Faya Wang

Study of radio frequency breakdown in pressurized L-band waveguide for the International Linear Collider

An L-band (1.3 GHz) radio frequency (rf) waveguide system was assembled at SLAC to test components of a high power distribution scheme proposed for the International Linear Collider (ILC). All parts were made of aluminum and pressurized with dry nitrogen. The rf breakdown rate measured in this resonantly powered system is presented as a function of field level at different gas pressures and rf pulse widths (typically, only breakdown thresholds are reported.). The data are compared to predictions of a simple model which relates the breakdown phenomenon to the rate at which the free electron density builds in the gas.

Modified TTF3 Couplers for LCLS-II

The LCLS-II 4 GeV SCRF electron linac will use 280 TESLA-style cavities with TTF-3 power couplers that are modified for CW operation with input powers up to about 7 kW. The coupler modifications include shortening the antenna to achieve higher Qext and thickening the copper plating on the warm section inner conductor to lower the peak temperature. Another change is the use of a waveguide transition box that is machined out of a solid piece of aluminium, significantly reducing its cost and improving its fit to the warm coupler window section. This paper describes the changes, plating and surface issues, simulations and measurements of the coupler operation (heat loads and temperatures) and RF processing considerations. INTRODUCTION The LCLS-II project has adopted the TTF-3 coupler design (see Fig. 1) to power the 1.3 GHz TESLA-style cavities in its 4 GeV SCRF linac [1]. These couplers were designed for pulsed operation with several hundred kilowatts of input power at a ~ 1% duty factor, so the peak fields for 7 kW operation at LCLS-II will be much smaller. The low duty required the static heat load to be kept small, which was done by using only a thin (1030 m) layer of copper plating on the inner stainless steel surfaces. For LCLS-II CW operation, however, the inner conductor in the ‘warm’ section (between the windows) would overheat. Also the Qext range of the coupler is too low for the small LCLS-II beam currents (< 300 A). Thus modifications were made as described below. MODIFICATIONS Thicker Copper Plating Simulations show that with 7 kW fully reflected input power (worst case), the peak temperature of the warm inner conductor decreases from about 700 K to 400 K if its plating thickness is increased from 30 m to 150 m. This reduces the temperature below the 450 K level at which the couplers are baked so the vacuum levels should be manageable. LCLS-II adopted this thickness, which increases the 45 K total (static + dynamic) cryogenic load by 14 %. As a test, several ILC coupler warm sections were modified by removing their 30 m plating and replating to 150 m. The photo in Fig. 1 shows the achieved plating uniformity in the bellows region, which meets requirements. So far these couplers have worked well and vendors are producing new ones with this plating thickness. Shorter Antenna The TTF-3 couplers can move over a 15 mm range, which changes Qext by about 19 % per mm. LCLS-II will run with Qext = 4.1×10 (about 10 times higher than XFEL), which is above the nominal TTF-3 Qext range. To shift up the range, the flared antenna tip will be shortened by 8.5 mm, keeping the same flare angle and 3 mm radius edges. For the first two cryomodules, ILC cold sections have been modified by milling down the existing antenna tips using a fixture that prevents the couplers from being damaged. Aluminium WG Box and Flex Rings To lower cost and improve performance, the copper waveguide box that attaches to the warm window will be replaced by one machined from a single block of aluminium, without RF matching posts (see Fig. 2). Also the capacitor ring that allows HV holdoff will be replaced by a copper flex ring to provide a better RF seal between the waveguide and the coupler body. ___________________________________________ *Work supported by the Department of Energy, Office of Science, Office of Basic Energy Science, under Contract No. DE-AC0276SF00515 #star@slac.stanford.edu Figure 1: (top) TTF-3 coupler, (bottom) sectioned inner bellows after plating to 150 m thickness. Figure 2: Pair of aluminium waveguide boxes being used for coupler RF processing. THPB077 Proceedings of SRF2015, Whistler, BC, Canada ISBN 978-3-95450-178-6 1306 C op yr ig ht

Surface RF magnetic field enhancement due to closely spaced cylindrical protrusions

The magnetic field enhancement caused by closely spaced, cylindrical, metallic protrusions in an rf cavity is studied theoretically and compared to field solver results for such micro-structures. It is found that the enhancement between the protrusions can be large when the magnetic field is perpendicular to the gap between them, and that the enhancement increases as this gap is reduced. This mechanism could help explain the observation of melting on normal-conducting cavity surfaces subjected to non-enhanced pulsed heating well below that required to melt the surface material. It could also help explain quenching in superconducting cavities when the nominal magnetic fields are below the critical value.

EXPERIMENTAL STUDY OF SURFACE RF MAGNETIC FIELD ENHANCEMENT CAUSED BY CLOSELY SPACED PROTRUSIONS

The RF magnetic field enhancement between two closely spaced protrusions on a metallic surface has been studied theoretically. It is found that a large enhancement occurs when the field is perpendicular to the gap between the protrusions. This mechanism could help explain the melting that has been observed on cavity surfaces subjected to pulsed heating that would nominally be well below the melting temperature of the surface material. To test this possibility, an experiment was carried out in which a pair of copper “pins” was attached to the base plate of an X-band cavity normally used to study pulsed heating. Melting was observed between the pins when the predicted peak temperature was near or exceeded the copper melting temperature. INTRODUCTION We have theoretically investigated the effect of a pair of identical, very small diameter, parallel, cylindrical, metallic, surface protrusions on an applied RF magnetic field in which the flux lines thread between them. The pins concentrate the flux lines, enhancing the magnetic field on the inner surfaces [1]. This is to be distinguished from the magnetic enhancement from general surface features, for example, those associated with sharp edges in normal conducting structures [2] and surface bumps and pits in superconducting cavities [3-6], where the radius of curvature of the surface is the main factor in determining the enhancement, as opposed to the gap opening between above-surface structures. To verify our theoretical findings and explore the surface damage due to high pulsed heating temperatures (close to the melting point), we fabricated 4 pairs of short, cylindrical, copper pins and brazed them to a sample plate of a TE01 cavity used for pulsed-heating studies [7]. The electric field on the surface of the plate is nominally zero so only magnetic field effects are observed. The pins have a height of 2.5 mm and radius of 0.5 mm. They were all located at the radius of maximum pulsed heating on the sample plate. Different gap distances were used for the pairs, two that should produce a magnetic field enhancement of about 3.3 and two with about 2.3. This is illustrated in Fig. 1, where the solid line is the result of a simulation for the pin size, and the squares indicate the enhancement deduced from the measured pin spacing. The surface field around the pin pair is shown in the color maps of Fig. 2. The magnetic field is enhanced almost uniformly from the bottom of the pin to the start of its rounded tip. The electric field, which nominally increases linearly above the surface, is enhanced in small areas at the rounded tips. The ratio of maximum electric field to the maximum magnetic field is about 190 Ω with a magnetic field enhancement of 3.2. Figure 1: Peak surface magnetic field enhancement between the pin pairs with different gap distances.

Performance Limiting Effects in X‐Band Accelerators

Acceleration gradient is a critical parameter for the design of future TeV‐scale linear colliders. The major obstacle to higher gradient in room‐temperature accelerators is rf breakdown, which is still a very mysterious phenomenon that depends on the geometry and material of the accelerator as well as the input power and operating frequency. Pulsed heating has been associated with breakdown for many years however there have been no experiments that clearly separate field and heating effects on the breakdown rate. Recently, such experiments have been performed at SLAC with both standing‐wave and travelling‐wave structures. These experiments have demonstrated that pulsed heating is limiting the gradient. Also, a dual‐moded cavity has been designed to better distinguish the electric field, magnetic field and pulsed heating effects on breakdown.