1. M. Furman, HHH2004 Session 6B: “Overview of EC Simulation Codes” p. 1 Overview of Electron-Cloud Simulation Codes Session 6B Miguel A. Furman LBNL First CARE-HHH APD Workshop on Beam Dynamics in Future Hadron Colliders and Rapidly Cycling High-Intensity Synchrotrons CERN, 8-11 November 2004 HHH 2004 Lawrence Berkeley National Laboratory

2. M. Furman, HHH2004 Session 6B: “Overview of EC Simulation Codes” p. 2 Acknowledgments My gratitude for collaboration or education over time to: A. Adelmann, G. Arduini, V. Baglin, M. Blaskiewicz, O. Brüning, Y. H. Cai, R. Cimino, I. Collins, O. Gröbner, K. Harkay, S. Heifets, N. Hilleret, J. M. Jiménez, R. Kirby, A. Kulikov, G. Lambertson, R. Macek, K. Ohmi, M. Pivi, G. Rumolo, D. Schulte, F. Zimmermann. I stole many slides from the ECLOUD’04 talks http://icfa-ecloud04.web.cern.ch/icfa-ecloud04/ Lawrence Berkeley National Laboratory My apologies for the incompleteness of this talk please bring omissions to my attention

3. M. Furman, HHH2004 Session 6B: “Overview of EC Simulation Codes” p. 3 Summary List of codes and features; contact persons, status,… Code features, sample results The CERN e-cloud comparisons center Current and future directions Lawrence Berkeley National Laboratory

4. M. Furman, HHH2004 Session 6B: “Overview of EC Simulation Codes” p. 4 Types of codes Lawrence Berkeley National Laboratory EC buildup codes: –beam is prescribed (not dynamical, except possibly for multibunch dipole motion) –electrons are dynamical (macroparticles) –vacuum chamber geometry, various electron sources Instability codes: –e-cloud is prescribed, at least initially; either lens or particle cloud –beam is dynamical (macroparticles) Self-consistent codes: –various degrees of self-consistency –both beam and e-cloud are dynamical –typically 3D ; may accept an input lattice description –may or may not describe e-wall collisions (SEY) –ultimately: model gas desorption, photoelectric effect, ionization, stray particles/wall collisions, secondary ionization Map code (MEC) (later)

5. M. Furman, HHH2004 Session 6B: “Overview of EC Simulation Codes” p. 5 Code table (incomplete; possible errors) SR=synchrotron rad. photoelectrons; SE=secondary electron emission; IZ=ionization of resid. gas; BPL=beam-particle losses SC=self-consistent; Lawrence Berkeley National Laboratory

6. M. Furman, HHH2004 Session 6B: “Overview of EC Simulation Codes” p. 6 Sample build up simulation:  e vs. time K. Ohmi, T. Koyama and C. Ohmori, PRSTAB 5, 114402 (2002) JPARC PSR ISIS SNSAGS Lawrence Berkeley National Laboratory

7. M. Furman, HHH2004 Session 6B: “Overview of EC Simulation Codes” p. 7 E-cloud sample simulation in a quad (CLOUDLAND) L. Wang, ECLOUD’04 Lawrence Berkeley National Laboratory

8. M. Furman, HHH2004 Session 6B: “Overview of EC Simulation Codes” p. 8 HEADTAIL simulation setup M. Pivi, ECLOUD’04 Lawrence Berkeley National Laboratory

9. M. Furman, HHH2004 Session 6B: “Overview of EC Simulation Codes” p. 9 QUICKPIC and HEADTAIL results:  vs. time Red : QuickPIC(Single Kick Mode) Blue : HEAD-TAIL Benchmarking: Single Kick QuickPIC vs. HEAD-TAIL Benchmarking: Single Kick QuickPIC vs. HEAD-TAIL (LHC params.) For accurate benchmarking, QuickPIC is modified to be in single kick regime Good agreement between the two codes. LHC parameters have been used for benchmarking purpose. A. Ghalam, ECLOUD’04 Lawrence Berkeley National Laboratory

10. M. Furman, HHH2004 Session 6B: “Overview of EC Simulation Codes” p. 10 QUICKPIC and HEADTAIL: more Growth rate changes with the number of kicks! QuickPIC Results for LHC HEAD-TAIL results for LHC Green : 4 Kicks/Turn Blue : 2 Kicks/Turn Red : 1Kick/Turn Aqua : 16Kicks/Turn A. Ghalam, ECLOUD’04 Lawrence Berkeley National Laboratory QuickPIC and HEADTAIL results for  vs. time E. Benedetto

11. M. Furman, HHH2004 Session 6B: “Overview of EC Simulation Codes” p. 11 Contemporary developments Do we need self-consistency? –Yes, in some cases: At PSR, electron-cloud signal is 10-100 times larger for unstable beam than for stable Do we need the 3rd dimension? –Yes, for long bunches (PSR) (see PSR quad movie) –Probably yes for long bunch trains and long/complicated machine lattices Lawrence Berkeley National Laboratory

12. M. Furman, HHH2004 Session 6B: “Overview of EC Simulation Codes” p. 12 PSR EC instability measurements R. Macek Lawrence Berkeley National Laboratory

13. M. Furman, HHH2004 Session 6B: “Overview of EC Simulation Codes” p. 13 PSR EC instability measurements R. Macek Lawrence Berkeley National Laboratory “For high intensity unstable beams the electrons saturate our electronics. Setting up unstable beam at lower beam intensities allows us to see the electrons without saturation.” R. Macek

14. M. Furman, HHH2004 Session 6B: “Overview of EC Simulation Codes” p. 14 Self-consistency plan R. Cohen, ECLOUD’04 Lawrence Berkeley National Laboratory roadmap for WARP+POSINST

15. M. Furman, HHH2004 Session 6B: “Overview of EC Simulation Codes” p. 15 Self-consistency plan A. Shishlo, ECLOUD’04 Lawrence Berkeley National Laboratory

16. M. Furman, HHH2004 Session 6B: “Overview of EC Simulation Codes” p. 16 Benchmarking ORBIT Y. Sato, ECLOUD’04 Lawrence Berkeley National Laboratory Theory: two-stream instability of coupled continuous beam-continuous ecloud: centroids as a f. of time (Koshkarev & Zenkevich; Keil & Zotter)

17. M. Furman, HHH2004 Session 6B: “Overview of EC Simulation Codes” p. 17 A map code (“MEC”) Lawrence Berkeley National Laboratory U. Iriso, ECLOUD’04 Relate ecloud density at time t to density at t-  t by a heuristic nonlinear map

18. M. Furman, HHH2004 Session 6B: “Overview of EC Simulation Codes” p. 18 CERN code comparisons center http://wwwslap.cern.ch/collective/ecloud02/ecsim/ Lawrence Berkeley National Laboratory Established by F. Zimmermann after ECLOUD’02 Input parameters for “standard” test cases are spelled out Everybody is invited to contribute!

19. M. Furman, HHH2004 Session 6B: “Overview of EC Simulation Codes” p. 19 CERN code comparisons center contd. Comparison of Build Up Simulations simulation results: electron line charge (total no. of e- per unit length) vs time # ECLOUD code (eps file) FZ,GR, 23.07.2002 # COUNTRYCLOUD code (eps file) Lanfa Wang, August 2002 # BNL code (eps file) Mike Blaskiewicz, August 2002 # PEI code (ps file) Kazuhito Ohmi, September 2002 # POSINST code (eps file) Mauro Pivi and Miguel Furman, September 2002; details of the LBNL simulation Last updated 23 August 2002, FZ Comparison of Instability Simulations simulation result: emittances vs. time # HEADTAIL code (ps file) Giovanni Rumolo, August 2002 # PEHTS code (ps file) Kazuhito Ohmi, November 2002, comments and additional studies (pdf) # QUICKPIC code (pdf file) Ali Ghalam, Tom Katsouleas, Giovanni Rumolo, November 2002 Last updated 29 November 2002, FZ Measurements and Parametrizations of Secondary Emission Secondary Electron Emission Data for the Simulation of Electron Cloud by N. Hilleret et al. (contribution to ECLOUD'02 Proceedings) Excel file by N. Hilleret Last updated 15 July 2002, FZ Lawrence Berkeley National Laboratory

20. M. Furman, HHH2004 Session 6B: “Overview of EC Simulation Codes” p. 20 Possible future developments More “benchmarking” –debugging (code should calculate what is supposed to calculate) –validation (results should agree with established analytic result for specific cases) –comparisons (two codes should agree if the model is the same) –verification (code should agree with measurements) ECLOUD simulations vs. SPS measurements POSINST simulations vs. APS and PSR measurements Others… Move in 2 opposite directions: –More complete, detailed, quantitative predictions Ultimately requires fully self-consistent 3D calculations –Simplified descriptions, few parameters, qualitative results with broad applicability Identify a few basic relevant variables and input parameters (MEC code very promising in this regard) Lawrence Berkeley National Laboratory

21. M. Furman, HHH2004 Session 6B: “Overview of EC Simulation Codes” p. 21 Extra material

22. M. Furman, HHH2004 Session 6B: “Overview of EC Simulation Codes” p. 22 BIM in the APS (Advanced Photon Source, Argonne) Lawrence Berkeley National Laboratory (Furman, Pivi, Harkay, Rosenberg, PAC01) time-averaged e – flux at wall vs. bunch spacing measured simulated e + beam, 10-bunch train, field-free region

23. M. Furman, HHH2004 Session 6B: “Overview of EC Simulation Codes” p. 23 BIM for long bunches: PSR bunch length ~60 m >>  t –a portion the EC phase space is in resonance with the “bounce frequency” –“trailing edge multipacting” (Macek; Blaskiewicz, Danilov, Alexandrov,…) Lawrence Berkeley National Laboratory ED42Y electron detector signal 8  C/pulse beam 435  A/cm 2 (simulation input) electron signal measured (R. Macek) simulated (M. Pivi) (  max =2.05)

24. M. Furman, HHH2004 Session 6B: “Overview of EC Simulation Codes” p. 24 Future computer  2004: NERSC: 8000 processors (power PC3), ~8 Tflops  2004: Red Storm: ~11600 processor Opteron-based MPP [>40 Tflops]  2005: ~1280-Processor 64-bit Linux Cluster [~10 TF]  2006 Red Storm upgrade ~20K nodes, 160 TF.  2008--9 Red Widow ~ 50K nodes, 1000 TF. (?) Lawrence Berkeley National Laboratory Each center will get one: Sandia ORNL Pittsburgh