Fabrication and cold test of dielectric assist accelerating structure

We present the detailed description of a successful design and cold testing of the dielectric assist accelerating (DAA) structure. The DAA structure consists of ultralow-loss dielectric cylinders and disks with irises which are periodically arranged in a metallic enclosure. The advantage of the DAA structure is that it has an extremely high quality factor and a very high shunt impedance at room temperature since the electromagnetic field distribution of accelerating mode can be controlled by dielectric parts so that the wall loss on the metallic surface is greatly reduced. A prototype of the five-cell DAA structure was designed and built at C-band (5.712 GHz), and cold tested. Three types of dielectric cell structure, “regular,” “end,” and “hybrid” dielectric cells, are fabricated by sintering high-purity magnesia. The prototype was assembled by stacking these cells in the hollow copper cylinder, whose two ends are closed by copper plates. The resonant frequency of the prototype was tuned to the desired frequency by machining only end copper plates. The unloaded quality factor of the accelerating mode was measured at 119,314 and the shunt impedance per unit length of the prototype was estimated from the experimental results of the bead pull measurement as $Z_{\mathrm{sh}}=617\mathrm{M}\mathrm{\Omega }/\mathrm{m}$, which were within 2 percent of the design values. The field distribution of accelerating mode was also measured by the bead pull method, and its results agreed well with simulation results.

Figure 1

(a) Conceptual illustration of a multicell DAA structure. (b) Longitudinal cross section of the DAA structure. Blue and red areas represent the regular cell and the end cell, respectively.

Figure 2

Conceptual illustrations of the end cell containing a frequency tuning structure. The inner and outer radii of the concentric circular groove are set to $r_1$ and $a_1$, respectively. The groove depth of it is defined in $d$.

Figure 3

Shifts of the resonant frequency $\mathrm{\Delta }{f}_0$ and shunt impedance $\mathrm{\Delta }{Z}_{\mathrm{sh}}$ as functions of the groove depth $d$. The shifts of resonant frequency are defined as $\mathrm{\Delta }{f}_0=f_0(d)-f_0(0)$. The shifts of shunt impedance are defined as $\mathrm{\Delta }{Z}_{\mathrm{sh}}/Z_{\mathrm{sh}}=\{Z_{\mathrm{sh}}(d)-Z_{\mathrm{sh}}(0)\}/Z_{\mathrm{sh}}(0)$.

Figure 4

Normalized longitudinal electric field strengths on the beam axis in the prototype models with $d=0.0\mathrm{mm}$ (blue line) and $d=4.0\mathrm{mm}$ (red line).

Figure 5

The fabricated dielectric cell structures: (a) regular dielectric cell, (b) end dielectric cell, (c) hybrid dielectric cell.

Figure 6

(a) Frequency dependence of the measured $|S_{11}|$. The horizontal axis indicates the frequency (GHz) and the vertical axis indicates the $S_{11}$ in decibel units (dB). (b) The measured $S_{11}$ in the format of the Smith chart. The coupling coefficient $\beta$ was approximately 0.19 during these measurements.

Figure 7

Dependence of (a) the resonant frequency $f_0$ and (b) the unloaded $Q$ factor $Q_0$ on the coupling coefficient $\beta$.

Figure 8

Conceptual illustration of the rf input coupler including a TE-TM mode converter.

Figure 9

Dependence of the resonant frequency $f_0$ (red dot line) and the external $Q$ factor $Q_{\mathrm{ext}}$ (blue dot line) of the DAA structure on the groove depth $d$. The radius of coupling slot $r_c$ is 9.8 mm.

Figure 10

Schematic half-cross section of the DAA structure assembly including the input coupler and the TE-TM mode converter.

Figure 11

Photograph of the fabricated DAA structure assembly including the input coupler and the TE-TM mode converter.

Figure 12

(a) Frequency dependence of the measured $|S_{11}|$. The horizontal axis indicates the frequency (GHz) and the vertical axis indicates the $S_{11}$ in decibel units (dB). (b) The measured $S_{11}$ in the format of the Smith chart.

Figure 13

(a) Comparison between the measured (blue dots) and the simulated (red line) values of $\sqrt{-\mathrm{\Delta }f}/f_0$. (b) Comparison between the simulated longitudinal field distribution of the accelerating mode along the beam axis for the DAA structure and the experimental result of the bead pull measurement. The experimental results (blue dots) show normalized $\sqrt{-\mathrm{\Delta }f}/f_0$, which are scaled by using the maximum value of $\sqrt{-\mathrm{\Delta }f}/f_0$. The simulated results (red line) show normalized longitudinal electric field strength of the accelerating mode along the beam axis, which are scaled by using a maximum value of $E_z(z)$.

Figure 14

Magnitude of (a) the longitudinal electric field distribution and (b) the rotating magnetic field distribution in the optimized DAA structure. The longitudinal cross section of the DAA structure is shown. This field intensity is shown in arbitrary unit.

Figure 15

Dependence of $\mathrm{\Delta }{f}_0$ and $\mathrm{\Delta }{Z}_{\mathrm{sh}}/Z_{\mathrm{sh}}$ on (a) the deviation of the outer radius $b_1$ at the dielectric cylinder, and on (b) the deviation of the thickness $D$ at the dielectric disk.