Peak temperature rise specification for accelerating structures: a review and discussion CLIC meeting

We currently specify a cooling water temperature rise in powered accelerating structures of 10 °C. Consequently we will find this ΔT between un-powered and powered conditions (there is another ΔT associated with unloaded and full beam loading but this can be compensated to first order with a T feedback). The resulting size of the thermal expansion after rf is turned on makes achieving the micron-precision transverse alignment and rf phase tolerances needed for beam dynamics quite challenging. On the other hand, delivering the increased water flow to reduce this temperature rise is potentially extremely expensive. Have we made the correct choice with 10 °C? There is also the effect of air temperature rise but we will deal with this another day. Introduction

Organization of this meeting We will start with four ten minute presentations to provide the essential background information 1.Overall Introduction and rf – Walter Wuensch 2.Beam dynamics tolerance requirements – Daniel Schulte 3.Accelerating structure thermo-mechanical behavior – Germana Riddone 4.Cooling system capacity – Mauro Nonis Then we will proceed with an open discussion, perhaps resulting in defining some actions and perhaps even a conclusion or two…

rf Accelerating structure position influences the beam via rf in two main ways (there are two aspects to each issue – the size of the effect and our ability to compensate through measurement): 1. Longitudinal movements result in a beam to fundamental mode phase change. Changes in the length of the connecting waveguide does the same thing. To guide our discussion - the wavelength of 12 GHz is 25 mm in free space and in 30 mm in WR-90 waveguide. Hence 1µm is degrees in free space and degrees in WR-90. We will pre-compensate dimensions for the nominal operating temperature. 2. Transverse movements affect the transverse wakefields and tilts give fundamental driven kicks. We do have wakefield monitors for average offsets but not enough to monitor tilts.

General layout in a DB Sector (~800m) DB deccelerator DB inDB out ………………………. MB linac RF MB inMB out Two types of PETS+2AS units will be installed: 1.Reference (black) (2 units at the beginning and 2 units at the end of a DB sector) It will have more signals and with higher resolution The signals will be time resolved: dt ~ 0.5 ns (pulse shape) 2.Regular (blue) (all the rest) It will have 1 or 2 signals Integral over the pulse (1 or 2 numbers per pulse) A. Grudiev

PETS+2AS unit: Reference PETS DB inDB out On Off Hybrid AS1AS2 Electronic box 12 GHz Amplitude and phase during the pulse There are two things to distinguish: 1.Beam control related issues (must be defined together with DB and MB beam control system) 1.RF power production (also single- and multi-bunch longitudinal DB (in)stability using reference units at the beginning and at the end of the deccelerator) 2.Energy measurement and Beam loading transient compensation. (requirement for bunch-to- bunch energy spread: dE/E~1e-4, we will go to 1e-3 level based on rf measurement and for the final adjustments we rely on the MB measurements at the end of the MB linac?) 2.RF diagnostic related (similar to Regular units but has more signals, higher amplitude and time resolution to look into details) 1.RF breakdown in PETS and AS 2.PETS on/off failure 3.Provide references for the regular PETS+2AS units Time sampling rate 0.5 ns or better MB synchronous rf phase requires independently measured MB phase Dynamic range: 40 dB Phase resolution: 0.1 deg Amplitude resolution: at least 1e-3 or maybe 1e-4 if it is not too expensive Again requirements are dominated by the bunch-to-bunch energy measurement (beam control). 12 GHz Ch1 Electronic box Ch2 Directional coupler MB inMB out PU MB phase Ch3 A. Grudiev

PETS+2AS unit: Regular 1 PETS DB inDB out On Off Hybrid AS1AS2 Electronic box 12 GHz DC Channel 1 addresses two things: 1.RF breakdown 2.PETS on/off failure Diagnostic is based on the comparison of the signals to the reference provided by the reference PETS+2AS units. If readout is different from the reference => something is wrong: If it is stable pulse to pulse => on/off failure If it is not => breakdown in PETS or one of the AS. Dynamic range : 20 dB Resolution: 1e-3 linear Accuracy pulse-to-pulse: 1e-3 Absolute accuracy: 1e-2 is OK, 1e-3 is under question Ch1 MB inMB out A. Grudiev

What alignment will be achieved From Igor and Alessandro: Choke mode flange pf phase change for different displacement: Transverse dx,dy dphi < ±0.01 deg Longitudinal dphi / dz = 2.4 deg/mm; dz dphi < ±0.12 deg From CLIC module WG : After CLIC module alignment procedure (without beam) is finished the accuracy of the PETS-to-2ASs alignment will be better than dx < ± 0.05mm dy < ± 0.05mm dz < ± 0.05mm The longitudinal displacement is the most critical for the rf phase change. It will be below ±0.12 deg. What can be measured is ~ deg. In conclusion, the measurement of the rf phase is not more accurate than mechanical alignment which will be achieved and this already satisfies BD requirements of ±0.1 deg correlated and ± 1 deg uncorrelated (Daniel). No rf phase measurement in PETS+2AS Regular units A. Grudiev

CLIC-ACE, 26 May 2009 Alexej Grudiev, Accelerating structures. Wake field of proposed structures

CLIC-ACE, 26 May 2009 Alexej Grudiev, Accelerating structures. Transverse impedance