X -band linac design

Abstract

In order to explore physics in the EUV to soft x-ray region, we have designed a machine which is capable of accelerating a $\ensuremath{\sim}250\text{ }\text{ }\mathrm{pC}$ electron bunch to an energy of $\ensuremath{\sim}1\text{ }\text{ }\mathrm{GeV}$. The front end of the CLARA (Compact Linear Accelerator for Research and Applications) system at Daresbury Labs will be used as an S-band injector of $\ensuremath{\sim}180\text{ }\text{ }\mathrm{MeV}/\mathrm{c}$, sub-ps FWHM, $\ensuremath{\sim}250\text{ }\text{ }\mathrm{pC}$ electron bunches into the XARA (X-Band Accelerator for Research and Application) system. A rf feasibility study has been carried out for a structure operating in the $2\ensuremath{\pi}/3$ mode at a frequency of 11.9942 GHz which is fed by a SLED klyston setup. The average cell of this structure has an iris radius of $⟨a⟩=3.2\text{ }\text{ }\mathrm{mm}$ and a shunt impedance of $⟨{R}_{s}⟩=106.55\text{ }\text{ }\mathrm{M}\mathrm{\ensuremath{\Omega}}/\mathrm{m}$. A high target gradient of $80\text{ }\text{ }\mathrm{MV}/\mathrm{m}$ for a single-bunch operation of the linac is necessary due to spatial constraints at Daresbury Labs. We have also implemented Gaussian detuning of the linac in order to future-proof the project for potential multibunch operation of the machine. After combining the rf study with an analysis of the uncoupled long-range wakefield and the short-range transverse wakefields, the optimal structure parameters are outlined as a compromise between the shunt impedance, electrical breakdown rate and wakefields in the structure. As novel designs will be tested using this free-electron laser (FEL) an increased beam charge may be useful. Therefore a beam dynamics study via the particle tracking code elegant has been performed to assess how the beam quality evolves while traveling through the XARA rf structures for different bunch charges and beam offsets. These simulations reveal how the bunch is disturbed for varying bunch charges and offsets and give an initial indication of how sensitive the beam parameters (beam centroid position, emittance, RMS beam size, etc.) are to the wakefields generated in XARA. An analytical formulation of the beam motion as it travels through the XARA linac has been utilized to calculate the emittance growth. This allows for comparison between analytical and numerical simulation of the beam dynamics to give confidence in the results. The beam dynamics study shows that for a bunch charge of ${Q}_{b}=250\text{ }\text{ }\mathrm{pC}$ and a beam offset of ${C}_{x}={\ensuremath{\sigma}}_{z}/2=0.253\text{ }\text{ }\mathrm{mm}$, the normalized emittance growth at the end of the XARA linac is $\mathrm{\ensuremath{\Delta}}{\ensuremath{\epsilon}}_{Nx}/{\ensuremath{\epsilon}}_{Nx}\ensuremath{\sim}270%$. This may be mitigated by the design of a magnetic lattice which would include kickers to recenter the beam on the electrical axis of the structure, reducing the effect of the wakefields on the bunch. Currently it is probable that a maximum bunch charge of $\ensuremath{\sim}250\text{ }\text{ }\mathrm{pC}$ will be utilized for this machine; however a future design of a magnetic focusing lattice may allow higher charges to be viable and will reduce the emittance growth.