3.1 Device and method of the RF measurement

Agilent E5061B VNA (Vector Network Analyzer) was used as the main measure instrument. The signal cables connected to the coupler and the pickup were attached to the port 1 and port 2 of the VNA, respectively. In the measurement of the electric field of the IH-DTL1, the perturbation method [12, 13] was adopted. According to the perturbation theory, the phase shift, ΔΦ, indicates the electric field, $\overline{E}$, by:

where P=1 mW is the average output power of the VNA, and ω₀ is the output circular frequency. Then if the size of the bead Δ V is small enough, we can approximately consider $E=\overline{E}$, where E is the electric field at the center of the bead.

Several PTFE (polytetrafluoroethylene) cylinder beads were considered, and towed by a high strength Nomex line with the diameter of 0.3 mm. This measurement system is driven by a step motor. The VNA records one phase datum in 0.02 s and the tempo of the bead is 21.42±0.02 mm/s. So we can transform the achieved data of equal time intervals into equal distance intervals. The time of VNA recording and the bead moving are both 32 s, and the two devices should be triggered at same time. Throughout the process of measurement, the temperature and humidity of the environment fluctuate in the range of 10.0∼11.5 ℃ and 50∼60% respectively. During the field distribution measurement, the electromagnetic interference and mechanical vibration should be avoided. By perturbation method, usually the phase shift below 5 can ensure the precision of the measurement. Finally, we choose a 5mm bead in length with Φ4.00 mm and the maximum phase shift is 3.5°.

3.2 RF performance of DTL1 cavity

The S₁₁ parameter of the power coupler has been tuned to better than -40 dB and the S₂₁ parameters of the two pickups are both about -50 dB finally. The frequency range of the cavity with coarse tuner is 1.02 MHz, as shown in Fig. 9, and completely covers the operation frequency, 53.667 MHz. The fine tuner can tune the resonant frequency about 140 kHz in 100 mm moving distance with no effect on field distribution along the beam axis. The measured Q₀ value of this cavity is 10400 at operation frequency finally, which is 85% of the simulation result by CST-MWS. Then the shunt impedance of the DTL cavity is 198 MΩ/m which was found by the measured Q₀ and the simulated R/Q value. Then we can evaluate that 18.4 kW RF power should be required if the U³⁴⁺ beam was accelerated successfully in our case.

Figure 9:Full-screen View

(Color online) The resonant frequency vs. the coarse tuner position.

3.3 Data analysis and beam dynamics benchmarks

In the measurement, background noise of the phase shift is less than ±0.04°. For the whole sets of the measured data, we used the Fourier transform filter high frequency noise. Because the bead is not infinitely small, the measurement accuracy in the low field strength area is not as good as the high field strength area. Lastly the data are averaged and compared with the designed field in the beam dynamic.

For the benchmarking, the measured field and the designed field on the central axis are both normalized by the maximum value of E-field on the beam axis, as shown in Fig. 10. As can be seen, the measured fields match the simulation field very well. Figure 11 shows the normalized cell voltage differences between measured and designed value. These voltage values can be imported into BEAMPATH code [14] for beam tracking. The beam dynamics simulation of ²³⁸U³⁴⁺ with 0.5 emA beam current was performed using the designed and measured field distribution respectively, and the beam phase space distributions at the exit of the IH-DTL1 were almost the same. So the field distribution of the IH-DTL1 can meet the beam dynamic design requirements.

Figure 10:Full-screen View

(Color online) Comparison between measurement and CST simulation result of the field distribution on the beam axis.

Figure 11:Full-screen View

(Color online) Cell voltage error between measurement and design value on beam axis.

3.4 The dynamics effects of off-axis fields analyze

Since the structure of the IH-DTL is symmetry in horizontal direction and asymmetry in vertical, it is hard to remove the vertical dipole field. In the RF design, the vertical dipole field has been optimized lower than 3.0% of the longitudinal field. As shown in Fig. 12, in the vertical center plane, the electric field distribution is asymmetrical about the centerline, and the longitudinal position of the field peak is variant with different vertical positions. The slope of the line formed by the peak field position can be used as the evidence of the dipole field quantity.

Figure 12:Full-screen View

(Color online) Electric field distribution on the vertical mirror plane calculated by CST-MWS.

So the fields on the vertical center plane were measured and the peak positions of the measurement data relative to the each gap center are compared with the CST-MWS simulation results, which are shown in Figs. 13 and 14. The dislocations of the peaks match the CST-MWS simulation results except those in the 9th gap. However, the dislocations in the 9th gap are smaller than the simulation data, which indicates the actual dipole field is not larger than the simulation result in the 9th gap. So the 3D simulation fields can represent the actual field in this IH-DTL for investigating the beam dynamics effects of the dipole field. This dipole field would cause the particle motion offset in the transverse plane for the average value of the vertical component of the E-field being not equal to 0 and should be investigated carefully for stable operation. The particle tracking of the beam center was performed using a 3D electromagnetic field and the transverse position offset is shown in Fig. 15. It can be found clearly that the position offset in the vertical direction is 0.27 mm, which is much larger than that of the horizontal plane, and the deflection angle is about 0.5 mrad in vertical direction. The beam phase space distribution at the exit of the IH-DTL1 is shown in Fig. 16. Because the dipole field component only exists in the vertical direction, the beam emittance of vertical direction is about 5% bigger than it in horizontal. By the vertical steering magnet at the entrance of the IH-DTL1, the correction of 0.3 mrad on the inject beam can make the beam center be close to the beam axis in the vertical, as shown in Fig. 17. But the beam emittance in the vertical is still bigger than it in the horizontal. Anyhow, to satisfy the operating requirement, the sets of correcting magnets need to be installed both at the entrance and exit of the IH-DTL1.

Figure 13:Full-screen View

(Color online) Peak positions of the simulation and measurement data relative to the each gap center in upper half vertical mirror plane.

Figure 14:Full-screen View

(Color online) Peak positions of the simulation and measurement data relative to each gap center in the lower half vertical mirror plane.

Figure 15:Full-screen View

(Color online) Transverse beam position offsets caused by the asymmetry of the E field.

Figure 16:Full-screen View

(Color online) Beam phase space distribution at the exit of the IH-DTL1 with the vertical dipole field component.

Figure 17:Full-screen View

(Color online) Transverse beam position offsets considering vertical dipole field component and the the correction of 0.3mrad on the injected beam.

3.5 Acceptance and output beam energy study

The energy acceptance range of the DTL1 must be studied carefully because the output energy of the RFQ may be a little higher or lower than the design value [15]. We set ±2% errors about the input energy of the IH-DTL1 in this study. Generally, when the input energy of the DTL is changed, the beam parameters at the exit should be different from the desired value. But for the separated function DTL, the RF phase and power can be tuned easily to rematch the requirement of the following acceleration structure.

As shown in Fig. 18, when the center energy of the IH-DTL1 inject beam is 140 keV/u and the beam inject phase at the original design value which corresponds the green square dot, the extraction beam’s energy is a little higher than the desired value, and the longitudinal phase space distribution of the extraction beam is shown in the left of Fig. 20. The transverse beam envelope is very small because the shortage of the inject energy lead to the beam phase in the acceleration gap slipping backward and the transverse RF focusing strength enhancing. The enhancing of the transverse RF focusing strength means the fading of the longitudinal constraint of the RF field. So the longitudinal emittance and bunch length are larger than the desired value. By tuning the inject phase forward and keeping the inject energy 140 keV/u, the energy, longitudinal emittance, and bunch length of the extraction beam become smaller and smaller, at the same time, the transverse envelope is growing. When the inject phase is tuned to -9, which corresponds the red square dot in the Fig. 18, the transverse envelope, longitudinal emittance, and bunch length of the extraction beam become acceptable and the beam energy can ensure the injection requirement of the following accelerators, as shown in the right side of Fig. 20.

Figure 18:Full-screen View

(Color online) Beam energy, envelope, longitudinal emittance and bunch length vs. the phase of IH-DTL1 respectively when the input energy is 3 keV/u below the design value. Green square dots correspond to the DTL design entrance phase setting, while red squares show an improvement of beam quality for this case.

Figure 20:Full-screen View

(Color online)The longitudinal phase space when the center energy of injected beam is 3 keV/u below the design value and inject phase at the original design value(left) and forward adjusted 9degree(right).

Same as above, Fig. 19 shows that when the injection beam’s energy is 146 keV/u, the the transverse envelope, longitudinal emittance, and bunch length of the extraction beam vary with the different inject phase, and the longitudinal phase space distributions of the extraction beam correspond the green and red square dots, which represent that the injection phase is the original value and the optimized one respectively. These are shown in Fig. 21. In brief, the energy acceptance of the IH-DTL is enough to overcome the ±2% error of output energy in RFQ.

Figure 19:Full-screen View

(Color online) Beam energy, envelope, longitudinal emittance and bunch length vs. the phase of IH-DTL1 respectively when the input energy is 3 keV/u above the design value. Green square dots correspond to the DTL design entrance phase setting, while red squares show an improvement of beam quality for this case.

Figure 21:Full-screen View

(Color online)The longitudinal phase space when the center energy of injected beam is 3 keV/u above the design value and inject phase at the original design value(left) and backward adjusted 5degree(right).

Moreover, it is easier to inject into the post accelerators if the output beam energy of the IH-DTL1 can be adjusted in a certain range. This can be realized by tuning the phase and global gap voltage. The output energy of the IH-DTL1 can be adjusted from 260 to 300 keV/u, with the longitudinal emittance and transmission efficiency in a acceptable range, for U³⁴⁺ as shown in Fig. 22. For the inject phase being -0.2 rad, the output beam energy can be tuned from 260 keV/u to 280 keV/u with the cavity power from 10 kW to 15 kW, and the corresponding longitudinal emittance and the transmission efficiency are both very good. In the same way, setting the inject phase to be 0.3 rad, the extraction beam’s energy can be larger than 300 keV/u and the corresponding longitudinal emittance and the transmission efficiency are both acceptable if the cavity power is above 20 kW. Moreover, keeping the injection phase and increasing the cavity power up to a value higher than desired, the output beam’s energy changes little if the cavity power continue increasing. This rule can be used for distinguishing whether the cavity power is too high for the design value. Considering the beam commissioning, the RF power can be decided distinctly for each kind of ion from above RF measurement result, but the accurate cavity phase must be tuned by manual work. Thus the simulation results can give the guide for the beam commissioning based on the beam energy spread and bunch length signal obtained by the beam diagnostics system.

Figure 22:Full-screen View

(Color online) Beam energy, longitudinal emittance and transmission efficiency at the exit of DTL by simulation under different injection RF phase and cavity power.

4.1 IH-DTL1 RF conditioning

The RF conditioning was started from pulse mode at the duty 1% when the DTL cavity vacuum was 1.0×10^-5 Pa. When we turned on the RF power at the level from 0.1 kW to 1 kW, the RF power unstable of the IH-DTL1 cavity began and the power reflection appeared, which indicated some discharge phenomenon occurring. In addition, the operating stability of the solid-state 40kW power amplifier of IH-DTL1 was poor in the begining. After about a half month of conditioning, the stability of the RF amplifier was improved. One month later, the RF power up to 18 kW with duty 20% was transmitted into the cavity. The cavity vacuum is continuously improving and after 2 months of pumping, the pressure is better than 1.0×10^-5 Pa. After then the IH-DTL1’s RF conditioning shifted to CW mode from lower power. It should be noted that the water cooling design was not perfect on the cavity wall and the highest temperature of the cavity wall became 67 ℃ in the up and bottom flanges where the magnetic fields were high, which was much higher than expected when the cavity power was 15 kW with CW mode. So we decided to suspend the higher power conditioning, and to design 4 additional cooling plates on the cavity wall. At the same time, we began to prepare the first beam commissioning of the IH-DTL1.

4.2 Experimental setup

The experimental setup used for the beam commissioning of the IH-DTL1 is shown in Fig. 23. It consists of an ECRIS, a low energy beam transport (LEBT) line, and a CW 4-rods RFQ. The LEBT system includes 90 bending magnets, 4 matching quadrupoles, and two solenoids for the injection matching of the RFQ. A beam diagnostic system including wire scanners and a Faraday cup was installed in the LEBT. The LEBT and the RFQ were used in our previous research work and had been described in Ref. [15]. The IH-DTL1 was connected to the RFQ through the medium energy beam transport (MEBT)line. A spiral stem rebuncher with 4 gaps was successuflly developed and installed in the MEBT. A beam diagnostic platform was installed after the IH-DTL1, which includes a Faraday cup, two slit-wire type transverse emittance measurement devices, and a silicon detector for energy spread measurement. A Low Level Radio Frequency (LLRF) control system was developed to operate the RF cavity in a self-constant mode including the frequency, phase, and power.

Figure 23:Full-screen View

(Color online) General layout of SSC-Linac for the beam commissioning of the IH-DTL1.

4.3 Beam commissioning results

When the beam was injected into the IH-DTL1, the power of the rebuncher and IH-DTL1 were fixed at the design values which were obtained by calculating through the RF measurement values. The operating phases were scanned carefully and fixed to insure the signals of the BPM behind the IH-DTL1 where most clear, as shown in Fig. 24. It should be noted that the combined beam with ¹²C³⁺ and ¹⁶O⁴⁺ ions were firstly used in the beam commissioning. The reason for accelerating the combined beam is to make preparations of the future energy spread detector’s calibration. The front end section including ECR, LEBT, and RFQ were tuned to transport approximately 78 euA this combined ions to the IH-DTL1 entrance, and the beam current was 63 euA on the Faraday cup at a distance of 1.2 m from the exit of the IH-DTL1, which correspond to the 80% transmission efficiency. The particle energy was measured with the time of flight method using two Beam Position Monitors (BPM), separated by a distance of 563 mm. By manually tuning the phase of IH-DTL1, the time of flight of beam between the two BPMs changed from 73.5 ns to 75.6 ns, which correspond the beam energy of 304.2 keV/u and 287.5 keV/u by calculating. Energy spectrums of C³⁺ and O⁴⁺ were measured, as shown in Fig. 25. With different RF power of IH-DTL1, output beam energy of the IH-DTL1 can be adjusted conspicuously. By tuning the RF phase of the IH-DTL1, the output beam energy can be changed slightly. The measured HWHM energy spread values of the IH-DTL1 is approximately 5%.

Figure 24:Full-screen View

(Color online) The BPM signal used for beam energy measurement behind the IH-DTL1.Drift lengths from the DTL exit to the monitors have been 260 mm and 823 mm, respectively.

Figure 25:Full-screen View

(Color online) The measured output beam energy spectrums of the RFQ and IH-DTL1.

Figure 26 shows emittance measurement results of the combined beam. The measured values of the normalized RMS emittance in the horizontal and vertical plane were 0.139 π.mm.mrad and 0.157 π.mm.mrad, respectively. The vertical emittance is a little larger than the horizontal one, because of the vertical dipole field component in the IH-DTL1 and the beam and lens alignment of the last lens. The detail information and result analysis of the beam commissioning will be published in the near future.

Figure 26:Full-screen View

(Color online) The measured emittance at the exit of the IH-DTL1.