Christopher Nantista ISG8 SLAC June 25, 2002

Transcription

1 Christopher Nantista ISG8 SLAC June 25, 2002

2 TM 01 Mode Launcher Development Developed for upcoming traveling-wave single- structure tests as part of R&D to solve rf breakdown problem. Launchers to be flanged and reuseable, with matching s designed for and bonded to each tested. matching 2π/3 TW matching Early designs: High fields. Requires extra circular irises. Match too narrow (~16 MHz).

3 TM 01 Mode Launcher for MC Coupler WC90 beampipe WR90 E s max = ~34 48 MW H s max = ~ MW (on inner edge of waveguide iris) pulsed heating ~3 Group velocity steps from to 0.82c to 0.48c. C. Nantista 01

4 atching the TM 01 Launcher to T53VG3 Adjust a and b of matching. Initial Approach Get in ballpark by using mode matching code on unrounded structure. Model in HFSS and tweek parameters to match. 4 s 2 s a m b m tweeked to perfection check hardly perfect

5 atching the TM 01 Launcher to T53VG3 After initial real, periodicize s. Adjust matching length, rather than diameter to allow use of the Tantawi method. matching Another Approach WC90 L m first from table periodic structure b 60.5 (b 60 +b 61 )/2 a m b 59 a 59 b 60 a 60 b 60.5 a 60 b 60.5 a 60 b 60.5 a 60 Match from diameter waveguide through a matching iris and a variable length 59 into 60 followed by a constant impedance structure with a 60 and b averaged between s 60 and 61.

6 Design Technique (Tantawi Method) Write transmission matrices in terms of scattering matrix T = 1 1 S 22. parameters, and cascade the mirror-symmetric structure S 21 S 11 dets matrix between the matching section and reversed matching section matrices. T tot = T (2) T (1) T (2(1 2)). The matching condition is tot tot = 0 ( or S 11 = 0) T 21 (2) S 11 ( ) = cos φ 11 (1) (2) + φ 22 (1) S 11 (1) (2) 2 ( +φ 22 ) (1) S 11 cos φ Using S (1) parameters from simulation, plot over the range where the value is real and positive. (2) S 11 (2) vs. φ 22 Do this for one, two, and three s (only two required, third is a check). The curves will intersect at a point which gives the unique match to a pure traveling wave in the periodic structure.

7 Unmatched S Matrices one : S = two s: S = three s: S =

8

9 Matched S Matrices one : S 1 = two s: S 2 = three s: S 3 = phasor average: S av 11= (S S S 3 11) R ( S av 11 + S 1 11-S av 11 + S 2 11-S av 11 + S 3 11-S av 11 )/4 coupler reflection: R db

10 Matched Field Plots 1 2 s 3 s 1 : db match. 2 s: db match, 0.9% field variation. 3 s: db match, 1.65% field variation. 12% higher field in coupling, but not on iris.

11 Agilent HFSS Electric Field Amplitude WC90 ~20% lower than

12 Check and Analysis with Ansoft HFSS On-axis electric field Electric field on axis Ez [MV/m] R (Kroll method) Local phase advance per R Phase advance per [deg] z [mm] z [mm] z [mm] Frequency GHz R = 0.05 Phase advance per = 123 C. Nantista, V.Dolgashev, 1 May 02.

13 urface Fields for 48MW (~75 MV/m average gradient) E [V/m] H [A/m] Surface electric fields Surface magnetic fields S [mm] S [mm] Maximum surface fields: 160 MV/m and 0.2 MA/m, Magnetic field at point of max. electric field ~0.1MA/m. Max. temperature rise 10 C for 400 ns pulse. C. Nantista, V.Dolgashev, 1 May 02.

14 Measurement of T53VG3MC S 11 S 11 = GHz Initial measurement of T53VG3MC with the couplers clamped to the body (unbrazed). The input coupler is the mode converter coupler, and the output coupler is of the standard geometry with thick, rounded irises. No resonances due to the circular waveguide and oversized matching s were observed. The time-domain plot shows that the small mismatch is predominantly due to the (pretuning) output coupler, attesting to the relative accuracy of the above method of coupler design. S 11 frequency (GHz) time (ns)

15 New Design The double step of matching into a circular waveguide and then matching this into the structure is unnecessary. One can match from an iris on the top of the rectangular waveguide directly into a matching to the structure. This can lead to a more compact coupler that more effectively uses beamline space. This also eliminates the need for matching irises in the rectangular waveguide, making the coupler more compact transversely as well and eliminating the hottest remaining spot from the mode launcher coupler. no matching elements in waveguide

16 Evolution of Broad Wall Coupler Mode Launcher Coupler circular waveguide matching waveguide iris mm By proper choice of matching b dimension, matching length can be made equal to standard length. New Coupler A New Coupler B matching matching mm mm b m = b of accelerator b m = 0.450

17 Comments This new coupler, along with the mode launcher (MC) design, is characterized by the fact that the interface between the rectangular waveguide and the structure is not in a plane perpendicular to the waveguide axis, but rather in a plane in or parallel to the broad wall of the waveguide, hence the name broad wall structure coupler. A pair of symmetric feeds forms a continuous waveguide in which a standing wave is set up with an anti-node at the beam axis. Here the power is sucked out through the top (broad wall) of the waveguide, through the matching, and into the structure. The region of waveguide/beamline intersection is seen as waveguide, not an accelerator. With the coupler high-field bottleneck removed, this small waste of real estate should be more than compensated by the higher gradient it enables us to sustain in the structure.

18 Acknowledgements Thanks to Sami Tantawi for motivating the design of a TM01 mode launcher for his idea for single- breakdown tests and for the useful matching method he came up with. Thanks also to Valery Dolgashev for help with matching to a pure traveling wave /structure, for lobbying for the adaptation of the mode launcher to a full structure coupler, and for his calculations on the design.