Design and Performance of Modern Storage Ring Light Sources - Robert Hettel, SLAC
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
Light source brightness FEL and storage ring light sources are complementary FELs: high peak brightness, strongly perturbs sample under study Storage rings: low peak brightness, minimal sample perturbation, high rep rate/stability, many simultaneous users, ….
Outline • Introduction • Diffraction limited emittance, brightness and coherence • Properties of 4th generation storage ring (4GSR) light sources and science applications • 4GSR lattice implementations • 4GSR challenges and solutions • Future of DLSRs • Appendices • 4GSR R&D Topics • 4GSR Accelerator and Science Reference • Low Emittance Ring and 4GSR Workshops
Acknowledgments Many appreciated contributions from: D. Robin and C. Steier, ALS M. Borland, L. Emery, APS R. Bartolini, Diamond P. Raimondi, ESRF M. Eriksson, S. Leemann, MAX-IV L. Liu, Sirius A. Streun, SLS L. Nadolski, Soleil H. Ohkuma, K. Soutome, H. Tanaka, SPring-8 Y. Cai, T. Rabedeau, Z. Huang SLAC and SPEAR3 Beam Physics Groups SLAC Directors, I. Lindau, C. Pellegrini, J. Stohr, H. Winick and participants in FLS, LowERing and DLSR workshops
4GSR (DLSR) Reference Journal of Synchrotron Radiation (2014).21 • 8 articles on accelerator physics and technology • 2 articles on MBA rings in construction (MAX-IV and SIrius) • 10 articles on scientific applications • 4 articles on X-ray beam line technology (optics, instrumentation, detectors, etc.) See Appendix 2
SESAME (Jordan) Environmental science & Physics Laboratory Archaeology Laboratory Energy; 2.5 GeV Circumference; 133m Emittance; 26 nm-rad BessyI 12 Insertion Devices 0.8 GeV injector 13 Bend Magnet beam lines Maximum beam line length; 37m Space for future full energy injector in main ring tunnel Cheap Bio-Medical Materials science Laboratory Laboratory
Sirius (Brazil) Energy; 3 GeV Circumference; 518 m Emittance; 0.2 nm-rad Up to 18 IDs Up to 18 BM beam lines Maximum beam line length: 150 m ~505 M USD
What is needed to design, build and operate a light source Definition of the science mission for the facility • areas of research Must be very clearly articulated in the • industrial partners user community and funding agencies • education goals • etc. Specification of photon parameters • spectral range • spectral brightness and flux • beam size requirements • photons/pulse • pulse length • etc.
What is needed - cont. Specification of beam line and source requirements • beam line applications • number of IDs and bend magnet lines • ID types and lengths • beam line optics and lengths • etc. Specification of accelerator parameters • electron energy, emittance, current, bunch length • ring circumference, number and length of straight sections • lattice design and accelerator physics studies • injector and injection • etc.
What is needed - cont. Accelerator engineering design • mechanical, vacuum, low and high power electrical (DC and pulsed power, etc), high and low power RF, electronics and instrumentation, computer control system, feedback systems process control, safety systems, etc. Beam line engineering design • X-ray optics, precision motion control, detectors, computer controls, etc. Radiation physics and safety • beam loss and radiation studies • shielding design • specification of radiation safety system requirements Facility engineering design • stable floor, building, water, air conditioning, electrical utilities, water cooling, gas and LN infrastructure, etc. Project management and business services • Extremely important!
The world is moving to ever brighter ring sources 2-‐bend achromat 7-‐ bend achromat 5-‐ bend achromat NSLS-‐II MAX-‐IV SIRIUS BNL: NSLS-‐II (2014): 3 GeV, Sweden: MAX-‐4 (2016): 3 GeV, Brazil: SIRIUS (2016/17): 3 GeV,
Brightness and coherence of present rings Best-available brightness and coherent fraction for selected presently-operating rings, with operating parameters and insertion devices. Parameters provided by facility contacts. Compiled by M. Borland for BESAC Sub-Committee meeting, July 2013.
Brightness and coherence of near-future rings Selected rings/upgrades now under construction, with anticipated parameters and insertion devices Parameters provided by facility contacts. Compiled by M. Borland for BESAC Sub-Committee meeting, July 2013.
Brightness and coherence of planned rings Selected upgrades now being planned (except APS-II), with anticipated parameters and ESRF-specified insertion devices. Parameters provided by facility contacts. Compiled by M. Borland for BESAC Sub-Committee meeting, July 2013. Notes: 1. ESRF-II: 6 GeV, 200mA, 150 pm 2. SPring-8-II: 6 GeV, 300 mA, 67pm
Brightness and coherence of future rings Selected diffraction-limited rings now being designed, with identical Nb3Sn super- conducting insertion devices and some PM devices. Parameters provided by facility contacts. Compiled by M. Borland for BESAC Sub-Committee meeting, July 2013. Notes: 1. 0.2km/2GeV: ALS-II, 52 pm 2. 1.1km/6GeV: APS-II, 80 pm 3. 1.4km/6GeV: SP8-II, 2nd stage, 34 pm 4. 2.2km/6GeV: PEP-X, 5 pm 5. 6.2km/9GeV: tauUSR, 3 pm 6. Except for 0.2km ring, uniform selection of SCUs and APS HPMs used.
Spectral brightness and coherence Spectral brightness: photon density in 6D phase space N ph (λ) Bavg (λ) ∝ (ε x (e−) ⊕ ε r (λ))(ε y (e−) ⊕ ε r (λ))(s ⋅ % BW) N ph (λ) B pk (λ) ∝ (ε x (e−) ⊕ ε r (λ))(ε y (e−) ⊕ ε r (λ))(σ t ⋅ % BW) Coherent fraction: ε r (λ ) ε r (λ ) f coh (λ) = ⋅ (ε x (e−) ⊕ ε r (λ)) (ε y (e−) ⊕ ε r (λ)) Coherent flux: 2 ⎛ λ ⎞ Fcoh (λ) = f coh (λ) ⋅ F(λ) = B avg (λ) ⋅ ⎜ ⎟ ⎝ 2 ⎠
Diffraction-limited emittance εr(λ) K-J Kim in Characteristics of Undulator Radiation, AIP 1989 Coherent beam of wavelength λ focused to spot size Δx will diffract with angle Δψ = ~λ/Δx In transversely coherent beam, spatial distribution Ek(x,z) for wavenumber k is related to angular distribution ℇk( , ) by Fourier transform (for 1-D in x): ℇ↓ ( , )=1/√2 ∫↑▒ ↓ ( , ) ↑ ↓ ( , )=1/√2 ∫↑▒ℇ↓ ( , ) − ↑ ≪1 ⇒ ↓ ↑ ( ) ↓ ( )= /4 = ↓ ( ) Diffraction limited emittance for coherent Gaussian photon distribution
Diffraction-limited emittance – cont. Fitting Gaussian profiles to spatial and angular profiles for undulator radiation at λ: P. Elleaume, in Wigglers, Undulators, and Their Applications, 2003. ↓ ↑ ( )≈1.9√2 /4 ↓ ↑ ( )≈√ /2 ⇒ ↓ ↑ ( ) ↓ ↑ ( )≈ /2 = ↓ ( ) Diffraction limited emittance for undulator radiation from single electron filament Other similar estimates for εr(λ): W. Joho, SLS Note 4/95; O. Chubar, FLS2012
Diffraction-limited emittance – cont. “Ultimate storage rings
X-ray emittance from electron source Transverse emittance Σx,y (λ) of X-ray beam from undulator (length L) is convolution of photon emittance εr from e- filament and e- emittance εx,y(e-) (Gaussian beams): ↓ ( )⊕ ↓ , ( −)=√ ↓ ↑2 ( )+ ↓ , ↑2 ( −) √ ↑′ ↓ ↑2 ( )+ ↑′ ↓ , ↑2 ( − ↓ ↑ ( )≈√2 /2 Here ′↓ ↑ ( )= ↓ ↑ ( )≈√ /2 ↓ , ↑ ( −)=√ ↓ , ↓ , +( ↓ , ′↓ , ↑ ( −)=√ ↓ , / ↓ , +( ′↓ , )↑2 (η, η’ =0 for )↑2 achromat) Transverse emittance Σx,y minimized when εx,y is minimized and photon and e- phase space orientations are matched: Note: many ↓ ↑ ( )/ ′↓ ↑ ( ) = ↓ , ↑ ( −)/ ↑′authors ↓ , ↑ ( cite −) ⇒ ↓ , ↑ = / ↓ , ↑ = / 2
Coherent fraction Coherent flux is important too: a low coherent fraction and high flux can yield the same coherent flux as a high coherent fraction and low flux Optimize trade- off between low of emittance vs. stored current • Many rings operate now with εy 2πεx,y(e-)
Storage ring emittance landscape CLS MAX-II PLSSPEAR3 SPEAR3 10 CANDLE SAGA-LS ALS ELETTRA SLS ASP BESSY-II ALBA ESRF SSRF SPring8 SOLEIL APS Diamond Emittance (nm) TPS PETRA-III 1 SPEAR3 QBA • NSLS-II SIRIUS MAX-IV ESRF-II Planned light source 0.1 • Existing light source APS-U • PEP-X 0.01 0 1 2 3 4 5 6 7 8 9 PEP-X Energy (GeV) • 24
Properties of 4GSRs • Brightness and coherence are as high as possible for given beam current • Small horizontal and vertical beam dimensions and the possibility of “round” beams – good for X-ray optics, minimal need for aperturing • Short bunches courtesy of C. Steier ~5-10 ps RMS from low momentum compaction factor – bunch lengthening usually needed to combat emittance growth from IBS and improve lifetime; synchrotron frequency < 1 kHz for large rings • “Long” lifetime: Touschek lifetime increases with small bunch dimensions Touschek lifetime for NSLS II assuming emittance can be • Large circumference for multi-GeV rings (km) arbitrarily reduced (M. Borland, • Damping wigglers used in some cases to lattice courtesy W. Guo) combat IBS and reduce emittance by ~x2-3 • Small dynamic aperture (~mm) for aggressive lattices (⇒ on-axis “swap- out” (A. Xiao et al. NA-PAC13) or possibly longitudinal injection (M. Aiba et al., IPAC14))
4GSR properties – coherence Transversely coherent x-rays - Uniform phase wavefronts: coherent imaging, holography, speckle, etc. - Focusable to smallest spot size: nano-focus - High flux (~1014-1015 photons/sec) in small spot: slits may not be required, etc. - Round beams: H-V symmetric optics, circular zone plates, flexibility in optics Some issues with coherence: - Reduced depth of focus – a problem for some forms of imaging - Speckle from coherent beams a problem for some applications - These problems can be resolved by “spoiling” beam on beamline
Fundamental challenge: science case (in the US) XDL 2011 Workshops for ERLs and DLSRs (Cornell, June 2011) : § Diffraction Microscopy, Holography and Ptychography using Coherent Beams § Biomolecular Structure from Nnanocrystals and Diffuse Scattering § Ultra-fast Science with “Tickle and Probe” § High-pressure Science at the Edge of Feasibility § Materials Science with Coherent Nanobeams at the Edge of Feasibility § Frontier Science with X-ray Correlation Spectroscopies using Continuous Sources (time resolution ∝ B2) BESAC Subcommittee on Future Light Sources (July 10-12, 2013) A consensus report on future opportunities from scientists at ALS, APS, NSLS-II, SSRL, together with a broad community of scientists at laboratories and universities. Applications address “Grand Challenge Science”
Low emittance lattice design Horizontal (natural) emittance determined by balance between radiation damping and quantum excitation due to synchrotron radiation in all magnets: How to minimize emittance? • Reduce dispersion and beta function in bend magnets (wigglers/undulators) • Achieved by refocusing beam ‘inside’ bending magnets -> need space • ‘Split’ bending magnets -> multi bend achromats 28
The path to low emittance rings Emittance scaling with energy and circumference: ↓0 = ( , ) ↑2 /( ↓ ↑ ↓ ↑ )↑3 ∝ ↑2 / ↑3 for fixed cell type ↓ =1/1+ ↓0 ↓ = /1+ ↓0 Ns = # sectors in ring, Nd = # dipoles/sector Emittance reduction with damping wigglers: εw 1+ f 1 = ≈ U0 = energy loss/turn in dipoles εo Lw ⎛ ρ o ⎞ 2 Uw 1+ 1+ UW = energy loss/turn in wigglers ⎜ ⎟ 4π ρo ⎝ ρ w ⎠ Uo Emittance reduction with damping partition: H Gradient dipoles 2 H(s)/⇢(s)3 ds ✏x = Cq H Robinson wigglers Jx 1/⇢(s)2 ds Amplitude bumps in quads Damping partition
Common lattice options Achromat Symmetry Point Achromat Symmetry Point Dispersion Function Dispersion Function 1/2 Insertion Straight 1/2 Insertion Straight 1/2 Insertion Straight 1/2 Insertion Straight Achromat Achromat Insertion Symmetry Point Insertion Symmetry Point Achromat Symmetry Point • Early 3rd generation SR sources all used double/triple bend achromats (some with gradient dipoles) • Later optimization included detuning from Dispersion Function achromatic condition (Optimizing effective 1/2 Insertion Straight 1/2 Insertion Straight emittance) • New designs (including DLSRs) employ MBA • Damping wigglers can help (emittance, Achromat Insertion Symmetry Point damping time, IBS) but trade energy spread 30
Other ring implementations “Double DBA” for PLS-II • Compact lattice that doubled the number of straight sections for PLS • Effective emittance in straight sections is larger due to dispersion η
History of TME / MBA Work in 1990 to find theoretical minimum emittance structures – Einfeld, et al. (NIM 1993, PAC 1995, EPAC 1996) MBA is a modification of this, with (detuned) TME structure in the middle of the arc and (short) matching sections at ends Originally considered challenging (“chromaticity wall”) Max-4 is first full implementation of MBA 32
The multibend achromat optimization cycle A. Streun, PSI
Example MBA Lattices Max-4 – 7BA, octupoles, ESRF-2- HMBA, APS-U - HMBA, swap-out, M. Eriksson P. Raimondi M. Borland ALS-U – 9BA, octupoles, PEP-X – 7BA, 4th order achromat, Y. Cai swap-out, H. Tarawneh, C. Steier 34
ESRF lattice conversion - hybrid MBA Present 2BA εx/y = 147/5 pm.rad @ 6 GeV, 200 mA εx = 4 nm.rad 17 magnets Space between D1-D2 and D6-D7: β-functions and dispersion allowed to grow to enable chromaticity correction with efficient sextupoles Dipoles D1, D2, D6, D7: Bleu: Dipoles Red: Quadrupoles Green: sextupoles longitudinally varying field to further Proposed Hybrid 7BA reduce emittance εx = 150 pm.rad 29 magnets Central section: combined dipole-quadrupoles D3-4-5; high-gradient focusing quadrupoles Source points for BM beamlines have same fields, positions, angles D1 D2 D6 D7 D3 D4 D5 Courtesy P. Raimondi, ESRF.
ESRF lattice conversion – cont. High gradient quadrupoles 85 T/m Quadrupole Around 50 T/m D7 •Spec:100 T/m x 335 mm D6 •Bore radius: 11 mm •Mechanical length: 360 mm D2 D5 •1 kW D3 D4 D1 Combined dipole quadrupoles Sextupoles 0.85 T / 45 T/m & 0.34 T / 50 T/m 1700 T/m-2 Permanent magnet dipoles longitudinal gradient 0.16 - 0.6 T, magnetic gap 22 mm 2 metre long, 5 modules With a small tuning coil 1% Courtesy P. Raimondi, ESRF.
APS-U lattice conversion - hybrid MBA εx/y = 65/8 pm.rad @ 6 GeV, 200 mA
Sirius 5BA εx/y = 190-270/3 pm.rad @ 3 GeV, 350 mA 38
ALS-U 9BA εx/y = 50/50 pm.rad @ 2 GeV, 500 mA superbend option includes octupoles
Longitudinal gradient dipoles βx βy η longitudinal gradient anti-bend bend ESRF-2 APS-U A. Streun • Allow reduction in emittance and larger dispersion in chromatic correction section • Potential draw backs: • Non-trivial to build • Longer • Highest field in line with straights (for HMBA) 40
Longitudinal dipoles and anti-bends dispersion: A. Streun, SLS anti-bend off / on βx βy relaxed TME cell, 5°, 2.4 GeV, Jx ≈ 2 Emittance: 500 pm / 200 pm • Very new proposal • Allows more control over momentum compaction factor as well as separate control of beta-function and dispersion 41
Damping wigglers NSLS-2 εw 1+ f 1 = ≈ εo Lw ⎛ ρ o ⎞ 2 U 1+ 1+ w X. Huang, PEP-X ⎜ ⎟ Uo 4π ρo ⎝ ρ w ⎠ • Help to reduce the equilibrium emittance – without reducing momentum compaction factor • Also decrease damping time (beneficial for IBS) • Possible tool to control vertical emittance (vertical wiggler, vertical dispersion) • Increase energy spread (reduces higher undulator harmonics) • Can impact dynamic aperture • Occupy space, increase RF need, need to handle high power synchrotron light 42
3PW, Superbends, … L. Liu, SIRIUS ESRF-II MBA lattices might have weak field dipoles ALS-U 3PW or Superbends allow to extend spectrum to harder photon energies Can also help to reduce the emittance (longitudinal gradient dipole) – if done right Need to create space in lattice, correct for potential symmetry distortion 43
Future large DLSRs? BAPS (China - Beijing) 5 GeV, 1-1.5 km,
BAPS IHEP/Beijing:
Other methods of emittance control If ring is diffraction limited in horizontal plane, vertical emittance does not need to be smaller • Very different from current rings which run with emittance ratios between 0.1% - few % • Intrabeam scattering and Touschek lifetime require the vertical emittance to not be too small User experiments sometimes prefer round beams Control of vertical emittance is needed • Coupling control (reduces horizontal emittance, fine with swap-out, but challenging with asymmetric apertures for off-axis injection, limits tune choices) • Vertical Dispersion (local/global) – routinely used in light sources right now, but only up to maybe 10% emittance ratio • Damping Wigglers (spurious or intrinsic vertical dispersion) • Moebius ring / emittance exchange / … 46
Scaling M. Borland • However, more and more magnet require magnets to become stronger (quadupoles about quadratically, sextupoles even quicker) – Engineering limits – Nonlinear dynamics • Energy scaling is complex (magnet strength, C) 47
Scaling – cont. ALS APS-U ∝ ↑2 / ↑3 • Slope of optimized parameter sets indicates R. Bartolini that this merit function is too simple 48
Scaling – cont. A USR-type Ring Design Study, X. Huang, SSRL, 8/30/2012 4GSR design involves packing as many TME cells as possible into a fixed circumference to reduce emittance. Quadrupole and sextupole strengths increase to keep proper cell phase advance. For linear optics, cell length L scaling: √ ↓1 =√ ↓1 / ∝ For circumference C, number of cells N related to cell length by NL = C For a given quadrupole gradient limit, beam energy E: ∝ ↑2 / ↑3 ∝ ↑7/2 / ↑3 ↓1 ↓↑3/2 from 2012 (dated)
Scaling – cont. energy circumference emittance gradient B1 Form factor [GeV] [m] [nm] T/m NSLS-II 3 780 2 20 1.815 Diamond 3 561.6 2.7 20 0.915 Normalized form SLS 2.4 288 5.5 22 0.633 factor: Soleil 2.75 354.1 3.74 21 0.463 ↓1 ↓↑3/2 ↑3 / SSRF 3.4 432 3.9 21 0.418 TPS 3 518.4 1.7 18 0.387 APS 7 1104 2.5 19 0.307 PLS-II 3 281.8 5.8 22 0.286 SPEAR3 3 234 9.8 22 0.277 MAX-IV 3 528 0.24 43 0.213 SIRIUS 3 518 0.28 40 0.211 ALBA 3 266 4.5 23 0.200 ALS 1.9 196.8 2 22 0.166 spring8-II 6 1435.95 0.067 56 0.157 ESRF upgrade 6 844.4 0.15 85 0.134 APS-U 6 1104 0.06 85 0.120 SPEAR3 MBA 3 234 0.5 85 0.107 ALS 9BA 2 200 0.1 100 0.071 SLS upgrade 2.4 288 0.073 85 0.064
Fundamental challenges of low emittance from M. Borland, § Inescapable fact GRC 8/13 – To reduce the amplitude of dispersive orbits, must focus more frequently and more strongly § Focusing (quadrupole) elements have chromatic aberrations – Sextupole magnets added to correct these – Introduces higher order aberrations – More sextupoles or octupoles added to correct these... § As Nd is increased to reduce emittance1 – Stronger chromatic correction sextupoles: strengths increase like Nd3 – Dynamic acceptance decreases like 1/Nd3 – Second order chromaticities increase like Nd3 – Dipole/quadrupole bore ~1/Nd2; sextupole bore ~1/Nd1.5 § Stronger focusing leads to difficult non-linear dynamics – Poor “momentum aperture” ⇒ reduced lifetime ⇒ frequent injection – Poor “dynamic aperture” ⇒ greater difficulty injecting ⇒ on-axis injection? 1: M. Borland, DLSR Workshop, SLAC, December 2013.
Fundamental challenges – cont. § Intra-beam scattering (IBS) – Multiple electron-electron scattering in a bunch – Leads to increased emittance and energy spread – Fights the beneficial E2 scaling of emittance – Mitigations: – Many low-intensity bunches - Round beams – Bunch lengthening system - Damping wigglers APS emittance at 200 mA as a function § Beam instabilities of energy with and without IBS – Transverse: resistive wall, ion trapping in multi-bunch mode, single bunch TMCI – Beam blow-up ⇒ brilliance reduction - transverse beam oscillations ⇒ beam losses – Longitudinal: primarily from cavity HOMs – Mitigations: mode-damped cavities, smooth chamber transitions, low-Z chamber material, low charge/bunch, longer bunches, feedback § X-ray optics and detectors – Advances in optics needed to preserve coherence, handle high power densities – Detectors with higher resolution and faster readout rates are needed
Intra Beam Scattering IBS is potentially a very significant effect for low emittance rings • Mitigation typically involves some combination of increased vertical emittance, bunch lengthening (harmonic cavities), faster damping times (damping wigglers, normal IDs, …). • Optimum solution for smallest quantum excitation does not necessarily minimize IBS emittance growth -> include in lattice optimization 5.5E-11 5.0E-11 Emittance εx,y [m.rad] 4.5E-11 4.0E-11 3.5E-11 3.0E-11 0 0.1 0.2 0.3 0.4 0.5 0.6 Beam Current [mA] Beam Current [A] Example: ALS-U, 2 GeV, harmonic cavities, full coupling, full set of insertion devices, no damping wigglers 53
Effects of harmonic cavity APS-U Bunch lengthening from |Z/n|=0.5 Ohm included Bunch lengthening for 3rd, 4th, 5th harmonic cavities M. Borland Computed with haissinski (L. Emery et al.) and ibsEmitance (A. Xiao et al.) Shift of bunch time-of-arrival with missing bunch in symmetric fill pattern
Injection Methods Accumulation • Traditional injection scheme for light sources • Requires sufficiently large dynamic aperture (at injection point) • Smallest emittance lattices might have sufficient momentum aperture (lifetime) but not sufficient dynamic aperture On-Axis Swap-out • Enables injection into small dynamic aperture • Bunch (or bunch-train) is replaces with fresh bunch (or bunch-train) • Recycle or dump spent bunches - Added complexity • Requires fast kickers to minimize fill pattern gaps - Might impact the range of possible fill patterns • In use for commissioning or damping rings • Many potential advantages 55
Off-axis injection and accumulation Traditional 4-kicker injection • Stored beam is kicked towards septum • Imperfect kicker bump results in transient orbit oscillations that can degrade beam quality for users during top-up injection Pulsed multipole injection • Injected beam is kicked when passing off-axis through multipole magnet • Stored beam is unperturbed passing through center of multipole magnet Sirius 56
On-axis injection inject here Longitudinal Injection requires fast kicker (width < bunch spacing) M. Aiba, M. Böge, Á. Saá Hernández, F. Marcellini and A. Streun, this conference Bunch Replacement (Swap-Out) Injection requires fast kicker (width ~ bunch spacing or longer for pulse trains) - M. Borland, L. Emery, Proc. PAC’03
On-axis injection (swap-out) APS-U • flat top: 10s ns • rise/fall: few ns • rise/fall times determine length Swap-out injection was first proposed by M. Borland of gaps between for possible APS upgrades bunch trains
On-axis injection with accumulator (ALS-U) storage ring bunches transferred to accumulator accumulator bunches transferred to storage ring Fast kicker magnets New ALS storage ring New accumulator ring
On-axis injection with accumulator (ALS-U) – cont. On-axis injection: • Further optimization of lattice (smaller emittance) • Round beams (more useful shape and reduced emittance growth) • Magnet field requirements relaxed (cost benefit) • Vacuum chambers with small and round apertures (better undulator performance) • Reduced injection losses (better performance)
iteration. The population is evolved generation by genera- A. Robust dynamic aperture tion until it converges or reaches the maximum number of tracking and quantification iterations. DA is usually defined as the maximum stable area in Figure 1 shows the last generation of objective func- transverse plane at injection point. Particles with initial tions, where we are maximizing both the on-momentum condition within this area will survive after a certain num- DA (horizontal axis) and the off-momentum DA (vertical ber of turns of tracking. axis). The trend of the optimization is that the whole Obviously, the area of this 2D bounded region alone population moves toward upper right, and converges to cannot represent the quality of DA. For different momen- 4GSRs: why now? – accelerator physics lower rank. By this generation, the Pareto optimal has tum deviation " ¼ ðp ( p0 Þ=p0 , the stable area or DA may only three candidates at the upper right corner. In a be different. Larger on-momentum DA may help the in- realistic optimization, the full convergence may take too jection, while off-momentum DA helps Touschek lifetime. long to achieve, while stopping at a set of reasonably good One common way to calculate DA is particle tracking solutions which are sufficient for further applications is along several radial lines (see Fig. 2) with fourth-order or desirable. In Fig. 1, although the solutions are not fully even higher symplectic integrator [22]. An ideal solution of converged, they have many candidates with DA DA would have an elliptical type of shape with no cut-in >120 mm2 . It is a good starting point for detailed unstable area (as the two red points shown in Fig. 2). Because of betatron oscillation, particles are crossing the transverse plan at different ðx; yÞ coordinate, and after long enough time, they will be lost at the cut-in unstable area. A good searching method should have good precision to detect these cut-in unstable areas while not requiring too SymplecTc Tracking based methods much trial tracking near the DA boundary, since the track- ing is expensive in computing time. Based on this, we from L. Nadolski, DA, MA separated DA, MA together ICFA LowEring, FIG. 1. The last generation of objective functions: DA of on- Oxford 7/13 momentum (horizontal axis) and off-momentum (vertical axis) FIG. 2. Larger DA area may not necessarily provide a better particles. Points are colored according to their rank. solution, unless it covers an ellipse fully. Direct tracking based opTmizaTon 054001-3 GLASS AnalyTcal based method GeneTc Algorithm Lie Algebra/DifferenTal Algebra MOGA Resonance Driving Terms Amplitude Tuneshia Frequency Maps RDT minimizaTon minimizaTon FMA Nonlinear Diffusion factor Canceling Phase advances “LOCO” Sextupole Interleaved Resonance idenTficaTon Resonances sextupoles Robustness to magneTc, alignment errors Robustness ID configuraTons Tracking codes: PTC MADX TRACY AT LEGO OPA ELEGANT
Online Optimization of Nonlinear Dynamics for DLSRs X. Huang, SLAC • Nonlinear dynamics challenges in DLSRs (4GSRs) make online lattice optimization and automatic tuning, based on beam measurements, desirable • Many optimization algorithms considered. Robust Conjugate Direction Search (RCDS) method selected for speed of convergence. • Beam-based RCDS correction of coupling and improvement of injection efficiency by nonlinear lattice optimization conducted on SPEAR. Coupling reduced to ~2 x 10-4; dynamic aperture increased from 15 mm to ~18 mm. Conclusions: • Online optimization provides an alternative to having an accurate simulation model. • Having an accurate nonlinear model would benefit 4GSR operation. • Having nonlinear optics monitors and a beam-based nonlinear optics calibration tool (non-LOCO) for 4GSRs would be useful.
4GSRs: why now? – cont. Compact magnet and vacuum technology • NEG-coated vacuum chambers enable small apertures to enable high magnet gradients Pioneered at CERN, used extensively at Soleil, and adopted for MAX-IV and Sirius MBA lattices • Precision magnet pole machining for small aperture magnets, combined function magnets, tolerance for magnet crosstalk (developed extensively at MAX-Lab) MAX-IV Courtesy S. Leemans heater tape for in-situ NEG SPring-8 bake-out Sirius concept K. Soutome
4GSRs: why now? – cont. MAX-IV
Laurent S. Nadolski Synchrotron SOLEIL - Beam Dynamics Group 4GSRs: why now? – cont. nadolski@synchrotron-soleil.fr Version 1.3, February 2011 Other advances in accelerator and light source technology NPAC 2010–2011 (v1.3) Laurent S. Nadolski Frequency Map Analysis 1 / 53 • Fast kickers for on-axis injection Introduction: Studying Nonlinear Dynamics Fast kickers (KEK ATF) • Sub-micron e- BPMs and orbit feedback Nonlinear dynamics • Accelerator and beam line component mechanical Complex dynamics I Higher order stabilizing systems ResonancesI Tunes shift with amplitudes I resonances detected by Chaos, instabilities • Micron resolution single pass BPMs (non-linear lattice I turn-turn BPMs Instability thresholds I (A. Franchi) tuning) Need of an accurate way to compute frequencies. Fast • “In-situ” magnet measurement and alignment methodsFourier Transform (FFT) is an efficient algorithm (Cooley-Tukey, 1965) and a powerful way yet limited. Precision is 1/T . To get (e.g. NSLS-II) high precision frequencies (tunes), very long integration, many data, or turns are needed. • Mode-damped RF cavities (fundamental and harmonic) NPAC 2010–2011 (v1.3) Laurent S. Nadolski Frequency Map Analysis 2 / 53 SPring-8 concept based on NSLS-II • Highly stable solid state RF power sources vibrating wire method - K. Soutome Delta undulator prototype - A. Temnykh • High performance IDs (superconducting, Delta, RF, etc.) SC undulator development at LBNL (S. Prestemon et al.), APS (E. Gluskin et al.) and elsewhere
SR design – short bunches • There is a storage ring user community interested in short bunch experiments that probe materials properties on the >~10 ps time scale. Bunch lengths of order 10 ps are desirable, but longer bunches are also useful for longer time scales. • Short single pulses are needed for pump-probe experiments with rep rates up to a few MHz (laser rep rate limit). • Single bunches are often isolated in a gap in the bunch fill pattern (“camshaft mode”), but some rings run in a “timing mode” with a few widely space bunches. Spacing is sufficient for material transient response to decay. Photon pulse optical gating can also be used to select single pulses in some cases. Charge per bunch is limited in 4GSR lattices by instabilities and IBS effects. • Bunch current is greatly reduced in low momentum compaction lattices. • Very short photon pulses (~100 fs) can be generated with laser slicing, but the number of photons per pulse is very low (~105?).
SR design – short bunches – cont. • ALS has developed a “pseudo single bunch” kicking mechanism the puts one bunch on an oscillating orbit where its radiation is viewed in an apertured beam line that only sees light when the bunch is at an oscillation maximum. • APS was considering a superconducting crab cavity scheme that would produce large charge, high rep rate short bunches in a localized part of the ring. • BESSY-II proposed a 2-freq RF system to generate alternating short and long bunches. High frequency (~1.5 GHz), high voltage RF systems are needed. • Exotic bunch manipulation schemes (e.g. emittance exchange, crab cavities, etc.) might be used in the future to produce temporary short bunches. • Long bunches are desirable for 4GSRs to reduce IBS effects and emittance (harmonic cavity bunch lengthening).
Short bunches with low a in SPEAR3 – synergy with LCLS 68
Simultaneously long & short bunches present nc-cavity (power) sc-cavity # 1 (focusing) sc-cavity # 1 & 2 (focusing) short short & long bunches long bunch Voltage / MV bunch sum voltage rel. long. phase position / ns 0.5 GHz, 1.5 MV 1.5 GHz, 25 MV 1.75 GHz, 21.4 MV V’=Vxfrf= 0.75 MVGHz V’=Vxfrf= 37.5 MVGHz V’ = Vxfrf= 75 MVGHz - flexible fill pattern, I
Short and long bunches with 2 RF frequencies
SPEAR3 future short bunch operation 71
Ultimate storage rings? • The term “ultimate storage ring” was first use in 2000: A. Ropert, J.M. Filhol, P. Elleaume, L. Farvacque, L. Hardy, J. Jacob, U. Weinrich, "Towards the Ultimate Storage Ring-Based Light Source", Proc. EPAC 2000, Vienna. • “Ultimate” inferred reducing emttance towards the diffraction limit for X-rays • “Ultimate” may have many meanings, including providing everything for every user • One way to make storage rings more “ultimate”: FELs becoming more ring-like: higher rep rate, reduced photons/pulse (e.g. NGLS) Can rings become more FEL-like: increased ph/pulse, reduced energy spread, short pulses, lasing? Longitudinal and transverse transform- limited beams?
Light source performance: other metrics J. Corlett, R. Hettel, “Performance Requirements and Metrics for Future X- ray Sources”, Proc. PAC09, Vancouver
SASE with transverse gradient undulator E = 4.5 GeV εx/y=160 / 1.6 pm δE/E = 1.6x10-3 rms Q = 0.75 nC ηy = Ver?cal 0.05 m βx/y = 16 / 50 m σβ = 52 mm ση = 78 mm λu = 3 cm K = 3.7 TGU σz= 1ps Ipk = 300 A λph = 1.5 nm Bunch switched into FEL bypass ID grad = 22.9 m-1 (10-‐100 kHz) Z. Huang, Y. Cai, Y. Ding IPAC 2013 1.5 nm pulse energy = 0.5 mJ Reduce longitudinal emiQance to reach high peak current – a challenge for future ring designers! 8 CEBAF SC caviTes in a cryomodule produce 108 MV for longitudinal focusing Hard XFEL oscillator? – K-J Kim
Towards the ultimate Will ERLs rise up? transform-limited beam in 3 dimensions
4GSR design – comments Brightness/coherence vs. flux • User community is divided – some need flux, not brightness • Figure of merit: number of “usable” photons per unit time in the spatial and energy bandwidth acceptance phase space of the experiment (e.g. protein crystal angular acceptance is quite large – moderate brightness is OK). “Brightness isn’t everything”. • Diminishing return on coherent fraction and flux as emittance is reduced • Cost-performance optimization needed for every light source design • Science case should drive the optimization (is 10 or 1 pm worth it? – maybe!) Lattice 16 beam lines in PEP-X with 7BA (left) and DBA/TME hybird (right) • ID straight section length is always an issue (canted IDs?) • Spacing between ID straights is an issue with large rings, leading to large, expensive experimental halls. Consolidating beam lines with hybrid lattice BRIGHTNESS may be more efficient (e.g. PETRA-III) • A relaxed, larger dynamic aperture mode for aggressive lattices?: “emittance knob”
A new generation of storage ring light sources
Appendix 1 4GSR R&D Topics
4GSR R&D: lattice design Low emittance, buildable lattices: Develop low emittance lattice designs having “reasonable” multipole gradients and magnet apertures. Explore benefit of using dipoles with longitudinal gradient. Design optimization: Develop algorithms for optimizing ring parameters (e.g. energy, emittance, circumference, beta functions, RF, etc) based on targeted spectral brightness, coherence, special operating modes (e.g. short bunches, lasing) and number of beam lines. Define a quality factor to gauge this optimization. Consolidated beam lines: Develop lattice geometries, potentially non-circular and/ or having hybrid lattices, that enable consolidating beam line straight sections in very large rings – a part of design optimization. Robinson wigglers: Are they a preferred replacement for conventional damping wigglers in reducing emittance? Round beams: Determine optimal ways to produce nearly round beams at source points. Test on existing machines if possible. Momentum compaction: Develop very low emittance lattices with increased momentum compaction as a way to increase bunch length (e.g. chicanes, etc.). 79
4GSR R&D: accelerator physics Simulation codes: Develop codes that account for close magnet spacing and include collective effects during lattice optimization. Improve simulation codes impedance, ion instability, CSR and other effects as needed. Benchmark codes on existing machines operating in special modes. Scaling laws: Develop general scaling laws that take into account as much as possible all the effects, including emittance (with collective effects), brightness, spectrum, circumference, magnet strengths, RF, running costs, etc. Non-linear lattice correction: Develop improved techniques to measure and correct higher order resonance driving terms to maximize dynamic apertures. Test on existing rings. RF frequency: Study the benefit of higher RF frequency for reducing longitudinal emittance, bunch length and operating costs, and the use of using 2 frequencies to generate alternating long and short bunches. 80
4GSR R&D: accelerator physics – cont. Reduced energy spread, longitudinal emittance: Explore ways to reduce energy spread (to the level of 0.05% or less), and longitiudinal emittance in general, to enable using high ID harmonics, short bunches and potential lasing. High peak current: Explore ways to produce >200 Apk with 10 pm-scale emittance to enable lasing. Lasing: Determine beam parameters and consequent ring designs that would enable X-ray FEL operation, either in a switched bypass or in the ring itself, including oscillator configurations. Beam manipulation: Explore ways to (locally) reduce emittance, bunch length, energy spread, etc. (e.g. emittance exchange, flat-to-round converter (ID in solenoid), RF and optical manipulation methods, etc.). Very short bunches and CSR: Explore ways to suppress CSR to reduce the lengthening and emittance increase of very short bunches propagating in the ring. Space charge: Determine if space charge is an issue for low-E USRs. 81
4GSR R&D: injection SIngle-shot top-up: Ways to restore charge to multiple arbitrary bunches in a single injection shot to reduce the duration of the top-up-related orbit transient, maintaining variation in charge for all bunches to ~20% or less for a uniform fill pattern. Pulsed multipole (PM) injection: Continued development of PM injection schemes, including schemes with septum and PM in the same straight. Accumulator/booster for swap-out injection: Study the practicality of implementing a combined accumulator/booster, possibly located in the main ring tunnel, for realizing multibunch single-shot swap-out injection. Investigate the possibility of recovering the beam kicked out from the ring in the accumulator/ booster for reinjection. Injection kickers: See Accelerator Engineering Longitudinal injection: Investigate practicality of longitudinal injection as a way to eliminated stored beam orbit transient. 82
4GSR R&D: accelerator engineering Magnets: Determine optimal magnet bore dimensions with respect to mechanical tolerances, multipole strengths, yoke saturation and vacuum system design. Investigate magnet material choice, solid versus laminated cores and compact combined function magnet designs. Vacuum system: Designs for small aperture vacuum systems with focus on chamber material, NEG coating and activation processes, heat absorption, synchrotron light extraction and BPM head stability. Stability: Develop site vibration specifications for USRs. Develop passive and active ways to minimize effects on the stability of the photon beam and critical accelerator and beam line components caused by ground motion, cooling water, machine- and temperature-induced motion and vibration. Develop stable builiding design concepts. Motion sensors: Develop affordable 100-nm-resolution component motion sensors. Alignment: Develop practical and simplified ways to achieve 10-µm alignment tolerances. 83
4GSR R&D: accelerator engineering – cont. RF system: Optimal frequency(s), improved cavity mode damping, solid state amplifiers, harmonic cavity systems (including passive vs. active), crab and other beam manipulation cavities, solid state RF power sources, continued improvements to LLRF. Power supplies: Not discussed. Pulsed multipole injection magnets: Designs that reduce the required separation of injected and stored beams. Fast kickers: Develop injection kicker and pulser designs having
4GSR R&D: instrumentation and feedback e- BPMs: Stable BPM designs for small aperture vacuum chambers having micron turn-turn resolution or better. BPM processors: A factor of 10 or more increase in turn-turn resolution than present state-of-the-art for measurement of higher order lattice resonance driving terms; reduced processing latency to be commensurate with 10-kHz digital feedback clock rates; improved stability and reduced current dependence. X-ray BPMs: Continued development of photon BPMs for IDs, especially EPUs and VUV. Orbit feedback: Integrated orbit and beam line component feedback systems to achieve maximal beam stability at the experiment using multiple sensor types (e.g. e-BPMs, X-BPMs, beam line sensor and detector information, motion monitors, etc). Beam size stabilization: Feedback and feedforward systems to stabilize beam size as IDs, especially EPUs, are varied. Multibunch feedback: Improved systems having higher resolution, reduced noise impact and capable of accommodating variable bunch fill patterns, including ones with single large bunches and many small ones. 85
4SR R&D: insertion devices New IDs: Continue ongoing R&D on CPMUs, SCUs, variable polarization and other new IDs will benefit USRs. ID length: Establish optimal lengths for IDs in USRs; straight section lengths should be determined accordingly. Small gaps: Determine minimum ID gaps. Vertically oriented IDs: Can they be accommodated (e.g. Delta-type, helical, TGUs, etc.)? Power on optics: Develop improved masking schemes and IDs that minimize unused power on optics. Dynamic effects: Establish ID tolerance requirements and study effects of present and anticipated future IDs and USR beam dynamics and properties and develop effective compensation schemes. ID commissioning: Develop new ID commissioning strategies as needed for USRs; test on existing machines. X-ray optics: Develop X-ray optical components capable of preserving photon beam properties, including coherence, from USR IDs. 86
Appendix 2 4GSR Accelerator and Science Reference: Journal of Synchrotron Radiation (2014) 21
JSR (2014) 21 - contents Diffraction limited storage rings - how and why? M. Eriksson, C. Quitmann, J.F. van der Veen DLSR projects and plans, an international overview R. Hettel Lattice design challenges for fourth-generation storage ring light sources M. Borland, G. Decker, L. Emery, V. Sajaev, Y. Sun and A. Xiao Magnet design for a low emittance storage ring M. Johansson, B. Anderberg, L-J Lindgren DLSR vacuum technology E. Al-Dmour, J. Ahlbäck, D. Einfeld, P. F. Tavares, M. Grabski First multi-bend achromat lattice consideration D. Einfeld, M. Plesko, J. Schaper Collective effects in DLSR R. Nagaoka, K.L.F. Bane The MAX IV storage ring project P. Tavares, S. Leemann, M. Sjöström, Å. Andersson The Sirius project L. Liu, N. Milas, A.H.C. Mukai, X. R. Resende, A.R.D. Rodrigues, F.H. de Sá Short bunches in diffraction limited storage rings X. Huang, T. Rabedeau, J. Safranek
JSR (2014) 21 – contents – cont. Coherent imaging at the diffraction limit P. Thibault, M. Guizar-Sicairos, A. Menzel Hard X-ray scanning microscopy Christian Schroer, G. Falkenberg XPCS and coherent imaging of fluctuating condensed matter O. Shpyrko X-ray nanoprobes and diffraction limited storage rings: opportunities for fluorescence tomography of biological specimens M. de Jonge. C. Jacobsen Structural Biology in the 21st Century R.F. Fischetti, B.K. Kobilka, J.L. Smith, W.I. Weis High resolution resonant X-ray inelastic scattering studies of materials, liquids and gases T. Schmitt, F. de Groot, J-E Rubensson NanoARPES investigations of heterogenous correlated electron systems E. Rotenberg X-ray spectroscopy for chemical and energy sciences A.I. Frenkel, J.A. van Bokhoven Reciprocal and real space spectromicroscopy and imaging with diffraction limited X-ray sources with focus on energy materials applications M. Toney, A. Hitchcock High pressure science on diffraction limited storage rings M.I. McMahon Pixel detectors for diffraction-limited storage rings B. Schmitt, P. Denes Optics for coherent X-ray applications M. Yabashi, K. Tono, Hidekazu Mimura, K. Yamauchi, K. Tamasaku, H. Ohashi, S. Goto, T. Ishikawa New challenges in beamline instrumentation for Diffraction Limited Storage Rings J. Susini, R. Barrett, C. Morawe, P. Fajardo On the characterization of ultra-precise X-ray optical components – advances and challenges in ex-situ metrology F. Siewert, J. Buchheim, T. Zeschke, M. Störmer, R. Sankari
Appendix 3 Low Emittance Ring and 4GSR Workshops
4GSR Workshops • ICFA Future Light Source Workshops (especially over last few years) • ICFA Low Emittance Rings Workshops (LowERing) • XDL 2011 Workshops for ERLs and DLSRs, Cornell, June 2011 • Beijing USR Workshop, Huairou, October 2012 • DLSR Workshop, SPring-8, December 2012 • DOE BESAC Subcommittee on Future Light Sources, July 2013 • Low Emittance Ring Workshop, Oxford, July 2013 • SLAC DLSR Workshop, SLAC, December 2013 • Workshop on Advances in Low Emittance Rings Technology (ALERT 2014), Valencia, May 2014 • Low Emittance Rings Workshop (LER2014), Frascati, September 2014 • DLSR Workshop, Argonne, November 2014 many other workshops on low emittance rings, including those in the past for ILC damping rings
SR design optimization Light Source Design ID reqmts SR properties impedance acceptance injection energy emittance dp/p lifetime BSC coupling stability geometry e-‐ beam dimen dyn ap current ID comp lattice fns orbit ctrl Install/align facilities I&C RF pwr supplies beam lines supports vac pump magnets IDs vac chamb/absorb R. Hettel Draft SPEAR 3 Design rev 1 Jan. 17 2003
You can also read