Kerr-Dold vortices in an axisymmetric stagnation point flow

Prabakaran Rajamanickam

REVIEW 3 major objections 5 minor 24 references

Kerr-Dold vortices in an axisymmetric stagnation point flow

T0 review · 3 major / 5 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read This paper shows that Kerr–Dold vortices—periodic counter-rotating vortex arrays with algebraically decaying tails—exist in axisymmetric stagnation point flow.

desk verdict Clean linear construction of Kerr-Dold-type vortices in axisymmetric stagnation flow, but the finite-Reynolds-number existence claim rests on an undocumented numerical continuation. read the letter →

arxiv 2506.07101 v2 pith:SOHNGAQI submitted 2025-06-08 physics.flu-dyn

Prabakaran Rajamanickam This is my paper

classification physics.flu-dyn

keywords Kerr–DoldvorticesaxisymmetricstagnationpointflowthickvortexstructuresBurgersmodifiedBesselfunctionsReynoldsnumberNavier–Stokesequationsalgebraicvorticitydecay

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 establishes that Kerr–Dold vortices—periodic arrays of counter-rotating, algebraically decaying (thick) stretched vortices previously known in planar stagnation flows—also exist in axisymmetric stagnation point flow. For small vortex Reynolds numbers the author derives explicit closed-form solutions for vorticity and streamfunction in terms of modified Bessel functions, parameterized by a sequence of Fourier amplitudes. For finite Reynolds numbers it is asserted that these linear solutions continue numerically, with computed vorticity contours shown for N = 2, 3, 4 branches. If correct, this adds a new family of steady thick vortex solutions to the Navier–Stokes equations alongside the classical Burgers vortex and vortex sheet.

What carries the argument

The construction is carried by reducing the steady nonlinear equations to a linear system at small $Re$, expanding vorticity $\omega(r,\theta)$ and streamfunction $\psi(r,\theta)$ in azimuthal Fourier harmonics $e^{\pm inN\theta}$. Each harmonic satisfies a decoupled second-order ODE; the solution regular at $r=0$ is a combination of modified Bessel functions $I_{(nN\pm1)/2}(r^2/4)$, and the streamfunction is obtained in closed form by variation of parameters. The far-field function $f(\theta)$ encodes the amplitude coefficients $\{A_n\}$, and the algebraic $r^{-2}$ tail is what makes these thick rather than thin vortices. Because the circulation diverges, the vortex Reynolds number is fixed by a normalization condition rather than by circulation.

What would settle it

Run a grid-converged numerical solve for the $N=4$, $Re=30$ branch starting from the linear solution of Section 3; if no nearby steady solution exists, the numerical-continuation claim behind the finite-$Re$ existence statement is false.

Watch

Extended reading notes

Core claim

Superposing a periodic disturbance on the axisymmetric stagnation flow $v_r=-r$, $v_z=2z$, the paper seeks steady solutions with azimuthal period $2\pi/N$, $N\ge 2$, and homogeneous boundary conditions at $r=0$ and infinity. At leading order in the vortex Reynolds number $Re$ the Fourier modes decouple, and the regular mode of the linearized vorticity equation is solved exactly in terms of modified Bessel functions; the streamfunction follows by variation of parameters. The vorticity decays as $r^{-2}$ at infinity, making the structure thick: the circulation around one vortex of the pair diverges logarithmically with domain radius. The paper asserts that this family, parameterized by Fourier amplitudes $\{A_n\}$ with $A_1\neq 0$, can be continued numerically to finite $Re$, with representative vorticity contours shown for $N=2,3,4$.

Load-bearing premise

The finite-$Re$ existence claim rests on numerical continuation that is described only by a sentence and a figure, with no numerical method, grid, convergence criteria, or error estimates reported.

Editorial extensions

If this is right

The axisymmetric stagnation point flow supports a previously missing class of steady thick vortex solutions, extending the planar Kerr–Dold family to a flow with both stretching and axial straining.
For small $Re$ the solutions are explicit: any choice of Fourier amplitudes $\{A_n\}$ with $A_1\neq 0$ gives a closed-form vorticity and streamfunction, so the family is as broad as the periodic far-field function $f(\theta)$.
The $N=1$ mode is excluded on physical grounds, since it would leave non-vanishing velocity at the symmetry axis.
Because the circulation of an individual vortex diverges logarithmically, the vortex Reynolds number cannot be defined through circulation; a normalization condition is required.
The numerical continuation shown in Figure 1 indicates that representative branches ($N=2,3,4$) survive at finite $Re$, at least for the amplitude choice $A_1=-i$.

Reading between the lines

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

A natural next step the paper does not take is a linear stability analysis; depending on the outcome, these thick vortices could be stable like Burgers vortices or unstable like Burgers vortex sheets, which would determine whether they are observable in experiments.
The far-field algebraic tail means numerical and experimental studies must handle a large or truncated domain carefully; the log-divergent circulation could be used as a diagnostic for how well a computation captures the true boundary condition at infinity.
By analogy with Kerr's planar construction, choices of $\{A_n\}$ beyond the single-mode $A_1$ branch should produce a wider zoo of axisymmetric Kerr–Dold vortices, including asymmetric or multi-vortex patterns, though the paper does not compute these.
The closed-form small-$Re$ solution could serve as an initial guess for continuation codes or as a benchmark for axisymmetric Navier–Stokes solvers.

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

The analytical core of this paper is solid and genuinely new. The author constructs explicit closed-form solutions (3.3) and (3.5) for Kerr-Dold-type vortices in an axisymmetric stagnation point flow, extending a family that previously existed only in planar geometry. I checked the linearized equations and the boundary behavior; they follow cleanly from the Navier-Stokes ansatz, and the N=1 exclusion and logarithmic divergence of circulation are handled correctly. The normalization condition (2.11) is a sensible way to define the vortex Reynolds number without relying on circulation. The free amplitudes A_n give an infinite family of solutions, consistent with Kerr's earlier work on manipulating vortex tails. This part deserves credit: it is a useful, citable construction for anyone working with vortex structures in stagnation flows.

The soft spot is exactly where the reader and stress-test put it. The abstract says existence is 'demonstrated' for Kerr-Dold-type vortices, and the summary repeats that. But for finite Reynolds numbers, the only evidence is a sentence in Section 3 saying the linear solutions 'can be continued numerically' plus Figure 1. There is no numerical method, no grid, no continuation scheme, no convergence criterion, no residuals. I cannot tell from the paper whether the plotted vorticity contours are genuine steady solutions of (2.4)-(2.5) or artifacts of an under-resolved computation. This is a missing-support problem, not an internal inconsistency, but it is load-bearing: the closed-form solutions are only leading-order in Re, and the central claim is that they persist.

The citation pattern is appropriate. The self-citations are relevant and modest. The paper is honest in its scope: it explicitly says it is showcasing a few computed examples rather than an exhaustive catalogue. If the author adds a brief description of the numerical scheme, a grid-resolution check, and perhaps a residual measure, the main gap closes. Even without that, the analytical contribution stands on its own.

This paper deserves a serious referee. It is short, clear, and the mathematics is correct. The referee should require the missing numerical details before accepting the finite-Re claim, but that is a normal revision request, not grounds for rejection. I would cite the closed-form formulas in my own work, and I'd bring it to a reading group that cares about vortex solutions, though it is niche enough that I would not insist.

Send it to peer review. The analytical result is worth publishing; the numerical claim needs to be made reproducible.

Referee Report

3 major / 5 minor

Summary. The paper studies steady axisymmetric stagnation-point flow with a superposed periodic array of vortex perturbations. After formulating the governing equations (2.4)-(2.5) in terms of a vortex streamfunction and vorticity, it linearizes the problem for small vortex Reynolds number Re and derives explicit Fourier-mode solutions for the vorticity and streamfunction, Eqs. (3.3) and (3.5), parameterized by arbitrary complex amplitudes A_n with A_1 nonzero. It then asserts that these linear solutions can be continued numerically to finite Reynolds numbers, presenting vorticity contours in Figure 1 for N=2,3,4. The central claim is that Kerr-Dold-type thick vortex solutions exist in axisymmetric stagnation-point flow, thereby extending the known class of such solutions.

Significance. If the finite-Reynolds-number continuation is genuine, the paper extends the Kerr-Dold family to an axisymmetric stagnation-point geometry and provides a new class of algebraically decaying vortex solutions. The small-Re analytical construction is explicit, internally consistent, and appears correct: the Fourier modes decouple cleanly after linearization, and the closed forms (3.3) and (3.5) have the expected boundary behavior (2.8)-(2.9). The family parameterized by the sequence {A_n} is a useful addition. The main weakness is that the finite-Re existence claim rests entirely on an undocumented numerical continuation and contour plots; this is a missing-support problem rather than an internal inconsistency.

major comments (3)

[Section 3, final paragraph; Fig. 1] The central claim that the linear solutions can be continued to finite Reynolds numbers is supported only by the sentence "These linear solutions can be continued numerically to finite Reynolds numbers" and by the vorticity contours in Figure 1. No numerical method is described: there is no discretization, no iteration or continuation scheme, no domain truncation, no boundary treatment, no grid resolution, and no convergence or residual criterion. I cannot verify that the plotted states are genuine solutions of (2.4)-(2.5) or that they are connected to the Re=0 linear branch. Please provide a full description of the numerical method, the computational domain, the discretization, the continuation procedure, and quantitative checks such as residual norms or grid-convergence studies, ideally with code or data.
[Fig. 1 and the A1=-i branch] The figure gives no quantitative validation. At Re=1 the numerical solution should be close to the explicit linear solution (3.3)-(3.5) with A1=-i and A2=A3=...=0; please provide a quantitative comparison, for example error norms or overlaid contour lines, together with a grid-convergence study. Without this, the plotted contours cannot be distinguished from numerical artifacts, and the claimed persistence to Re=10, 20, and 30 is not established.
[Eq. (2.11) and Section 3] The normalization condition (2.11) defines the vortex Reynolds number, but its numerical implementation is not described. It is unclear whether (2.11) is enforced exactly in the computation, how the amplitude A1=-i is set during continuation, and how the reported Reynolds numbers correspond to the normalization. Because the meaning of the values Re=1, 10, 20, and 30 in Figure 1 depends on this choice, the missing implementation detail affects the interpretation of all numerical results.

minor comments (5)

[Eq. (2.2)] The symbol Re is used both for the vortex Reynolds number and for the operation of taking the real part. This overloaded notation is confusing; for example, use a different font for the real-part operator or rename the Reynolds number.
[Eqs. (3.4)-(3.5)] The reduction from the integral form (3.4) to the closed form (3.5) is stated without showing the relevant Bessel function identities; please include the key identity or a citation so that the closed form can be independently verified.
[Section 3, first paragraph] The phrase "Since the problem is linear" is imprecise: the full problem (2.4)-(2.5) is nonlinear, and the Fourier decoupling occurs only for the leading-order small-Re linearized system. Please rephrase to avoid implying that the nonlinear equations decouple.
[Eq. (2.11)] The integrand written as (omega/r) r dr dtheta is dimensionally and algebraically just omega dr dtheta; the current form is unnecessarily confusing and could be simplified.
[Fig. 1 caption] The caption should state the computational domain, contour levels, and the numerical method used to produce the panels, as none of these details are currently provided.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the linear solution family is derived from the governing equations, and the finite-Reynolds continuation claim, though under-supported, is not circular.

full rationale

The central linear construction in Section 3 is self-contained: the paper starts from the axisymmetric Navier-Stokes reductions (2.4)-(2.5), imposes the boundary conditions (2.6)-(2.7) and regularity at r=0, solves the decoupled Fourier ODEs (3.2), and obtains the closed-form amplitudes (3.3) and streamfunctions (3.5). No target quantity is fitted: the constants A_n parametrize the family, and the normalization (2.11) fixes a convention for the vortex Reynolds number rather than being tuned to a desired output. The finite-Reynolds existence claim is asserted by the sentence 'These linear solutions can be continued numerically to finite Reynolds numbers' at the end of Section 3, with Figure 1 as illustration but no numerical method, grid, or convergence check described; this is a missing-support or reproducibility gap, not circularity, because the claim is not assumed as an input and does not reduce to the linear solutions by definition. The only self-citation (reference 10, Rajamanickam and Weiss) is background on radial stagnation-point vortices and is not load-bearing. No uniqueness theorem is imported to force the construction, and the solutions are not a renamed version of a known result in a new coordinate system. Accordingly, no circular step is exhibited.

Assumptions & free parameters 1 free parameters · 4 assumptions · 0 invented entities

The main assumptions are standard mathematical tools and the specific flow ansatz. The only non-standard and load-bearing assumption is the numerical continuation to finite Reynolds numbers, which is not documented. No new physical entities are introduced.

free parameters (1)

Vortex amplitudes A_n (with A_1 ≠ 0) = A1 = -i in the illustrative branch; otherwise arbitrary
These complex constants parametrize the family of linear solutions and are not fitted to data; the existence claim holds for any choice with A1 ≠ 0.

assumptions (4)

standard math Navier-Stokes equations govern the incompressible fluid flow
The entire derivation assumes the Navier-Stokes equations are the correct model, which is standard in fluid mechanics.
domain assumption The vortex perturbation is independent of z and periodic in θ with period 2π/N
The solution ansatz (2.2) restricts to perturbations that do not depend on the axial coordinate and are periodic in the azimuthal direction, which is the class of solutions being sought.
standard math Fourier series in θ converge and can be differentiated termwise
The linear solutions use a Fourier series representation (3.1), assuming standard convergence properties for the periodic functions considered.
ad hoc to paper The linear solutions can be continued to finite Reynolds numbers
The paper states this continuation but provides no proof or numerical details, so it is an unverified assumption on which the finite-Re existence claim rests.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Kerr-Dold vortices in an axisymmetric stagnation point flow." pith.science (2026). https://pith.science/paper/SOHNGAQI

@misc{pith2026250607101,
  author       = {Pith},
  title        = {Pith review of: Kerr-Dold vortices in an axisymmetric stagnation point flow},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/SOHNGAQI}},
  note         = {Machine review of arXiv:2506.07101}
}

read the original abstract

The existence of Kerr-Dold-type vortices in axisymmetric stagnation point flow is demonstrated, extending the class of known thick vortex solutions.

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

24 extracted references · 24 canonical work pages

[1]

H. K. Moﬀatt, S. Kida, and K. Ohkitani. Stretched vortices–the sin ews of turbulence; large-Reynolds-number asymptotics. J. Fluid Mech. , 259:241–264, 1994

work page 1994
[2]

Leibovich and P

S. Leibovich and P. Holmes. Global stability of the Burgers’ vortex. Phys. Fluids , 24(3):548–549, 1981

work page 1981
[3]

A. C. Robinson and P. G. Saﬀman. Stability and structure of stretc hed vortices. Stud. Appl. Math. , 70(2):163–181, 1984

work page 1984
[4]

Prochazka and D

A. Prochazka and D. I. Pullin. On the two-dimensional stability of the axisymmetric Burgers vortex. Phys. Fluids , 7(7):1788–1790, 1995

work page 1995
[5]

D. G. Crowdy. A note on the linear stability of Burgers vortex. Stud. Appl. Math. , 100(2):107–126, 1998

work page 1998
[6]

P. J. Schmid and M. Rossi. Three-dimensional stability of a Burgers v ortex. J. Fluid Mech., 500:103–112, 2004

work page 2004
[7]

Y. A. Aryshev, V. A. Golovin, and S. A. Ershin. Stability of colliding ﬂow s. Fluid Dyn. , 16(5):755–759, 1981

work page 1981
[8]

R. R. Lagnado, N. Phan-Thien, and L. G. Leal. The stability of two-d imensional linear ﬂows. Phys. Fluids , 27(5):1094–1101, 1984

work page 1984

Show all 24 references

[9]

K. N. Beronov and S. Kida. Linear two-dimensional stability of a Burg ers vortex layer. Phys. Fluids , 8(4):1024–1035, 1996

1996
[10]

Rajamanickam and A

P. Rajamanickam and A. D. Weiss. Steady axisymmetric vortices in ra dial stagnation ﬂows. Q. J. Mech. Appl. Math. , 74(3):367–378, 2021

2021
[11]

Prochazka and D

A. Prochazka and D. I. Pullin. Structure and stability of non-symme tric Burgers vortices. J. Fluid Mech. , 363:199–228, 1998

1998
[12]

O. S. Kerr and J. W. Dold. Periodic steady vortices in a stagnation-p oint ﬂow. J. Fluid Mech., 276:307–325, 1994. 6 P. Rajamanickam

1994
[13]

P. G. Drazin and N. Riley. The Navier–Stokes equations: a classiﬁcation of ﬂows and exact solutions . Number 334. Cambridge University Press, 2006

2006
[14]

R. R. Lagnado and L. G. Leal. Visualization of three-dimensional ﬂow in a four-roll mill. Exp. Fluids , 9(1-2):25–32, 1990

1990
[15]

Andreotti, S

B. Andreotti, S. Douady, and Y. Couder. An experiment on two asp ects of the interaction between strain and vorticity. J. Fluid Mech. , 444:151–174, 2001

2001
[16]

V. N. Kalashnikov and M. G. Tsiklauri. Ordered three-dimensional st ructures resulting from instability of two-dimensional ﬂow in crossed channels. Fluid Dyn. , 26(2):161– 165, 1991

1991
[17]

Burshtein, A

N. Burshtein, A. Q. Shen, and S. J. Haward. Controlled symmetry b reaking and vortex dynamics in intersecting ﬂows. Phys. Fluids , 31(3), 2019

2019
[18]

Burshtein, K

N. Burshtein, K. Zografos, A. Q. Shen, R. J. Poole, and S. J. Hawa rd. Periodic ﬂuctuations of streamwise vortices in inertia-dominated intersect ing ﬂows. Phys. Fluids, 33(1), 2021

2021
[19]

O. S. Kerr. Periodic steady vortices in a stagnation point ﬂow II. arXiv preprint arXiv:2409.09695, 2024

2024 arXiv
[20]

S. Towns. Roll-Up of a Burgers vortex layer . The University of Manchester (United Kingdom), 1999

1999
[21]

Xiong and S

Z. Xiong and S. K. Lele. Distortion of upstream disturbances in a Hiem enz boundary layer. J. Fluid Mech.s , 519:201–232, 2004

2004
[22]

A Migdal

A. A Migdal. Asymmetric vortex sheet. Phys. Fluids , 33(3), 2021

2021
[23]

Shariﬀ and G

K. Shariﬀ and G. E. Elsinga. Viscous vortex layers subject to more g eneral strain and comparison to isotropic turbulence. Phys. Fluids , 33(3), 2021

2021
[24]

M. Z. Bazant and H. K. Moﬀatt. Exact solutions of the Navier–Stok es equations having steady vortex structures. J. Fluid Mech. , 541:55–64, 2005

2005

Pith tools

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

[1] [1]

H. K. Moﬀatt, S. Kida, and K. Ohkitani. Stretched vortices–the sin ews of turbulence; large-Reynolds-number asymptotics. J. Fluid Mech. , 259:241–264, 1994

work page 1994

[2] [2]

Leibovich and P

S. Leibovich and P. Holmes. Global stability of the Burgers’ vortex. Phys. Fluids , 24(3):548–549, 1981

work page 1981

[3] [3]

A. C. Robinson and P. G. Saﬀman. Stability and structure of stretc hed vortices. Stud. Appl. Math. , 70(2):163–181, 1984

work page 1984

[4] [4]

Prochazka and D

A. Prochazka and D. I. Pullin. On the two-dimensional stability of the axisymmetric Burgers vortex. Phys. Fluids , 7(7):1788–1790, 1995

work page 1995

[5] [5]

D. G. Crowdy. A note on the linear stability of Burgers vortex. Stud. Appl. Math. , 100(2):107–126, 1998

work page 1998

[6] [6]

P. J. Schmid and M. Rossi. Three-dimensional stability of a Burgers v ortex. J. Fluid Mech., 500:103–112, 2004

work page 2004

[7] [7]

Y. A. Aryshev, V. A. Golovin, and S. A. Ershin. Stability of colliding ﬂow s. Fluid Dyn. , 16(5):755–759, 1981

work page 1981

[8] [8]

R. R. Lagnado, N. Phan-Thien, and L. G. Leal. The stability of two-d imensional linear ﬂows. Phys. Fluids , 27(5):1094–1101, 1984

work page 1984