Multiscale numerical method for nonlinear Maxwell equations (Q2388640): Difference between revisions

From MaRDI portal
Set OpenAlex properties.
Import241208061232 (talk | contribs)
Normalize DOI.
 
Property / DOI
 
Property / DOI: 10.3934/dcdsb.2005.5.631 / rank
Normal rank
 
Property / DOI
 
Property / DOI: 10.3934/DCDSB.2005.5.631 / rank
 
Normal rank

Latest revision as of 08:36, 18 December 2024

scientific article
Language Label Description Also known as
English
Multiscale numerical method for nonlinear Maxwell equations
scientific article

    Statements

    Multiscale numerical method for nonlinear Maxwell equations (English)
    0 references
    0 references
    0 references
    15 September 2005
    0 references
    The authors consider a numerical approximation of nonlinear Maxwell equations which describe how high-intensity light interacts with matter. The equations represent here a laser-matter interaction where the validity of the so-called paraxial approximation is not clear. The simulation requires to handle processes involving several orders of magnitude. The evolution of a laser beam in a nonlinear, noncentrosymmetric crystal is described by nonlinear Maxwell-Lorentz equations. An asymptotically equivalent semilinear model is derived that substitutes the nonlinear equations. The solution of the semilinear system relies on ideas of \textit{J.-L. Joly, G. Métivier} and \textit{J. Rauch} [Duke Math. J. 70, 373--404 (1993; Zbl 0815.35066); J. Differ. Equations 166, 175--250 (2000; Zbl 1170.78311)] where singular equations are used. Finite differences in space and time and an approximate Fourier basis for the phase are applied. The nonlinear Maxwell equations can be solved using a step size of order of the wave length. It is proved that ``the approximation is asymptotic in the sense that a multi-scale expansion can be performed on the discrete solution and the result of this expansion is a discretization of the continuous limit''(from the authors' summary). Moving computational windows are implemented in the algorithm in order to reduce the required memory space.
    0 references
    nonlinear medium
    0 references
    laser-matter interaction
    0 references
    crystal
    0 references
    asymptotics
    0 references
    singular equations
    0 references
    finite differences
    0 references
    Fourier basis
    0 references

    Identifiers

    0 references
    0 references
    0 references
    0 references
    0 references
    0 references