipschitz theorem fail.

Watch

Extended reading notes

Core claim

The paper proves Theorems 3.1 and 4.1: under the transversality condition $|\partial_x\varphi(b_i)| > 0$ at each initial phase-switch point, with $\varphi \in W^{2-2/q}_q(\Omega)$, $q > 3$, there exists $T > 0$ such that the initial-boundary problem (1.2) admits a solution $u \in W^{2,1}_q(Q_T)$, and each branch $x = s_i(t)$ of the phase boundary is a monotone Hölder function of exponent $1/2$. If, in addition, $\varphi \in W^2_\infty(\Omega)$, then each branch is Lipschitz with constant $4N_2/m$. The new ingredient is a pair of one-sided difference-quotient estimates that bound, in the switching rectangles, how fast the solution can move across the threshold; together with the transversality lower bound $\partial_x u \ge m/4$ near the interface, these estimates force $|s_i(t_2) - s_i(t_1)| \le (4N_2/m)|t_2 - t_1|$. For non-transversal data, Theorem 5.1 shows that any continuous piece of the switching-time curve $r(x)$ must be constant and that left and right limiting values must agree, which rules out the level-set and sleeping-boundary interface structures used in the existence construction.

Load-bearing premise

The construction requires that the initial data cross every switching threshold at nonzero spatial slope ($|\partial_x\varphi(b_i)|>0$); if a switching point has vanishing slope, the level-set curves cannot be defined, the fixed-point map collapses, and the existence proof yields nothing.

Editorial extensions

If this is right

  • For transverse data in $W^{2-2/q}_q(\Omega)$ with $q>3$, the free-boundary problem (1.2) is solvable on a time interval whose length depends only on $q$ and the initial datum, and each interface branch is monotone Hölder-1/2.
  • If the initial datum lies in $W^2_\infty(\Omega)$, the interface speed is uniformly bounded: every branch satisfies $|s_i(t_2)-s_i(t_1)| \le (4N_2/m)|t_2-t_1|$, so the phase boundary cannot jump; flat, vertical-in-time segments remain possible.
  • For non-transversal data, no interface of the level-set or sleeping-boundary type can exist: any continuous piece of the switching-time curve is constant, so the free boundary cannot be a graph of a strictly monotone function.
  • The constructed interface branches are monotone envelopes of level curves, so they are nondecreasing (or nonincreasing) and alternate between strict motion and rest, giving an explicit structural description of the phase boundary.

Reading between the lines

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

  • The explicit dependence of the Lipschitz constant on $1/m$ suggests that as the transversality slope $m$ tends to zero, the maximal existence time $T$ should shrink and the interface may become more irregular; a natural conjecture left unproved in the paper is that regularity degenerates no faster than $O(m^{-1})$.
  • Theorem 5.1 leaves open the possibility of pathological solutions with a switching set of positive measure and empty interior, as the authors explicitly note; a testable extension is to check whether a regularized multi-valued hysteresis operator produces such solutions and whether they propagate with finite speed.
  • The one-sided difference-quotient technique may extend to two-sided estimates under a two-sided non-degeneracy condition, which would upgrade the interface branches to $C^1$; the paper does not pursue this.
  • The same level-set plus monotone-envelope plus fixed-point strategy should transfer to multi-dimensional radial configurations, where transversality becomes the non-vanishing of the radial derivative of the initial datum at the switching surface.
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

1 major / 4 minor

Summary. The paper studies the one-dimensional parabolic problem (1.2) with a discontinuous hysteresis source term. For initial data φ ∈ W^{2−2/q}_q(Ω), q > 3, satisfying the transversality condition at each initial phase-switch point, it constructs a solution in W^{2,1}_q(Q_T) by a Schauder fixed-point argument, and shows that each branch of the phase boundary is the graph of a nondecreasing (or nonincreasing) Hölder-1/2 function (Theorems 2.1 and 3.1). When φ ∈ W^2_∞(Ω), Theorem 4.1 upgrades this to Lipschitz continuity of each branch with constant 4N_2/m. The last section addresses non-transversal data: Theorem 5.1 shows that, under the assumption r < T_1 on an interval, a continuous switching-time function r must be constant and its left and right upper limits coincide, but the authors explicitly leave open the existence of pathological solutions. The abstract, however, claims a stronger non-existence statement.

Significance. The positive results are a genuine improvement over the earlier construction in [12]: the existence proof is simplified, the Hölder exponent 1/2 is better than the previously obtained exponent, and the Lipschitz estimate for W^2_∞ initial data is new. The proofs are largely self-contained and rest on standard linear parabolic estimates, comparison and maximum-principle arguments, and a clean one-sided difference-quotient technique in Lemmas 4.2–4.3. The precise negative statement about regular interface branches in the non-transversal case is also a useful contribution, although it is considerably weaker than the headline claim in the abstract.

major comments (1)
  1. [Abstract and §5] The abstract's final sentence states that for non-transversal initial data 'solutions with an interface boundary do not exist.' This is not what Theorem 5.1 proves. Under the additional assumption r(x) < T_1 on an interval, Theorem 5.1 shows (i) that a continuous r on that interval is constant and (ii) that the left and right upper limits of r coincide at every point. It does not exclude solutions whose switching set has positive measure and is nowhere dense, nor solutions with u = β on an open interval for some time, as the authors themselves explicitly say in the paragraph following Theorem 5.1. The abstract and the introductory paragraph of §5 should be revised to claim only that no interface branch of the regular type constructed in §§2–3 can exist for non-transversal data, not that solutions with an interface boundary do not exist.
minor comments (4)
  1. [§4, Lemma 4.1] The proof of Lemma 4.1 claims one may take N_0 = sup_Ω |Δφ|. But the functions w = t ± ε(v−φ) satisfy ∂_t w − Δw = 1 ± ε(f + Δφ), and both expressions are nonnegative only if ε is at most of order 1/(1 + sup |f + Δφ|). The argument therefore yields a constant of the form 1 + sup |Δφ| (up to a factor depending on the bound on f), not sup |Δφ|. The existence of some constant N_0 is unaffected, so the statement of the lemma is correct, but the displayed constant is wrong.
  2. [§5, Definition 5.1] The definition of topological non-transversality is clear from the two displayed alternatives, but the phrase 'for some neighborhood (x_0 − ε, x_0 + ε)' makes ε overloaded with the small positive parameter used elsewhere in the paper for the difference-quotient range; this is harmless but should be clarified.
  3. [§4, Lemma 4.2] In the proof of Lemma 4.2, the decomposition u^{(h)} = w_1 + w_2 is only sketched: the boundary values of w_1 on the lateral sides of Q^{**}_{i,ρ} are described as 'constant' but the actual values are not specified. The reader must infer that w_1 is chosen to match u^{(h)} (up to a constant) on those sides. Please spell out the exact boundary conditions for w_1 and w_2 so that the heat equations asserted for both are unambiguous.
  4. [§3, Theorem 3.1] The phrase 'необходимое условие очередности начальных фаз' is not formalized; the intended alternating condition (a phase-I/II/… sequence at the bi) should be stated explicitly so that the hypotheses of the theorem are self-contained.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the Hölder and Lipschitz regularity theorems are proven directly from linear parabolic estimates, comparison principles, and a self-contained Schauder fixed-point argument.

full rationale

The derivation chain is self-contained and not circular. For the existence and Hölder 1/2 result, the paper does not assume the interface regularity it wants to prove: for each monotone interface candidate ξ it solves the linear problem (1.5) with right-hand side (2.11), extracts the level-set curve a(t), forms the monotone envelope a0(t), proves the envelope is Hölder with exponent 1/2 uniformly ((2.14)), proves continuity of the map ξ → a0 ((2.15)), obtains compactness via Arzelà–Ascoli, and then applies Schauder. The fixed point s(t) is exactly the interface, and its Hölder modulus is inherited from the uniform estimate already proved, not fed into the argument as an assumption. The Lipschitz result in §4 is likewise derived forward: Lemma 4.1 bounds |v−φ| linearly in t; Lemma 4.2 bounds time-difference quotients away from the interface using the maximum principle and hypoelliptic derivative estimates; Lemma 4.3 gives one-sided difference-quotient estimates inside the switching rectangles using monotonicity of s_i and the sign of f^(h); combining these with the transversality lower bound yields (4.9). None of these steps uses Theorem 4.1 as an input. The paper does follow the construction of [12], which is coauthored by one of the present authors, but it reproduces the construction and proves the new regularity claims directly; the self-citations to [12]–[16] are contextual and not load-bearing for the new theorems. The non-transversal negative claim is overstated in the abstract relative to Theorem 5.1: the authors explicitly concede that pathological solutions with nowhere-dense switching sets or with u=β on an open interval cannot be excluded. This is a scope/correctness gap, not a circularity. A minor technical slip in Lemma 4.1 (the displayed constant should be of order 1+sup|Δφ| rather than sup|Δφ|) is also unrelated to circularity.

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

No data-fit parameters exist; this is a pure proof. The constants σ, T, m, N_0, N_1, N_2 are existential choices in the proof depending on φ and q, not model parameters. The axioms are standard PDE tools plus the domain assumptions of the theorem.

assumptions (6)
  • standard math Standard parabolic L_q estimates and Sobolev embedding for W^{2,1}_q, q>3 (Ladyzhenskaya-Solonnikov-Uraltseva)
    Used in (1.6)-(1.9) to get uniform bounds and Hölder continuity of v and ∂_x v; invoked in §1 and throughout §§2-4.
  • standard math Schauder fixed point theorem
    Used in §2 and §3 to obtain a fixed point of the monotone-envelope map R on a convex compact set.
  • standard math Strong maximum principle and Hopf boundary point lemma for parabolic inequalities
    Used in §5 (Propositions 5.2-5.3) to rule out regular interface branches for non-transverse data.
  • domain assumption Initial data regularity φ ∈ W^{2-2/q}_q, q>3, with ∂_x φ(0)=∂_x φ(1)=0 and transversality at finitely many switch points
    Core assumptions of Theorems 2.1, 3.1, 4.1; they guarantee C^1 regularity and the lower bound (2.3).
  • domain assumption The hysteresis operator H is fully determined by the initial phase distribution and the threshold rule (1.3)-(1.4)
    This is the model definition of the problem; the solution concept relies on it in (1.3)-(1.4) and Figure 1.
  • domain assumption For the Lipschitz result, additional assumption φ∈W^2_∞(Ω)
    Needed in Lemma 4.1 to control |v(x,t)-φ(x)| ≤ N_0 t and hence the initial difference quotients (4.4).

how reviews work

0 comments
Cite this review

Pith. "Pith review of Properties of the phase boundary in the parabolic problem with hysteresis." pith.science (2026). https://pith.science/paper/CQPRNQXN

@misc{pith2026241117512,
  author       = {Pith},
  title        = {Pith review of: Properties of the phase boundary in the parabolic problem with hysteresis},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/CQPRNQXN}},
  note         = {Machine review of arXiv:2411.17512}
}
abstract

We study solutions of parabolic equations with a discontinuous hysteresis operator, described by a free interface boundary. It is established that for spatially transverse initial data from the space $W^{2-2/q}_q$ with $q > 3$, there exists a solution in the space $W^{2,1}_q$, where the interface boundary exhibits Holder continuity with an exponent $1/2$. Furthermore for initial data from the space $W^2_\infty$, it is proven that the interface boundary satisfies the Lipschitz condition. It is shown that for non-transversal initial data, solutions with an interface boundary do not exist.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

32 extracted references · 32 canonical work pages

  1. [12]

    Brokate, J

    M. Brokate, J. Sprekels. Hysteresis and Phase Transitions. Springer, Berlin, 1996

    work page 1996

  2. [1]

    |a0(t2) − a0(t1)| ⩽ 4c1 m |t2 − t1|1/2, [t1, t2] ⊂ [0, T], (2.14)

    work page

  3. [2]

    (2.15) Доказательство

    max t∈[0,T ] |a0(t) − ea0(t)| ⩽ max t∈[0,T ] |a(t) − ea(t)|. (2.15) Доказательство. Докажем утверждение 1). Еслиt1 и t2 попали на один интервал постоянства, то оценка (2.14) тривиальна. Иначе, положим t2 = max {τ : τ ⩽ t2 и a0(τ ) = a(τ )} , t1 = min {τ : τ ⩾ t1 и a0(τ ) = a(τ )} . Ясно, что t1 ⩽ t1 ⩽ t2 ⩽ t2. Тогда справедлива следующая оценка 0 ⩽ a0(t2)...

    work page

  4. [3]

    одно из значенийv(bi + σ, t) − φ(bi) и v(bi − σ, t) − φ(bi) больше чемm 4 σ, а другое меньше чем− m 4 σ

    work page

  5. [4]

    вне σ-окрестностей точек bi функции v(x, t) удовлетворяют нера- венству v(x, t) > α, если φ(x) находится в фазе II, или неравенству v(x, t) < β, если φ(x) находится в фазе I

    work page

  6. [5]

    Кроме того, вσ-окрестности каждой точки bi функция ∂xv(x, t) имеет тот же знак, что и∂xφ(bi)

    |∂xv(x, t)| > m 4 при |x−bi| ⩽ σ. Кроме того, вσ-окрестности каждой точки bi функция ∂xv(x, t) имеет тот же знак, что и∂xφ(bi). В дальнейшем мы считаем положительные постоянныеσ и T зафикси- рованными (как и в предыдущем параграфе, их значения зависят только от φ). Рассуждая так же как и в §2, нетрудно показать, что для всех индек- сов iлинии уровняv(x, t...

    work page

  7. [6]

    φ(bi) = α и ∂xφ(bi) ⩾ m неубывающие f (x, t) = ( 1, при x ⩽ ξi(t), −1, при x > ξi(t)

    work page

  8. [7]

    φ(bi) = α и ∂xφ(bi) ⩽ −m невозрастающие f (x, t) = ( 1, при x ⩾ ξi(t), −1, при x < ξi(t)

    work page

Show all 32 references
  1. [8]

    φ(bi) = β и ∂xφ(bi) ⩾ m невозрастающие f (x, t) = ( −1, при x ⩾ ξi(t), 1, при x < ξi(t)

  2. [9]

    спящей границе

    φ(bi) = β и ∂xφ(bi) ⩽ −m неубывающие f (x, t) = ( −1, при x ⩽ ξi(t), 1, при x > ξi(t). 16 та w = (w1, . . . , wn) ∈ Bопределяется формулой ∥w∥B = max 1⩽i⩽n ∥wi∥C([0,T ]). Из определения пространстваB следует, что утверждение о выпуклости (замкнутости, компактности) множестваE ...

  3. [10]

    Красносельский, А.В

    М.А. Красносельский, А.В. Покровский.Системы с гистерезисом. М.: Наука, 1983

    1983

  4. [11]

    Visintin

    A. Visintin. Differential Models of Hysteresis.Springer-Verlag. Berlin — Heidelberg, 1994

    1994

  5. [13]

    Hoppensteadt, W

    F.C. Hoppensteadt, W. J¨ ager. Pattern formation by bacteria. In Biological growth and spread (Proc. Conf., Heidelberg, 1979), volume 38 of Lecture Notes in Biomath., pages 68–81. Springer, Berlin-New York, 1980. 29

    1979

  6. [14]

    Hoppensteadt, W

    F.C. Hoppensteadt, W. J¨ ager, C. P¨ oppe. A hysteresis model for bacterial growth patterns. In Modelling of patterns in space and time (Heidelberg, 1983), volume 55 ofLecture Notes in Biomath., pages 123–

    1983

  7. [15]

    Gurevich, S

    P. Gurevich, S. Tikhomirov. Rattling in spatially discrete diffusion equations with hysteresis. Multiscale Model. Simul., 15(3):1176–1197, 2017. 30

    2017

  8. [16]

    Apushkinskaya, N.N

    D.E. Apushkinskaya, N.N. Uraltseva. Free boundaries in problems with hysteresis. Philos. Trans. Roy. Soc. A, 373(2050):20140271, 10, 2015

    2015

  9. [17]

    Curran, P

    M. Curran, P. Gurevich, S. Tikhomirov. Recent advance in reaction- diffusion equations with non-ideal relays. InControl of self-organizing nonlinear systems, Underst. Complex Syst., pages 211–234. Springer, [Cham], 2016

    2016

  10. [18]

    Visintin

    A. Visintin. Evolution problems with hysteresis in the source term. SIAM J. Math. Anal., 17(5):1113–1138, 1986

    1986

  11. [19]

    H.W. Alt. On the thermostat problem.Control Cybernet., 14(1-3):171– 193, 1985

    1985

  12. [20]

    Kopfov´ a

    J. Kopfov´ a. Nonlinear semigroup methods in problems with hysteresis. Discrete Contin. Dyn. Syst., pages 580–589, 2007

    2007

  13. [21]

    Visintin

    A. Visintin. Ten issues about hysteresis.Acta Appl. Math., 132:635–647, 2014

    2014

  14. [22]

    Gurevich, R

    P. Gurevich, R. Shamin, S. Tikhomirov. Reaction-diffusion equations with spatially distributed hysteresis.SIAM J. Math. Anal., 45(3):1328– 1355, 2013

    2013

  15. [23]

    Gurevich, S

    P. Gurevich, S. Tikhomirov. Uniqueness of transverse solutions for reaction-diffusion equations with spatially distributed hysteresis. Nonlinear Anal., 75(18):6610–6619, 2012

    2012

  16. [24]

    Apushkinskaya, N.N

    D.E. Apushkinskaya, N.N. Uraltseva. On regularity properties of solutions to the hysteresis-type problem. Interfaces Free Bound., 17(1):93–115, 2015

    2015

  17. [26]

    Gurevich, S

    P. Gurevich, S. Tikhomirov. Spatially discrete reaction-diffusion equations with discontinuous hysteresis.Ann. Inst. H. Poincar´ e C Anal. Non Lin´ eaire, 35(4):1041–1077, 2018

    2018

  18. [27]

    Ладыженская, В.А

    О.А. Ладыженская, В.А. Солонников, Н.Н. Уральцева.Линейные и квазилинейные уравнения параболического типа. М.: Наука, 1967

    1967

  19. [28]

    Канторович, Г.П

    Л.В. Канторович, Г.П. Акилов.Функциональный анализ. М.: Наука,

  20. [29]

    М.А. Шубин. Лекции об уравнениях математической физики. М.: МЦНМО, Изд. 2е, исправ. edition, 2003

    2003

  21. [30]

    Крылов.Нелинейные эллиптические и параболические уравне- ния второго порядка

    Н.В. Крылов.Нелинейные эллиптические и параболические уравне- ния второго порядка. Наука, 1985

    1985

  22. [31]

    Apushkinskaya, A.I

    D.E. Apushkinskaya, A.I. Nazarov. On the boundary point principle for divergence-type equations. Atti Accad. Naz. Lincei Rend. Lincei Mat. Appl., 30(4):677–699, 2019

    2019

  23. [32]

    A.I. Nazarov. A centennial of the Zaremba-Hopf-Oleinik lemma.SIAM J. Math. Anal., 44(1):437–453, 2012. Российский университет дружбы народов, г. Москва, E-mail: apushkinskaya@gmail.com Pontif´ ıcia Universidade Cat´ olica do Rio de Janeiro - PUC-Rio, г. Рио де Жанейро E-mail: ...

    2012

  24. [134]

    Springer, Berlin, 1984

    1984

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.