pith. sign in

arxiv: 1907.04851 · v1 · pith:6CNPE5FYnew · submitted 2019-07-10 · 🧮 math.NA · cs.NA· physics.comp-ph

Uniformly accurate methods for three dimensional Vlasov equations under strong magnetic field with varying direction

Philippe Chartier , Nicolas Crouseilles , Mohammed Lemou , Florian Mehats , Xiaofei Zhao This is my paper

Pith reviewed 2026-05-24 23:22 UTC · model grok-4.3

classification 🧮 math.NA cs.NAphysics.comp-ph
keywords Vlasov equationstrong magnetic fieldmultiscale methodsparticle-in-cellhomogenized modelgyro-motion
0
0 comments X

The pith

Multiscale methods solve the 3D Vlasov equation with strong varying magnetic fields at accuracy and cost independent of field strength.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper develops integrators for the three-dimensional Vlasov-Poisson system when a strong magnetic field points in a spatially varying direction. When the field strength is constant the particle motion contains fast periodic oscillations; averaging over one period produces a homogenized model that removes the stiff scale. The authors discretize this model with particle-in-cell methods and obtain schemes whose global error and wall-clock time remain essentially unchanged as the field intensity grows. Large time steps become admissible, and the unresolved gyro-motion is recovered afterward by linear interpolation along each trajectory. Extensions to fields whose intensity also varies are given in the linear setting.

Core claim

When the magnetic field has constant intensity the oscillations generated by the stiff term are periodic, permitting derivation of a homogenized model; multiscale methods built on this model and combined with particle-in-cell discretization remain uniformly accurate and efficient regardless of field strength.

What carries the argument

The homogenized model obtained by averaging over the periodic gyro-oscillations induced by a constant-intensity magnetic field.

Load-bearing premise

The oscillations generated by the stiff term are periodic whenever the magnetic field has constant intensity.

What would settle it

A computation in which the observed error of the proposed scheme grows linearly with magnetic-field strength, for a test problem with constant field intensity, would falsify the uniform-accuracy claim.

Figures

Figures reproduced from arXiv: 1907.04851 by Florian Mehats, Mohammed Lemou, Nicolas Crouseilles, Philippe Chartier, Xiaofei Zhao.

Figure 1
Figure 1. Figure 1: Errors of MRC, TSF and MM with respect to time steps M under different ε (left) or with respect to ε under different M (right) for example 5.1. Let us remark that the restart strategy is used at every time step. In [PITH_FULL_IMAGE:figures/full_fig_p017_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Error of TSF and MM with respect to Nτ under different ε for example 5.1. 10−2 10−1 100 10−5 10−4 10−3 10−2 ε=1/2 cpu time error MM TSF MRC 10−2 10−1 100 10−5 10−4 10−3 10−2 ε=1/214 cpu time error MM TSF MRC [PITH_FULL_IMAGE:figures/full_fig_p018_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Efficiency comparison of TSF, MM and MRC in classical (left) or as￾ymptotic regime (right) of ε for example 5.1: error versus computational time. 0 50 100 0 0.002 0.004 0.006 0.008 0.01 t error TSF M=1024 M=2048 0 20 40 60 80 100 0 0.005 0.01 0.015 t error MM M=1024 M=2048 0 20 40 60 80 100 0 2 4 6 x 10−5 MRC t error M=64 M=128 [PITH_FULL_IMAGE:figures/full_fig_p018_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Energy error of TSF, MM (with restart every T0 = 8π) and MRC for example 5.1 under ε = 1/2 14 till T = 32π. ∆t = 0.0982 or 0.0491 for TSF and MM. In [PITH_FULL_IMAGE:figures/full_fig_p018_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Left: exact trajectory of the particle in example 5.1 till T = π under ε = 1/2 5 . Right: numerical solution of MM under ∆t = 0.0982 (red curve) and the fully recovered trajectory with fine linear interpolation. 101 102 10−7 10−6 10−5 10−4 10−3 M error x M −2 101 102 10−7 10−6 10−5 10−4 10−3 M error |v| M −2 101 102 10−7 10−6 10−5 10−4 10−3 M error v // ε=1/2 ε=1/22 ε=1/23 ε=1/24 ε=1/25 ε=1/26 ε=1/27 ε=1/2… view at source ↗
Figure 6
Figure 6. Figure 6: Error of MM under different M = ∆s −1 in x, |v| and vk at T = 1 in example 5.2 of varying intensity. |v⊥| 2 = |v| 2 − |vk| 2 . The deviation of the magnetic moment (5.5) behaves as |I(t) − I(0)| = O(ε) in the simulation, which is consistent with the results obtained in [25]. Our scheme captures this adiabatic quantity even when ∆s ε whereas the scheme used in [25] needs ∆s < ε. 5.2. Simulation of the Vlaso… view at source ↗
Figure 7
Figure 7. Figure 7: Energy error of MM (restart each step) for ε = 1/2, 1/2 5 , 1/2 14 till t = 100 and the evolution of t(s) in example 5.2. 0 50 100 150 200 0 0.5 1 1.5 2 2.5 s Deviation ε=1/29 0 50 100 150 200 0 0.5 1 1.5 2 2.5 s Deviation ε=1/210 0 50 100 150 200 0 0.5 1 1.5 2 2.5 s Deviation ε=1/211 [PITH_FULL_IMAGE:figures/full_fig_p020_7.png] view at source ↗
Figure 8
Figure 8. Figure 8 [PITH_FULL_IMAGE:figures/full_fig_p020_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: Vlasov-Poisson case: pseudo-color snapshots of ρ ε under ε = 1/2 5 at t = 0, 16π, 32π, 64π with initial condition 5.6 with α = 0. Acknowledgements This work is supported by the French ANR project MOONRISE ANR-14-CE23-0007-01. This work has been carried out within the framework of the French Federation for Magnetic Fusion Studies (FR-FCM) and of the Eurofusion consortium, and has received funding from the E… view at source ↗
Figure 10
Figure 10. Figure 10: Vlasov-Poisson case: pseudo-color snapshots of ρ ε under ε = 1/2 5 at t = 0, 16π, 32π, 64π with initial condition 5.6 with α = 0.003 [PITH_FULL_IMAGE:figures/full_fig_p022_10.png] view at source ↗
Figure 11
Figure 11. Figure 11: Vlasov-Poisson case: pseudo-color snapshots of ρ ε under ε = 1/2 5 at t = 88π, 128π with initial condition 5.6 with α = 0.003. [7] Ph. Chartier, M. Lemou, F. Mehats, G. Vilmart ´ , A new class of uniformly accurate methods for highly oscillatory evolution equations, hal-01666472, 2017. [8] Ph. Chartier, J. Makazaga, A. Murua, G. Vilmart, Multi-revolution composition methods for highly oscil￾latory differe… view at source ↗
Figure 12
Figure 12. Figure 12: Vlasov-Poisson case. Left: time history of the energy error with initial condition 5.6. Right: difference between (1.1) and the limit model (2.4) (maximum error of |ρ ε (t = π, x) − ρ(t = π, x)|/|ρ ε (t = π, x)|). [13] N. Crouseilles, M. Lemou, F. Mehats, X. Zhao ´ , Uniformly accurate forward semi-Lagrangian methods for highly oscillatory Vlasov-Poisson equations, SIAM J. Multiscale Model. Simul. 15 (201… view at source ↗
read the original abstract

In this paper, we consider the three dimensional Vlasov equation with an inhomogeneous, varying direction, strong magnetic field. Whenever the magnetic field has constant intensity, the oscillations generated by the stiff term are periodic. The homogenized model is then derived and several state-of-the-art multiscale methods, in combination with the Particle-In-Cell discretisation, are proposed for solving the Vlasov-Poisson equation. Their accuracy as much as their computational cost remain essentially independent of the strength of the magnetic field. The proposed schemes thus allow large computational steps, while the full gyro-motion can be restored by a linear interpolation in time. In the linear case, extensions are introduced for general magnetic field (varying intensity and direction). Eventually, numerical experiments are exposed to illustrate the efficiency of the methods and some long-term simulations are presented.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 1 minor

Summary. The manuscript develops multiscale numerical methods for the 3D Vlasov-Poisson equation with strong inhomogeneous magnetic fields of constant intensity but varying direction. It states that stiff oscillations are periodic under constant |B|, derives a homogenized model from this, proposes combinations of multiscale schemes with Particle-In-Cell discretization, claims that both accuracy and computational cost remain essentially independent of |B| strength (allowing large time steps with optional restoration of gyro-motion via linear interpolation), provides extensions to general B in the linear case, and presents numerical experiments to illustrate efficiency and long-term behavior.

Significance. If the periodicity assumption holds and uniform accuracy is achieved, the work would enable efficient long-time simulations of strongly magnetized plasmas without resolving fast gyromotion, which is of practical value in plasma physics. The combination with PIC discretization and the interpolation feature for recovering full orbits are concrete strengths. Presentation of long-term simulations is also positive, though the lack of any mentioned error analysis or bounds reduces the assessed significance relative to a fully rigorous treatment.

major comments (2)
  1. [Abstract] Abstract: the statement that 'the oscillations generated by the stiff term are periodic' whenever |B| is constant (even with varying direction) is asserted without derivation, proof, or external reference. This periodicity is the explicit basis for the homogenized model and is load-bearing for the subsequent claim of |B|-independent accuracy and cost; in the nonlinear 3D case, particles sample spatially varying directions, so exact closure after one gyroperiod 2π/|B| is not immediate and requires justification.
  2. [Abstract] Abstract: the central claim that 'Their accuracy as much as their computational cost remain essentially independent of the strength of the magnetic field' is presented without any error bounds, convergence analysis, or description of the numerical experiments (e.g., tested |B| values, measured errors, or data-exclusion criteria). The claim therefore rests solely on unverified numerical evidence.
minor comments (1)
  1. The abstract would be clearer if it indicated the specific range of |B| strengths over which independence was observed in the experiments.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the careful reading and constructive comments on our manuscript. We address the two major comments point by point below.

read point-by-point responses

  1. Referee: [Abstract] Abstract: the statement that 'the oscillations generated by the stiff term are periodic' whenever |B| is constant (even with varying direction) is asserted without derivation, proof, or external reference. This periodicity is the explicit basis for the homogenized model and is load-bearing for the subsequent claim of |B|-independent accuracy and cost; in the nonlinear 3D case, particles sample spatially varying directions, so exact closure after one gyroperiod 2π/|B| is not immediate and requires justification.

    Authors: When |B| is constant the Lorentz force induces a rotation of the velocity vector at fixed angular speed |B|, so that the fast-scale velocity returns to its initial value after each interval of length 2π/|B|. The spatial variation of the direction is treated as a slow modulation; the guiding-center motion and the slow evolution of the distribution are obtained by averaging the periodic fast oscillation over one gyroperiod. This averaging step is carried out explicitly in Section 2 of the manuscript to obtain the homogenized model. We agree that a brief sentence recalling this local periodicity would improve the abstract and will add it in the revised version. revision: partial

  2. Referee: [Abstract] Abstract: the central claim that 'Their accuracy as much as their computational cost remain essentially independent of the strength of the magnetic field' is presented without any error bounds, convergence analysis, or description of the numerical experiments (e.g., tested |B| values, measured errors, or data-exclusion criteria). The claim therefore rests solely on unverified numerical evidence.

    Authors: The uniform accuracy and cost follow from the design of the multiscale integrators, which advance the slow variables without resolving the fast gyromotion; the Particle-In-Cell discretization inherits this property. Section 4 presents numerical tests in which |B| is increased over several orders of magnitude while the time step is held fixed; the reported L2 errors and wall-clock times remain essentially constant. Although a complete a-priori error analysis is not included, the numerical evidence is quantitative and reproducible. We will enlarge the experimental description in the revised manuscript to list the precise |B| values, measured errors, and acceptance criteria used. revision: partial

Circularity Check

0 steps flagged

No significant circularity; derivation rests on external periodicity assumption and independent multiscale construction

full rationale

The paper states that oscillations are periodic when |B| is constant and derives the homogenized model from this property before constructing multiscale schemes. This periodicity is presented as an input assumption rather than fitted from or defined by the numerical outputs or validation data. No step reduces a claimed prediction or uniform-accuracy result to a parameter fit on the same data, a self-citation chain, or a renaming of an input. The central claims of |B|-independent accuracy and cost therefore remain independent of the paper's own fitted quantities or self-referential definitions. The derivation chain is self-contained against external benchmarks.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

Central claim rests on the domain assumption that oscillations are periodic for constant-intensity fields, allowing a homogenized model to be derived; no free parameters, invented entities, or additional axioms are mentioned in the abstract.

axioms (1)
  • domain assumption Oscillations generated by the stiff magnetic term are periodic when magnetic-field intensity is constant
    Invoked in the abstract as the prerequisite for deriving the homogenized model that underpins the uniformly accurate schemes.

pith-pipeline@v0.9.0 · 5687 in / 1343 out tokens · 38841 ms · 2026-05-24T23:22:22.242750+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

36 extracted references · 36 canonical work pages · 1 internal anchor

  1. [1]

    Abdulle, W

    A. Abdulle, W. E, B. Engquist, E. Vanden-Eijnden , The heterogeneous multiscale method, Acta Numer. 21 (2012) pp. 1-87

    work page 2012

  2. [2]

    Birdsall, A.B

    C.K. Birdsall, A.B. Langdon , Plasma Physics via Computer Simulation, Adam Hilger, 1991

    work page 1991

  3. [3]

    Bostan, The Vlasov-Maxwell system with strong initial magnetic field

    M. Bostan, The Vlasov-Maxwell system with strong initial magnetic field. Guiding-center approximation, SIAM J. Multiscale Model. Simul. 6 (2007), pp.1026-1058

    work page 2007

  4. [4]

    Bostan, A

    M. Bostan, A. Finot , The effective Vlasov-Poisson system for the finite Larmor radius regime, SIAM J. Mul- tiscale Model. Simul. 14 (2015), pp. 1238-1275

    work page 2015

  5. [5]

    Averaging of highly-oscillatory transport equations

    Ph. Chartier, N. Crouseilles, M. Lemou, An averaging technique for transport equations, arXiv:1609.09819v1, preprint, 2018

    work page internal anchor Pith review Pith/arXiv arXiv 2018

  6. [6]

    Chartier, N

    Ph. Chartier, N. Crouseilles, M. Lemou, F. M ´ehats, Uniformly accurate numerical schemes for highly oscillatory Klein-Gordon and nonlinear Schr¨ odinger equations, Numer. Math. 129 (2015), pp. 211-250. 22 PH. CHARTIER, N. CROUSEILLES, M. LEMOU, F. M ´EHATS, AND X. ZHAO Figure 10. Vlasov-Poisson case: pseudo-color snapshots of ρε under ε = 1/25 at t = 0, ...

    work page 2015

  7. [7]

    Chartier, M

    Ph. Chartier, M. Lemou, F. M ´ehats, G. Vilmart , A new class of uniformly accurate methods for highly oscillatory evolution equations, hal-01666472, 2017

    work page 2017

  8. [8]

    Chartier, J

    Ph. Chartier, J. Makazaga, A. Murua, G. Vilmart , Multi-revolution composition methods for highly oscil- latory differential equations, Numer. Math. 128 (2014), pp 167-192

    work page 2014

  9. [9]

    Chartier, F

    Ph. Chartier, F. M´ehats, M. Thalhammer, Y. Zhang, Convergence analysis of multi-revolution composition time-splitting pseudo-spectral methods for highly oscillatory differential equations of Schr¨ odinger equations, ESAIM:M2AN 51 (2017) pp. 1859-1882

    work page 2017

  10. [10]

    Chartier, N

    Ph. Chartier, N. Crouseilles, M. Lemou, F. M ´ehats, X. Zhao , Uniformly accurate methods for Vlasov equations with non-homogeneous strong magnetic field, to appear on Math. Comp. (2019)

    work page 2019

  11. [11]

    Chartier, N

    Ph. Chartier, N. Crouseilles, X. Zhao, Numerical methods for the two-dimensional Vlasov-Poisson equation in the finite Larmor radius approximation regime, J. Comput. Phys. 375 (2018) pp. 619-640

    work page 2018

  12. [12]

    Crouseilles, M

    N. Crouseilles, M. Lemou, F. M ´ehats, Asymptotic preserving schemes for highly oscillatory Vlasov-Poisson equations, J. Comput. Phys. 248 (2013) pp. 287-308. VLASOV-POISSON EQUATION IN 3D WITH STRONG MAGNETIC FIELD 23 0 50 100 150 2000 0.005 0.01 0.015 t energy error α=0 α=0.003 10 −210 −2 10 −1 10 0 ε difference ε−1 Figure 12. Vlasov-Poisson case. Left:...

    work page 2013

  13. [13]

    Crouseilles, M

    N. Crouseilles, M. Lemou, F. M ´ehats, X. Zhao , Uniformly accurate forward semi-Lagrangian methods for highly oscillatory Vlasov-Poisson equations, SIAM J. Multiscale Model. Simul. 15 (2017), pp. 723-744

    work page 2017

  14. [14]

    Crouseilles, S.A

    N. Crouseilles, S.A. Hirstoaga, X. Zhao, Multiscale Particle-In-Cell methods and comparisons for long time two-dimensional Vlasov-Poisson equation with strong magnetic field, Comput. Phys. Commun. 222, pp. 136-151 (2018)

    work page 2018

  15. [15]

    Crouseilles, M

    N. Crouseilles, M. Lemou, F. M´ehats, X. Zhao, Uniformly accurate Particle-in-Cell method for the long time two-dimensional Vlasov-Poisson equation with uniform strong magnetic field, J. Comput. Phys. 346 (2017), pp. 172-190

    work page 2017

  16. [16]

    Degond, F

    P. Degond, F. Filbet , On the asymptotic limit of the three dimensional Vlasov-Poisson system for large magnetic field: formal derivation, J. Stat. Physicists. 65 (2016), pp. 765-784

    work page 2016

  17. [17]

    Filbet, C

    F. Filbet, C. Yang, Numerical simulations to the Vlasov-Poisson system with a strong magnetic field, preprint, (2018)

    work page 2018

  18. [18]

    Filbet, T

    F. Filbet, T. Xiong, E. Sonnendr ¨ucker, On the Vlasov-Maxwell system with a strong magnetic field, SIAM J. Applied Mathematics 78 (2018), pp. 1030-1055

    work page 2018

  19. [19]

    Filbet, M

    F. Filbet, M. Rodrigues , Asymptotically stable particle-in-cell methods for the Vlasov-Poisson system with a strong external magnetic field, SIAM J. Numer. Anal. 54 (2016), pp. 1120-1146

    work page 2016

  20. [20]

    Filbet, M

    F. Filbet, M. Rodrigues, Asymptotically preserving particle-in-cell methods for inhomogeneous strongly mag- netized plasmas, SIAM J. Numer. Anal. 55 (2017), pp. 2416-2443

    work page 2017

  21. [21]

    Fr ´enod, F

    E. Fr ´enod, F. Salvarani and E. Sonnendr ¨ucker, Long time simulation of a beam in a periodic focusing channel via a two-scale PIC-method, Math. Models Methods Appl. Sci. 19 (2009), pp. 175-197

    work page 2009

  22. [22]

    Fr ´enod, E

    E. Fr ´enod, E. Sonnendr ¨ucker, Long time behavior of the two-dimensional Vlasov equation with a strong external magnetic field, Math. Models Methods Appl. Sci. 10 (2000), pp. 539-553

    work page 2000

  23. [23]

    Fr´enod, S.A

    E. Fr´enod, S.A. Hirstoaga, M. Lutz, E. Sonnendr ¨ucker, Long time behavior of an exponential integrator for a Vlasov-Poisson system with strong magnetic field, Commun. in Comput. Phys. 18 (2015), pp. 263-296

    work page 2015

  24. [24]

    Golse, L

    F. Golse, L. Saint-Raymond, The Vlasov-Poisson system with strong magnetic field, J. Math. Pures Appl. 78 (1999), pp. 791-817

    work page 1999

  25. [25]

    Hairer, Ch

    E. Hairer, Ch. Lubich, Long-term analysis of a variational integrator for charged-particle dynamics in a strong magnetic field, preprint 2018

    work page 2018

  26. [26]

    Jacobs, J.S

    G.B. Jacobs, J.S. Hesthaven , Implicit-Explicit time integration of a high-order particle-in-cell method with hyperbolic divergence cleaning, Comput. Phys. Comm. 180 (2009), pp. 1760-1767

    work page 2009

  27. [27]

    Jin, Efficient asymptotic-preserving (AP) schemes for some multiscale kinetic equations, SIAM J

    S. Jin, Efficient asymptotic-preserving (AP) schemes for some multiscale kinetic equations, SIAM J. Sci. Comput. 21 (1999), pp. 441-454

    work page 1999

  28. [28]

    Kraus, K

    M. Kraus, K. Kormann, P. Morrison, E. Sonnendr ¨ucker, GEMPIC: geometric electromagnetic Particle In Cell methods, Journal of Plasma Physics 83 (4), (2017)

    work page 2017

  29. [29]

    G. Latu, M. Mehrenberger, Y. G ¨uc ¸l¨u, M. Ottaviani, E. Sonnendr ¨ucker, Field-Aligned Interpolation for Semi-Lagrangian Gyrokinetic Simulations, J. Sci. Comput. 74 (2018), pp. 1601-1650

    work page 2018

  30. [30]

    Lee, Gyrokinetic approach in particle simulation, Phys

    W.W. Lee, Gyrokinetic approach in particle simulation, Phys. Fluids 26 (1983)

    work page 1983

  31. [31]

    Northrop, The adiabatic motion of charged particles

    T.G. Northrop, The adiabatic motion of charged particles. Interscience Tracts on Physics and Astronomy, Vol

    work page

  32. [32]

    Interscience Publishers John Wily & Sons New York-London-Sydney, 1963

    work page 1963

  33. [33]

    Qin et al , Canonical symplectic particle-in-cell method for long-term large-scale simulations of the Vlasov- Maxwell system, Nucl

    H. Qin et al , Canonical symplectic particle-in-cell method for long-term large-scale simulations of the Vlasov- Maxwell system, Nucl. Fusion, 56, 014001 (2016)

    work page 2016

  34. [34]

    Saint-Raymond, The gyro-kinetic approximation for the Vlasov-Poisson system, Math

    L. Saint-Raymond, The gyro-kinetic approximation for the Vlasov-Poisson system, Math. Models Methods Appl. Sci. 10 (2000), pp. 1305-1332

    work page 2000

  35. [35]

    Sanders, F

    J.A. Sanders, F. Verhulst , Averaging methods in nonlinear dynamical systems, volume 59 of Applied Math- ematical Sciences. Springer-Verlag, New York, 1985. 24 PH. CHARTIER, N. CROUSEILLES, M. LEMOU, F. M ´EHATS, AND X. ZHAO

    work page 1985

  36. [36]

    Sonnendr¨ucker, Numerical Methods for Vlasov Equations, Lecture notes, 2016

    E. Sonnendr¨ucker, Numerical Methods for Vlasov Equations, Lecture notes, 2016. Ph. Chartier: Univ Rennes, INRIA-MINGuS, CNRS, IRMAR-UMR 6625, F-35000 Rennes, France E-mail address : Philippe.Chartier@inria.fr N. Crouseilles: Univ Rennes, INRIA-MINGuS, CNRS, IRMAR-UMR 6625, F-35000 Rennes, France E-mail address : nicolas.crouseilles@inria.fr M. Lemou: Uni...

    work page 2016