Скачать презентацию Wrap-up and future plans Goals Remarks Summary Siegfried Скачать презентацию Wrap-up and future plans Goals Remarks Summary Siegfried

9f0994f8b69539942e16beb85a35cd33.ppt

  • Количество слайдов: 21

Wrap-up and future plans Goals Remarks Summary Siegfried Schreiber Workshop on Linac Operations on Wrap-up and future plans Goals Remarks Summary Siegfried Schreiber Workshop on Linac Operations on Long Bunch Trains DESY 24 -February-2010 FLASH. Free-Electron Laser in Hamburg

FLASH. FLASH layout – after this upgrade Diagnostics Accelerating Structures Bunch Compressor 5 Me. FLASH. FLASH layout – after this upgrade Diagnostics Accelerating Structures Bunch Compressor 5 Me. V 160 Me. V Bunch Compressor 500 Me. V s. FLASH 1200 Me. V Free-Electron Laser in Hamburg Undulators Bypass FEL Experiments 315 m Siegfried Schreiber | Workshop on Linac Operation an Long Bunch Trains | 22 -24 Feb 2010

Workshop attendance FLASH. Free-Electron Laser in Hamburg > 49 people registered > 14 plenary Workshop attendance FLASH. Free-Electron Laser in Hamburg > 49 people registered > 14 plenary talks, 41 working group talks in 4 working groups > This shows a broad interest from all three regions into our common goal > The workshop is an important step towards the common goal of the superconducting machines FLASH, XFEL, and ILC → to explore the full potential of the superconducting technique → providing thousands of bunches per second to users (FELs) → to go as close as possible to the full beam loading limit to maximize integrated luminosity (ILC) Siegfried Schreiber | Workshop on Linac Operation an Long Bunch Trains | 22 -24 Feb 2010

FLASH. Workshop goals Free-Electron Laser in Hamburg Goals for the workshop: > Review results FLASH. Workshop goals Free-Electron Laser in Hamburg Goals for the workshop: > Review results and lessons learnt from the most recent 9 m. A studies period in Sept 09 > Explore strategies for efficiently re-establishing stable long bunch train operation with currents from 1 m. A to 9 m. A > Explore strategies for characterizing LLRF stabilization of beam energy at the limits of RF power and cavity gradient > Develop proposals for future studies at FLASH and identify associated preparatory work Siegfried Schreiber | Workshop on Linac Operation an Long Bunch Trains | 22 -24 Feb 2010

FLASH. Working Groups Free-Electron Laser in Hamburg > Working Group #1: FLASH setup, tuning, FLASH. Working Groups Free-Electron Laser in Hamburg > Working Group #1: FLASH setup, tuning, and operation § Leaders: B. Faatz, J. Carwardine > Working Group #2: FLASH feedback and control § Leaders: H. Schlarb, V. Ayvazyan > Working Group #3: ILC studies at FLASH § Leaders: N. Solyak, S. Michizono > Working Group #4: DAQ and data analysis § Leaders: T. Wilksen, N. Arnold) Siegfried Schreiber | Workshop on Linac Operation an Long Bunch Trains | 22 -24 Feb 2010

Review and lessons learned FLASH. Free-Electron Laser in Hamburg Review results and lessons learnt Review and lessons learned FLASH. Free-Electron Laser in Hamburg Review results and lessons learnt from the most recent 9 m. A studies period in Sept 09 > That’s what we did most during this workshop > The analysis is certainly not complete, sometimes controversial, but I think we have a fair picture now > to that extend, this is a good start – we certainly need follow-up meetings and a follow-up workshop Siegfried Schreiber | Workshop on Linac Operation an Long Bunch Trains | 22 -24 Feb 2010

September 2009 9 m. A Run: Reality • First week effectively in short-pulse mode September 2009 9 m. A Run: Reality • First week effectively in short-pulse mode (<30) – Fighting beam loss – LLRF issues (including DOOCs server) • Both hardware and firmware • ‘on-the-fly’ development – Various (and numerous) other hardware failures • Klystron interlocks • Coupler events • Laser problems • … • Second week pushed to achieve long pulse trains – Rocky road due to never-ending fight with beam losses – Calibration issues (what is the beam energy? ) – Continued LLRF software development – TPS/MPS issues identified 22 -02 -10 N. Walker - Long Bunch Train Workshop Typical postshutdown startup issues (but not all!) Focus now on just achieving basic goals. No ‘experimental’ programme

Major achievements (Sept 2009 studies) Metric Bunches per pulse Goal Achieved Beam power Gradients Major achievements (Sept 2009 studies) Metric Bunches per pulse Goal Achieved Beam power Gradients close to quench 800 x 3 n. C 2400 x 3 n. C (3 MHz) Charge per pulse 800 x 3 n. C (1 MHz) 1800 x 3 n. C 2100 x 2. 5 n. C ~2400 x 2 n. C 7200 n. C @ 3 MHz 5400 n. C @ 3 MHz 36 k. W 22 k. W (7200 n. C, 5 Hz, 1 Ge. V) (5400 n. C, 5 Hz, 800 Me. V) Up to 32 Mv/m Several cavities above 30 Mv/m at end of long pulse • • • 15 contiguous hours running with 3 m. A and 800 us bunch trains • • Energy deviations within long bunch trains: <0. 5% p-p (7 m. A beam) Running at ~9 m. A with bunch trains of 500 -600 us for several hours Full pulse length (800 us, ~2400 bunches) at ~6 m. A for shorter periods Energy jittter pulse-pulse with long bunch trains: ~0. 13% rms (7 m. A) 22 -02 -10 N. Walker - Long Bunch Train Workshop

Last week of studies Bunch charge • Bunch charge was consistently between ~2. 7 Last week of studies Bunch charge • Bunch charge was consistently between ~2. 7 n. C and ~3 n. C • Rapid progress increasing number of bunches during the last 3 days (but very unstable) Number of bunches 15 hr run (800 bunches) 22 -02 -10 N. Walker - Long Bunch Train Workshop Almost 2400 bunches

Performance V. Ayvazyan • Stable operation at 1 MHz beam repetition rate was achieved, Performance V. Ayvazyan • Stable operation at 1 MHz beam repetition rate was achieved, resulting in stable 3 m. A running with a full 800 us pulse for over ten contiguous hours. • Quick start-up after machine access (40 min) • Several hours of operation at ~9 m. A but with short trains (300 -500 us) • Achieved full pulse length (800 us, 2400 bunches) at 6 m. A for short period

Some issues I picked up > Beam loss hampered us quite a bit § Some issues I picked up > Beam loss hampered us quite a bit § darkcurrent! § injector set-up – difficult matching with low ACC 1/gun gradients § energy measurement § optics/orbit: need to take space charge effects into account § explore and justify trip limits, implement 1 st fault informations > LLRF § too many different hardware, new not fully tested firmware § many changes in the hard- and firmware during run § tuning strategy controversial § adaptive feedforward, exception handling not mature § absolute phase/amplitudes required Siegfried Schreiber | Workshop on Linac Operation an Long Bunch Trains | 22 -24 Feb 2010 FLASH. Free-Electron Laser in Hamburg

Some issues I picked up FLASH. Free-Electron Laser in Hamburg > Hardware problems § Some issues I picked up FLASH. Free-Electron Laser in Hamburg > Hardware problems § 3 MHz laser thermal effects – too little time left to fix this § TPS system difficult to set-up and only partly working § ACC 1 coupler trips > Dump line § repair successfully, new line and new diagnostics worked well > DAQ § has been shown to be extremely useful - ~10 THz data stored § many tools exist to retrieve and analyze data, more to come (survey) – hampered by the complexity of the data § many improvements suggestions came up and are planned to be implemented > Operation sc cavities at their limits, HOMs § studies have started, though operational time with full beam loading was limited § studies to optimize operation with different gradients in modules Siegfried Schreiber | Workshop on Linac Operation an Long Bunch Trains | 22 -24 Feb 2010

V. Ayvazyan Main Limitation: ACC 1 Coupler Interlocks Background activities at ACC 1 couplers V. Ayvazyan Main Limitation: ACC 1 Coupler Interlocks Background activities at ACC 1 couplers Working point was close to the limits

Dump Line Upgrade 2009 New Loss Monitors BPM 16 DUMP BPM 13 DUMP BPM Dump Line Upgrade 2009 New Loss Monitors BPM 16 DUMP BPM 13 DUMP BPM 9 DUMP BPM 10 DUMP BPM 5 DUMP Toroid 9 DUMP OTR screen 9 DUMP Ionization Chambers I 14 R. DUMP 8 x BHM I 14 L. DUMP 16 DUMP I 14 U. DUMP I 14 D. DUMP Cerenkov monitors N. Baboi, MDI BLM 14 R. DUMP BLM 14 L. DUMP 14 U. DUMP 14 D. DUMP BLM 13. 1 DUMP 13. 2 DUMP BLM 6 DUMP BLM 1. 1 DUMP 1. 2 DUMP BLM 9 DUMP Workshop on Linac Operation with Long Bunch Trains, Feb. 22 -24, 2010

Dump Chamber BPM magnetic loop BPM BHM for glass fibres for ionization chambers » Dump Chamber BPM magnetic loop BPM BHM for glass fibres for ionization chambers » T. Wohlenberg, M. Schmitz, MDI 3 N. Baboi, MDI Workshop on Linac Operation with Long Bunch Trains, Feb. 22 -24, 2010

FLASH. Workshop goal Free-Electron Laser in Hamburg Explore strategies for efficiently re-establishing stable long FLASH. Workshop goal Free-Electron Laser in Hamburg Explore strategies for efficiently re-establishing stable long bunch train operation with currents from 1 m. A to 9 m. A > WG 1 discussions have been lively and did not lead – to my point of view – to a commonly accepted conclusion yet > Reproducibility § magnets restoring fine § but many other parameters need a complicated and time consuming restoring, like phases, amplitudes § need to restore machine fingerprint or machine state instead – more procedures and tools required Siegfried Schreiber | Workshop on Linac Operation an Long Bunch Trains | 22 -24 Feb 2010

Save. Restore Sequencer Siegfried Schreiber | Workshop on Linac Operation an Long Bunch Trains Save. Restore Sequencer Siegfried Schreiber | Workshop on Linac Operation an Long Bunch Trains | 22 -24 Feb 2010 FLASH. Free-Electron Laser in Hamburg

FLASH. Workshop goal Free-Electron Laser in Hamburg Explore strategies for characterizing LLRF stabilization of FLASH. Workshop goal Free-Electron Laser in Hamburg Explore strategies for characterizing LLRF stabilization of beam energy at the limits of RF power and cavity gradient > new llrf controller firmware with better performance – same/very similar hardware and server for all RF-stations § with the new set-up, adaptive ff algorithms should work now § MIMO controller, shadow table and other features > revised exception handling and beam loading compensation > more effort needed on stabilization with full beam load > 3 rd harmonic cavity exceptions need to be addressed > piezo-tuners: automated procedures in preparation → many of this is being prepared for the upcoming FLASH runs Siegfried Schreiber | Workshop on Linac Operation an Long Bunch Trains | 22 -24 Feb 2010

Energy deviation along bunch train (examples) 80 bunches, 100 k. Hz, ~3 n. C/bunch Energy deviation along bunch train (examples) 80 bunches, 100 k. Hz, ~3 n. C/bunch (0. 3 m. A) Along pulse: 0. 035% p-p 1. 8 Me. V 800 us 2100 bunches, 3 MHz, ~2. 5 n. C/bunch (7. 5 m. A) 30 Me. V Along pulse: 0. 5% p-p Pulse-pulse: 0. 13% RMS 700 us 22 -02 -10 N. Walker - Long Bunch Train Workshop (Rev)

FLASH. Future plan Free-Electron Laser in Hamburg Develop proposals for future studies at FLASH FLASH. Future plan Free-Electron Laser in Hamburg Develop proposals for future studies at FLASH and identify associated preparatory work: > We need more experience in many fields identified during the workshop > I would propose to participate in the FLASH effort to provide long pulse trains – though not with full beamloading yet > many study issues can be addressed: § adaptive feedforward § feedbacks § trip studies, recovery from trips, exception handling > Establish a task force § to continue and intensify the effort started § to prepare experiments at FLASH (and elsewhere) § prepare the next full beam loading experiment Siegfried Schreiber | Workshop on Linac Operation an Long Bunch Trains | 22 -24 Feb 2010

FLASH. Future plans Free-Electron Laser in Hamburg Develop proposals for future studies at FLASH FLASH. Future plans Free-Electron Laser in Hamburg Develop proposals for future studies at FLASH and identify associated preparatory work: > Appealing would be a joint experiment with photon scientists: § offering as many bunches as possible – with the risk of low efficiency § give them extra beamtime to compensate for this > A next full beam loading experiment I would schedule after achieving stable long trains for the FLASH users > Many proposals have been made (especially from WG 3), which partly can be included into regular accelerating beam time slots – where full beamloading is not required > We will have a follow-up workshop Siegfried Schreiber | Workshop on Linac Operation an Long Bunch Trains | 22 -24 Feb 2010