1. Low Emittance Rings 2014 Workshop INFN-LNF, 18. September 2014 Low Emittance Studies at 3 GeV at PETRA III Joachim Keil DESY

2. Joachim Keil | Low emittance studies at 3 GeV at PETRA III | 18. September 2014 | Page 2 Outline > Motivation > Overview of PETRA III > Energy choice for low emittance studies > Beam diagnostics (emittance & bunch length) > Measurements with flat beam > Measurements with round beam > Summary

3. Joachim Keil | Low emittance studies at 3 GeV at PETRA III | 18. September 2014 | Page 3 Motivation > Machine performance of low emittance rings is limited by collective effects (Intra beam scattering, Touschek effect, etc.) > Tests with reduced energy on a running machine (like PETRA III) can be used to reach ultra-small emittances and test conditions of LERs > Do measurements and predictions of formulas/codes fit? > Low emittance studies of intra beam scattering have been carried out between July 2013 and Feb. 2014 during MD times > End of studies due to beginning of work for the PETRA III extension project (P3X)

4. Joachim Keil | Low emittance studies at 3 GeV at PETRA III | 18. September 2014 | Page 4 PETRA III at DESY, Hamburg Parameters > Circumference: 2304 m > Energy: 6 GeV > Emittance: 1 nm·rad > Current: 100 mA > Bunches: 40 / 960 PETRA III N E W S Experimental Hall > 3 rd generation light source > In operation since 2009 > FODO/DBA structure > Top-up operation > 14 user beam lines > Now: P3 extension P3X 10 more beam lines in 2 new halls

5. Joachim Keil | Low emittance studies at 3 GeV at PETRA III | 18. September 2014 | Page 5 Layout of PETRA III 9 DBA cells 14 beam lines 20 damping wigglers (wiggler length 4 m) FODO cells Straight sections N & W New octant Arcs and straight sections PETRA III: Hybrid lattice of FODO and DBA cells

6. Joachim Keil | Low emittance studies at 3 GeV at PETRA III | 18. September 2014 | Page 6 Scaling of the Emittance with Energy at PETRA III The natural emittance with damping wigglers can be written as: with the contributions The factor F depends on the energy loss ratio Emittance minimum is ≈2.55 GeV → Choice for experiments: 3 GeV (and in addition 5 GeV) Emittance vs. Energy For dipoles: I 2 d and I 5 d are independent of  For wigglers: I 2 w ∝ 1/  2 and I 5 w ∝ 1/  5. dipoles wigglers F =3.59 at 6 GeV

7. Joachim Keil | Low emittance studies at 3 GeV at PETRA III | 18. September 2014 | Page 7 How much Emittance Reduction can be expected? 158 pm·rad @ 3 GeV 580 pm·rad @ 5 GeV > Edge focusing of wiggler poles is getting stronger at lower energies > Matching of optics in wiggler sections for different energies is necessary > Exact calculation (MAD-X):  158 pm·rad at 3 GeV  580 pm·rad at 5 GeV > Main emittance contribution for low energies from wigglers → Good optics model of wigglers necessary Improved wiggler model

8. Joachim Keil | Low emittance studies at 3 GeV at PETRA III | 18. September 2014 | Page 8 Improved Damping Wiggler Model > 20 permanent magnet damping wigglers decrease the natural emittance from 4.3 to 1.0 nm·rad > Damping wiggler model used so far: Hard edge model with one bending magnet and two drifts per period > Optics model was improved: more slices per period, an accurate model of end poles and fitting to measured magnetic field data > Better modelling of SR integrals I 2, I 5 and edge focusing ParameterValue Total Length L4 m Period length w 0.2 m Fixed gap g24 mm Peak field B 0 1.57 T Number of poles37 (main) + 4 (end) ⨜ B 2 ds 4.16 T 2 ·m Wiggler parameters

9. Joachim Keil | Low emittance studies at 3 GeV at PETRA III | 18. September 2014 | Page 9 Beam Diagnostics: Bunch Length Measurements > Bunch length measurements with a streak camera system (Hamamatsu C5680) > Installed in the optical beam line using visible light from a dipole magnet (NL 50 m) > Extraction of light using a water cooled copper mirror; optical relay system (25 m length) guides light into an experimental hut > No parallel use with the SR interferometer! synchrotron light from bending magnet Mirror boxes visible light beam pipe

10. Joachim Keil | Low emittance studies at 3 GeV at PETRA III | 18. September 2014 | Page 10 Beam Diagnostics: Emittance Measurements > Emittance measurements at 6 GeV uses imaging of a compound refractive lens (CRL) optics installed in X-ray beam line. Intensity is too low at 3 GeV. > Used a double slit wavefront-division interferometer in the optical beam line > SR interferometer can measure either x or y plane; no simultaneous use of the streak camera > The modulation depth of interference pattern (visibility) is a function of the beam size > Ver. plane: For very small beam sizes the visibility is nearly 1 (even for maximum slit distance) → The minimum of measureable emittance is reached at 3 GeV Ü alignment mirror lens double slit magnification lens Glan- Thomson polariser bandpass filter & CCD G. Kube SR Interferometer

11. Joachim Keil | Low emittance studies at 3 GeV at PETRA III | 18. September 2014 | Page 11 Studies at 5 GeV > Established an 5 GeV optics from scratch (booster, transport channel, first turn steering in PETRA III) > Corrected the optics: ORM, global decoupling, chromaticity, dispersion function, … > A horizontal emittance of  x = 570 pm·rad was measured with the SR interferometer > Good agreement with theoretical emittance of 580 pm·rad for zero current → wiggler model ok! > Achieved 100 mA in 480 bunches; did some test runs for users > Moderate emittance growth due to IBS observed Horizontal plane

12. Joachim Keil | Low emittance studies at 3 GeV at PETRA III | 18. September 2014 | Page 12 Studies at 3 GeV > Same procedure to set up the machine at 3 GeV: optics corrections (one iteration), etc. > Hor. emittance measured for zero current:  x = (155 ± 10) pm·rad > Natural emittance without IBS (MAD-X):  x = 158 pm·rad > Higher requirements for residual dispersion in wiggler straight section: More iterations of dispersion correction are needed > Strong emittance growth with bunch current observed Horizontal plane 16.7.2013: I = 5 mA, 480 bunches, I b =10.4 µA

13. Joachim Keil | Low emittance studies at 3 GeV at PETRA III | 18. September 2014 | Page 13 Flat Beam Measurements – July 2014 Simulation parameters:   x0 = 130 pm·rad,  y0 = 4.55 pm·rad   = 1.2%, r=0.34   z0 = 32.6 ps  Ver. interferometer resolution ≈7.5 pm · rad?  No bunch length increase due to PWD measured Resolution corrected → Main contribution to  y is ver. dispersion

14. Joachim Keil | Low emittance studies at 3 GeV at PETRA III | 18. September 2014 | Page 14 Comparison: Bunch Length Growth at 6 GeV > IBS negligible at 6 GeV > Bunch length growth only due to potential-well distortion (PWD) > Measured bunch length: s z0 = (41.7 ± 0.5) ps > Theory: 40.8 ps (19 MV) > Current dependency: ~0.65 ps/mA > Fit of effective inductance impedance from Zotter’s formula for PWD:  Im(Z/n eff ) = -0.038  /m Flat beam; 6 GeV 0.65 ps/mA

15. Joachim Keil | Low emittance studies at 3 GeV at PETRA III | 18. September 2014 | Page 15 IBS Growth Times Calculations

16. Joachim Keil | Low emittance studies at 3 GeV at PETRA III | 18. September 2014 | Page 16 IBS Calculations with Synchrotron Radiation > Compute time development of emittances self- consistent by solving (K. Bane): between IBS emittance growth, radiation damping and quantum excitation > Find equilibrium emittances: > Assumption: Rad. damping times IBS growth times Emittances at I b =0 with ver. dispersion Betatron coupling

17. Joachim Keil | Low emittance studies at 3 GeV at PETRA III | 18. September 2014 | Page 17 Round Beam Measurements – Jan 2014 > Injection was not possible on difference coupling resonance → injected at normal tunes and moved tunes in steps to coupling resonance > Horizontal/vertical emittance at zero current:   x = (87 ± 10) pm·rad   y = (83 ± 3) pm·rad Horizontal plane Vertical plane ≈ “round beam”

18. Joachim Keil | Low emittance studies at 3 GeV at PETRA III | 18. September 2014 | Page 18 Round Beam - Emittance Growth > Measured current dependency of emittance on the difference coupling resonance Q x – Q y = 6 > Zero current emittance:   x0 = (125 ± 10) pm·rad   y0 = (120 ± 10) pm·rad  Emittances corrected by  x and  y at the interferometer (from ORM) > Emittance growth:  d  x /dI b = (120 ± 15) pm·rad/mA  d  y /dI b = (138 ± 15) pm·rad/mA > Emittance growth of round beam is lower compared to a flat beam > Emittance growth is equal in both planes for round beams

19. Joachim Keil | Low emittance studies at 3 GeV at PETRA III | 18. September 2014 | Page 19 Round Beam – Bunch Length Growth > Bunch length growth with current:  Flat beam: ~2.5 ps/mA  Round beam: ~0.7 ps/mA Round beam; 3 GeVFlat beam; 3 GeV The bunch length growth of a round beam is smaller ~0.7 ps/mA ~2.5 ps/mA

20. Joachim Keil | Low emittance studies at 3 GeV at PETRA III | 18. September 2014 | Page 20 Round Beam - Simulations Simulation parameters:   x0 =  y0 =105 pm · rad   = 100% (full coupling)   z0 = 31.7 ps  No bunch length increase due to PWD

21. Joachim Keil | Low emittance studies at 3 GeV at PETRA III | 18. September 2014 | Page 21 Summary > The natural emittance at zero current at 3 and 5 GeV is in good agreement with theoretical expectations. > For flat beams an hor. emittance of 155 pm · rad was measured for zero current. The vertical emittance is near the resolution limit of the interferometer. > The vertical emittance for flat beams is dominated by vertical dispersion. > For round beams emittances of 87/83 pm · rad were measured for zero current at the coupling resonance. > The emittance growth with round beams was experimentally verified to be smaller compared to a flat beam and equal. > IBS simulations with the Bjorken-Mtingwa formula are in agreement with measurements. A tail-cut applied to the Coulomb logarithm seems to be necessary.