USTRAT/SUPA, Glasgow
A new un-averaged 3D numerical model has been developed that will allow investigation of previously unexplored FEL physics. Previous 1D models have allowed exploration of the effects of amplification of coherent spontaneous emission and non-localised electron dynamics (see e.g. [1] and refs. therein.) A 3D model was also developed based upon a mixed finite element\Fourier method [2]. However, due to some limitations in the parallel routines, this restricted somewhat the FEL systems the model could describe. A significantly modified version of this model is presented here which does not require a finite element description and uses only transforms in Fourier space. This allows more effective and consistent data organization across multiple parallel processors enabling larger, more complex FEL systems to be studied. Furthermore, unlike the previous 3D model which uses commercially produced numerical packages, the new simulation code uses only open-source routines which will ultimately allow it to be freely available (open-source.)
[1] BWJ McNeil, GRM Robb & MW Poole, MPPB060, PAC, Portland, USA, 2003
[2] C.K.W. Nam, P. Aitken, & B.W.J. McNeil, MOPPH025, FEL Conference, Korea, 2008
STFC/DL/ASTeC, Daresbury, Warrington, Cheshire
USTRAT/SUPA, Glasgow
Cockcroft Institute, Warrington, Cheshire
Start-to-end modelling of a SASE mode-locked FEL amplifier scheme [1] is presented using a superconducting re-circulating linac design[2]. Locking of the modes is achieved by modulating the electron beam energy at the mode frequency spacing. Previous studies [3] have shown that in a High Harmonic Generation (HHG) seeded mode-coupled FEL amplifier scheme (no electron beam energy modulation), although the attosecond pulse train structure of the seed is amplified through to saturation, temporal broadening of the individual pulses occurs. An HHG seeded mode-locked FEL amplifier scheme is modelled and it is seen that the temporal spikes of the HHG seed must be correctly phase-matched with the electron beam energy modulation for successful operation. By using a filtered HHG seed, which removes the seed’s attosecond pulse train structure, no such phase matching is required. Despite the absence of an initial attosecond pulse structure, a modal structure develops and is subsequently amplified to generate an attosecond pulse train with the good temporal coherence properties of the seed, significantly shorter individual pulse widths and higher peak powers than may be achieved in the other schemes.
[1] N.R. Thompson, B.W.J. McNeil, Phys. Rev. Lett. 100, 203901 (2008)
[2] P.H. Williams et al, WE5RF, 23rd PAC, Vancouver (2009)
[3] B.W.J. McNeil et al, MOCAU04, 30th Int. FEL Conf., Gyeongju (2008)
STFC/DL/ASTeC, Daresbury, Warrington, Cheshire
JAI, Oxford
USTRAT/SUPA, Glasgow
The New Light Source (NLS) project was launched in April 2008 by the UK Science and Technology Facilities Council (STFC) to consider the scientific case and develop a conceptual design for a possible next generation light source based on a combination of synchronised conventional laser and free-electron laser sources. The requirement identified for the FELs was continuous coverage of the photon energy range 50-1000eV with variable polarisation, 20fs pulse widths and good temporal coherence to as high a photon energy as possible. This paper presents a design study of three separate FELs which in combination satisfy these requirements. It is proposed to use an HHG seed source tunable from 50-100eV giving direct seeding at the fundamental FEL wavelength up to 100eV, then one or two stages of harmonic upconversion within the FEL to reach the higher photon energies. FEL simulations using realistic electron beam distributions tracked from the gun to the FEL will be presented, illustrating the predicted coherence properties of the FEL output at different photon energies.
