Conceptual Design Reports
|
|
Conceptual Design Report
Original
Published Summer 2007 |
|
Conceptual Design Report 2.0A Conceptual Design Report focused on the upgrade strategy is available at the link: FERMI 2.0 Conceptual Design Report. |
CDR chapter 1 - Executive Summary
|
|
Synchrotron radiation is a fundamental and indispensable tool for the study of materials which encompasses a wide spectrum of sciences, technologies and applications, from life sciences to nanotechnologies, from environmental sciences and geochemistry to archaeology. Synchrotron radiation has seen an explosive growth in its application to research and development and in the number of facilities built to serve its users, covering a large range of radiation wavelengths, extending from the infrared down to hard X-rays, in the form of radiation pulses with time duration down to the few picoseconds range. The number of facilities in operation worldwide is close to eighty, serving tens of thousands of users per year. The main figure of merit of radiation sources is brilliance, which defines the intensity of radiation, within a given bandwidth around the desired wavelength, that can be focused unto a sample of given area. Typical brilliance values for the highest performance “third generation” light sources are around 1019 to 1021 photons/s/mm2/mrad2/0.1% bandwidth. Another important characteristic is the pulse duration: ultra short, sub-picosecond radiation pulses are needed to open up the new investigation field covering not only the structure of a sample but also its dynamics during irradiation. ... |
CDR chapter 2 - Overwiew
 |
The FERMI project at the Elettra Laboratory of Sincrotrone Trieste (ST) will be a national and interna-tional user facility for scientific investigations of ultra-fast and ultra-high resolution processes in materi-al science and physical biosciences with high brilliance X-ray pulses. The underlying technology of this new facility is the Free Electron Laser (FEL) preferably employing a master-oscillator-power-amplifier configuration with high gain harmonic cascades of wigglers. The full FERMI facility will consist of a linear accelerator plus two principal FEL beamlines in a new experimental hall in the complex environ-ment of a multi-beamline user facility provided by the Elettra synchrotron light source. The FERMI Project utilizes a 1.2 GeV linear accelerator, part of which is the current injector of the Elet-tra storage ring, and a new electron source based on photo-injector technology. In the near future, with the start of operation of the booster ring as the new Elettra injector, the existing linac will become avail-able as part of a dedicated driver of the free-electron laser. The addition of seven accelerating sections brings the linac energy to ~ 1.2 GeV. At this energy, and with state-of-the-art undulator technology, the free-electron laser described in this report can operate in the 100-40 nm energy region in the initial phase (FEL-I) and down to 10 nm in a subsequent phase (FEL-II).This chapter gives a broad overview of the facility and of the issues relevant to achieving the design objectives. |
CDR chapter 3 - The Scientific Case
|
Chemical, physical and biological processes are intrinsically dynamic in nature since they are related to electronic and atomic structures that evolve with time. The characteristic time scales span from a few femtoseconds, in the case of electronic processes, to a few tens or hundreds of femtoseconds, as in the case of atomic and molecular processes. Furthermore other phenomena, which control the behavior of critical systems, may happen at relatively longer time scales, ranging from a few picoseconds to a few hundreds of picosecond or more. These include phe-nomena produced by phase transitions, such as the processes related to magnetic order or to superconductivity. The capability of measuring these processes at the relevant time scales will open completely new perspectives and analyses. In particular, the direct observation of electronic processes, of structural dynamics and of dynamical crit-ical phenomena (such as phase transitions) represents an unexplored landscape in the study of condensed matter physics. These possibilities were already evident to the inventors of the first coherent sources of femtosecond opti-cal pulses. Ultra-sort pulses of coherent light have generated remarkable scientific progress that was recognized in the 1999 the Nobel Prize for Chemistry awarded to Ahmed Zewail for his pioneering work on the application of ultra-short laser infrared spectroscopy to the study of the dynamics of chemical bonds. Currently available, coherent light sources emit radiation only in a limited wavelength range. Their use is limited to optical and spectroscopic techniques in the infrared, visible and near-ultraviolet range, excluding all the meas-urements needing photons of energy higher that a few eV. There is therefore a strong scientific need for a tunable light source with an energy range from the vacuum ultraviolet (VUV) to the X-ray with a stable and well-charac-terized temporal structure in the femtosecond and picosecond time domain. To this end, international research is moving in three main directions: 1) laser driven light sources which use non-linear processes to create very high harmonics, 2) interaction between an ultra-short laser pulse and an electron bunch in an storage ring (laser “bunch-slicing technique”), 3) free electron lasers (FEL). The first two techniques are able to produce radiation pulses in the femtosecond time domain and in the soft X-ray region with low brilliance (i.e. a low useful photon flux on the material under investigation). In contrast FELs can produce light pulses with peak brilliance as much as ten orders of magnitude higher than the pulses generated in present third generation synchrotron light sources and with pho-ton energies spanning from the VUV to the hard X-ray, i.e. from about 10 eV (120 nm) to 10 keV (0.12 nm). |
CDR chapter 4 - FEL Physics
|
The FERMI@ELETTRA project is based on the harmonic up-shifting of an initial “seed” signal in a single-pass FEL amplifier employing multiple undulators. The basic principles which underlie this approach are: the energy modulation of the electron beam via the resonant interaction with an external laser seed in a first undulator (modu-lator); the use of a chromatic dispersive section to then develop a strong density modulation with large harmonic overtones; the production of coherent radiation by the microbunched beam in a downstream undulator (radiator). The first stage of the project, FEL-1, aims at generating coherent output radiation in the 40-100 nm spectral range. For these wavelengths, users require short (<100 fs) pulses with adjustable polarization and high temporal and spatial reproducibility. The goal of project’s second stage, FEL-2, is to extend the spectral range down to 10 nm. Present users’ requirements point to long (narrow-bandwidth) pulses with high peak brilliance and adjustable polarization. FEL-1 relies upon a single-stage scheme (i.e., modulator-dispersive section-radiator), like the one already operational at Brookheven [refs]. As for FEL-2, a two-stage harmonic cascade is necessary for reaching short wavelengths. The selected configuration is based on the so-called “fresh bunch” approach [ref], in which the radiation from the first radiator is used to energy-modulate in a subsequent modulator a part of the electron beam that did not interact with the external seed. This chapter is organized as follows. After a general Introduction to the schemes adopted for the design of both FEL-1 and FEL-2 (section 4.1), in section 4.2 an overview is given of FEL basic output requirements and of main phenomena which may affect the FEL performance. Section 4.3 is devoted to the design of undulators and transport lattice. Design and expected performance of FEL-1 and FEL-2 are addressed in Sections 4.4 and 4.5, respectively. Calculations, which rely both upon time-steady input parameters and full start-to-end time-dependent simulations, have been performed using the 3D numerical codes Genesis [ref] and Ginger [ref]. As a benchmark for FEL-2 “fresh bunch” scheme, calculations are also presented for a possible alternative configuration, named “whole bunch”, in which the entire electron bunch is seeded and the second modulator is eliminated. Section 4.6 contains conclusions and perspectives. |
CDR chapter 5 - Photoinjector
|
The front-end injection systems of the FERMI@Elettra linac produce the high brightness electron beams that define the performance of the FEL and the quality of the x-ray beams delivered to the users. The injector mainly consists of the RF gun, based on the BNL/UCLA 1.6 cell design [1] and scaled to European S-band frequency, its compensation solenoid and two traveling wave S-band rf sections (called SØA and SØB) which accelerate the beam up to 100 MeV. The slice emittance at the end of the injector is specified to be less than 1.5 mm mrad. The injector must provide a linearly current ramped bunch in order to linearize wakefield effects in the linac sections [2]. This equirement translates into finding the best laser pulse shape at the cathode that produces an electron bunch evolving into the desired current profile along the drift between the gun and the first booster section. Two main bunch configuration have been studied: a 0.8 nC/9 ps long “medium length bunch (MLB)” configuration and a 1 nC/11 ps “long bunch (LB)”. It is shown that the optimum laser temporal profile for both the MLB and the LB regimes is a quadratic ramp, which is transformed to a linear ramp by space charge forces in the injector. As the charge density varies from head-to-tail in the bunch, a best compromise has to be found for the emittance compensation process. As a consequence it is shown that slice emittance as well as charge distribution must be ramped along the bunch (see Paragraph 5.6.1): slice emittance values range from 0.7 to 1.1 mm mrad for the MLB regime and from 0.8 to 1.2 mm mrad for the LB regime while the current increases from 40 A up to 80 A. Furthermore, since a seeded harmonic cascade FEL is very sensitive to shot-by-shot variations in beam characteristics, the effects of jitters on the FEL performance have been studied in the ramped current scenario. In particular, time jitters are critical because they are translated into energy jitter in the two chicanes, thus affecting the FEL output power stability. The main source of time jitter comes from jitter in the drive laser arrival time at the cathode with respect to the gun RF waveform, specified to be equal to or less than 200 fs and corresponding to a 0.1 deg jitter in the RF phase seen by the beam. Hundreds of simulation of the injector output beam quality (Section 5.6.2) with input parameters randomly picked, within tolerance values dictated by present technology and supported by measurements, have been performed; they show that the overall expected rms time jitter value at the injector exit is 350 fs. Sensitivity studies and tolerance budgets concerning injector output current, energy, emittance and injector optics parameters are included in the discussion. |
CDR chapter 6 - Accelerator
|
The FERMI harmonic cascade FEL operates within a range of wavelengths from 100 to 10 nm, covered by two distinct undulator chains at the fixed energy of ~1.2 GeV. Two electron beam scenarios have been developed: the medium length bunch (MLB) mode with a bunch length of ~ 700 fs and the long bunch (LB) mode with a bunch length of ~1.4 ps. The electron peak current is 800 or 500 A, respectively. The accelerator was designed with sufficient flexibility to accommodate such variations in bunch parameters. Other important electron beam parameters include the normalized slice electron beam emittance and the slice energy spread, which are about 1.5 μm rad and 150 keV, respectively. A challenging aspect was the demand to produce an electron beam with as uniform as possible peak current and energy distributions along the bunch. For this reason, a new parameter, the “flatness”, defines the value of the quadratic component of energy variation along the bunch for which the increase in bandwidth of the x-ray signal due to this variation becomes equal to the Fourier transform limited bandwidth defined by the bunch length. Tracking results predict flatness of 0.8 MeV/ps2 for the MLB mode and of 0.2 MeV/ps2 for the LB mode. Since the RF photocathode gun produces 0.8 nC and 1 nC for the two options distributed over bunch length of 9 ps and 11 ps respectively, the bunch has to be compressed by a total factor of about 9 before it enters the undulator. The acceleration and compression is done in the main S-band linac. The two bunch compressors (BCs) consist of symmetric magnetic chicanes, each 8.0 m long. They include trim quadrupoles for a fine tuning of the dispersion bump. The locations and compression factor of the two chicanes were fixed in order to minimize the 6-dimensional emittance dilution of the electron bunch in presence of space charge forces and wake fields. The electron energy at BC1 is ~230 MeV in order to avoid space charge effects, while compressing the bunch early enough in the linac to reduce the effects of transverse wake fields. The energy of the second compressor is about 580 MeV, which balances the conflicting requirements of minimizing the transverse and longitudinal emittance dilution by coherent synchrotron radiation (CSR) and that of canceling the final correlated energy spread by means of the downstream longitudinal wake field. By using a weak chicane with a bending angle smaller than 0.07 rad per dipole and a large initial correlated energy spread within the range of 1.0% - 2.5%, the CSR effects can be reduced, but the chromatic aberrations make the tolerances on the magnets field quality tighter. ... |
CDR chapter 7 - Undulators
|
The FERMI undulators are based on PPM arrays, a choice dictated mostly by the need to provide variable polarization. The present design is based on a remanent field Br=1.2 T, which allows to choose material grades with close to the maximum available coercitivity. High coercitivity makes the material highly resistant to radiation, a feature of paramount importance for its long lifetime. The undulator coefficient K must be greater than 1 in order to provide sufficient FEL gain. The minimum acceptable gap height is 10 mm, dictated by residual gas pressure and energy losses induced by resistive wall wakefields in the undulator vacuum chambers. The modulators for both FEL-1 and FEL-2 have fixed, linear polarization. They must be tunable in the 240-360 nm range, a requirement that can be satisfied by a wide range of period lengths among which the shortest acceptable, ~ 10 cm, is favoured for FERMI. The first stage radiator and the second stage modulator for FEL-2 are also designed for fixed, linear polarization. All the other undulators are of the variable polarization type, based on the widely used APPLE-II (Advanced Planar Polarized Light Emitter) configuration, the most efficient one for this application. All polarizations are tunable over the full design tuning ranges of 100 to 40 nm and 40 to 10 nm of FEL-1 and FEL-2 respectively. Because no analytical expression is known for the field amplitude as a function of the geometrical parameters of such devices, a semi-empirical formula is used derived by fitting the results of 3D magnetostatic calculations performed on a number of different special cases. ... |
CDR chapter 8 - Photon Beam Transport and Diagnostics
|
FERMI@Elettra includes two separate undulator sections named FEL-1 and FEL-2 delivering radiation in the 100 ÷ 40 nm and 40 ÷ 10 nm ranges, respectively. For FEL-1 a 50 ÷ 100 fs pulse is delivered with a peak power of about 1-5 GW, and ~ 1014 photons per pulse are expected. FEL-2 is characterized by a 200 fs pulse carrying about 1 GW and 1012 (fresh bunch mode). In order to characterize, select, and carry the photon beam to the experimental endstations, a set of optical systems is placed after the undulators. Characteristics such as energy, energy resolution, pulse length, intensity, arrival time, polarization, and so on, are determined by means of several diagnostics located between the undulators and the experimental hall. Gas-based systems such as absorbers and intensity monitors are mounted within window-less in-vacuum sections. The gases intercepting the radiation axis serve as natural absorbers reducing the overall photon flux. Additionally, the gas ionization signal gives information about the relative intensity of the beam. A system of slits removes unwanted off-axis radiation mainly coming from spontaneous emission. It also works as an angular collimator making possible spectral-angular filtering. At the entrance of the experimental hall radiation coming from each FEL impinges on a plane mirror used as a power absorbing element upstream of the more delicate elements along the beamlines. Moreover, this optical element is important from the radio-protection point of view. ... |
CDR chapter 9 - Timing and Synchronization
|
The FERMI timing and synchronization (T&S) scheme is based on a hybrid system utilising both “pulsed” and continuous wave (CW) timing techniques. The “pulsed” technique has been originally developed at MIT [1]: an ultra-low phase noise pulsed laser, called optical master oscillator (OMO), is locked to the radiofrequency reference generator. Its pulses distributed over dispersion compensated fiber links (FO) give the time reference to all the “pulsed” timing clients, such as lasers and diagnostics devices The “CW” timing technique, developed by LBNL at Berkeley [2] is based on a frequency stabilized CW laser amplitude modulated by the radio frequency (RF) of CW timing clients, such as low level RF systems. In this scheme, the FO links are stabilized using the optical mixing concept which fully exploits a carrier frequency 105 times higher than the repetition rate of the pulsed system. The FERMI timing system is compatible with both the European (fS-band-EU=2.998010 GHz) and U.S. (fS-band-US = 2.856 GHz) S-band frequencies, a ecessary condition since the fourth harmonic (X-band) linearizer, that is part of the FEL design, will work at the US frequency. The greatest common divisor of these two frequencies is the coincidence frequency fCOIN (15.779 MHz) used to generate the "bunch clock" at the FEL repetition rate frequency fbunch=10-50 Hz. At each period of the coincidence frequency waveform the rising slopes of the (EU) S-band and the (US) aX-band waveforms overlap, thus providing the required phase coincidence. ... |
CDR chapter 10 - Laser Systems
|
Laser systems will undoubtedly be crucial to determining the overall performance of FERMI. As can be seen from the previous chapters, the laser systems will include the photoinjector laser (PIL), laser heater (LH), seed laser (SL) and beam-line lasers (BLL). In addition, the timing and synchronization system will contain a mode-locked optical master oscillator (OMO). This chapter emphasizes the PIL and SL, as the technology for the other two systems will be very similar; the OMO has already been considered in Chapter 9. The main issues related to the FERMI laser systems have been thoroughly studied during the last year, some of the important points and preliminary data have been summarized and presented in [1]. We note that nearly all the features required of the laser systems by the FERMI FEL design are within reach of readily available laser technology. In contrast, a few characteristics, such as pulse/beam shaping for the photoinjector laser as well as wavelength stability and pulse quality for the seed laser are challenging and will require additional R&D. As it will be shown later, the comparison of the existing laser technologies for obtaining the required parameters indicated the basic unit in all cases to be a Ti:Sapphire chirped pulse amplifier pumped by diode pumped solid state lasers. The photoinjector laser includes two amplifier stages – a regenerative stage followed by a multipass stage – to reach a pulse energy of 20 mJ in the IR. Pulse shaping is done partially in the IR, by an acoustooptic dispersive filter (DAZZLER), and is completed in UV in a transmission grating-based stretcher or Fourier-system. Beam shaping is done either in the IR or in the UV by an aspheric shaper. A small part (~400 μJ) of the IR beam is split away and transported for use by the laser heater. ... |
CDR chapter 11 - Electron Beam Diagnostics
|
The FERMI beam diagnostics includes a complete set of instruments specifically designed to completely characterize the FERMI free electron beams. Measurements to be performed at different machine sections are presented, starting from the photo-injector and moving downstream, through the linac and the FEL. The characterization of the photo-injector, given the electron bunch physical properties, is based on a set of traditional instruments. The bunch charge and the transverse and longitudinal profiles are measured by means of a current transformer, movable Faraday cups, Yttrium Aluminium Garnet Cerium crystal (YAG:Ce) screens and a Cherenkov radiator coupled to a single sweep streak camera respectively. This set-up ensures a detailed characterization of the bunch non-gaussian longitudinal profile, one of the new features of the FERMI photo-injector. A movable slit plus screen assembly measures the emittance of the space charge dominated, low energy bunch, while a dispersive beamline is foreseen for energy, energy spread and longitudinal phase space measurements. The uncorrelated energy spread can also be measured exploiting the bunch correlation between energy and longitudinal position. The following linac sections are equipped with standard intra-section diagnostics stations (X-Y position, profile and charge). Two cavity beam position monitors (BPMs) with micrometer resolution measure the bunch transverse position at the entrance of the first ELETTRA type accelerating section (S1) for beam centering. ... |
CDR chapter 12 - Controls
|
The control system provides operators, machine physicists and scientists with a comprehensive and easy‑to‑use set of tools to control machine components and experimental beam lines. It is designed to be robust and reliable in order to insure long periods of operation without failures or malfunctions. Diagnosis and possibly repair capabilities are implemented in order to allow for remote recovery from malfunctions of both the equipment and the control system itself, with minimum impact on the facility operation. In general, the control system design is flexible enough to accommodate the specific requirements of a large variety of both conventional and highly specialized devices that are installed and controlled on the accelerator, the beamlines and the experimental stations. The control system consists of several computers distributed around the facility that interface with the different equipment and acquire data. A number of PC-based consoles allow to remotely operate the machine from the control room. Similar consoles in the experimental hall are used to control the experiments. A switched Ethernet network connects all the control system computers. State-of-art software technologies are employed, based on open standards and free open-source packages. A uniform and homogeneous software environment using the GNU/Linux operating system and the Tango control system software is adopted for the whole control system. A high level software framework supports model based design of machine physics applications. General purpose control room applications (graphical panels, synoptics, alarms, archiving, logging, etc.) are implemented using the Tango package software tools. ... |
CDR chapter 13 - Alignment
|
In order to ensure that the machine components are placed according to the physics requirements, FERMI needs well defined survey and alignment procedures and techniques. Surveys are necessary for placement of components and for support of conventional facilities. ELETTRA has a solid history of quality surveying and engineering. The needs of FERMI will be well served by building on the available infrastructure of reference points and the current observation techniques. All surveys will be based on a single coordinate system integrating the existing surface network and storage ring surveys. Components that have been carefully fiducialized can then be installed and aligned using a series of steps from the ground up and later checked with a survey map. The linac area needs special attention, both for the quality control of the accelerator sections and for the alignment methods to be applied in the field. The alignment coordinate system is a Cartesian right-handed system, with the origin placed at a particular point in the beamline. A three-dimensional network methodology is proposed, with the FERMI network integrated with that of the existing ELETTRA storage ring. The instrumentation will include theodolites, total stations, laser trackers and precise levels. The areas to be aligned are the injector, linac, undulator, photon beamlines and experimental hall. The tolerances for the injector, linac and undulator components, 100 μm rms, are achievable with established techniques. The fine adjustment to the BPM and quadrupole positions in the undulator will be done via the beam-based alignment technique described in Chapter 7. The components of the experimental hall have relatively loose alignment requirements, and traditional measuring techniques like those currently in practice at ELETTRA can be used for the placement of components. |
CDR chapter 14 - Civil Engineering
|
The FERMI@Elettra project requires the excavation and construction of several new buildings to house the accelerator, FEL’s, Experimental area and associated offices and laboratories. These new infrastructures will be furnished and supplied with conventional systems. Furthermore the project will use most of the existing well-established infrastructures at ELETTRA as a development platform. The project will evolve from the existing infra-structure that houses the storage ring linac injector. This will require a physical extension of the present linac tunnel in both upstream and downstream directions, excavation and construction of the FEL and experimental hall. New conventional systems will be provided as will upgrading of the existing ones. The linac is situated below ground at a depth of ~5 m in a tunnel 110 m long. The linac tunnel will be extended backwards by an additional ~80 m for a total length of 195 m and a surface area of ~750 m2. The tunnel will conform to health physics regulations. Adjacent to this backward extension a laboratory area will be built up that will serve for the laser systems used by the Linac. This area, at the same depth as the linac, will also house support laboratories and storage rooms. The control room for the FEL will be situated at ground level approximately midway along the accelerator. The surface building (Klystron gallery) housing the RF power sources and ancillary equipment for the linac will also be extended above the tunnel to the same length. ... |
CDR chapter 15 - Radiation Protection
 |
The main goals of Radiation Protection are to evaluate the ionizing radiation sources produced in the linac and undulator tunnel and in the beamline hutches, to estimate the needed biological shielding barriers and to adopt appropriate safety systems for personnel protection. In the design of the shielding for FERMI, the different sources of ionizing radiation produced inside the accelerator tunnel have been considered. The radiation intensity depends strongly on the electron beam energy and current and is related to the beam loss frequency and distribution inside the machine. The evaluation of the shielding thickness has been performed on the basis of semi-empirical expressions available from the literature and that have been validated by experimental measurements performed at ELETTRA during radiation protection surveys. The goals of dose limits for the various areas around the facility (free, supervised, and controlled areas) are lower than those established by Italian regulations in compliance with the European/Euratom directives. According to our project goals, only the linac and undulator tunnel will be classified as controlled area. The undulator Service Area and seeding laser room will be considered supervised areas while the experimental hall outside the beamline hutch will be a free access area. Only personnel involved in machine operation will be classified as radiation workers and will be monitored using personal dosimeters. The experimental users will be not classified from the radiation protection point of view. Radiation monitoring outside the tunnels will be based on passive dosimeters and on a network of active gamma and neutron detectors located at various points of the experimental hall and of the undulator Service Area. Similarly to the ELETTRA storage ring, radiation monitors will be connected to the radiation safety system and beamlines operation will be prevented if predetermined gamma doserate thresholds is exceeded. |
