1 / 16

Kinetic & Fluid descriptions of interchange turbulence

Kinetic & Fluid descriptions of interchange turbulence. Y. Sarazin, E. Fleurence, X. Garbet, Ph. Ghendrih, V. Grandgirard, M. Ottaviani Association Euratom-CEA, CEA/DSM/DRFC Cadarache, France P. Bertrand LPMIA-Université Henri Poincaré, Vandœuvre-l è s-Nancy, France. c . fluid. kinetic.

Download Presentation

Kinetic & Fluid descriptions of interchange turbulence

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Kinetic & Fluid descriptionsof interchange turbulence Y. Sarazin, E. Fleurence, X. Garbet, Ph. Ghendrih, V. Grandgirard, M. Ottaviani Association Euratom-CEA, CEA/DSM/DRFC Cadarache, France P. Bertrand LPMIA-Université Henri Poincaré, Vandœuvre-lès-Nancy, France

  2. c fluid kinetic Motivations • Strong discrepancies between kinetic & fluid descriptions of turbulence: Linear thresholds Non linear fluxes [Beer '95, Dimits '00] • New type of non collisionnal closures: Non local [Hammet-Perkins '90, Snyder-Hammet-Dorland '97, Passot-Sulem '03] Non dissipative [Sugama-Watanabe-Horton '01,'04]

  3. Outline of the talk Standard closure assumes weak departure from local thermodynamical equilibrium (F Maxwellian)  small number of moments required Aim: Compare kinetic & fluid approaches (linear & non-linear) in a simple turbulence problem: • Same instability (2D interchange) • Same numerical tool • Closure based on entropy production rate

  4. 2D+1D interchange instability v//=0 ions  E  v2 Adiabatic electrons Constant curvature drift: Evd ey Slab geometry (x,y) Limit kri 0 Hamiltonian: H = vdEx +f Drift kinetic eq. Quasi-neutrality

  5. Fluid description 2 first moments of Vlasov  evolution of density & pressure Closure: =2  neglected

  6. Different linear stability diagrammes at W*n=0 • Threshold instability: W*Tc =(1 + k2) wd kinetic (-1) (1+k2) wd fluid • Vanishing relative discrepancy for large density gradients (W*n)

  7. Fluid  F at 2 energies cf [V. Grandgirard, 2004 & this conference] 2 distributions at energies Constraint: same numerics to treat fluid & kinetic descriptions Ensures dissipation at small scales stability of Equivalent to =1 closure in the limit e <<1

  8. at W*n=0 Adjustable linear properties Linear fluid properties can mimic kinetic ones: 3 degrees of freedom in fluid: T0, e and D • Unstable spectrum width (or maximum growth rate) • Linear threshold: W*Tc =wd(1 + k2)  F(T0, e, D) adequate choice

  9. Non linear discrepancy: Qfl>> Qkin Heat turbulent transport larger in fluid than kinetic by orders of magnitude Suggest non linear threshold (Dimits upshift ?) Transition not understood: ZF unchanged (amplitude & dynamics) (analogous to "Dimits graph" with the same code)

  10. Zonal Flows DO NOT explainthe whole difference • Larger turbulent flux when ZF artificially suppressed • Difference still present between kinetic & fluid (orders of magnitude) Note similar T profiles for similar fluxes (analogous to "Dimits graph" with the same code)

  11. Quantifying the departure from FMaxwell Projection on the basis of Laguerre polynomials Lp (standard approach for neoclassical transport) with Correspondance kth Fluid moment  Polynomes L1 … Lk

  12. 2 fluid moments are not enough Ortho-normal basis Lp(x) Slow convergence towards 0 Suggest any fluid description of the problem should account for high order moments Mk (k>2) May explain why fluid & kinetic results are still different w/o ZF

  13. Alternative closure: entropy production rates . Main ideas: SQL governed by QL transport Closure fulfils 2nd principle Fluid closure: Q =  ( P + Peq s) T operator:sr(ky) + i si(ky) Weights WQL : Kinetic: infinity of resonances Fluid: only 2

  14. Similar linear behaviour w/o ad-hoc dissipation • Same threshold as in kinetic: W*Tcfl= kyvd(1+k2) =W*Tckin • Stability of small scales: impliessi / ky< 0 Similar linear spectra  what about non linear behaviour ?

  15. Conclusions • Looking for adequate fluid closures: what degree of convergence kinetic-fluid is requested? (c, spectrum, dynamics, …) • Same numerical tool applied to 2D interchange model • 1st closure: weak departure from FM (Q =  P T) • Linear properties can be made comparable (D required) • Fluid transport >> Kinetic transport • Non linear upshift not captured by ZF only ( Dimits) • Possible explanation: large number of fluid moments required • 2nd closure: Q =  (P + Peq s) T • Target: balance entropy production rates   • Linear properties are similar

  16. time t + Dt t Dx Phase space Semi-Lagrangian numerical scheme • Semi-Lagrangian scheme: Fixed grid in phase space Follow the characteristics backward in time • Total distribution function F • Global code • Damping at radial ends to prevent numerical instabilities at boundaries • Good conservation properties (e.g.Error on energy < 1%) [Grandgirard et al. 2004]

More Related