Download CHESS Technical Memo 01-003 and more Study notes Nuclear Physics in PDF only on Docsity!
i
CHESS Technical Memo 01-
JLAB-ACT-01-
Study
for a proposed
Phase I Energy Recovery Linac (ERL) Synchrotron
Light Source at Cornell University
4 July 2001
Sol M. Gruner & Maury Tigner, eds.
Cornell University
Ithaca, NY 14853-
Ivan Bazarov 1,3^ , Sergey Belomestnykh 3 , Don Bilderback 1,2^ , Ken Finkelstein 1 , Ernie Fontes^1 , Steve Gray 3 ,Sol M. Gruner 1,4,6^ , Geoff Krafft 5 , Lia Merminga 5 , Hasan Padamsee 3,4^ , Ray Helmke^3 , Qun Shen 1 , Joe Rogers 3,4^ , Charles Sinclair 5 , Richard Talman 3,4^ , Maury Tigner, 3,
(^1) Cornell High Energy Synchrotron Source (CHESS), Cornell University, Ithaca, NY (^2) Applied and Engineering Physics, Cornell University, Ithaca, NY (^3) Laboratory for Nuclear Studies, Cornell University, Ithaca, NY (^4) Physics Department, Cornell University, Ithaca, NY (^5) Thomas Jefferson National Accelerator Facility, Newport News, VA (^6) Laboratory of Solid State and Atomic Physics, Cornell University, Ithaca, NY
Errata
Date Page Comment 7/23/01 39 2 nd^ line from bottom: Table 2.1.1! Table 2.3.3- 7/26/01 I Author list error corrected 9/26/01 69 2 nd^ paragraph: Changed ì Er =20MeVî to ì Er =100MeVî. Changed ìthe threshold current is approximately 45 mAî to ìthe threshold current is approximately 225 mAî 10/1/01 114 & 115
References to Table 4.3.8-1 and Table 4.3.8-2 were removed from figure captions. These tables were in earlier draft version.
10/28/01 95 Table 4.2-1 entries changed: Longitudinal emittance, rms [keV*deg] ì20î → ì40î Energy spread rms [fractional] ì1E-4 to 4E-3î →î1E-4 to 3E-3î 12/27/01 118 Corrected Eq. 4.3.8- 1/17/02 77 128
2 nd^ para. ì3.8îMWî → ì2.8MWî Table 5.2.1-1: ìKCsSbî → ìKCsTeî 2/24/02 57- 121- 122
As now noted in the footnote on pg 57, it was found that two literature references have errors in the formula for CSR. This leads to numerous small changes on pp. 57-59 and 121-122. Basically, the mistaken formula overestimates the CSR power by about x5.
- Executive Summary Acknowledgements vi
- Motivation for an ERL Synchrotron SR Sources
- 2.1. General Description of ERL Source
- 2.2. Comparison of ERL, SR, and XFEL Sources
- 2.3. ERL Science Case
- 2.3.1. Outstanding Scientifically Usable Features of the ERL
- 2.3.2. X-ray Optics Frontiers and Challenges
- 2.3.3. X-ray Microbeams
- 2.3.4. Time-resolved X-ray Diffraction and Scattering
- 2.3.5. Non-linear X-ray Spectroscopy
- 2.3.6. Coherent Flux and Hard X-ray Microscopy
- 2.3.7. X-ray Photon Correlation Spectroscopy (XPCS)
- 2.3.8. Coherent Scattering from Non-crystalline Materials
- 2.4. Section 2 References
- Vision & Parameters of a Phase II ERL Facility
- 3.1. Phase II ERL Accelerator Physics & Technology Issues
- 3.1.1. Design Principles
- 3.1.2. Electron Source
- ERL 3.1.3. General Observations about Linac Optics for Phase II
- 3.1.4. Transverse Beam Break Up Stability
- 3.1.5. Transport Loop Design
- 3.1.6. Single Bunch Dynamics
- 3.1.7. Longitudinal Stability
- 3.1.8. Higher Order Modes, Cryomodules, and Refrigeration
- 3.1.9. RF Power, Coupling Optimization and Stability
- 3.1.10. Ion Effects
- 3.1.11. Insertion Devices
- 3.1.12. Design Considerations for ERL X-ray Optics
- 3.2. Section 3 References
- Motivation for the Phase I ERL
- 4.1. Why is a Phase I ERL Needed?
- 4.2. Phase I ERL Parameters
- Experiments 4.3. Phase I ERL Accelerator Physics Issues, Technology Issues &
- 4.3.1. Coherent Synchrotron Radiation and Space Charge
- 4.3.2. Ions in the Phase I ERL Injector
- 4.3.3. Gun Performance
- 4.3.4. Injector Performance
- 4.3.5. Linac Transverse Stability iii
- 4.3.6. RF Stability
- 4.3.7. Higher Order Mode Cooling
- 4.3.8. Emittance Preservation in the Phase I ERL
- 4.4. Section 4 References
- Technical Description of the Phase I ERL
- 5.1. Overall Machine Layout & Siting
- 5.2. Technical Systems
- 5.2.1. Electron Source
- 5.2.2. Injector
- 5.2.2.1. Booster & Control
- 5.2.2.2. Merging Optics
- 5.2.3. Main Linac
- 5.2.3.1. SRF Cavities and HOM Structures
- 5.2.3.2. RF Power System
- 5.2.4. Transport System
- 5.2.4.1. Magnet System
- 5.2.4.2. Vacuum System
- 5.2.4.3. Feedback System
- 5.2.4.4. Beam Dump
- 5.2.5. Controls and Diagnostics
- 5.2.5.1. Operational Control
- 5.2.5.2. RF Control
- 5.2.5.3. Diagnostics - Logged 5.2.5.3.1. Variables to be Measured and
- 5.2.5.3.2. Performance Evaluation
- 5.2.5.3.3. Operational Diagnosics
- 5.3. Phase I Infrastructure
- 5.3.1. Building
- 5.3.2. Cryosystem
- 5.3.3. Electrical Power
- 5.3.4. Evaporative Cooling Tower
- 5.3.5. Air Conditioning
- 5.3.6. SRF Structure
- 5.3.7. Ancillary Infrastructure
- 5.3.8. Safety Systerms
- 5.3.8.1. Radiation Safety
- 5.3.8.2. Cryogenic Hazards
- 5.4. Section 5 References
iv
List of Acronyms
ALS Advanced Light Source APS Advanced Photon Source BBU Beam Break Up BNL Brookhaven National Laboratory CEBAF Continuous Electron Beam Accelerator Facility CHESS Cornell High Energy Synchrotron Source CSR Coherent Synchrotron Radiation DESY Deutsches Elektronen Synchrotron ERL Energy Recover Linac ESRF European Synchrotron Radiation facility FEL Free Electron Laser HOM Higher Order Mode ID Insertion Device IRFEL Infra-Red Free Electron Laser JLAB Thomas Jefferson National Accelerator Facility LCLS Linear Coherent Light Source LNS Laboratory of Nuclear Studies MAFIA (computer code name) PARMELA (computer code name) RF Radio Frequency SASE Self-Amplified Spontaneous Emission SCHEFF (subroutine in PARMELA) SPring8 Super Photon ring- SR Synchrotron Radiation SRF Superconducting RF TDBBU Three Dimensional Beam Break Up (computer code name) TESLA Tera electron volt Energy Superconducting Linear Accelerator TTF TESLA Test Facility XFEL X-ray Free Electron Laser
vi
Acknowledgements
We thank Mike Billing, Ricky Campisi, David Douglas, Gerry Dugan, Steve Gray, Don Hartill, Yulin Li, Dave Rice, Stefan Simrock, Charles Strohman, and Byung Yunn for advice and help in preparing this report.
Research on accelerator physics at Cornell is supported by NSF Cooperative Agreement PHY-9809799, research on synchrotron radiation at Cornell is supported by NSF Cooperative Agreement DMR-9713424. Research at the Thomas Jefferson National Accelerator Facility is supported by the DOE under contract number DE-AC05- 84ER40150.
Section 1: Executive Summary
1 Executive Summary
Synchrotron radiation (SR) has become an essential and rapidly growing tool across the sciences and engineering. World-wide, about 70 SR sources are in various stages of operation, construction, or planning, representing a cumulative investment on many billions of dollars and serving a growing research community well in excess of 10,000 scientists. To date, all major SR x-ray facilities are based on electron (or positron) storage rings. Given the expected continued growth, importance and expense of SR sources, it is important to ask if there are alternatives to the storage ring SR source which offer advantages of capability or cost. A step in this direction is being taken by the SR community with the proposed developments of linac-based x-ray free-electron lasers (XFELs) utilizing the self-amplified spontaneous emission process (SASE). However, the versatility of modern developments in accelerator physics, as applied to synchrotron radiation, is not limited to storage rings or XFELS. New developments in laser driven photoinjectors and superconducting linac technology open new and exciting possibilities for novel SR-generating machines which offer extraordinary capabilities and promise to catalyze whole new areas of SR-based science.
This is a Project Description of a SR machine technology which has been studied by a collaboration of accelerator and synchrotron light physicists at Cornell University and the Thomas Jefferson National Accelerator Facility (JLAB). We believe that SR machines based on low-emittance photoinjectors and superconducting linacs run in an energy recovery mode will significantly change the conduct and future progress of SR science. These machines, henceforth called Energy Recovery Linacs (ERLs), have captured the attention of the SR community, are the subject of a great deal of discussion and excitement within the SR community, and have already catalyzed plans for ERLs in several laboratories. Although there is no question that ERL SR sources are feasible, there are a great many open questions which have to be resolved before a large ERL machine can be built to produce x-rays beams of the projected brilliance, flux, time structure and size described in this document and in the recent literature.
We propose to construct a Phase I high current, low emittance ERL at Cornell University to experimentally test and develop ERL technology. The machine energy of 100 MeV is too low to produce x-rays; rather, the Phase I machine is designed to answer critical questions about ERL photoinjectors, characteristics and optimizations which must precede the design of a high-energy ERL. This Phase I ERL is a necessary first step towards designing a full-scale Phase II ERL hard x-ray source which we hope to build at Cornell University. The information acquired from the Phase I ERL will also be important for ERL plans at other laboratories and will serve as a vehicle for the training of accelerator physicists in ERL technology.
This Project Description is organized into 4 additional sections after this executive summary: Section 2 describes the motivation for building ERLs, their potential characteristics and the science that they enable. Section 3 paints a vision of the full-scale Phase II ERL we eventually hope to build. Section 3 includes a summary of the machine parameters of the Phase II ERL followed by detailed discussion of the accelerator and x-
Section 2: Motivation for an ERL
2. Motivation for an ERL Source
Synchrotron radiation has proven to be immensely important throughout the physical, biological, and engineering sciences. World-wide, about 70 SR sources are in various stages of operation, construction, or planning, representing a cumulative investment on many billions of dollars and serving a growing research community of more than 10,000 scientists.
The demand for SR continues to grow, with new uses opening all the time. On a global scale, the demand for new sources is immediate. Europe, with roughly the same size research community as the U.S., is in the process of building several new rings (DIAMOND in England, SOLEIL in France). Not counting Russia, Europe already has 12 SR storage rings of ≥ 1.5 GeV energy (built or under construction), and is planning at least 2 more. By contrast there are 5 such machines in the U.S. The question is not if the U.S. will need more SR sources in the future, but rather when the U.S. will need new SR resources. Given the cost and lead-time in designing a new SR machine, it is important to start preparing for this need now.
Currently, all major SR sources are based on storage rings. Given the importance of SR, it is necessary to ask:
- Are storage rings the optimal SR technology for the future?
- Are there alternatives to storage rings which enable new science?
We propose that a SR source based on closed-loop energy recovery with superconducting linacs offers significant advantages over storage ring sources, both in terms of the possible x-ray beams and, once the technology is developed, cost- effectiveness (Gruner 2000; Bilderback 2001). The basic idea behind an Energy Recovery Linac was suggested long ago (Tigner 1965) and the feasibility of operating an ERL has recently been demonstrated with a highly successful infra-red free electron laser (IRFEL) at JLAB (Neil 1998; Benson 1999; Benson 2000; Neil 2000). Our long term goal is to build a high energy (~5 - 7 GeV) SR source at Cornell, both as a development laboratory for ERL technology and as a unique user resource. As explained below in section 3, a high current, high brilliance SR machine will push ERL technology to new limits. Before committing to specific designs for a large and expensive machine, it is absolutely essential that accelerator and technology issues be explored on a brilliant, high current prototype machine. This document is a study for construction of the prototype, which is the first step in a two-phase project to a high-energy ERL SR source.
The advantages of an ERL x-ray source are best understood by first considering storage ring sources. The characteristics of the x-ray beams which may be produced by a SR source will always be limited by the qualities of the electron beams used to produce the SR. Specifically, it is desired to have
(1) Low electron beam emittance in 6 dimensions to increase the brilliance and coherence of the resultant SR; (2) Very short electron bunches to enable fast time-resolved experiments;
Section 2: Motivation for an ERL
(3) Ultra-small round beams; (4) A SR output which does not decay over time; and (5) Flexibility of operation to enable easy tailoring of x-ray beams to specific science applications. (6) An easy upgrade path as limiting components (e.g., the electron source) improve.
Each of these electron beam characteristics is limited by the well-understood physics of storage rings (Sands 1970; Limborg 1998; Ropert 2000). A storage ring is an equilibrium device in which the stored electron beam characteristics result from the competing factors of damping and excitation which determine the stable distribution of the electrons. Importantly, these characteristics develop slowly as equilibrium sets in during the first several thousand orbits around the storage ring, as determined by the energy, ring radius, and lattice. To choose a specific example, it is possible to use a laser photoinjector to make a round, very low emittance, sub-picosecond long bunch. If injected into a storage ring, however, by the time the beam reaches equilibrium, the ring will have imposed its own constraints on the bunch shape, emittance and length. In other words, in a storage ring, the important beam characteristics are a function of the entire ring. These characteristics are near in-principle limits in existing 3rd^ generation rings, such as the ESRF, APS and SPring8. Although some improvement is possible, the technology of storage rings is mature and further performance gains will be limited and will come at enormous cost. So for example, an ìultimate storage ringî has been proposed which would be 2 km in circumference (Ropert 2000) and, because of the strong Touschek effect at high particle bunch densities, would have to be refilled (or ìtopped offî) at least each hour.
An entirely different approach is taken in an ERL machine. In an ERL, the electrons are not stored, so constraints of beam equilibrium never become limiting. As described in section 3.3, photoinjectors can produce bunches with emittances, sizes and lengths which are superior to the equilibrium bunches stored in storage rings. These bunches are then accelerated to high energy via a superconducting linac (SC linac), which can preserve the salient bunch characteristics. These superior high energy bunches are then passed though undulators to produce SR beams with unprecedented characteristics. For these reasons, ERL SR sources have recently become the focus of a number of next- generation SR source efforts (CHESS 2000a; CHESS 2000b; Ben-Zvi 2001; ESRF 2001; Padmore 2001; SRI2001 2001; Bazarov 2001b; Bazarov 2001c).
The ERL approach has the enormous advantage that the beam quality is limited by the photoinjector, rather than by the machine as a whole. Compared to a storage ring, photoinjectors are small and relatively inexpensive. Below, we detail how present photoinjector technology can already provide electron beams with qualities superior to even the best 3rd^ generation storage ring sources. Moreover, photoinjectors are being intensively developed for a variety of applications and are likely to continue to be improved (Nuhn 2000).
A difficulty with any SR source is that the required beam currents carry enormous power, e.g., a 5 GeV, 100 mA electron beam carries 500 MW. Therefore, it is
Section 2: Motivation for an ERL
- The main linac serves to both accelerate the bunches to high energy and to recover energy from bunches which have already been used to produce SR. In the simplest configuration (Figure 2-1) a single linac is used. Alternatives, such as multi-loop acceleration, in which the beam is accelerated through the linac several times, have been proposed. The advantage of multi-loop acceleration is that a shorter SC linac is required. Disadvantages include complexity and possible degradation of emittance. Most importantly, the steady state accelerating current in the linac of an N-loop machine is 2N times the injector current in the SR transport loop. Since beam break up (BBU) issues are major concerns at high currents, multi-pass ERLs require especially careful exploration of BBU limits. Although the proposed Phase I ERL uses a single pass linac, it will allow investigation of the limiting physics that would apply to a high energy multi-pass machine.
Note that the accelerating bunches are interleaved with the decelerating bunches returning from the SR transport loop, so the actual current in the linac in even single loop ERLs is twice the current in the SR return loop.
- The SR transport loop includes the undulators and beam lines required to produce SR. Optimal use of an ERL will require undulators with a very large number of periods and will tax the capabilities of present day x-ray optics. Fortunately, undulator and beamline optics developments are being driven by the 3rd^ generation sources and XFEL efforts (Nuhn 2000) and are expected to evolve significantly over the next few years. The Phase I ERL will be of too low an energy to produce hard x-rays, so there is only very limited funding requested in this proposal for undulator and x-ray optical development. The transport loop also includes beam compressors to shorten bunches for science applications requiring fast timing. Because of wake-field and coherent synchrotron radiation effects, it will probably not be desirable to circulate the shortest possible bunches around the entire machine. Again, this is an important issue which will be further explored within the Phase I ERL plan. Compressed bunches will likely have to be uncompressed before being reinjected in the main linac to reduce excitation of higher order modes.
- The beam dump is a source of noticeable inefficiency. For example, a dumped current of 100 mA at 5 MeV dissipates half a megawatt of power. Fortunately, the power dissipation in the beam dump is determined by the photoinjector beam power and not the main linac. The Phase I ERL will be used to investigate the minimum required photoinjector energy. In a Phase II machine, consideration will be given to utilizing the dumped energy.
Each of the main components of the ERL machine described above require support systems, e.g., RF, cryogenics, etc., as described in subsequent sections.
Section 2: Motivation for an ERL
Figure 2-1. Schematic representation of a simple, single-loop ERL, consisting of the injector, the main SC linac, the SR transport loop and the energy recovered beam dump. The actual shape of the SR return transport loop can vary, depending on requirements.
Section 2: Motivation for an ERL
Table 2.2-1: Comparison of the Cornell ERL source with other existing and proposed synchrotron light sources. Assuming high duty-cycle operations
ERL hi- flux
ERL hi- coh.
APS und. A
APS upgrade
ESRF U35 SPring 5m
SPring 25m
LCLS spont.
LCLS SASE
TESLA spont.
TESLA SASE Energy E G (GeV) 5.3 5.3 7 7 6 8 8 15 15 25 25 Avg. Current I (mA) 100 10 100 300 200 100 100 72∑10 -6^ 72∑10 -6^ 0.063 0. Charge q (nC/bunch) 0.077 0.008 14 14 0.85 0.29 0.29 1 1 1 1 ε x (nm-rad) 0.15 0.015 8 3.5 4 6 6 0.05 0.05 0.02 0. ε y (nm-rad) 0.15 0.015 0.08 0.0035 0.01 0.003 0.003 0.05 0.05 0.02 0. Bunch fwhm τ (ps) 0.3 0.3 73 73 35 36 36 0.23 0.23 0.188 0.
Machine design
of bunches f (Hz) 1.3∑10 9 1.3∑10 9 7.3∑10 6 22∑10 6 2.3∑10 8 3.4∑10 8 3.4∑10 8 120 120 56575 56575
Undulator L (m) 25 25 2.4 4.8 5 4.5 25 100 100 30 87 Period λ u (cm) 1.7 1.7 3.3 3.3 3.5 2.4 3.2 3 3 3.81 5
of period N (^) u 1470 1470 72 145 142 187 781 3300 3300 787 1740
Horizontal β x (m) 12.5 4.0 15.9 4.0 35 24 24 18 18 14.7 33. Vertical β y (m) 12.5 4.0 5.3 4.0 2.5 3.9 15 18 18 14.7 33. Und. K (@ E 1 ) 1.38 1.38 1.24 1.24 0.67 2.08 1.66 3.9 3.9 2.28 4.
Insertion device
1 st^ harmonic E 1 (keV) 8.0 8.0 8.0 8.0 8.0 8.0 8.0 8.26 8.26 10 12. H. div. fwhm (μrad) 9.1 6.2 54.3 70.2 26.8 38.4 37.4 4.9 1 6.7 1. V. div. fwhm (μrad) 9.1 6.2 16.2 9.7 10.4 10.0 4.3 4.9 1 6.7 1. H. source fwhm (μm) 103 24.5 839 277 879 892 890 82 78 60 60 V. source fwhm (μm) 103 24.5 48.6 11.4 13.9 10.6 22.8 82 78 60 60
Beamline & opticsPower^ P^0 (kW)^ 33.9^ 3.4^ 1.2^ 7.2^1 15.7^ 31.2^ 0.0027^ 0.003^ 0.070^ 1. DP/dA @20m (W/mm^2 ) 2600 260 180 1080 194 1830 4568 0.45 63 336 2∑10 5 Ave. flux F (^) n (p/s/0.1%) 1.5∑10 16 1.5∑10 15 7.0∑10 14 4.2∑10 15 1.3∑10 15 2.4∑10 15 9.0∑10 15 3.3∑10 10 2.4∑10 14 6.4∑10 12 4∑10 17 Ave. brilliance B (p/s/0.1%/mm^2 /mr^2 ) 1.3∑^
(^22) 5.2∑10 22 1.5∑10 19 1.5∑10 21 3.1∑10 20 5.0∑10 20 2.2∑10 21 1.6∑10 17 4.2∑10 22 3.6∑10 19 8∑10 25
Coh flux Fc (p/s/0.1%) (^) 8.1∑10 13 3.1∑10 14 0.9∑10 11 9.0∑10 12 1.8∑10 12 3.0∑10 12 1.3∑10 13 9.0∑10 8 2.4∑10 14 1.4∑10 11 4∑10 17 DC experiments Coh. fraction pc (%) 0.52 20 0.013 0.22 0.14 0.13 0.14 2.7 100 2.1 100 Photons / bunch 1.2∑10 7 1.2∑10 6 9.6∑10 7 1.9∑10 8 5.7∑10 6 7.1∑10 6 2.7∑10 7 2.8∑10 8 2∑10 12 1.1∑10 8 7∑10 12 Peak brilliance (p/s/0.1%/mm^2 /mr^2 ) 3.0∑10^25 1.2∑10^26 2.5∑10^22 8.3∑10^23 3.3∑10^22 3.6∑10^22 1.6∑10^23 4.8∑10^27 1.2∑10^33 3.4∑10^27 7∑10^33 Peak flux (p/s/0.1%) 3.9∑10 19 3.9∑10 18 1.3∑10 18 2.6∑10 18 1.6∑10 17 1.9∑10 17 7.4∑10 17 1.2∑10 21 7.2∑10 24 6.0∑10 20 3∑10 25
Pulsed expts.Pk coh. flux (p/s/0.1%) 2.1∑10 17 7.9∑10 17 1.7∑10 14 5.6∑10 15 2.2∑10 14 2.5∑10 14 1.1∑10 15 2.7∑10 19 7.2∑10 24 1.4∑10 19 3∑10 25 Peak degen. par. δ D 95 368 0.078 2.6 0.103 0.113 0.49 1.3∑10 4 3.3∑10 9 4.7∑10 3 8∑10 9 Ave. coh. power (W) 0.10 0.40 1.2∑10 -4^ 0.011 0.0023 0.0038 0.017 1.2∑10 -6 0.32 2.2∑10 -4^794 Peak coh. power (W) 269 1011 0.22 7.2 0.28 0.32 1.4 3.8∑10 4 9∑10 9 2.2∑10 4 60∑10 9 A coh dP/dA (W/mm^2 ) 12.0 848 0.0029 3.5 0.19 0.40 0.84 2.3∑10 -4^ 0.0077 0.078 2.8∑10 5 P coh dP/dA (W/mm^2 ) 3.2∑10 4 2.2∑10 6 5.4 2280 22.9 33.8 69.0 7.2∑10 6 1.9∑10 12 7.8∑10 6 2.1∑10 13 Nonlinear exptAve. E -field (V/m) (^) 1.0∑10 5 8.0∑10 5 1479 5.1∑10 4 1.2∑10 4 1.7∑10 4 2.5∑10 (^4 416 2410 7670) 1.5∑10 7 Peak E -field (V/m) (^) 4.9∑10 6 4.1∑10 7 6.4∑10 4 1.3∑10 6 1.3∑10 5 1.6∑10 5 2.3∑10 5 7.4∑10 (^7) 3.8∑10 10 7.7∑10 (^7) 1.3∑10 11
Section 2: Motivation for an ERL
Notes for Table 2.2-1 and Figures 2.2-1 to -5 :
All flux and brilliance calculations are performed at 8 keV fundamental energy except for the proposed XFEL sources LCLS and TESLA, which are at 8.26 and 12. keV, resp. Thus in some cases the value of the brilliance may be somewhat lower than the value at the peak fundamental energy of the undulator. This is done for the purpose of proper coherence comparisons since coherence is very sensitive to x-ray wavelength. The XFEL numbers are obtained from the (LCLS 1998) and the TESLA (Brinkmann
- design reports. Power density numbers at 20m for SASE are for average coherent power only and include the effects from source size and beam divergence, while all other numbers for that row are calculated using the formula listed at the end of ìFlux, Brightness and Powerî in (Shen 2001).
All calculated results are for high-duty cycle operations only (APS: 6+21 singlets; ESRF: 2x1/3 filling of 992 RF buckets; SPring8: 2/3 filling of 2436 RF buckets) which respresent the most common running modes at the existing storage-ring sources. The machine parameters are obtained from the respective web sites. APS parameters are obtained from http://www.aps.anl.gov/xfd/calendar/fp_2000-4.html on 1/8/01, and from e-mail communications with Dr. Dennis Mills. Upgraded parameters for APS are based on Arthur (2000). ESRF parameters are obtained from http://www.esrf.fr/machine/myweb/machine/brill.html on 12/18/00, and from communications with Dr. Pascal Elleaume (2000). SPring8 parameters are obtained from SPring8 Annual Report 1998 , available at http://www.spring8.jp/, and from Dr. Don Bilderbackís personal communication with Dr. Kitamura at SPring8.
For the purposes of the calculations, we assume β = 12.5 m in the high-flux
option and β = 4 m in the high-coherence option for a 25 m undulator with segmentation.
The relatively short β assumes segmentation due to focusing requirements.
One of the greatest uncertainties in Table 2.2-1 and the accompanying figures stems from a fundamental lack of knowledge about the minimum bunch length and emittances which can be simultaneously achieved. To account for this uncertainty we have inserted in the peak brilliance plot a curve that corresponds to an ERL with 0.15nm emittances at 100mA but with a more conservative 4.7ps FWHM bunch length. Furthermore, a very conservative peak brilliance value is calculated assuming 300 fs bunches but 1.5nm transverse emittances at 100 mA. This result is shown in Figures 2.2- 3 and 2.2 ñ 6 and compared with other existing and proposed storage-ring based sources.
Resolution of these uncertainties is one of the most important objectives of the prototype ERL machine proposed in this document. Nonetheless, as Figure 2.2-6 shows, even with conservative numbers the ERL will push the synchrotron x-ray brilliance and pulse length into new territories that are completely uncharted by the existing sources.
The peak brilliance and coherence of the ERL will be greatly exceeded by the proposed XFEL sources, such as TESLA and the LCLS. Why then, one might ask, should the ERL be pursued? The answer is straightforward and compelling: Although XFEL
Section 2: Motivation for an ERL
Figure 2.2-1: Comparison of calculated average spectral brilliance of the various sources. The parameters used for undulator sources are listed in Table 2.2-1. The CHESS wiggler sources assume 5.3 GeV 300 mA operation and a FWHM source size of d (^) x = 5.5 mm and d (^) y = 0.9 mm for the 24-pole wiggler at F-line and dx = 3.3 mm and dy = 0.85 mm for the 49-pole wiggler at A/G-line.
10 100
1013
1014
1015
10 16
10 17
10 18
1019
1020
1021
1022
1023
ESRF U
Sp8 5m
Sp8 25m APS 4.8m
APS 2.4m
0.15nm 100mA
ERL 25m 0.015nm 10mA
LCLS SASE
LCLS spont.
CHESS 24p wiggler
CHESS 49p wiggler Average Brilliance (ph/s/0.1%/mm
2 /mr
2 )
Photon Energy (keV)
Section 2: Motivation for an ERL
Figure 2.2-2: Comparison of calculated average photon flux of the various sources per second. The parameters used for all sources are listed in Table 2.2-1.
14
15
16
LCLS SASE
APS 2.4m
ESRF U APS 4.8m
Sp8 5m
Sp8 25m
ERL 25m 0.015nm 10mA
ERL 25m 0.15nm 100mA
Average Flux (photons/s/0.1%)
Photon Energy (keV)
