1 / 21

ITM General Meeting Lisbon, November 2013

ITM General Meeting Lisbon, November 2013. IMP5 2013 overview D Farina, T Jonsson, G Vlad on behalf of contributors to IMP5 TF Leader : G. Falchetto, Deputies: R. Coelho, D. P. Coster EFDA CSU Contact Person: D. Kalupin. IMP5 activities. IMP5 goal

osimpson
Download Presentation

ITM General Meeting Lisbon, November 2013

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. ITM General Meeting Lisbon, November 2013 IMP5 2013 overviewD Farina, T Jonsson, G Vladon behalf of contributors to IMP5TF Leader : G. Falchetto, Deputies: R. Coelho, D. P. CosterEFDA CSU Contact Person: D. Kalupin

  2. IMP5 activities • IMP5 goal • Codes in the field of ECRH, LHCD, ICRH, NBI, alpha particles and fast particle interaction with instabilities integrated in the ITM framework • 2013 main achievements • EC benchmark updated • IC wave codes adaptation and benchmark • NBI codes adaptation and first results • Fast particle codes 2

  3. EC codes benchmark • EC codes benchmark refinement • A few bugs found thanks to the first benchmark, additional cases considered (low/high edge density, low/high injected ECW power) • Still few discrepancies, but generally excellent match. Could be interesting to extend the benchmark to more varied test cases: high temperature (DEMO?), real shots (AUG, TCV,…) • Some arbitrary effect can still arise due to data discretization and lack of a “unified” approach to interpolation (vacuumplasma transition, possible singularities at magnetic axis, etc.)

  4. EC codes in ITM • Standard inductive H-mode ITER "Scenario 2” • (B0 = 5.3 T, Ip= 15 MA) • Ordinary mode at 170 GHz from the Equatorial (EL) and Upper Launcher (UL): • Divergent beam from EL at small and large toroidal launching angles, core andoff-axis H&CD • Focused beam from the UL 4

  5. EC benchmark (2012) • First EC codes benchmarking results for ITER Equatorial Launcher25 deg (top) and 40 deg (bottom) toroidal launching angle dP/dV R-R_vac z-z_vac

  6. EC benchmark (2013) Updated results dP/dV R-R_vac z-z_vac

  7. ICRF wave benchmark A benchmark of wave codes has been performed during 2013. Strong emphasis on documentation and tracebility of the benchmark. • Note: many differences between codes • FLR effects • parallel dynamics • electric field representation / parallel electric field • representations of the magnetic fields… • Also, potential differences in ITM implementation

  8. ICRF wave benchmark: results ITER full field 3% He3 minority heating in D-T plasma. ITER half field, non-activated phase. H-minority heating in He4 plasma. • Still initial state of the benchmark. • Both agreements and discrepancies observed • Detailed investigatrion require analysis of the physics models, numerical implementation and ITM adaptation

  9. NBI codes Monte Carlo FEM / FD 9

  10. NBI sources – first benchmark • The NBI source is source at beam ionization in plasma • source term in Fokker-Planck equation • Comparison of two modules: • BBNBI (Monte Carlo) • NEMO (finite difference) • Good agreement! Benchmark using the IMP5HCD workflow for a JET shot #77922 (IMP3 shot database) with 4 injectors. NEMO BBNBI

  11. New NBI source actor SNBI(not yet in IMP5HCD workflow) 0.0s 0.2s 0.4s 0.8s 1.0s 0.6s Z, m 1.6s 1.4s 1.2s 1.6s R, m

  12. NBI Fokker-Planck, towards a benchmark • RISK: Finite element solver, bounce averaged FP • ASCOT: 5/6D Monte Carlo FP • Fully implemented within the IMP5 workflow (4.10a) • Benchmark will be performed in 2014 Results using the IMP5 workflow (JET 77922): RISK ASCOT

  13. FIDIT: orbit averaged Fokker-PlanckNBI ions distribution functions at r/a=0.15 D (8 PINIs Oct.4) T (8 PINIs Oct.8) 0.2s 0.4s 1.3s

  14. IMP5HCD workflow IMP5 has an actor, IMP5HCD, that combines all heating schemes • Each CPOs is filled by one composite actor svn: https://gforge.efda-itm.eu/svn/keplerworkflows/trunk/4.10a/imp5/imp5hcd/imp5hcd.xml website: https://https://www.efda-itm.eu/ITM/html/imp5_imp5hcd.html python graphs ETS output Physicsmodules CORESOURCE DISTRIBUTION WAVES DISTSOURCE NOTE: this is an actor – to be plugged into any workflow

  15. Multilevel structure of workflow Output graphs Physics 15

  16. More user friendly interface • Selecting HCD systems: • Access to Machine Description Databases • Preprepared input CPO • Code parameters

  17. Fast particle code activities • First period of 2013 spent in transition to new Gateway (Garching): • environment, compilers, etc. • Benchmark of new MHD-GK code HYMAGYC with HMGC (EPS 2013 poster) • Benchmark of MHD module MARS with MARS-F, KINKX (in conjunction with IMP12) • Parallel solver for MHD part of HYMAGYC (HLST project ParFS (etention of 2012 project) • Tests on complex number definitions in new UAL 17

  18. Benchmark HMGC-HYMAGYC • The benchmark between HMGC and HYMAGYC has continued in order to verify the new code • TAE and EPM solutions comparison continued (EPS 2013 Poster P4.151): circular equilibrium, a/R0=0.1, q0=1.1, qa=1.9, n=2, ρH/a=0.01, vH0/vA0=1. • to better investigate the quantitative differences observed in the growth-rates (in particular for the EPM (“upper”) mode), the GK module of HYMAGYC has been “dismounted” and plugged into the MHD module of HMGC: tests are under way. 18

  19. Benchmark MARS-MARS-F-KINKX • The benchmark between MARS (the original, eigenvalue MHD code from which the initial value MHD solver for HYMAGYC has been derived) and MARS-F and KINKX (in collaboration with IMP12) • EPS 2013 Poster P5.162: benchmark on a JET reconstructed equilibrium, external kink, using the equilibrium&stability chain Fig. 3. Poloidal Fourier components (m=1-10) of the normal displacement ξn. Blue(solid) – KINX, red(dashed) – MARS-F, green(dashed-dotted) – MARS. 19

  20. Parallel MHD solver • HLST project ParFS (Parallel Field Solver): short extension of 2012 project (to be continued in 2014…) • while GK module of HYMAGYC is already parallelized to be able to evolve a large number of energetic particles (Particle-in-cell code), the MHD part is still serial (and replicated on each node) • this could pose strong limitations in view of ITER scenarios, where Alfvénic modes modes up to n~40 are foreseen to be relevant (in particular with respect to memory available on a single node) • in the frame of a HLST project, a first parallel solver (MUMS) has been successfully tested; nevertheless, the scalability, in terms of memory and time execution, of MUMPS is not yet satisfactory and continuation of the project to 2014 has been proposed. • A byproduct of this project could be the availability of a parallel eigensolver for MARS in view of ITER 20

  21. Extension of MHD CPOs to complex • Actual 4.10a CPOs (e.g. MHD CPOs) has Fourier components of eigenvectors defined as two real arrays (e.g., perturbed contravariant s component vs): • mhd_out(1)%plasma%v_pert%coord1%re(nrp1,nntor,msmax) • mhd_out(1)%plasma%v_pert%coord1%im(nrp1,nntor,msmax) • It would be desirable to have it defined as a single complex array: • mhd_out(1)%plasma%v_pert%coord1(nrp1,nntor,msmax) • Some tests on a test version of the standard 4.10a UAL release has been recently performed using the MARS code 21

More Related