Expected properties of the radiation from VUV-FEL / femtosecond mode of operation / E.L. Saldin, E.A. Schneidmiller, M.V. Yurkov TESLA Collaboration Meeting, September 6-8, 2004 Experience from TTF FEL, Phase I Two-step compression with BC2 and BC3 Expected properties of the radiation An outlook for commissioning procedure
Nominal mode of operation of VUV FEL Low-emittance injector Linearization of longitudinal phase space: 3rd harmonic RF cavity Two bunch compressors: BC2 and BC3 FEL saturation down to 6 nm expected Relatively "long" SASE pulse (200 fs FWHM) Advantages of linearized compression: Longer high-current (lasing) part of the bunch is less sensitive to CSR and LSC: Standard diagnostics (BPMs, screens) is better suited for such a regime since a larger fraction of total charge contributes to the lasing 3rd harmonic rf cavity is not available for the next few years
TTF FEL, Phase I Radiation wavelength 80-120 nm Radiation pulse energy at saturation 60 µj Radiation pulse duration (FWHM) 40 fs Radiation peak power 1.5 GW rf-gun bunch compressor undulator Bunch compressor forms bunches with strongly non-gaussian shape Short spike in the head of electron bunch produces SASE FEL radiation A technique of bunch tailoring possesses significant potential for further reduction of the FEL pulse length
TTF FEL, Phase I Two modes of operation were tested: BC2 only. Beam dynamics is nontrivial: space charge dominated (see DESY 03-197). Operation with BC1 and BC2 gave more possibilities to control beam parameters. Electron beam on the screen 1EXP3 under lasing conditions (TTF1): only BC2 is on (left); both BC1 and BC2 are on (right). Full width of the screen corresponds to approx. 8 MeV
To what extent the TTF1 experience (femtosecond mode) can be extrapolated? Optimized injector: better emittance Final compression at higher energy (350-400 MeV) in BC3 With mild preliminary compression in BC2 one can get reasonable parameters of the leading peak (spike) The most critical collective effects (CSR and LSC) can be tolerated this way At present level of understanding of the beam dynamics we do not see any effects preventing FEL from successful operation (down to 6 nm)
Simulation results (undulator entrance) Astra: up to the end of ACC1 Elegant: up to the end of ACC4 (CSR included) Astra: up to the undulator entrance (simplified lattice, no dogleg) 4 ps Gaussian laser pulse, charge 0.5 nc and 1 nc For the slice with a maximal current (1.3 ka) For the slice with a maximal current (2.2 ka) mean geometric of x- and y-normalized the mean geometric of x- and y- normalized ittances is about 1.5 mm-mrad, and the local emittances is about 3.5 mm-mrad, and the local rgy spread is about 300 kev energy spread is about 300 kev
SASE simulation with code FAST Top: energy in the radiation pulse versus undulator length. Bottom: fluctuations of the energy in the radiation pulse versus undulator length. Left column: the case of bunch charge 0.5~nC. Right column: the case of bunch charge 1~nC. Solid and dashed lines refer to the radiation wavelength 30~nm and 6~nm.
SASE simulation with code FAST: Bunch charge 0.5 nc, saturation length is 18 m for wavelength 30 nm Left column: temporal and spectral structure of the radiation pulse. Three single shots are shown with different line shapes (solid, dashed, and dotted). Grey line shows profile of the bunch current. Right column: averaged distribution of the radiation intensity in the near and far zone
SASE simulation with code FAST: Bunch charge 0.5 nc, saturation length is 22 m for wavelength 6 nm Left column: temporal and spectral structure of the radiation pulse. Three single shots are shown with different line shapes (solid, dashed, and dotted). Grey line shows profile of the bunch current. Right column: averaged distribution of the radiation intensity in the near and far zone
Summary Conclusion of our study is optimistic: femtosecond mode of operation will allow to overlap complete operating wavelength range of the VUV FEL. Minimum wavelength will be mainly limited by the available energy of the electron beam. With six undulator modules we can go down well below design value of 6 nm wavelength. In principle, saturation at 6 nm can be achieved with five undulator segments. Replacement of the last section with frequency doubler will allow to reach 3 nm wavelength with GW level of output power. Commissioning of the VUV FEL at a longer wavelength is preferable. Starting of the FEL commissioning procedure at the wavelength around 30 nm will allow to use all experimental experience collected during operation of TTF FEL, Phase I. Temporal and spectral properties of the radiation are close to those obtained at TTF FEL, Phase I. Jump in the radiation wavelength is about a factor of 3 only which allows to use photon diagnostic tools and methods developed at the TTF FEL, Phase I. The most critical step will be obtaining the first indication of the amplification process.
An outlook for FEL comissionning procedure: To get lasing one has to satisfy three conditions at the same time: The local quality of the spike (current, emittance etc.) must be sufficient for lasing, The orbit of the spike in the undulator must be straight enough The spike must be optically matched to the undulator (with some accuracy) Since the spike contains some 10 % of total charge, and the local properties (emittance, Twiss parameters, orbit etc.) may differ dramatically from the projected ones, and there is no diagnostics on a 10 fs scale, very precise optimization of the projected parameters is not required. This simplifies the first steps of commissioning ("quick-and-dirty" set up is sufficient) and complicates the last one, tuning SASE. First two steps: injector commissioning and beam through the undulator. On-crest in all modules, only standard beam diagnostics is required (BPMs, screens, possibly pyrodetectors and wirescanners). Last step: empirical search and optimization of SASE