BOOSTER CAVITY AND FUNDAMENTAL POWER COUPLER DESIGN ISSUES FOR bERLinPro∗

HZB has started building the 50MeV, 100mA demonstrator energy-recovery-linac (ERL) facility bERLinPro. The high power injector system needs to deliver this beam at 6.5MeV by combining the energy gain of a 1.4 cell SRF photo-injector and three Cornell style 2-cell booster cavities. One booster cavity will be operated at zero-crossing for bunch energy chirping. Thus two booster cavities have to deliver 2MV each requiring a strong coupling with a loadedQ of 105. To house the two envisaged KEK fundamental power couplers (FPC) with the cavity, the geometry was slightly modified. Further, to increase coupling and reduce transverse kick effects to the beam, a ”golf-tee” antenna tip was designed. This paper summarizes the SRF challenges for the booster cavities, the operational conditions and the modification to the KEK couplers, including tracking calculations to estimate the coupler kick effect to higher order. INTRODUCTION For the bERLinPro (see Fig. 1) ERL [1] the injector section will consist of a 1.4 cell SRF photo-injector cavity [2] delivering a 100mA 2.3MeV beam at a normalized emittance of ≤ 1 mm mrad and a module with 3 SRF booster 2-cell cavities of Cornell design [3]. The latter has to preserve the low emittance of the beam, accelerate it to the injection energy of the recirculator of 6.5MeV and imprint an energy chirp for bunch compression. Table 1 summarizes the expected beam parameters from the SRF photo-injector cavity and the operating parameter for the booster cavities. To allow for an energy gain of 2.1MeV per cavity the heavily beam-loaded structures have to be operated at a strong input coupling with a loaded Q of 105. To account for the high required forward power of 230 kW, it was decided to ∗ Work supported by German Bundesministerium für Bildung und Forschung, Land Berlin, and grants of Helmholtz Association † [email protected] Figure 1: Scheme of the bERLinPro ERL with the injector section on the top right of the picture. Table 1: Beam parameters as expected from the SRF photoinjector cavity as input parameter for the Booster coupler kick calculations and Booster cavity operating parameters. n,y,x 1 mm mrad σt 5 ps rms σy,x ≈1.0mm rms Iavg,beam 100mA Ekin,ini 2.3MeV Vacc 2.1MV E0 7, 19, 19MV/m Φacc -90, 0, 0 deg equip the cavity with two high power KEK-style coaxial couplers [4]. The coupler will have a fixed coupling providing matching conditions for full beam current. In order to reduce dipole-like coupler kicks and emittance growth and the power load limit of the coupler, two opposed couplers will provide the power to the cavity. However, to achieve the mandatory operational parameters, the cavity geometry and the coupler design had to be modified. To improve the coupling strength and to reduce the effects of higher order field deviations caused by the coupler, the antenna tip was modified. In the following sections the modified design will be discussed as well as its influence on the beam quality by tracking a bunch using 3D field maps. THE MODIFIED BOOSTER CAVITY DESIGN To house the large coaxial coupler with an outer conductor diameter of 82mm, it was tapered to 70mm and the beam tube on the coupler side was increased from 70 to 88mm. Introducing a nose transition between the exit iris and the coupler beam tube, the reduction of the fundamental’s R/Q could be limited to 0.4%. Further the level of peak field ratios remains unchanged compared to the original design. By that the penetration depth of the inner conductor into the beam tube would still be 14mm for the targetQL. To achieve Figure 2: New ”golf tee” shaped coupler tip to increase the coupling strength and reduce coupler kick effects. WEPRI007 Proceedings of IPAC2014, Dresden, Germany ISBN 978-3-95450-132-8 2490 C op yr ig ht © 20 14 C C -B Y3. 0 an d by th e re sp ec tiv e au th or s 07 Accelerator Technology Main Systems T07 Superconducting RF a stronger coupling at less penetration the antenna tip was modified to a so called ”golf tee” shape which is described by Viviani’s curve, see Figure 2. By optimizing the ellipse aspect ratio of the tip’s surface, its thickness and the blending edge of the opening to the outer conductor the target QL was achieved at 3 mm penetration. The distance between coupler and cavity exit iris was limited by the need to house the helium vessel of the cavity in between. In Table 2 the calculated RF properties of the modified structure are given. The shown Qext of the first two dipole bands assume perfect absorption by beam tube HOM absorbers downstream of the cavity. Table 2: Calculated RF parameters of the Booster cavity with two coaxial modified KEK type couplers. Number of cells 2 R/Q ‖ 219 Ω fRF 1.3 GHz Epeak/Eacc 2.0 Bpeak/Eacc 4.4 mT/MVm-1 Qext TE111 dipole 130, 370 Qext TM110 dipole 170, 7300 QL at 3mm penetration depth 1.05 · 105 Pforward 230 kW COUPLER KICK AND EMITTANCE DILUTION CALCULATIONS To evaluate the new coupler design and prove their feasibility to maintain the high beam quality, coupler kick calculations were performed using 3D field maps calculated by CST MWS [7] using the new hexahedral mesh of version 2014. The fields were calculated with electric and magnetic boundary conditions to not only simulate the standing wave (SW) regime of the low beam current mode of operation, but also the traveling wave (TW) scenario with full beam current accelerated. Further the on axis fields were crosschecked with Superfish [8] to check the mesh quality and to rule out any unwanted offsets. Figure 3: Misalignment scenarios studied for the modified KEK coupler as attached to the 2-cell SRF booster cavity. The upper inner conductor is shifted with respect to the opposing FPC port in x and y and (for the ”golf tee” only) rotated in φ. Two methods were used to calculate the kick to the beam and the emittance dilution neglecting space charge. The obtained field maps were imported within CST into the PS Particle In Cell (PIC) tracking solver and a bunch with bERLinPro phase space parameter created externally by ASTRA [9] served as the initial particle distribution. The PIC results were compared with the outcome of ASTRA runs using 3D ASCII exported (step 1 mm) field maps of the same CST eigenmode simulation. Several scenarios were studied for acceleration phases of +90, 0 and -90 deg.: • The bare cavity structure as reference • Twin original KEK couplers, aligned and misaligned • Twin modified ”golf tee” couplers, aligned and misaligned + angle (rotation in φ) • Single coupler with design QL Figure 3 shows the misalignment scenarios studied with simulations. The greatest influence is by an error in the relative penetration depth of the two antennas and for the ”golf tee” type the rather improbable shift of the inner conductors angle with respect to the coupler axis. For this study a rather pessimistic alignment error of 3 mm vertical and 2 mm in horizontal direction was assumed. Figure 4 shows the on−150 −100 −50 0 50 100 150 200 250 300 350 400 −2 −1 0 1 2 3 4 5 6 x 10 7 z(mm) E z , E y ⋅1 e3 ,c ⋅B ⋅1 e3 ( V /m ) Golf Tee misaligned