Abstract
We construct solutions of the magnetohydrostatic (MHS) equations in bounded domains and on the torus in three spatial dimensions, as infinite time limits of Voigt approximations of viscous, non-resistive incompressible magnetohydrodynamics equations. The Voigt approximations modify the time evolution without introducing artificial viscosity. We show that the obtained MHS solutions are regular, nontrivial, and are not Beltrami fields.
Similar content being viewed by others
Data Availability
Data sharing not applicable to this article as no datasets were generated or analysed during the current study.
Notes
We use the following definition, whenever \(s \in {\mathbb {R}}\): \(D((-\Delta )^{s/2}):= \{v \in ({\dot{H}^s})^3: \text { div } v = 0\}\).
F.P. thanks Tarek Elgindi for pointing out this example.
We say that \(f \in D((-\Delta )^{\frac{\beta -}{2}})\) if for all sufficiently small \(\eta > 0\), \(f \in D((-\Delta )^\frac{\beta - \eta }{2})\).
These formal estimates can be made rigorous by Galerkin approximation.
References
Alfvén, H.: Existence of electromagnetic-hydrodynamic waves. Nature 150(3805), 405–406 (1942)
Batchelor, G.K., Moffatt, H.K., Worster, M.G. (eds.): Perspectives in Fluid Dynamics: A Collective Introduction to Current Research. Cambridge University Press, Cambridge (2000)
Beekie, R., Friedlander, S., Vicol, V.: On Moffatt’s magnetic relaxation equations. Comm. Math. Phys. 390(3), 1311–1339 (2022)
Biskamp, D.: Nonlinear Magnetohydrodynamics. Cambridge Monographs on Plasma Physics, vol. 1. Cambridge University Press, Cambridge (1993)
Boyer, F., Fabrie, P.: Mathematical Tools for the Study of the Incompressible Navier–Stokes Equations and Related Models. Applied Mathematical Sciences, vol. 183. Springer, New York (2013)
Brenier, Y.: Topology-preserving diffusion of divergence-free vector fields and magnetic relaxation. Comm. Math. Phys. 330(2), 757–770 (2014)
Bruno, O.P., Laurence, P.: Existence of three-dimensional toroidal MHD equilibria with nonconstant pressure. Comm. Pure Appl. Math. 49(7), 717–764 (1996)
Cao, Y., Lunasin, E.M., Titi, E.S.: Global well-posedness of the three-dimensional viscous and inviscid simplified Bardina turbulence models. Commun. Math. Sci. 4(4), 823–848 (2006)
Constantin, P., Foias, C.: Navier–Stokes equations. University of Chicago Press, Chicago (1988)
Constantin, P., La, J., Vicol, V.: Remarks on a paper by Gavrilov: Grad–Shafranov equations, steady solutions of the three dimensional incompressible Euler equations with compactly supported velocities, and applications. Geom. Funct. Anal. 29(6), 1773–1793 (2019)
Constantin, P., Drivas, T.D., Ginsberg, D.: Flexibility and rigidity in steady fluid motion. Comm. Math. Phys. 385(1), 521–563 (2021)
Constantin, P., Drivas, T.D., Ginsberg, D.: On quasisymmetric plasma equilibria sustained by small force. J. Plasma Phys. 87(1), 905870111 (2021)
Elgindi, T.M., Masmoudi, N.: \(L^\infty \) ill-posedness for a class of equations arising in hydrodynamics. Arch. Ration. Mech. Anal. 235(3), 1979–2025 (2020)
Enciso, A., Luque, A., Peralta-Salas, D.: MHD equilibria with nonconstant pressure in nondegenerate toroidal domains. Preprint: https://arxiv.org/abs/2104.08149 (2021)
Gavrilov, A.V.: A steady Euler flow with compact support. Geom. Funct. Anal. 29(1), 190–197 (2019)
Hill, M.J.M.: On a spherical vortex. Philos. Trans. R. Soc. London A 185, 213–245 (1894)
Kiselev, A., Šverák, V.: Small scale creation for solutions of the incompressible two-dimensional Euler equation. Ann. of Math. (2) 180(3), 1205–1220 (2014)
Kraus, B.F., Hudson, S.R.: Theory and discretization of ideal magnetohydrodynamic equilibria with fractal pressure profiles. Phys. Plasmas 24(9), 092519 (2017)
Kruskal, M.D., Kulsrud, R.M.: Equilibrium of a magnetically confined plasma in a toroid. Phys. Fluids 1, 265–274 (1958)
Larios, A., Titi, E.: On the higher-order global regularity of the inviscid Voigt-regularization of three-dimensional hydrodynamic models. Discr. Contin. Dyn. Syst. B 14, 1531–3492 (2010)
Larios, A., Titi, E.S.: Higher-order global regularity of an inviscid Voigt-regularization of the three-dimensional inviscid resistive magnetohydrodynamic equations. J. Math. Fluid Mech. 16(1), 59–76 (2014)
Levant, B., Ramos, F., Titi, E.S.: On the statistical properties of the 3D incompressible Navier–Stokes–Voigt model. Commun. Math. Sci. 8(1), 277–293 (2010)
Linshiz, J.S., Titi, E.S.: Analytical study of certain magnetohydrodynamic-\(\alpha \) models. J. Math. Phys. 48(6), 065504 (2007)
Moffatt, H.K.: Magnetostatic equilibria and analogous Euler flows of arbitrarily complex topology. I. Fundamentals. J. Fluid Mech. 159, 359–378 (1985)
Moffatt, H.K.: Magnetostatic equilibria and analogous Euler flows of arbitrarily complex topology. Part 2. Stability considerations. J. Fluid Mech. 166, 359–378 (1986)
Oskolkov, A. P. : The uniqueness and solvability in the large of boundary value problems for the equations of motion of aqueous solutions of polymers. Zap. Naučn. Sem. Leningrad. Otdel. Mat. Inst. Steklov. (LOMI), 38:98–136, 1973. Boundary value problems of mathematical physics and related questions in the theory of functions, 7
Pasqualotto, F.: Nonlinear Waves in General Relativity and Fluid Dynamics. PhD thesis, Princeton University, 2020
Ramos, F., Titi, E.S.: Invariant measures for the 3D Navier–Stokes-Voigt equations and their Navier–Stokes limit. Discrete Contin. Dyn. Syst. 28(1), 375–403 (2010)
Taylor, J.B.: Relaxation and magnetic reconnection in plasmas. Rev. Mod. Phys. 58, 741–763 (1986)
Vaĭnshteĭn, S.I.: Force-free magnetic fields with constant alpha. In: Topological Aspects of the Dynamics of Fluids and Plasmas (Santa Barbara, CA, 1991). NATO Adv. Sci. Inst. Ser. E: Appl. Sci., vol. 218, pp. 177–193. Kluwer Acad. Publ, Dordrecht (1992)
Yudovich, V.I.: On the loss of smoothness of the solutions of Euler’s equations with time. Dinamika Sploshnoi Sredy (Dyn. Contin. Media) 16, 71–78 (1974)
Acknowledgements
We thank Theo Drivas and Huy Nguyen for several insightful comments. P.C.’s research was partially supported by NSF Grant DMS-2106528 and by Simons Foundation Grant 601960.
Funding
The Project is funded by National Sciewnce Foundation Grant no. 601960.
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflict of interest
The authors have no competing interests to declare that are relevant to the content of this article.
Additional information
Communicated by A. Ionescu.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Appendices
Appendix A. Periodic Boundary Conditions: Setup
Let \({\mathbb {T}}^3\) be the three-dimensional torus.
1.1 A.1 Norms and spaces
We work with the spaces \(H^s({\mathbb {T}}^3)\) and \({\dot{H}^s}({\mathbb {T}}^3)\), which we define as follows.
Definition A.1
Let \(f \in {\mathscr {P}}'\), where \({\mathscr {P}}'\) is the space of periodic distributions on the 3-dimensional torus \({\mathbb {T}}^3\). Then, consider the Fourier coefficients \({\hat{f}}(k)\), where \(k \in {\mathbb {Z}}^3\). We say that \(f \in H^s({\mathbb {T}}^3)\) if the Fourier coefficients of f satisfy:
We also say that \(f \in {\dot{H}^s}({\mathbb {T}}^3)\), if the Fourier coefficients of f satisfy:
We moreover denote
Remark A.2
In the remainder of Sect. 2 (and only restricted to this section), we denote
Remark A.3
For conciseness, we denote \(\Vert f \Vert _{s}:= \Vert f \Vert _{{\dot{H}^s}}\), the homogeneous Sobolev norm.
If \(f, g \in {\dot{L}}^2({\mathbb {T}}^3)\), we let
the usual \(L^2\) inner product.
The dual of \({\dot{H}}^1\) is naturally identified with \({\dot{H}}^{-1}\). If \(f \in {\dot{H}}^{-1}\) and \(g \in {\dot{H}}^1\), we let \(\langle f, g \rangle \) be the dual pairing between f and g. Note that, if \(f\in {\dot{L}}^2\), then \(\langle f, g \rangle = (f,g)\).
These scalar spaces extend in a straightforward manner to their vectorial counterparts. We shall use a superscript to denote vectorial spaces, i.e. if v is a 3 dimensional vector field whose components lie in a space W, then we write \(v \in W^3\).
Definition A.4
We denote \(H:= \{v \in L_0^2({\mathbb {T}}^3): \text {div} \ v = 0 \}\), and \(V:= \{v \in {\dot{H}}^1({\mathbb {T}}^3): \text {div} \, v = 0 \}\).
Definition A.5
We define the space \(D((-\Delta )^{s/2})\) for \(s \in {\mathbb {R}}\) as the domain of \((-\Delta )^{s/2}\):
Without ambiguity, we denote the homogeneous Sobolev s-norm on \(D((-\Delta )^{s/2})\) also by \(\Vert \cdot \Vert _s\).
1.2 A.2 Leray projection, the fractional Laplacian, and classical estimates
Let \({\mathbb {P}}: L_0^2 \rightarrow H\) be the Leray projector. In terms of Fourier coefficients, it can be expressed as
Here, \(\textrm{Id}_3\) is the \(3\times 3\) identity matrix.
If \(v, w \in V\), we define
Let us also recall the following inequality, proved in [9]. There exists a constant \(C > 0\) such that, if \(u,v,w \in V\) and are all divergence-free,
We also recall that elements \(u \in {\dot{H}}^1\) enjoy the following Poincaré inequality:
where \(\lambda > 0\) is the Poincaré constant.
We define the fractional Laplacian as the operator \((-\Delta )^\alpha : {\dot{H}^s} \rightarrow \dot{H}^{s-2\alpha }\) whose action on the Fourier coefficients is as follows:
We define
We have the following lemma:
Lemma A.6
Let \({\mathbb {L}}:= (-\Delta )^\alpha \), and assume that \(\alpha \ge 1\). Then, \({\mathbb {L}}\) maps \({\dot{H}^s}\) onto \(\dot{H}^{s-2\alpha }\), is injective and has a well defined inverse \({\mathbb {L}}^{-1}: \dot{H}^{s-2\alpha }\rightarrow {\dot{H}^s}\). Furthermore, the composition operators \(\Delta {\mathbb {L}}^{-1} = {\mathbb {L}}^{-1} \Delta \) are well defined, and are both bounded operators from \({\dot{H}^s}\) to \(\dot{H}^{2\alpha - 2 +s}\). In particular, there exists a constant \(C > 0\) such that, for all \(v \in {\dot{H}^s}\),
Remark A.7
The lemma above holds true replacing all instances of \(\dot{H}^\gamma \) with \(D((-\Delta )^{\gamma /2})\).
Proof of lemma A.6
The proof follows from the spectral characterization of the operator \((-\Delta )^\alpha \). \(\square \)
Remark A.8
By the Fourier characterization of \((-\Delta )^\alpha \) and of \({\mathbb {P}}\), it is evident that both these operators commute with partial derivatives. Furthermore, we have that
We also recall the definition of time-dependent spaces. Let \((X, \Vert \cdot \Vert _X)\) be a Banach space, and let \(T > 0\). We say that a function \(f: [0,T] \rightarrow X\) is such that \(f \in L^\infty (0, T; X)\) if the following holds:
Let \(f: [0,T] \rightarrow X\) be Bochner integrable, and define \(f'\) as the weak time derivative of f. We then say that \(f \in {\mathcal {C}}^1(0, T; X)\) if there holds
1.3 A.3 The viscous Voigt–MHD system
We focus our attention on the Voigt regularized system (VMHD-1)–(VMHD-4) (recall that \({\mathbb {L}} = (- \Delta )^\alpha \)):
Here, u(x, t) and B(x, t) are three-dimensional vector fields depending on position \(x \in {\mathbb {T}}^3\) (the three-dimensional flat torus) and time t. Furthermore \(\nu >0\) is a fixed parameter, and q(x, t) is the pressure term.
Let \(\alpha \ge 1\). We set up initial data \(u_0\) and \(B_0\) such that \(u_0, B_0 \in D((-\Delta )^{\alpha /2})\) (see Definition A.5). In particular, \(u_0\), \(B_0\) satisfy:
and moreover
1.4 A.4 Well-posedness, global existence and regularity
We show that the Voigt–MHD system (32)–(34) is locally well posed for strong solutions, and it moreover admits global solutions for large initial data. In the Voigt case, strong regularization easily implies well-posedness. Note that the issue is generally more involved in the case of non-regularized systems: for a discussion of local ill-posedness for a wide range of hydrodynamical systems we refer the reader to [13].
We first prove a local existence statement with a suitable continuation criterion.
Proposition A.9
(Local existence of solutions to (32)–(34)). Let \((u_0, B_0)\) in \(D((-\Delta )^{\alpha /2})\) with \(\alpha \ge 1\), and let \(u_0\) and \(B_0\) satisfy the divergence-free condition (35) and the mean zero condition (36). Then, there exists a time \(T > 0\) and \(u, B \in L^\infty (0,T; D((-\Delta )^{\alpha /2})\) which solve (32)–(34) in the strong sense, and such that
Furthermore, if \(T_*\) is the maximal time of existence, we necessarily have either
Proof of Proposition A.9
The proof follows by Picard iteration carried out in the space
We follow the approach in [20] (Theorem 6.1).
We consider the following evolution equations, which are equivalent to system (32)–(34):
We then let \(v = {\mathbb {L}}u\), \(Z = {\mathbb {L}}B\), and we rewrite the system as follows:
The idea is to show that \(N_1\) and \(N_2\) are Lipschitz mappings from \(D((-\Delta )^{-\alpha /2}))\) to itself, and then the Picard–Lindelöf theorem will apply. We start from \(N_1\), using the conventions \(v_i = {\mathbb {L}}u_i\), \(Z_i = {\mathbb {L}}B_i\):
Here, we used inequality (30) to go from the second to the third line, the Poincaré inequality (31), the fact that, for \(\alpha \ge \gamma \), we have \(\Vert f \Vert _\alpha \ge \Vert f\Vert _\gamma \), and the fact (proved in Lemma A.6) that, for \(z \in D((-\Delta )^{\alpha / 2})\),
since \(1-2\alpha \le -\alpha \) if \(\alpha \ge 1\). Similarly, we have, for the terms in \(N_2\),
We conclude that the mapping \((N_1, N_2)\) is locally Lipschitz in \(D((-\Delta )^{-\alpha /2}))\), which concludes the existence proof by an application of the Picard–Lindelöf theorem. Finally, the continuation criterion is evident from the fact that
This concludes the proof. \(\square \)
We show that high norms are propagated by the system (32)–(34).
Theorem A.10
(Global existence of solutions to (32)–(34)). Let \((u_0, B_0)\) in \(D((-\Delta )^{\alpha /2}))\), with \(\alpha \ge 1\), satisfying (35) and (36). Then, there exist \((u, B) \in L^\infty (0,\infty ; D((-\Delta )^{\alpha /2})))\) which solve (32)–(34) in the sense of distributions, and such that
Proof of Theorem A.10
Due to the continuation criterion of Proposition A.9, we only need to show that \(\Vert u\Vert _{\alpha }+\Vert B\Vert _{\alpha }\) is bounded a-priori in terms of initial data. We have, taking the \(L^2 \) inner product of the momentum equation with u,
On the other hand, we have, taking the \(L^2\) inner product of the induction equation with B,
Recalling that \(({\textbf{B}}(u,B),B)= 0\), and summing the two previous equations, we finally get, for all times \(t_2 \ge t_1\ge 0\),
which provides the required a-priori control. These a-priori estimates are only formal, but can be made rigorous by Galerkin approximation. \(\square \)
We show that the system (32)–(34) propagates higher regularity.
Proposition A.11
(Higher regularity of solutions to (32)–(34)). Let \((u_0, B_0)\) in \(D((-\Delta )^{\beta /2}))\), with \(\beta \ge \alpha \ge 1\). Then, the solution (u, B) to the system (32)–(34) constructed in Theorem (A.10) satisfies the stronger bounds:
Sketch of proof
We only provide a sketch of the proof in the case \(\beta = \alpha + k\), with \(k \in {\mathbb {N}}\). Let us deal with the case \(k = 1\). We formallyFootnote 4 differentiate Eqs. (32)–(34) by \(\partial _i\), and obtain
Multiplying the first equation by \(\partial _i u\) and the second equation by \(\partial _i B\), summing over i, and integrating we get, using the fact that \(({\textbf{B}}(a,b),c) = - ({\textbf{B}}(a,c),b)\),
Now, by Sobolev embedding, it is clear that
Using the a-priori energy estimate, in conjunction with Grönwall’s inequality, we obtain
i.e. the required bound.
The proof for larger k is similar, and we omit it here. \(\square \)
1.5 A.5 Magnetic potential
We state and prove a lemma on the existence of the magnetic potential in the periodic case.
Lemma A.12
(Existence of the magnetic potential). Let (u, B) such that
solving the system (32)–(34), according to Theorem A.10, with divergence-free initial data \((u_0, B_0) \in D((-\Delta )^{\alpha /2})\). Let us consider the following initial value problem, for an unknown vector field \(\Psi \):
Here, \(\Psi _0\) is the unique \(\dot{H}^{\alpha +1}({\mathbb {T}}^3)\) vector field satisfying the following two properties:
Under these conditions, we have that the system (37) has a global solution \(\Psi \in L^\infty (0,\infty ; D((-\Delta )^{(\alpha +1)/2}))\) which satisfies (37) and (38). Furthermore, the following equality holds true for all \(t \ge 0\) and \(x \in {\mathbb {T}}^3\).
Proof of Lemma A.12
The existence and regularity parts are standard. To prove relation (39), we take the curl of the evolution equation in (37), in order to obtain
Integrating in time and using the initial conditions, this gives \({\mathbb {L}}(\nabla \times \Psi ) = {\mathbb {L}}B\) at all times \(t \ge 0\). We conclude by the fact that the kernel of \({\mathbb {L}}\) in \(D((-\Delta )^{\alpha /2})\) is trivial. \(\square \)
Appendix B. Proof of Proposition 1.5
In this section, we provide a particular solution to the 3D MHD equation which, in the infinite time limit, creates discontinuities of B. The example is classical, and it is essentially obtained by decoupling the momentum and the induction equation, imposing that the B field is always vertical.
This example should be contrasted with the result obtained in Sect. 2, and it shows that, in general, for the full MHD system may not relax to a regular MHS equilibrium in the infinite time limit. This is different from the situation in the Voigt–MHD system.
Proof of Proposition 1.5
Let us consider the three dimensional torus obtained from the box \([-1,1] \times [-1,1] \times [-1,1]\) with opposite sides identified.
Suppose that u has the following form: \(u = (u_1(x,y), u_2(x,y),0)\), and that B has the following form: \(B= (0,0,B_3(t,x,y))\).
Let us consider the following stream function for u:
and we let \(u_1:= \partial _y \Psi \), \(u_2 = -\partial _x\Psi \). Under these conditions, u satisfies the steady Euler equations:
Let us moreover evolve \(B_3\) according to the following transport equation:
With these choices, we have that the pair (u, B) satisfies the ideal 3D MHD equations:
Since u is constant in time, we only need to specify initial data for B. We set \(B_0 (x,y) = \chi (y)\), where \(\chi \) is a smooth and periodic function, \(\chi : [-1,1] \rightarrow {\mathbb {R}}\), with the property that \(\chi (y) = y\) for \(|y| \le 1/2\). We show that, locally around the origin, the gradient of \(B_3\) grows exponentially in time.
Restricting Eq. (40) to the y-axis, we have
Let us define a function y(t, a) by the ODE (a is the Lagrangian label): \(y'(t,a) = - \sin (y(t,a))\), with \(y(0) = a\). Then, it can be easily checked that, for all positive t and for all a such that \(|a| \le 1/2\)
Note that have the following relation satisfied by y(t, a):
Differentiating relation (41) and calculating the result at \(a = 0\), we have that
for a positive constant \(C'\), thereby proving the claim. \(\square \)
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Constantin, P., Pasqualotto, F. Magnetic Relaxation of a Voigt–MHD System. Commun. Math. Phys. 402, 1931–1952 (2023). https://doi.org/10.1007/s00220-023-04770-1
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00220-023-04770-1