1 / 8

TRANSP Users Meeting - Agenda and Developments

This document contains the agenda for the TRANSP users meeting, including discussions on TRANSP development, TORIC, TrICH utility, ADAS reaction data, website updates, and ongoing MPI and nubeam issues.

lhope
Download Presentation

TRANSP Users Meeting - Agenda and Developments

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. TRANSP users meeting22.04.2010 - Agenda • TRANSP development in 2010/2011, user’s suggestions to facilitate/accelerate TRANSP submissions. • TRANSP work: systematic efforts on TRANSP validation • TORIC • How to run it ( TR.DAT changes ) • QLO renormalisation – (known) problems • PTRANSP: recently installed transport models (TRANSP users meeting 17.12.2009) IV

  2. Recent Developments • TrICH utility – Combines A2 & ILA antenna data • Correction to A2 antenna position • TrFile more user friendly • ADAS reaction data now available ( needs further testing ) • Code now distribiuted via SVN, not CVS ( transparent to users.. ) • Jetdsp / transp interface mended & documented – ( use ~pshare/jetdsp ; no jetdsp RO at present .. ) • JETtransp – Archive step no longer hides your files.. • Website – help file & Ufiles list updated JEC

  3. Ongoing MPI / nubeam • PPPL - uses 700Mb of virtual memory, ~ constant throughout run • JET – grows steadily to around 3Gb, then dies Memory leak (?) • Valgrind – doesn’t work properly with Lahey fortran • Totalview ( interactive ) – found nothing obvious, but limited run duration • Totalview ( batch ) – BUG – reported to Totalview • ( Very many ) extra diagnostic messages added after array allocations, MPI operations, but cause of crashes not revealed Move to 64bit compilers might be the ( long term ) answer JETtransp Needs changes to submit MPI jobs JEC

  4. Planned Developments • JETtransp Upgrade pppl submission to use MPI+ user requests ?? • Diagnostics, heating system & vessel parameters • Check derivation of values • Store values centrally ( XML ? Or .. ?) • Transp code • Continue to improve diagnostics & error checking • Integrate RF antenna merging • Store beam parameters etc in ‘plasma state’ data files ? JEC

  5. Physics • Improve regression testing • Larger selection of runs • Look at TEQ solver again ? • More systematic check of outputs • Toric verification • Against experimental data ( How ? ) • Revisit D-T campaign ? • Much improved hardware • New code JEC

  6. Proposals/suggestions/requests 1. TRANSP as research tool: - possibility to run TRANSP for other tokamaks: MAST (done), ITER, DIII-D - possibility to manipulate with NBI: change energy and energy fractions, number of beams, power Easily changeable: beam species, geometry? 2. Scripts for fast and easy submission to Fusion Grid? 3. Starting new run – what parameters are modified automatically in the namelist (NPTCLS, NICHA)? 4. Problem with “step arch” – unable when the browser is open, require “force”. Are these files archived properly? 5. Disk space problem: the files we do not need (*sh, *mk, *_*, dir _mpi, others?), should they be removed automatically at the end of the run? IV

  7. Wishlist / Minutes Database • Problems if browser is already running; cant exit to terminate step • No facility to edit ( or delete ) entries. > Raise CFR ( but no RO for the database at present.. ) Inputs • Verify input geometries – vessel, diagnostics, heating systems • Investigate pppl ( & other ) methods of storing this data ( XML schemes preferred ) • Neutral Beams • Are the actual beam directions significantly different from the nominal directions • If so, can we access ( & use ) more accurate geometry ? • ICRH – antenna geometry – upsets pretr if RF is off (?) JEC

  8. Wishlist / Minutes 2 JETtransp • Add option to use MPI at pppl • Replace ( rather flakey ) IDL procedure which generates the MDSplus input trees • Archiving / Disk space • Delete some larger files ASAP ( eg restart directory, if not needed ) • Save minimal set of files & run directory under users work area Transp for ITER ( & DIIID ?? ) No equivalent of the PPF system for ITER data, so will need to use the alternative Ufiles system – to run at JET or pppl Also need infrastructure ( shot & run number assignment, Ufile archiving .. ) to support this JEC

More Related