Introduction

The Chinese initiative Accelerator Driven Subcritical System (CiADS), currently under development at the Institute of Modern Physics, Chinese Academy of Sciences (IMP, CAS), is a multi-MW proton source for energy generation and long-half-life nuclear waste transmutation [1-3]. This national major science and technology infrastructure adopts the continuous wave (CW) and superconducting technical route, capable of accelerating a 5 mA proton beam to 500 MeV in energy. It comprises an intense-beam driver linac, a high-power spallation target station, and a subcritical-mode nuclear reactor. The driver superconducting radio frequency (SRF) linac accelerates the high-intensity proton beam to high energy, which then bombards a spallation target, producing neutrons with a wide spectrum and high flux capable of driving a chain reaction in a subcritical reactor. This process releases nuclear energy and transmutes long-lived nuclear waste. The linac accelerator’s front end features a four-vane Radio Frequency Quadrupole (RFQ) cavity operating at a frequency of 162.5 MHz, with a proton energy at the cavity outlet of 2.1 MeV. Injecting 96 kW of driving radio frequency (RF) power into the RFQ cavity, it spans a length of 4883.8 mm.

In the RFQ cavity design, four couplers share the required driving power and are situated at two different longitudinal positions, with each of the two couplers placed at the same longitudinal position. Each coupler necessitates a nominal RF power of 26 kW for CW operation and a maximum transmitted power of 60 kW. The primary function of the power input coupler is to link the RFQ cavity to the power source, which supplies microwave power, while also utilizing a ceramic window to isolate the atmosphere from the ultra-high vacuum environment within the RFQ cavity. Consequently, the ceramic window stands as one of the critical components of the accelerator, determining the long-term stable operation of the RFQ cavity [4, 5]. Several factors contribute to the cracking of the ceramic window. Firstly, the secondary electron multipactor effect around the window can cause it to overheat, leading to damage and ultimately cracking [6-9]. Secondly, improper high-frequency structure of the power input coupler can result in excessive electric field around the window, thus causing RF electrical breakdown. Thirdly, inadequate design of the mechanical structure can lead to overheating of the window, resulting in excessive thermal stress. Moreover, unreasonable mechanical structure design can also cause excessive temperature of the coupling loop, reducing coupling strength and increasing power reflection [10]. Additionally, defects in the processing of the power coupler and the smelting of materials produce dust and metal burrs on the inner surface of the power coupler, leading to cavity contamination. Hence, offline RF conditioning and power tests are necessary before the coupler is connected to the RFQ cavity to verify the structural design’s reasonability and reduce cavity contamination from the coupler.

At least two couplers are required for offline high-power tests of the power input couplers, typically connected by either a resonant cavity or coaxial connector [11-18]. The coaxial connector requires inner conductors for connection, making it suitable for electrical coupling, such as with high-power couplers for superconducting cavities. In contrast, the resonant cavity does not necessitate inner conductors for connection and can be used for both electrical and magnetic coupling. Magnetic coupling is employed to supply power to the RFQ cavity, utilizing a coupling loop structure where inner conductors cannot be connected. Consequently, the resonant cavity is commonly utilized for offline RF conditioning and high-power tests. For instance, the RFQ cavity of the European Spallation Source (ESS) operates at a frequency of 352.1 MHz, with two high-power input couplers utilized to feed 1.6 MW of peak power [19]. During offline tests, a resonant cavity was employed, with each coupler subjected to a target power of 1MV peak power and a duty cycle of 5%. The S11 and S21 parameters of the test cavity were -27 dB and -0.2 dB, respectively, with a power loss of 4%. Similarly, the RFQ cavity of the International Fusion Materials Irradiation Facility (IFMIF) operates at a frequency of 175 MHz, with eight high-power input couplers adopted to feed 1.28 MW continuous wave power. To conduct offline tests of these high-power input couplers, the IFMIF project also employed a resonant cavity [20]. The target power for each coupler during tests was 200kW continuous wave power. The S11 and S21 parameters of the test cavity were -29 dB and -0.31 dB, respectively, with a power loss of 7%.

Based on the preceding discussions, resonant cavities are typically preferred as connections for high-power couplers during offline RF conditioning and power tests, particularly for those high-power input couplers that supply power to the cavity utilizing magnetic coupling. For cavities operating at high frequencies, pill box structures [15-18] or rectangular resonators [11-14] are commonly employed for offline tests of high-power couplers. However, given that the frequency of the CiADS RFQ cavity is 162.5 MHz, using the aforementioned types of cavities would result in larger cavity dimensions and lower transmission efficiency. Hence, this paper proposes a low-loss resonant and compact structure cavity designed specifically for offline tests of high-power input couplers for the CiADS RFQ. The cavity was designed using numerical calculation methods, and subsequent high-power tests were conducted.

Design of high power input coupler structure

This study focuses on the high-power testing of power couplers; therefore, this section briefly describes the structural design of the high-power coupler. The primary function of the high-power coupler is to transfer RF power from the power source to the RFQ cavity [4, 5]. Additionally, it acts as a barrier between the air-side transmission line and the RFQ cavity, maintaining a vacuum. The high-power input coupler should sustain specified power under different operating conditions (continuous wave or pulsed) without degrading cavity performance or beam quality. Hence, during the design phase, efforts must be directed towards the following areas. Firstly, achieving excellent radio frequency transmission performance is essential to reduce RF power loss. Secondly, optimizing the electromagnetic field distribution is desired to minimize RF electric breakdown around the ceramic window. Thirdly, to ensure the long-term stable operation of the power coupler, multipacting should be avoided at nominal power. The RF window maintains a high vacuum and ultra-clean environment inside the cavity, so it’s essential to ensure the window’s reliability from various perspectives throughout the development process. Therefore, comprehensive safety analysis for the window is required during the design phase, and a comprehensive interlock system should be implemented to monitor all activities during offline conditioning and routine operations. These safety analyses typically include radio frequency performance, thermal-mechanical stress at nominal radio frequency power and pressure, and the design of the cooling channel. Monitors often include vacuum levels, arc discharging, and overheating. Generally, when the processing of a new coupler structure is completed, offline conditioning is necessary to validate the design rationality of the power coupler.

The general layout of the high power coupler for the CiADS RFQ cavity is depicted in Fig. 1. It comprises a coupling loop module, coaxial transmission line module (RF window and coaxial line), and impedance match module. Each module is detachable via a flange connection. The inner conductor of the coaxial transmission line module is crafted from oxygen-free copper and cooled by deionized water. The impedance match module is constructed from aluminum.

Fig. 1Full-screen View

(Color online) High power input coupler structure schematic

Due to the high power input coupler’s design power of 60 kW, it is essential to cool both the inner and outer conductors of the ceramic window. The IFMIF RFQ cavity utilized a λ/4-long cooling water port for heat dissipation by the high power input coupler, with the inlet/outlet of cooling water positioned at the end cooling port. However, this design led to the power coupler occupying excessive space [21]. To ensure a compact structure for the CiADS RFQ cavity’s coupler, a T-box structure was adopted for the inner conductor cooling channel of the ceramic window (see Fig. 1). This T-box structure also functions as a matching section between the power source and the coaxial transmission line module, enabling the high power input coupler to maintain a low reflection coefficient. One end of the T-box connects to a standard 6-1/8” coaxial transmission line for input power, while the other end connects to the coaxial transmission line module of the coupler for output power. The third end serves as a short-circuit section, allowing adjustment of the power coupler’s reflection coefficient and providing a cooling channel for the inner conductor of the ceramic window.

Based on the frequency of the RFQ cavity, a coaxial structure with a 50 Ω impedance was chosen for the coaxial transmission line module of the high power input coupler. The outer and inner conductor diameters connected to the cavity port were 81 mm and 24 mm, respectively. This module includes the coaxial line, ceramic window, and Teflon. The ceramic window divides the coaxial transmission line module into two parts. The first part connects to the coupling loop module, with its inner conductor linked to the ceramic window’s inner conductor via a high-frequency inner bullet, and its outer conductor connected to the ceramic window using stainless steel flanges. The second part connects to a T-box at one end and to the ceramic window at the other, supported by Teflon polyethylene to sustain the outer conductor. To facilitate the replacement of only the RF window components in case of a ceramic window breakage, the ceramic window component connects to the outer conductor of the coaxial transmission line using knife-edge flanges.

The design of the RF window was based on the CiADS HWR010 162.5 MHz input power coupler [22], as depicted in Fig. 1. A coaxial disk-type ceramic, composed of 97.6% Al₂O₃, served as the vacuum barrier. However, the introduction of ceramic windows led to impedance changes in the coaxial line, resulting in excessive power losses for the input power coupler. Thus, during the high-frequency design phase, optimization of the ceramic window components was necessary to minimize thermal stress and ensure excellent impedance matching. Variables optimized included the ceramic window’s thickness, inner and outer diameters, and taper length. The ceramic’s final thickness was 8 mm. Impedance matching was achieved by increasing the outer diameter and reducing the inner diameter of the ceramic window, resulting in an outer diameter of 128 mm. With the outer diameter of the cooling channel of the inner conductor less than 10 mm, a ceramic window with an inner diameter of 40 mm was chosen. To prevent multipactoring discharge on the surface, a 10 nm thickness TiN coating was applied to the ceramic window surface using sputtering technology [23-26]. A 10 nm thickness was selected for several reasons. Firstly, RF loss due to the TiN layer can be neglected at 162.5 MHz. Secondly, this thickness could be accurately measured for this dimension, ensuring uniformity of the ceramic window to within a few nanometers.

The performance targets of the CiADS RFQ coupler and those of other laboratories are summarized in Table 1 [21, 27, 28]. The structural schematic of the coupling loop is depicted in Fig. 1. The coupling loop is joined to the inner and outer conductors of the coaxial transmission line module via vacuum brazing, with an O-ring made of fluorine rubber used to seal the vacuum. The power loss of the coupling loop itself is inversely proportional to the diameter of the coupling loop conductor. Thus, the smaller the cross-sectional area of the coupling loop conductor, the higher the power loss of the coupling loop, resulting in greater thermal stress. Excessive thermal stress can alter the area of the coupling loop (coupling strength), thereby affecting its proper operation. Conversely, the larger the cross section of the coupling loop conductor, the greater the volume of the coupling loop in the RFQ cavity, which can significantly alter the distribution of high-frequency electromagnetic fields within the cavity. Based on the analysis above, the parameters of the coupling loop are as follows: the length of the coupling loop inserted into the cavity was 32.5 mm, the cross-section of the coupling loop was 180 mm², the inner radius of the coupling loop was 14.5 mm, and the area of the coupling loop was 745 mm². Coupling strength, magnetic field strength around the loop, and surface loss density on the loop were evaluated using CST code. Considering that there may be three couplers when the RFQ cavity is in operation, the coupling strength was designed accordingly during the design stage. When the coupler is connected to the cavity, the mounting angle of the coupling loop can be adjusted to meet the demand for coupling strength of either four or three couplers. The coupling strength of the cavity is [29]. $\beta =\frac{Q_0}{Q_{\text{ext}}}=\frac{P_{\text{ext}}}{P_{\text{cav}}}=\frac{\left(P_{\text{cav}}+P_{\text{beam}}\right)}{P_{\text{cav}}}$ (1) The equivalent coupling strength of each coupling port is expressed as follows: ${\beta }_i^{*}=\frac{{\beta }_i}{1+{\displaystyle {\sum }_{i\ne j}^4{\beta }_j}}.$ (2) The total coupling strength of the coupler is 1.11. When three couplers are used, the equivalent coupling strength of each coupler is 0.216, with an S11 of -3.75 dB. If four couplers are used, the equivalent coupling strength of each coupler is 0.1514, with an S11 of -2.65 dB.

In this coupler, there were three cooling channels: two for cooling the inner and outer conductors of the ceramic window, and one for cooling the coupling loop. The water flow rate in all cooling channels was designed to be 1.6 m/s, with a pressure of 8 atm during the pressurization test.

Electromagnetic and structural design of the cavity

3.1

Design considerations for cavity

Due to the coupler’s design power of 60 kW continuous wave, the following design criteria had been established to reduce cavity manufacturing costs and power losses:

(a) Compact structure and low costs. Large cavity dimensions impede transportation, low-temperature baking, and conditioning system assembly, while also increasing manufacturing costs. Therefore, a rigid coaxial transmission line with a diameter of 9-3/16" was used to process the inner and outer conductors of the cavity.

(b) High transmission efficiency and low power losses. High power losses in the cavity would elevate cooling requirements and gas exhaust, consequently prolonging test time and increasing costs.

(c) Tunable. Theoretical calculations show that the power loss rises from 4.6% to 10% (Fig. 5) as the cavity frequency deviates from the center frequency by 110 kHz. Cavity frequency shifts may occur due to cavity processing, gas pumping vibrations, and temperature increases. Thus, the cavity needs sufficient tuning capability to maintain low power losses during operation.

(d) Minimize K_p value and maximize coupling strength. If the RF electric field is sufficiently high, the room temperature copper cavity may experience RF electrical breakdown or sparking. The K_p value defines the maximum electric field achievable within the cavity without experiencing RF electrical breakdown or sparking at a given frequency. Hence, a smaller K_p value can decrease the likelihood of RF electrical breakdown or sparking.

The coupling strengths of the upstream and downstream couplers are β₁ and β₂, respectively. The power loss of the cavity (P_cav) [30]is as follows: $P_{\text{cav}}=\frac{4P_{\text{in}}{\beta }_1}{{\left(1+{\beta }_1+{\beta }_2\right)}^2}.$ (3) Eq. (3), P_cav represents cavity loss, andP_in represents the input power of the upstream coupler.

The dimensions of the coupling loops of the two couplers were determined before the tests. Thus, β₁ is equal to β₂. Then the expression of P_cav is transformed into the following equation: $P_{\text{cav}}=\frac{4P_{\text{in}}\beta }{{\left(1+2\beta \right)}^2}.$ (4) The power extracted by the downstream coupler can be represented as $P_{\text{t}}={\beta }_2{P}_{\text{cav}}$. Substituting the value of Eq. (4) into this equation yields $P_{\text{t}}=\frac{4P_{\text{in}}{\beta }^2}{{\left(1+2\beta \right)}^2}.$ (5) Based on Eq. (4) and Eq. (5), cavity loss decreases and transmission power increases with higher coupling strength. A significant coupling strength suggests that the majority of the cavity’s stored energy is efficiently transmitted to the external load via the power coupler, resulting in minimal cavity power loss. Thus, for cavities tasked with power transmission, a high coupling degree is desirable.

The high-power input coupler, supplying power to the superconducting cavity at a frequency of 166.6 MHz, utilized a QWR-type cavity for offline high-power conditioning [5]. Therefore, the quarter-wave resonator (QWR) cavity was chosen for the offline testing of the high-power coupler for the CiADS RFQ cavity, as depicted in Fig. 2. One side of the cavity, with a stronger magnetic field, was equipped with two ports for connecting the coupler, while the other side was also designed with two ports for connecting the tuner.

Fig. 2Full-screen View

(Color online) Cavity structure schematic

3.2

Electromagnetic design

The electromagnetic design and optimization of the offline test cavity were conducted using the three-dimensional electromagnetic simulation software CST. Since a rigid coaxial transmission line with a diameter of 9-3/16" was employed in manufacturing the cavity, the diameters of the inner and outer conductors were predetermined. The cavity’s outer and inner conductor diameters were 240 mm and 100 mm, respectively. Consequently, the cavity’s frequency could be adjusted to 162.5 MHz by altering the length of the inner conductor, as illustrated in Fig. 3a. From the figure, it is evident that increasing the length of the inner conductor leads to a decrease in frequency. This can be attributed to the increase in capacitance with length, resulting in a lower frequency. Beyond a length of 400 mm, the frequency decrease rate accelerates.

Fig. 3Full-screen View

a Variation of cavity frequency with the inner conductor; b Variation of cavity frequency with tuner insertion depth

Figure 3b illustrates the cavity frequency as a function of the insertion depth of one of the tuners while the insertion depth of the other tuner remains constant. The tuning range of the tuner is 3.2 MHz. As depicted in the figure, increasing the insertion depth results in a lower frequency. However, excessive insertion depth reduces the gap between the inner conductor and the tuner, raising the maximum surface electric field of the cavity. This phenomenon increases the risk of rf electrical breakdown. To mitigate this, two tuners were designed. Upon cavity installation, the insertion depths of both tuners were adjusted to align the cavity frequency with the operating frequency. During operation, one tuner remains fixed while the other tuner’s insertion depth is adjusted to maintain the cavity frequency around 162.5 MHz, thereby ensuring cavity losses remain within acceptable limits.

The position of the coupling loop during installation influences the coupling strength, which subsequently impacts S11 and cavity losses [31]. To assess how cavity and coupler processing errors, as well as mounting errors, affect cavity power losses, the effect of the coupling loop’s rotation angle on the S parameter was investigated. The results are depicted in Fig. 4. When the installation angle is 180 °, as shown in Fig. 4a, the magnetic flux within the area enclosed by the coupling loop is maximized, resulting in an S11 value of -28.77 dB. Using this angle as the reference point, S11 is maximized at a rotation of ±90 °, minimizing the magnetic flux within the enclosed area of the coupling loop. Within the range of 180±5 °, S11 reaches its maximum, with a value of -2 dB (Fig. 4b). Therefore, adjusting the mounting angle of the coupling loop during assembly can minimize cavity power losses.

Fig. 4Full-screen View

Coupling ring rotation angle a 180; b 180±5 °

The final optimized S-parameters and electric field distribution are illustrated in Fig. 5. The simulated value of S21 for the cavity is -0.19 dB, corresponding to power losses of 4.3%. In comparison, the power losses for the offline test cavities of the IFMIF and ESS RFQ cavity couplers are 7% [20] and 4% [19], respectively. According to the conversion relationship between power losses and frequency [30], they equate to 7.26% and 5.9%, respectively. Therefore, the power losses of the offline test cavity for the CiADS RFQ cavity couplers are superior to the results reported in the literature. The simulated value of S11 for the test cavity is -28.77 dB, which closely aligns with values documented in the literature [19, 20] and can be considered acceptable. At an input power of 60 kW, the maximum electric field is 4.9 MV/m, the K_p value of the cavity is 0.42, and the coupling strength is 24.

Fig. 5Full-screen View

(Color online) a S parameter; b electric field distribution; c magnetic field distribution

3.3

Mechanical stability and thermal analysis

Due to the cavity’s wall thickness being 5 mm, it’s necessary to study the impact of temperature variation on its frequency. Thus, thermal and mechanical analysis of the cavity was conducted using the CST Multiphysics module. In this study, the ambient temperature was set at 22 ℃, and both the coupling loop and tuner, along with the inner conductor, were water-cooled at a flow rate of 1.6 m/s. A pressure of 0.1 MPa was applied to the cavity’s outer wall. The results, depicted in Fig. 6, reveal that the cavity’s maximum temperature, reaching 53.5 °C, occurs at the outer conductor due to its lack of cooling. The maximum stress within the cavity is 62 MPa, which is lower than the yield strength of oxygen-free copper [32]. The maximum displacement is 0.17 mm, resulting in a negligible change in frequency. These thermal and mechanical analysis findings confirm that the 5 mm wall thickness of the cavity meets the requirements for offline testing.

Fig. 6Full-screen View

(Color online) a Temperature distribution; b stress distribution; c displacement distribution

High power offline test

4.2

Preparation conditions

Upon completion of cavity fabrication, the cavity underwent visual inspection, ultrasonic cleaning, drying with high-purity nitrogen gas, a 24-hour placement in a Class 100 cleanroom for drying, and vacuum leak detection. If the vacuum leak rate was better than 5×10^-10 mbarL/s, the cavity was transported to the offline conditioning cleanroom and connected to two couplers as depicted in Fig. 7a.

Fig. 7Full-screen View

(Color online) a S parameter measurement diagram; b S parameter measurement results; c offline conditioning platform

The transmission characteristics of the entire offline conditioning system were tested using the R&S ZNB network analyzer from Rohde & Schwarz. The results are shown in Fig. 7b. To minimize power losses of the offline conditioning cavity during the test, the mounting angle of the rotating coupling loop was adjusted. The measured value of S21 for the entire conditioning system is -0.3 dB, corresponding to a loss of 6.3%, compared to losses of 7% [20] and 4% [19] reported in the literature, respectively. Therefore, the power loss of the offline conditioning system in this paper is within an acceptable range. The simulated value of S21 for the cavity is -0.19 dB, corresponding to a loss of 4.3%, as shown in Fig. 7b. The increase in measured power losses is attributed to the fact that the simulation was conducted under ideal conditions, without considering transmission line or coupler power losses. When the power loss of the conditioning system is 10%, the frequency variation of the offline conditioning cavity is 132 kHz. During the test, we can adjust the frequency variation using the tuner to minimize power losses of the conditioning system.

After completing the transmission performance test of the offline conditioning system, it was vacuumed and baked at 120 ℃ (held for 48 hours) to remove any residual moisture from the cleaning process. Subsequently, the high power input couplers were installed on the offline conditioning platform, awaiting high power conditioning, as depicted in Fig. 7c.

4.3

Offline conditioning

The high-power coupler offline conditioning system is depicted in Fig. 8. It comprises the coupler conditioning platform, power source system, low-level radio frequency, and peripheral auxiliary equipment. The conditioning platform features two couplers and a QWR cavity. The low-level radio frequency facilitates vacuum data acquisition, ARC detection, implementation of conditioning mode, and safety interlock protection. The power source system encompasses a power source, signal source, directional coupler, and coaxial transmission lines. The power source is a 1 kW solid-state unit. Peripheral auxiliary components include a vacuum pump system and water chiller. Power output from the source enters the directional coupler and then passes through the upstream coupler, conditioning cavity, downstream coupler, and another directional coupler. This cycle continues multiple times. The entire conditioning system forms a resonant system, with the system frequency adjusted to 162.5 MHz by altering the length of the coaxial transmission line.

Fig. 8Full-screen View

High power coupler offline conditioning system

Due to the coupler being more likely to outgas under low-power conditions, the travelling wave conditioning mode was employed at power levels below 1 kW. In this mode, the downstream coupler was connected to a matched load. When the solid-state power source exceeded its power limit, the standing wave resonance conditioning system was used for the test. After processing the power coupler and conditioning cavity, there are usually burrs, dust, and small metal particles on their inner surfaces. Therefore, at the beginning of conditioning, outgassing is typically severe. Generally, a pulse conditioning mode was used, with the duty cycle gradually increasing from 1% to 100%. Real-time monitoring of the coupler and cavity vacuum, as well as coupler ARC discharge, was carried out during conditioning. The vacuum protection threshold was set to 5×10^-4 mbarL/s. After 60 hours of conditioning, the standing wave power reached 26.5 kW in continuous wave mode. The maximum temperature for the upstream and downstream coupler on the T-box inner conductor was 55.1 ℃ and 47.5 ℃, respectively. The maximum temperature of the outer conductor of the cavity was 42.9 ℃, and the inner conductor temperature was 22.1 ℃ (Fig. 9).

Fig. 9Full-screen View

(Color online) Conditioning result: a power; b temperature distribution

To adequately condition the power coupler, the standing wave point of the power coupler was adjusted by modifying the phases of the entire standing wave resonance conditioning system. During the conditioning process, the phase difference between the two standing wave points of the conditioning system is 15 °. Upon completion of the conditioning, the vacuum and power change curves during the loading process from zero power to the target power are depicted in Fig. 10. The phase of the standing wave resonance conditioning system was set at 180 °, with a duration of 1 h. From this figure, it is evident that the maximum standing wave power is 25 kW in continuous wave mode, which exceeds the target power.

Fig. 10Full-screen View

High power conditioning results at a phase angle of 180 °

Conclusion

To verify the rationality of the design of the power coupler for the CiADS RFQ cavity and reduce cavity contamination, this study designed a low-loss offline conditioning cavity and conducted a high-power test. The conclusions are as follows:

The conditioning cavity utilized a QWR type cavity, featuring two coupling ports and two tuners. Minimizing cavity loss was achieved through adjusting the installation angle of the coupling loop and the insertion depth of the tuner. Electromagnetic structural and multiphysics simulations of the cavity were conducted using 3D microwave simulation software CST. The transmission efficiency was studied by varying tuner insertion depth, inner conductor length, and coupling strength. The optimal theoretical power loss was 4.3%, outperforming reported literature values. At a frequency variation of 110 kHz, the theoretical power loss increased to 10%, necessitating continuous tuner adjustment during conditioning. With a coupling strength of 24, the cavity exhibited minimal power loss, indicating efficient power transfer through the coupling loop. Multiphysics simulations demonstrated that cavity temperature variations did not affect frequency. Upon completion of the offline high power conditioning platform, the transmission performance of the entire system was measured, yielding a power loss of 6.3%, slightly higher than the theoretical calculation. This discrepancy may be attributed to processing, material properties, and transmission line effects, which were not accounted for in the theoretical model. Conditioning utilized efficient automatic range scanning and standing wave resonant methods. Travelling wave mode was employed for conditioning under 1 kW to address vacuum outgassing. Adjusting the phase difference between two standing wave points by 15 ° ensured thorough conditioning of the power coupler. The maximum continuous wave power exceeded 20 kW, surpassing the target, thereby ensuring stable operation of the CiADS RFQ accelerator cavity.