Analytic regularity for a fourth-order singularly perturbed boundary balue problem with two small parameters

C. Xenophontos; I. Sykopetritou

REVIEW 3 major objections 6 minor 18 references

Analytic regularity for a fourth-order singularly perturbed boundary balue problem with two small parameters

T0 review · 3 major / 6 minor · reviewed 2026-08-01 · deepseek-v4-flash

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

Pith's one-line read For a fourth-order two-parameter singularly perturbed boundary value problem, the solution splits into a smooth part, two distinct-width boundary-layer pairs, and an exponentially small remainder, with derivative bounds explicit in the diff

desk verdict The paper targets a real gap and has a plausible strategy, but the main theorem's layer estimates are swapped relative to the functions constructed from (8)–(10), and the proof omits the chain-rule and stability steps, so the central claim is not currently supported. read the letter →

arxiv 2607.16781 v1 pith:SR7UO266 submitted 2026-07-18 math.NA cs.NA

I. Sykopetritou , C. Xenophontos This is my paper · ORCID

classification math.NAcs.NA MSC 65L1134E15

keywords fourthordersingularlyperturbedproblemtwosmallparametersboundarylayersanalyticregularitymatchedasymptoticexpansionshpfiniteelementmethodderivativeestimates

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 proves that the solution of a fourth-order boundary value problem involving two small parameters, under analytic input data and the regime where one parameter is much smaller than the square of the other, admits a decomposition into a smooth part, two boundary layers of different widths, and a negligible remainder. For each part it establishes derivative bounds that are explicit in the order of differentiation and in both parameters, showing that the solution is analytic but that differentiating introduces negative powers of the parameters. These bounds are exactly what a high-order numerical method such as the p/hp finite element method needs to prove convergence rates. The result extends earlier one-parameter and second-order two-parameter analyses to the fourth-order two-parameter setting, where the two layer scales couple nontrivially.

What carries the argument

The method of matched asymptotic expansions with two stretched variables, x/ε2 and x ε1/ε2^2, separates the problem into a slow smooth part and two boundary-layer families with different widths. The recursive coefficient systems become coupled through both ε1 and ε2, and the paper's key technical tool is an abstract lemma (Lemma 4) that gives entire-function bounds with explicit factorial growth for solutions of coupled second- and fourth-order ODEs on the half-line; this lemma is then iterated by induction on the expansion indices to control every derivative of every layer term.

What would settle it

Take the constant-coefficient case b=c=f=1, where the solution can be written explicitly, and set ε1 = ε2^2. If the computed derivative u'(x) near the endpoints does not exhibit the two separate exponential scales claimed in Theorem 7, or if the remainder term fails to satisfy the e^{-δ/ε2} bound, then the regime condition is essential and the theorem does not extend beyond it.

Watch

Extended reading notes

Core claim

The central claim is Theorem 7: for ε1 ≪ ε2^2, the solution u can be written as u = u_M + ṽ^BL_M + ū^BL_M + ˇu^BL_M + ˆu^BL_M + r_M, where the smooth part satisfies ∥u_M^{(n)}∥ ≤ C K_1^n n!, the two left layers satisfy |(ṽ^BL_M)^{(n)}| ≤ C K̃^n ε2^{-1} (ε1/ε2)^{1-n} e^{-β ε2 x/ε1} and |(ū^BL_M)^{(n)}| ≤ C K̄^n ε2^{-n} e^{-β x/ε2}, with analogous right-end layers, and the remainder satisfies ∥r_M∥_{∂I} + ε2∥r'_M∥_{∂I} + ∥r_M∥_{1,I} + ε1∥r''_M∥_{0,I} ≤ C e^{-δ/ε2} for suitable truncation index M. The proof builds a full asymptotic expansion by matched asymptotic expansions with two stretched variables, derives coupled recursive systems for the smooth and layer coefficients, and bounds each ter

Load-bearing premise

The analysis rests on the scaling ε1 ≪ ε2^2; if the two parameters are comparable, the claimed two-layer decomposition and remainder estimate are not proven.

Editorial extensions

If this is right

If correct, the theorem supplies the exact analytic regularity estimates needed to prove uniform exponential convergence of p/hp finite element methods for this two-parameter fourth-order problem.
The two distinct layer widths identified in the decomposition directly suggest how a graded mesh should be constructed near each endpoint for high-order discretizations.
The classical differentiability result (Theorem 1) confirms that analytic data yield an analytic solution, but derivative growth inherits negative powers of ε2 and of ε1/ε2, quantifying the loss of regularity as the parameters vanish.
The alternative one-parameter reformulation (Theorem 8) shows that in the regime ε1 ≪ ε2^2, the problem can also be treated by rescaling into a single-parameter reaction-diffusion form with ε2-dependent coefficients, giving a second route to similar bounds.
The remainder estimate e^{-δ/ε2} is exponentially small, so the truncated expansion can serve as a reliable approximation for post-processing or for constructing initial guesses in iterative solvers.

Reading between the lines

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

The complementary regime where ε1 is comparable to or larger than ε2^2 is not covered by the main theorem; there one likely expects a single dominant layer scale and a reduction to a one-parameter reaction-diffusion problem, but this would need a separate proof to be made rigorous.
The bound structure suggests that for constant-coefficient versions the exact solution should satisfy the same layer-width separation only when ε1/ε2^2 is small; checking the exact solution in that regime could serve as a direct test of the predicted factor (ε1/ε2)^{1-n}.
A natural extension would be to vector-valued or higher-order analogues: the same two-scale expansion and the coupled-ODE lemma might carry over, but the interplay of boundary conditions at the two endpoints would become more intricate.
The estimates are phrased in one dimension; transferring them to two-dimensional domains (e.g., a rectangle with layers along edges) would require tensor-product layer representations and would likely need additional assumptions on the geometry.

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.

Desk Editor's Note

read the letter

Short version: this paper addresses a genuine gap — analytic regularity for a fourth-order two-parameter SPP — but the main theorem as stated has the two layer widths backwards. The proof also skips the chain-rule factors when converting derivatives from stretched variables to x, and the remainder bound is asserted without a stability argument. These are fixable, but they are not cosmetic.

What is new and good: the problem is the right one, and the earlier results genuinely don't cover this case. The asymptotic expansion in §2.1 and the appendix is detailed, and Lemmas 5 and 6 follow the established Melenk/Constantinou–Xenophontos machinery in a plausible way. Lemma 4's handling of the coupled boundary conditions via Green's functions is a useful piece. Theorem 1 on classical differentiability is fine. The paper is also honest that it only treats ε1 ≪ ε2^2.

Now the problems. The stretched variables at the start of §2.1 are stated as ~x = x/ε2 and ¯x = xε1/ε2, while the appendix derives (9) using ¯x = xε2/ε1 — the reciprocal. That inconsistency alone is serious, but the deeper issue is that (8) gives ~u_{0,0} = C e^{-√(c0/b0) x/ε2}, a wide layer, and (9) gives ¯u_{0,0} = C e^{-√b0 x ε2/ε1}, a narrow layer. Theorem 7 assigns the narrow bound e^{-β ε2 x/ε1} to ~u^BL_M and the wide bound e^{-β x/ε2} to ¯u^BL_M. That is exactly the opposite of what the construction yields. The proof does not fix this: it bounds derivatives in the stretched variable and never applies the chain rule to convert to x-derivatives, so the factors ε2^{-n} and (ε1/ε2)^{1-n}/ε2 in the theorem do not follow from the displayed estimates. This matters for the intended hp-FEM application, since the mesh construction depends on which layer has which width.

The remainder bound in Theorem 7 is also asserted with “from which the desired result follows” after an L∞ estimate on L r_M. There is no stability estimate connecting the L∞ residual to the H^1 + ε1 H^2 energy norm plus the exponentially small boundary values. Coercivity on H^2_0 should make this work, but it needs to be written out.

What kind of paper is this? If the scaling issue is fixed, it is exactly the sort of result the hp-FEM people need. I would send it to a serious referee, but the referee should be asked to check the layer definitions carefully and to require a rewritten Theorem 7 with correct exponents and a real stability argument.

Referee Report

3 major / 6 minor

Summary. The paper considers the fourth-order two-parameter singularly perturbed boundary value problem (1)–(2) with analytic data and the standing assumption ε1 ≪ ε2^2. The authors construct an asymptotic decomposition u = u_M + \tilde u_M^BL + \bar u_M^BL + \check u_M^BL + \hat u_M^BL + r_M, prove algebraic/analytic estimates for the smooth coefficients (Lemmas 2–3), establish entire-function bounds for the boundary-layer profiles with the aid of a Green's-function lemma (Lemmas 4–6), and state in Theorem 7 that the layer terms satisfy two-width estimates with an exponentially small remainder. Theorem 8 offers an alternative one-parameter-type decomposition in which the wide layer is absorbed into the smooth part. The stated purpose is to provide the explicit, differentiation-order-dependent regularity needed for hp-FEM convergence proofs, and the paper is positioned as an extension of the one-parameter fourth-order results [4,5] and the two-parameter second-order results [17].

Significance. If the main theorem were correct, the paper would fill a genuine gap: existing one-parameter results do not cover fourth-order two-parameter problems in which two boundary-layer scales interact, and the derivative bounds are exactly of the form needed for hp-FEM analysis. The proof strategy — matched asymptotic expansions, Cauchy estimates, and a coupled Green's-function lemma — is plausible, and the constants are not fitted to data. However, the main theorem as stated is internally inconsistent with the construction in Section 2.1: the layer names and the two width regimes are interchanged, so the theorem's bounds do not apply to the functions actually defined by (8)–(10). In addition, the proof of the remainder bound in Theorem 7 stops at residual estimates without the required stability argument or a correct choice of M. These are load-bearing issues for the paper's central claim. The underlying approach appears repairable, but the current version cannot be used as a basis for the claimed hp-FEM application in [16].

major comments (3)

[§2.1, Eqs. (8)–(10), Theorem 7] The layer assignments in Theorem 7 are incompatible with the functions constructed in §2.1. The Appendix derives (8) using \tilde x = x/ε2, so the leading profile is e^{-√(c0/b0) x/ε2}; the derivation of (9) uses b(x)=b(ε1 \bar x/ε2), i.e. \bar x = x ε2/ε1, giving the leading profile e^{-√{b0} x ε2/ε1}. Thus \tilde u is the wide layer and \bar u is the narrow layer. Theorem 7, however, states the narrow-layer bound for \tilde u_M^BL and the wide-layer bound for \bar u_M^BL. Concretely, \tilde u_{0,0}^{BL} from (8),(10) is O(1)e^{-√(c0/b0)x/ε2}; at n=1 and x=ε2 its derivative is O(ε2^{-1})e^{-O(1)}, whereas the theorem's \tilde-bound is O(ε2^{-1})e^{-β ε2^2/ε1}, which is exponentially smaller under (5). The stated bound is therefore false for the constructed function. The theorem statement must be reconciled with (8)–(10) by assigning the correct width estimates to the correct layer funct
[§3, proof of Theorem 7] The remainder estimate is asserted rather than proved. After bounding \|L_{ε1,ε2} r_M\|∞,I by algebraic terms such as ε2(ε2 M K2)^M, (ε2\barγ eM)^{M+1}, etc., and the boundary values by C max{e^{-β/ε2}, e^{-β ε2/ε1}}, the proof says 'from which the desired result follows.' To obtain \|r_M\|_{∞,∂I}+ε2\|r'_M\|_{∞,∂I}+\|r_M\|_{1,I}+ε1\|r''_M\|_{0,I} ≤ C e^{-δ/ε2}, one needs (i) a stability estimate for the fourth-order operator with the small nonhomogeneous boundary data, and (ii) an explicit choice of M — typically M ∼ 1/ε2 with a sufficiently small proportionality constant — under which every residual term becomes exponentially small. The condition imposed on M (ε2 e^{24} M max{·}<1) only guarantees convergence of the geometric series and allows fixed M, for which the residual is algebraically small, not exponentially small. This missing argument is load-bearing for Theorem 7.
[Lemma 4] The proof of Lemma 4 solves a 4×4 system for v(0), w(0), v'(0), w'(0) and divides by κ−λ1. The admissible data include κ=λ1, e.g. b(0)=c(0)=1 gives λ1=√(c0/b0)=1 and κ=√b0=1. The underlying half-line boundary value problem is still well-posed in that case, but the displayed formulas degenerate. Since Lemma 4 is used to prove the entire-function estimates of Lemma 5, the proof should be amended either by treating κ=λ1 separately or by a limiting/continuity argument with bounds independent of |κ−λ1|^{-1}. Without this, the main theorem is not proved for all data satisfying (3)–(5).

minor comments (6)

[Title/Abstract] Typo in the title and abstract: 'boundary balue problem' should be 'boundary value problem'. Similar typos appear in the Conclusions ('regults') and in reference [16] ('Singulalry').
[§2.1] The displayed definition of the stretched variables gives \bar x = x ε1/ε2, but the text just before (9) and the Appendix use \bar x = x ε2/ε1 (since b(x)=b(ε1 \bar x/ε2)). This inconsistency must be fixed consistently; it is directly related to the major comment on the layer assignments.
[Theorem 7] The hypotheses on M are stated using e^{24} M max{·}<1, but the proof uses different exponential factors and separates the constants (e.g., ε2 4M \tildeγ e^2<1, (ε1/ε2^2)M \tildeγ e<1). Align the statement with the proof or explain the cruder sufficient condition.
[§3, Theorem 7 proof] The notation |(\tilde u_M^BL)^{(n)}(x)| is ambiguous because \tilde u_M^BL is defined as a function of \tilde x in (14). State explicitly that the derivative is with respect to x and indicate how powers of ε2 result from the chain rule; the current proof ends with bounds in the stretched variable.
[§4, Numerical illustration] Figure 1 uses ε1=ε2=10^{-2}, which does not satisfy the standing assumption ε1 ≪ ε2^2. If the figure is meant to illustrate the complementary regime, this should be stated; otherwise choose parameters satisfying (5). The numerical section only plots exact solutions and does not test the decomposition or the estimates of Theorem 7.
[§3, Theorem 8] Theorem 8 is proved only as a sketch: it invokes [5, Lemma 1] and 'a standard energy argument' without details. Since Theorem 8 is cited as being relevant for the numerical analysis, either include the missing steps or clearly refer to a complete treatment in [16].

Circularity Check

0 steps flagged · score 2.0 of 10

No significant circularity: the main estimates are derived from the recursive asymptotic layer problems; self-citations are technical, not load-bearing.

full rationale

The derivation chain is not circular. The smooth-part bounds (Lemmas 2-3) are proved by induction and Cauchy estimates; the layer bounds (Lemma 5) are proved by induction from the recursive problems (8)-(10), using Lemma 4 whose proof is given in the text via Green's function representations; Theorem 7 then derives the bounds for u_M, the four layer sums, and r_M directly from these lemmas and the geometric-series convergence conditions on M. No parameter is fitted to the solution, and no claimed prediction is a renamed fit. The self-citations to [4,5] are used for the one-parameter base case and for standard proof techniques, not for the two-parameter conclusion: Lemma 6 is a direct Cauchy-estimate corollary of Lemma 5, and Theorem 8 explicitly revisits the proofs in [5] rather than importing the target theorem. There is no imported uniqueness theorem, and no ansatz is justified solely by the authors' prior work. I do note an internal correctness issue, unrelated to circularity: Section 2.1 defines the stretched variable as x-bar = x epsilon1 / epsilon2, whereas the derivation of (9) and the surrounding text use x-bar = x epsilon2 / epsilon1; this could affect the attribution of layer widths in Theorem 7, but it is an inconsistency in the derivation, not a reduction of the theorem to its own inputs.

Assumptions & free parameters 0 free parameters · 5 assumptions · 0 invented entities

No entities are invented. The free-parameter list is empty: constants in the proofs are existential and not fitted to data. The main load-bearing assumptions are analyticity, positivity, the regime ε1 ≪ ε2^2, and the validity of the matched asymptotic expansion ansatz.

assumptions (5)

domain assumption Analyticity of input data b, c, f with explicit derivative bounds ||f^{(n)}||∞ ≲ n! γ_f^n, and similarly for b and c (equation (3)).
Essential for the analytic regularity claims; all derivative estimates depend on these bounds.
domain assumption Positive lower bounds b(x) ≥ b* > 0, c(x) ≥ c* > 0 (equation (4)).
Ensures well-posedness and exponential decay of boundary layers; used to define β = sqrt(˜c0/˜b0) and sqrt(¯b0).
domain assumption Parameter regime ε1 ≪ ε2^2 (equation (5)).
The entire analysis is restricted to this scaling; the complementary case is not analyzed.
standard math The solution admits the decomposition into smooth and boundary-layer parts with the indicated bounds for n=1,2,3, quoted from O'Malley [12,13].
Used as the base case for Theorem 1's induction and for the formal decomposition.
domain assumption The formal matched asymptotic expansion ansatz (6) u ∼ Σ ε2^i (ε1/ε2^2)^j (u_i,j + layer terms) is legitimate.
The expansion is used to derive the recursive systems (7)–(9); its validity is only verified a posteriori through the remainder bound, which itself has a derivation gap.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Analytic regularity for a fourth-order singularly perturbed boundary balue problem with two small parameters." pith.science (2026). https://pith.science/paper/SR7UO266

@misc{pith2026260716781,
  author       = {Pith},
  title        = {Pith review of: Analytic regularity for a fourth-order singularly perturbed boundary balue problem with two small parameters},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/SR7UO266}},
  note         = {Machine review of arXiv:2607.16781}
}

abstract

We consider a fourth order singularly perturbed boundary value problem with two small parameters, in one dimension, under the assumption of analytic input data. We show that the solution may be decomposed into a smooth part, two different width boundary layers, and a negligible remainder. We provide estimates for arbitrary order derivatives of each term of the decomposition, which are explicit in the differentiation order and the singular perturbation parameters, and are needed for proving the convergence of high order numerical methods, such as the $p/hp$ versions of the Finite Element Method. We also provide classical differentiability results, which show that the solution will be analytic, if the data are analytic, but negative powers of the singular perturbation parameter(s) show up once we start differentiating.

Figures

Figures reproduced from arXiv: 2607.16781 by the authors.

**Figure 2.** The exact solution of (1), (2) and its derivative, with [PITH_FULL_IMAGE:figures/full_fig_p027_2.png] view at source ↗

**Figure 3.** The exact solution of (1), (2) and its derivative, with [PITH_FULL_IMAGE:figures/full_fig_p028_3.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

18 extracted references · 4 canonical work pages

[17]

Sykopetritou and C

I. Sykopetritou and C. Xenophontos, Analytic Regularity for a Singularly Perturbed Reaction-Convection-Diﬀusion Boundary Value Problem wit h Two Small Parame- ters, Commun. Math. Res., 40 (2024) 125–153

2024
[16]

Sykopetritou, The hp Finite Element Method for Singulalry Perturbed probl ems with two small parameters , Ph

I. Sykopetritou, The hp Finite Element Method for Singulalry Perturbed probl ems with two small parameters , Ph. D. Dissertation, Department of Mathematics and Statistics, University of Cyprus, 2025

2025
[1]

C. J. Amnick and J. B. McLeod, A Singular Perturbation Problem in Needle Crystals , Arch. Rational Mech. Anal., 109 (1990) 139–171

1990
[2]

Capriz, and G

G. Capriz, and G. Cimatti, On some singular perturbation problems in the theory of lubrication , Appl. Math. Optim., 4 285–297 (1977). https://doi.org/10.1007/BF01442145

work page doi:10.1007/bf01442145 1977
[3]

G. F. Carrier, Singular Perturbation Theory and Geophysics , SIAM Review, 12 (1970) 175–193, https://doi.org/10.1137/1012041

work page doi:10.1137/1012041 1970
[4]

P. Constantinou, The hp Finite Element Method for fourth order singularly per - turbed problems, Doctoral Dissertation, Department of Mathematics and Statist ics, University of Cyprus, (2019)

2019
[5]

Analytic regularity for a singularly perturbed fourth order reaction-diffusion boundary value problem

P. Constantinou and C. Xenophontos, Analytic regularity for a sin- gularly perturbed 4th order reaction-diﬀusion boundary value problem , https://doi.org/10.48550/arXiv.2311.06849 (2023)

work page Pith review arXiv doi:10.48550/arxiv.2311.06849 2023
[6]

L. S. Frank, On a Singular Perturbation in the Kinetic Theory of Enzymes and the Corresponding Nonlinear Time Optimal Control , In: Bensoussan, A., Lions, J.L. (eds) Analysis and Optimization of Systems. Lecture N otes in Control and Information Sciences, vol 83. Springer, Berlin, Heid elberg. https://doi.org/10.1007/BFb0007598

work page doi:10.1007/bfb0007598

Show all 18 references

[7]

Kokotovic, R.E

P.V. Kokotovic, R.E. O’Malley Jr. and P. Sannuti, Singular perturbations and order reduction in control theory – An overview , Automatica, 12 (1976) 123–132

1976
[8]

C. G. Lange, H. J. Weinitschke, Singular Perturbations of Limit Points with Ap- plication to Tubular Chemical Reactors , Studies in Applied Mathematics, 84 7–42 (1991)

1991
[9]

J. M. Melenk, hp Finite Elements for Singular Perturbations , Lecture Notes in Math- ematics 1987, Springer-Verlag, 2002

1987
[10]

J. J. H. Miller, E. O’Riordan and G. I. Shishkin, Fitted Numerical Methods Singular Perturbation Problems, World Scientiﬁc, 1996

1996
[11]

K. W. Morton, Numerical Solution of Convection-Diﬀusion Problems , Volume 12 of Applied Mathematics and Mathematical Computation, Chapman & Hall, 1996

1996
[12]

R. E. O’Malley, Introduction to Singular Perturbations , Academic Press, New York, 1974. 33

1974
[13]

R. E. O’Malley, Singular perturbations of boundary value problems for line ar ordinary diﬀerential equations involving two parameters , J. Math. Anal. Appl. 19 (1967) 291– 308

1967
[14]

R. E. O’Malley (ed.), The Method of Matched Asymptotic Expansions and Its Gener- alizations, Historical Developments in Singular Perturbations, Springer Inte rnational Publishing, pp. 53–121, ISBN 978-3-319-11924-3, 2014

2014
[15]

H.-G. Roos, M. Stynes, and L. Tobiska. Robust numerical methods for singularly per- turbed diﬀerential equations, volume 24 of Springer Series in Computational Mathe- matics. Springer-Verlag, Berlin, second edition, 2008. Convection-diﬀus ion-reaction and ﬂow problems

2008
[18]

Xenophontos, S

C. Xenophontos, S. Franz and I. Sykopetritou, A mixed hp FEM for fourth order singularly perturbed boundary value problems with two smal l parameters , Numer. Meth. PDEs, 38, (2022) 1543–1555. 34

2022

Pith tools

Reviewed August 1, 2026 · model on record in the stance chip above.

[1] [17]

Sykopetritou and C

I. Sykopetritou and C. Xenophontos, Analytic Regularity for a Singularly Perturbed Reaction-Convection-Diﬀusion Boundary Value Problem wit h Two Small Parame- ters, Commun. Math. Res., 40 (2024) 125–153

2024

[2] [16]

Sykopetritou, The hp Finite Element Method for Singulalry Perturbed probl ems with two small parameters , Ph

I. Sykopetritou, The hp Finite Element Method for Singulalry Perturbed probl ems with two small parameters , Ph. D. Dissertation, Department of Mathematics and Statistics, University of Cyprus, 2025

2025

[3] [1]

C. J. Amnick and J. B. McLeod, A Singular Perturbation Problem in Needle Crystals , Arch. Rational Mech. Anal., 109 (1990) 139–171

1990

[4] [2]

Capriz, and G

G. Capriz, and G. Cimatti, On some singular perturbation problems in the theory of lubrication , Appl. Math. Optim., 4 285–297 (1977). https://doi.org/10.1007/BF01442145

work page doi:10.1007/bf01442145 1977

[5] [3]

G. F. Carrier, Singular Perturbation Theory and Geophysics , SIAM Review, 12 (1970) 175–193, https://doi.org/10.1137/1012041

work page doi:10.1137/1012041 1970

[6] [4]

P. Constantinou, The hp Finite Element Method for fourth order singularly per - turbed problems, Doctoral Dissertation, Department of Mathematics and Statist ics, University of Cyprus, (2019)

2019

[7] [5]

Analytic regularity for a singularly perturbed fourth order reaction-diffusion boundary value problem

P. Constantinou and C. Xenophontos, Analytic regularity for a sin- gularly perturbed 4th order reaction-diﬀusion boundary value problem , https://doi.org/10.48550/arXiv.2311.06849 (2023)

work page Pith review arXiv doi:10.48550/arxiv.2311.06849 2023

[8] [6]

L. S. Frank, On a Singular Perturbation in the Kinetic Theory of Enzymes and the Corresponding Nonlinear Time Optimal Control , In: Bensoussan, A., Lions, J.L. (eds) Analysis and Optimization of Systems. Lecture N otes in Control and Information Sciences, vol 83. Springer, Berlin, Heid elberg. https://doi.org/10.1007/BFb0007598

work page doi:10.1007/bfb0007598