Juwen Wang

SLAC/CERN high gradient tests of an X-band accelerating section

High frequency linear collider schemes envisage the use of rather high accelerating gradients: 50 to 100 MV/m for X-band and 80 MV/m for CLIC. Because these gradients are well above those commonly used in accelerators, high gradient studies of high frequency structures have been initiated and test facilities have been constructed at KEK, SLAC and CERN. The studies seek to demonstrate that the above mentioned gradients are both achievable and practical. There is no well-defined criterion for the maximum acceptable level of dark current but it must be low enough not to generate unacceptable transverse wakefields, disturb beam position monitor readings or cause RF power losses. Because there are of the order of 10,000 accelerating sections in a high frequency linear collider, the conditioning process should not be too long or difficult. The test facilities have been instrumented to allow investigation of field emission and RF breakdown mechanisms. With an understanding of these effects, the high gradient performance of accelerating sections may be improved through modifications in geometry, fabrication methods and surface finish. These high gradient test facilities also allow the ultimate performance of high frequency/short pulse length accelerating structures to be probed. This report describes the high gradient test at SLAC of an X-band accelerating section built at CERN using technology developed for CLIC.

Room temperature accelerator structures for linear colliders

Early tests of short low group velocity and standing wave structures indicated the viability of operating X-band linacs with accelerating gradients in excess of 100 MeV/m. Conventional scaling of traveling wave traveling wave linacs with frequency scales the cell dimensions with /spl lambda/. Because Q scales as /spl lambda//sup 1/2/, the length of the structures scale not linearly but as /spl lambda//sup 3/2/ in order to preserve the attenuation through each structure. For the NLC we chose not to follow this scaling from the SLAC S-band linac to its fourth harmonic at the X-band. We wanted to increase the length of the structures to reduce the number of couplers and waveguide drives which can be a significant part of the cost of a microwave linac. Furthermore, scaling the iris size of the disk-loaded structures gave unacceptably high short range dipole wakefields. Consequently, we chose to go up a factor of about 5 in average group velocity and length of the structures, which increases the power fed to each structure by the same factor and decreases the short range dipole wakes by a similar factor. Unfortunately, these longer (1.8 m) structures have not performed nearly as well in high gradient tests as the short structures. We believe we have at least a partial understanding of the reason and will discuss it below. We are now studying two types of short structures with large apertures with moderately good efficiency including: 1) traveling wave structures with the group velocity lowered by going to large phase advance per period with bulges on the iris, 2) /spl pi/ mode standing wave structures.

RF Design of X-Band RF Deflector for Femtosecond Diagnostics of LCLS Electron Beam

We designed a successful constant impedance traveling wave X-band rf deflector for electron beam diagnostics at the 14 GeV SLAC Linac Coherent Light Source (LCLS). This is the first practical deflector built with a waveguide coupler. The 1-meter rf deflector produces 24 MeV peak transverse kick when powered with 20 MW of 11.424 GHz rf. The design is based on our experience with high gradient X-band accelerating structures. Several deflectors of this design have been built at SLAC and are currently in use. Here we describe the design and distinguishing features of this device.

RDDS cell design and optimization for the linear collider linacs

Each of the JLC/NLC main linacs will consist of /spl sim/1 million complex 3D accelerating cells that make up the 1.8 meter Rounded Damped Detuned Structures (RDDS) along its eight kilometer length. The RDDS is designed to provide maximum accelerating gradient to the beam while being able to suppress the long-range transverse wakefields to a satisfactory level. Using the 2D finite element code, Omega2, a 15% improvement in shunt impedance is found by changing the basic cell shape from a straight cylinder to a round outer wall contour that connects to slightly bulging circular disk noses. The HOM damping manifold is then designed around this optimal cell shape to improve the cell-to-manifold coupling for the dipole mode and the vacuum conductance under the frequency and minimal Q-reduction constraints for the fundamental mode. We use both MAFIA and the 3D finite element Omega3 code for this step to obtain a manifold geometry that consists of a round waveguide with additional narrow coupling slots that cut into the cell disks. As a time and cost saving measure for the JLC/NLC, the RDDS cell dimensions are being determined through computer modeling to within fabrication precision so that no tuning may be needed once the structures are assembled. At the X-band operating frequency, this corresponds to an error of a few microns in the cell radius. Such a level of resolution requires highly accurate field solvers and vast amount of computer resources. We will present calculations with the parallel code Omega3P that utilizes massively parallel computers such as the Cray T3E at NERSC. The numerical results will be compared with cold test measurements performed on RDDS prototypes that are diamond-turned with dimensions based on Omega3P simulations.

High gradient experiments on NLCTA accelerator structures

This paper presents new results of high-gradient studies performed on a 1.8 m traveling-wave accelerator section with detuned high-order deflecting modes. This structure was designed initially for studies of detuned structures and will be installed in the Next Linear Collider Test Accelerator (NLCTA). The paper describes the test set-up in the Accelerator Structure Test Area (ASTA) including electron gun, prebuncher, pre-accelerator, spectrometer, Faraday cups, 200 MW SLED-II power compression system, Magic-T type phase shifters and attenuators. Rf processing, detailed dark current analysis, radiation problems, and beam acceleration measurements are discussed.

Recent results & plans for the future on SLAC Damped Detuned Structures (DDS)

The cells in the SLAC DDS are designed in such a way that the transverse modes excited by the beam are detuned in a Gaussian fashion so that destructive interference causes the wake function to decrease rapidly and smoothly. Moderate damping provided by four waveguide manifolds running along the outer wall of the accelerator is utilised to suppress the reappearance of the wake function at long ranges where the interference becomes constructive again. The newly developed spectral function method, involving a continuum of frequencies, is applied to analyze the wake function of the DDS 1 design and to study the dependence of the wake function on manifold termination. The wake function obtained with the actually realized manifold terminations is presented and compared to wake function measurements recently carried out at the ASSET facility installed in the SLAC LINAC.

Traveling wave structure optimization for the NLC

The JLC/NLC linac must accelerate multi-bunch beams in order to obtain a luminosity >3/spl times/10/sup 34/ cm/sup -2/ sec/sup -1/ at a center of mass energy of 1 TeV. It is essential for the structure design to minimize the long and short-range dipole wakefields to prevent emittance degradation and the beam breakup instability (BBU). In addition, the structures must operate at a high RF gradient to minimize the linac cost. High-power testing of prototype structures at SLAC has shown noticeable damage. The damage is largest in the front of the structure, where the group velocity is high, and there is minimal or no damage at the back end, where the group velocity is low. Theoretical analysis using a simple circuit model suggests using structures with a lower group velocity, on the order of a few percent, would be a way of avoiding damage. For the standard 2/spl pi//3 accelerating mode, it is difficult to lower the group velocity without losing efficiency or increasing the wakefields. With this in mind, we have taken the phase advance as an additional parameter in structure optimization. We found that a low group velocity structure at higher phase advance can maintain high RF efficiency and low wakefields. In this paper, we study the impact of phase advance on structure performance. We then optimize the NLC S-band and X-band structures to meet design requirements.

Analysis and application of manifold radiation in DDS 1: First experiences

The cells in the SLAC DDS are designed in such a way that the transverse modes excited by the beam are detuned in a Gaussian fashion so that destructive interference causes the wake function to decrease rapidly and smoothly. Moderate damping provided by four waveguide manifolds running along the outer wall of the accelerator is utilized to suppress the reappearance of the wake function at long ranges where the interference becomes constructive again. The newly developed spectral function method, involving a continuum of frequencies, is applied to analyze the wake function of the DDS 1 design and to study the dependence of the wake function on manifold termination. The wake function obtained with the actually realized manifold terminations is presented and compared to wake function measurements recently carried out at the ASSET facility installed in the SLAC LINAC.

R&D of a Super-compact SLED System at SLAC

We have successfully designed, fabricated, installed and tested a super compact X-Band SLED system at SLAC. It is composed of an elegant 3dB coupler / mode converter / polarizer and a single spherical energy storage cavity with high Q0 of 94000 and diameter less than 12 cm. The available RF peak power of 50 MW can be compressed to a peak average power of more than 200 MW in order to double the kick for the electron bunches in a RF transverse deflector system and greatly improve the measurement resolution of both the electron bunches and the Xray FEL pulses. High power operation has demonstrated the excellent performance of this RF compression system without RF breakdown, sign of pulse heating and RF radiation. The design physics and fabrication as well as the measurement results will be presented in detail. INTRODUCTION This diagnostics for X-ray temporal measurement based on the transverse deflectors, the magnetic spectrometer and the Ce:YAG screen located downstream of the FEL undulator has been intensively used at the LCLS operation [1]. Its layout is shown in Figure 1. Figure 1: Diagnostics layout of the X-ray temporal measurement at the LCLS. In order to improve temporal resolution, we have designed and fabricated a novel super compact SLED system to double the peak deflection. [2] This paper describes its principle, design and technical advances. DESIGN OF SUPER COMPACT SLED SYSTEM More than forty years ago, SLAC developed the SLED system to obtain high peak RF power in exchange for the RF pulse length reduction [3]. The key components of a SLED system include a 3dB coupler with two 90° apart power ports and two high Q energy storage cavities. The LCLS deflector is made of two traveling structures, Each one is a 1.0 m long, constant impedance structure with transverse impedance of 41.9 MΩ/m, filling time Tf=106 ns (group velocity of -3.165 % speed of light) and attenuation factor τ=0.62 Neper. If we assume the similar average RF power of 106 ns pulses for both the SLEDed pulse and non SLEDed flat pulses to feed a backward wave constant impedance deflector, the corresponding kick voltages are shown in Fig. 2. Figure 2: Kick voltage along the deflector structure. We need to optimize the SLED system by calculating its total gain for various coupling coefficients for the high Q0 cavity, its Q values and pulse length. Figure 3 shows that the highest gain of larger than factor of 2 can be obtained for Q0 ~9x10 and 1μs pulses if the over-coupling coefficient β is optimized to be 7-8 for 11424 MHz. Figure 3: SLED gain as function of pulse widths and Q0 values of energy storage cavity. Unified 3dB Coupler / Mode convertor / Polarizer Having all the basic functions of a 3dB coupler, a much more compact and elegant dual-mode circular polarizer was developed to transform the TE01 mode in a rectangular waveguide into two polarized TE11 modes in quadrature in a circular waveguide as shown in Fig. 4. Figure 4: Schematic view of the dual-mode polarizer. The input TE01 mode converts to both TE01 and TE02 modes in a widened rectangular waveguide region, and their magnetic field components will couple to two per___________________________________________ * Work supported by DOE contract DE-AC03-76SF00515. † Email address jywap@slac.stanford.edu Proceedings of IPAC2016, Busan, Korea MOOCA01 07 Accelerator Technology T06 Room Temperature RF ISBN 978-3-95450-147-2 39 C op yr ig ht

Energy dispersion compensation and beam loading in X-band linacs for the JLC/NLC

The shape of an RF pulse is distorted upon propagating through an X-band accelerator structure due to dispersive effects. This distortion together with beam loading introduce energy spread between 192 bunches. In order to minimize this energy spread we modify the input RF pulse shape. The pulse propagation, energy gain, and beam loading are modelled with a mode-matching computer code and a circuit model. A 2D model and a circuit model of a complete 60 cm structure, consisting of 55 cells and input and output couplers is analyzed. This structure operates with a 5/spl pi//6 phase advance per cell. Dispersive effects for this structure are more significant than for previously studied 2/spl pi//3 phase advance accelerating structures. Experimental results are compared with the theoretical model and excellent agreement is obtained for the propagation of an RF pulse through the structure.