RFQ Exit Bunch Modelling Simon Jolly 25 th July 2012.

Slides:



Advertisements
Similar presentations
Imperial College 1 Progress at the RAL Front End Test Stand J. Pozimski Talk outline : Overview Ion source development LEBT RFQ Beam Chopper MEBT.
Advertisements

MEBT Design Considerations The beam energy in the MEBT is sufficiently low for the space charge forces to have a considerable impact on the beam dynamics.
To figure out which tubes are working fine… Compare 24 to 2.6 and 2.51 to 2.6 by taking the ratio of counts. In general, 2.4 had the most problems, and.
R. Miyamoto, Beam Physics Design of MEBT, ESS AD Retreat 1 Beam Physics Design of MEBT Ryoichi Miyamoto (ESS) November 29th, 2012 ESS AD Retreat On behalf.
Pepperpot Emittance Measurements of the FETS Ion Source
AGS pp Status Feb. 6, 2015 RSC Meeting Haixin Huang.
RFQ CAD Model Tolerance Studies Simon Jolly 14 th December 2011.
Section 2.3 Gauss-Jordan Method for General Systems of Equations
SPACE CHARGE EFFECTS IN PHOTO-INJECTORS Massimo Ferrario INFN-LNF Madison, June 28 - July 2.
Resampling techniques Why resampling? Jacknife Cross-validation Bootstrap Examples of application of bootstrap.
Medical Image Analysis
The Physics and Applications of High Brightness Electron Beams - Erice, October 9-14, 2005 Simulations of coherent synchrotron radiation effects on beam.
Introduction to Longitudinal Phase Space Tomography Duncan Scott.
AWAKE Electron Spectrometer Design Simon Jolly 4 th September 2012.
RFQ Integrated Design Would like to have a method of designing RFQ where all steps are integrated: –Engineering design. –EM modelling. –Beam dynamics simulations.
, WP1 Meeting, RAL J. Pasternak Status and Recent Progress in the Muon FFAG Designs for the NF J. Pasternak, Imperial College, London / RAL STFC.
AGS Polarized Proton Development toward Run-9 Oct. 3, 2008 Haixin Huang.
MICE pencil beam raster scan simulation study Andreas Jansson.
S.J. Brooks RAL, Chilton, OX11 0QX, UK Options for a Multi-GeV Ring Ramping field synchrotron provides fixed tunes and small.
Simulation of direct space charge in Booster by using MAD program Y.Alexahin, N.Kazarinov.
Simulation of Positron Production and Capturing. W. Gai, W. Liu, H. Wang and K. Kim Working with SLAC & DESY.
Automated Face Detection Peter Brende David Black-Schaffer Veni Bourakov.
Position Reconstruction in Miniature Detector Using a Multilayer Perceptron By Adam Levine.
The set of files includes : Tcl source of the POLYGON program The database (file obtained initially by P.Afonine from using phenix.model_vs_data.
RFQ CAD Model Beam Dynamics Studies Simon Jolly 3 rd August 2011.
1 One Function of Two Random Variables Given two random variables X and Y and a function g(x,y), we form a new random variable Z as Given the joint p.d.f.
Learning to perceive how hand-written digits were drawn Geoffrey Hinton Canadian Institute for Advanced Research and University of Toronto.
Imperial College 1 Plans and Costs Front End Test Stand Aim is to demonstrate a 60 mA, 2 ms, 50 pps chopped beam at 3 MeV RAL/ISIS (ION SOURCE, CHOPPER,
Simulation of direct space charge in Booster by using MAD program Y.Alexahin, A.Drozhdin, N.Kazarinov.
11 Sep 2009Paul Dauncey1 TPAC test beam analysis tasks Paul Dauncey.
NA62 Trigger Algorithm Trigger and DAQ meeting, 8th September 2011 Cristiano Santoni Mauro Piccini (INFN – Sezione di Perugia) NA62 collaboration meeting,
Simulations and Diagnostics for the Front End Test Stand Simon Jolly Imperial College 18 th April 2007.
Design of the Turnaround Loops for the Drive Beam Decelerators R. Apsimon, J. Esberg CERN, Switzerland.
Linda R. Coney – 5 November 2009 Online Reconstruction Linda R. Coney 5 November 2009.
FETS RFQ Beam Dynamics Simulations for RFQSIM, CST and Comsol Field Maps Simon Jolly 2 nd June 2010.
RFQ 3Dtree Space Charge Studies Simon Jolly 6 th June 2012.
3D codes: TraFiC4 and CSRtrack Torsten Limberg DESY Zeuthen 2003.
FCC-hh: First simulations of electron cloud build-up L. Mether, G. Iadarola, G. Rumolo FCC Design meeting.
Electron Spectrometer: Status July 14 Simon Jolly, Lawrence Deacon 1 st July 2014.
RFQ GPT Input Beam Distributions Simon Jolly 22 nd August 2012.
S. Bettoni, R. Corsini, A. Vivoli (CERN) CLIC drive beam injector design.
Beam Measurements After the Sources at the Ion Source Test Stand Including H 2, N 2 and Kr Gas Injection R Scrivens, Linac 4 10/04/2014 Lots of input from.
Status of the Simulations on Photo Injector Optimization for Low Charges Yauhen Kot BD Meeting,
Light Spectrum: Photosynthesis Matt, Mike, Julia.
R. Miyamoto, MEBT Lattice Optimization, ESS AD Beam Physics Internal Review 1 MEBT Lattice Optimization Ryoichi Miyamoto (ESS) For Beam Physics Group,
ELENA RF Manipulations S. Hancock. Apart from debunching before and rebunching after cooling, the principal role of the rf is to decelerate the beam and.
Ionization Injection E. Öz Max Planck Institute Für Physik.
FETS Ion Source Diagnostics: The Pepperpot
F Sergei Nagaitsev (FNAL) Webex meeting Oct ICD-2 chopper requirements and proposal #1.
RFQ CAD Model Tolerance Studies Simon Jolly 2 nd May 2012.
CLIC Frequency Multiplication System aka Combiner Rings Piotr Skowronski Caterina Biscari Javier Barranco 21 Oct IWLC 2010.
RF manipulations in SIS 18 and SIS 100 O. Chorniy.
CSC 108H: Introduction to Computer Programming
M. Migliorati, C. Vaccarezza INFN - LNF
GPT Simulations of the Ion Source Beam
Observation of a “cusp” in the decay K±  p±pp
Pepperpot Emittance Measurements of the FETS Ion Source
Time-Reversed Particle Simulations In GPT (or “There And Back Again”)
Jean Ballet, CEA Saclay GSFC, 31 May 2006 All-sky source search
Optimisation of the FETS RFQ
Data Analysis in Particle Physics
CSCI1600: Embedded and Real Time Software
All our thanks to Kevin Li and Michael Schenck BE/ABP-HSC CERN
Beam dynamics requirements after LS2
Regression testing Tor Stållhane.
commissioning and measurements
CSCI1600: Embedded and Real Time Software
Injector for the Electron Cooler
JLEIC Ion Integration Goals
Simon Jolly UKNFIC Meeting 25th April 2008
Presentation transcript:

RFQ Exit Bunch Modelling Simon Jolly 25 th July 2012

RFQ Exit Bunches To correctly model MEBT, need a single, discrete exit bunch from RFQ. Problem: single “injected” bunch (simulation of 1 RF period of DC input beam) separates into several “exit” bunches. Solution: recombine several bunches into single exit bunch. Problem: how do we do the recombination… 27/06/12Simon Jolly, University College London2

Exit Bunch Recombination There are 4 methods we could use to recombine our multiple exit bunches into a single bunch for the MEBT simulation: –Select a single timeslice and move bunches in space by N RF periods in space to overlap bunches. –Select bunches from several timeslices, each separated by 1 RF period, and select the bunch at the same point in space. –Select all bunches from a single screen, reconstructing the spatial positions from the time distribution. –Only select a single bunch with the most particles in it. 27/06/12Simon Jolly, University College London3

1: Single Time Step In a single simulation timestep, all bunches are present but at different points in space. To recombine them into a single bunch, move the fast/slow bunches by N RF periods so all the bunches overlapp. Good: –All data present in single data slice. –Gets space charge right using SCtree3D. Bad: –Particle dynamics changes along beam path. –Not overlapping “same” bunch but bunches that have undergone more/less dynamics. 27/06/12Simon Jolly, University College London4 1 RF Period

2: Single Region in Space Select a region of space that you wish to extract the exit bunch from. Choose a timestep where the first bunch is exactly centred in this region. 27/06/12Simon Jolly, University College London5

2: Single Region in Space Add the particles in this bunch to the “exit” bunch. 27/06/12Simon Jolly, University College London6

2: Single Region in Space Move on N timeslices, where N  t = 1 RF period. 27/06/12Simon Jolly, University College London7

2: Single Region in Space Add the particles in the next bunch to the “exit” bunch. 27/06/12Simon Jolly, University College London8

2: Single Region in Space Repeat for all bunches with particles above threshold energy (2.8 MeV). Good: –Gets beam dynamics exactly right. Bad: –Computationally intense: need all simulation data. –Have to select exit bunch by hand (only need to do this once though). –Space charge not quite right: need to select sub-region to avoid unphysical distortions in particle density. 27/06/12Simon Jolly, University College London9

3: Single Screen Screen contains all particle data at a single position in space. To reconstruct beam distribution, create longitudinal distribution in time. 27/06/12Simon Jolly, University College London10

3: Single Screen Add/subtract N RF periods from each particle time to create single bunch. Use  x,  y and  z values for each particle to reconstruct single bunch eg. z = ct  z. Good: –All data present in single screen. –Much more economical to extract data: only output screens. Bad: –Particle dynamics changes along beam path due to space charge. –Distortions more pronounced due to evolving space charge. –Greater need to select particle subset: requires more simulated particles. 27/06/12Simon Jolly, University College London11 1 RF Period

4: Single Bunch Only select single bunch containing the most particles for exit bunch. 27/06/12Simon Jolly, University College London12 Good: –Simplest! No reconstruction required. Bad: –Still have to select time slice by hand. –Space charge wrong: parts of distribution are missing, so need large particle numbers. –Missing parts may contain different dynamics.

Results Ran single simulation with 100k particles. Extract 3 sets of data: –Several bunches in single timeslice 1571 with central bunch outside RFQ exit datum (method 1). –One bunch after RFQ exit datum from several timeslices 10 steps apart (1 timestep = 0.1 RF period) and overlap (method 2). –Several bunches from single screen at RFQ exit datum (method 3). RFQ Exit Datum fixed at outer face of main exit flange, NOT insert. 27/06/12Simon Jolly, University College London13 Output plane for RFQ GPT simulations

Single Timestep: X-Y 27/06/12Simon Jolly, University College London14

Multiple Timesteps: X-Y 27/06/12Simon Jolly, University College London15

Single Timestep: X-X’ 27/06/12Simon Jolly, University College London16

Multiple Timesteps: X-X’ 27/06/12Simon Jolly, University College London17

Single Timestep: Y-Y’ 27/06/12Simon Jolly, University College London18

Multiple Timesteps: Y-Y’ 27/06/12Simon Jolly, University College London19

Single Timestep: Z-Y 27/06/12Simon Jolly, University College London20

Multiple Timesteps: Z-Y 27/06/12Simon Jolly, University College London21

Single Timestep: Z-E 27/06/12Simon Jolly, University College London22

Multiple Timesteps: Z-E 27/06/12Simon Jolly, University College London23

Preliminary Conclusions Difference between several bunches at single timeslice and single bunch from several timeslices clear. Bunches from single timestep don’t overlap: difference from drift too great. “Fast” bunches have different dynamics (hollow z-E distribution): –Need to include this data (about 2% of transmitted beam). –Single bunch not good enough. 27/06/12Simon Jolly, University College London24

Multiple Timesteps: X-Y 27/06/12Simon Jolly, University College London25

Single Screen: X-Y 27/06/12Simon Jolly, University College London26

Multiple Timesteps: X-X’ 27/06/12Simon Jolly, University College London27

Single Screen: X-X’ 27/06/12Simon Jolly, University College London28

Multiple Timesteps: Y-Y’ 27/06/12Simon Jolly, University College London29

Single Screen: Y-Y’ 27/06/12Simon Jolly, University College London30

Multiple Timesteps: Z-Y 27/06/12Simon Jolly, University College London31

Single Screen: Z-Y 27/06/12Simon Jolly, University College London32

Multiple Timesteps: Z-E 27/06/12Simon Jolly, University College London33

Single Screen: Z-E 27/06/12Simon Jolly, University College London34

Conclusions Whatever method is chosen, we need to include multiple bunches at the same point in space. This leaves 2 options: –Multiple timeslices with manual selection of single bunch from each. –Automatic selection of bunches from single screen. Both methods give broadly similar results: –Former has slightly more accurate distribution since latter doesn’t (and can’t) use space charge in extrapolating particle positions. –Latter requires much smaller output file, less computing to extract data and no manual intervention. I will still need to run simulations with all timesteps for particle dynamics/losses, so take your pick… 27/06/12Simon Jolly, University College London35