Fabrication and testing of corrugated waveguides for a collinear wakefield accelerator

Significant progress has been made at Argonne National Laboratory in the development of a compact wakefield accelerator based on a cylindrical corrugated waveguide with a 2-mm ID and fine corrugations on the wall. The fabrication process of 10-cm-long corrugated waveguide structures has been established and a high quality of the final product has been confirmed by precision metrology. Several samples have been tested using the electron beam at Brookhaven National Laboratory’s Accelerator Test Facility. The frequency and group velocity of the fundamental monopole ${\mathrm{TM}}_{01}$ mode of sub-terahertz Čerenkov radiation produced by a 1.5-ps-long electron bunch propagating in the corrugated waveguide, as well as the frequencies of the dipole ${\mathrm{HEM}}_{11}$ and the quadrupole ${\mathrm{HEM}}_{21}$ modes have been measured and found to be in good agreement with the design values calculated using CST Microwave Studio. The energy modulation of a 5-ps-long electron bunch caused by the self-induced wakefield has also been measured and found to be in good agreement with the calculated values.

Figure 1

Dimensions of the cylindrical corrugated waveguide: $\mathrm{a}=1\mathrm{mm}$, p (period) $=340\mu \mathrm{m}$, g (gap) $=180\mu \mathrm{m}$, d (depth) $=264\mu \mathrm{m}$, R (corner radius) $=80\mu \mathrm{m}$.

Figure 2

Figures show various fabrication techniques that were investigated for production of the corrugated waveguide: (a) die stamping, (b) pressing, (c) electroforming, (d) optical lithography, (e) laser ablation, (f) electrical discharge machining.

Figure 3

Brazing assembly of the 0.5-meter-long waveguide.

Figure 4

(a) A fragment of the mandrel viewed under the microscope with a 150 times magnification. (b) Radius profile of the mandrel “tooth” that will form the CW “groove.” (c) Longitudinal profile of the corrugation indicating uniform periodicity. (d) Corrugation detail indicating groove depth. All dimensions are in microns.

Figure 5

Drawing of the end of the plated mandrel showing the diameter of the masked aluminum in millimeters and the location of the reference groove (1).

Figure 6

(a) A fragment of the CWG with one half open. (b) Longitudinal profile of the corrugations measured along the line at the bottom of the CWG. (c) Cross-sectional profile measured in the middle of the tooth. All dimensions are in microns.

Figure 7

The corrugation radii for (a) groove and for (b) tooth. All dimensions are in microns.

Figure 8

Photographic images of (a) a diagonal cut CWG segment at 100 times magnification, (b) a 3D rendering of the diagonal cut segment, and (c) a profile graph of the corrugation. All dimensions are in microns.

Figure 9

The electron beam image at the energy collimator screen with a slit in the middle that cuts off bunch tails, transmits electrons only within a fixed energy range, and produces the electron bunch after the collimator with a predominantly flat-top charge distribution.

Figure 10

A schematic of the experiment: (1) the device under test installed on the translation stage, (2) off-axis parabolic mirror, (3) collimator, (4) spectrometer magnet, (5) diagnostic screen, (6) TPX window, (7) Michelson interferometer, (8) bolometer, (9) a dog-leg lattice, (10) energy collimator.

Figure 11

(a) A snapshot of the radiation field calculated with CST Microwave Studio. The transition radiation is emitted forward. Panels (b–e) show diagrams of an angular distribution in the horizontal plane for the emission of the sub-THz Čerenkov radiation formed inside the CWG: (b) monopole mode, (c) dipole mode with vertical polarization, (d) dipole mode with horizontal polarization, and (e) quadrupole mode with vertical polarization.

Figure 12

A setup of the experiment with corrugated waveguides: (a) side view, (b) front view. See text for details.

Figure 13

(a) A qualitative illustration of the anticipated radiation. (b) The interferogram obtained when the electron bunch propagated through the CWG with the length $L=34\mathrm{mm}$. (c) The spectrum of the radiation emitted when the electron bunch propagated through the CWG with the length $L=96\mathrm{mm}$.

Figure 14

Spectra normalized to the peak obtained during the horizontal translation sweep of the electron bunch trajectory.

Figure 15

Comparing measurements and simulations for the cases of the electron bunch propagating the tube with smooth wall (left column of plots) and the corrugated waveguide with $L=74\mathrm{mm}$ (right column of plots). Plots (a) and (e) show the simulated distribution of electrons in the longitudinal phase space using energy on the horizontal axis and time on the vertical axis. Plots (b) and (f) show the simulated bunch image on the diagnostic screen using energy on the horizontal axis and the vertical coordinate $y$ normalized on the rms vertical size ${\sigma }_y$ on the vertical axis. Plots (c) and (g) show the simulated bunch energy distribution using energy on the horizontal axis and the charge in arbitrary units on the vertical axis. Plots (d) and (h) are similar to plots (b) and (f) and show the measured bunch image on the diagnostic screen using energy on the horizontal axis and the normalized vertical coordinate on the vertical axis.

Figure 16

(a) The difference between two diagnostics screen images, one taken when the electron bunch propagated the CWG with $L=74\mathrm{mm}$ and one taken when the electron bunch propagated the tube with the smooth wall. (b) The measured energy distribution of the electrons shown in the above plot (orange curve) and the calculated analogous energy distribution (black curve).

Figure 17

The diagnostics screen images taken when the electron bunch propagated the CWG2 with $L=93\mathrm{mm}$ and the CWG3 with $L=100\mathrm{mm}$.