Japan Proton Accelerator Research Complex Linac Research Articles

J-PARC linac and RCS – operational status and upgrade plan to 2-MW

The linac and the 3 GeV rapid-cycling synchrotron (RCS) at the Japan Proton Accelerator Research Complex (J-PARC) were designed to provide 1-MW proton beams to the following facilities. Due to the improvement of the accelerator system, we accelerated a 1-MW beam with a small beam loss. The lack of anode current in the radiofrequency (RF) cavity, rather than beam loss, limits the RCS beam power. Recently, we developed a new acceleration cavity that can accelerate a beam with a low anode current. This new cavity enables us to reduce the requirement for the anode power supply and accelerate a beam of more than 1 MW. We considered how to achieve beam acceleration beyond 1 MW. So far, a beam of up to 1.5 MW is expected to be accelerated after replacing the RF cavity. We also studied to achieve an up to 2-MW beam in J-PARC RCS.

Investigation of random beam trips in a linear accelerator at the Japan proton accelerator research complex for the development of an accelerator-driven nuclear transmutation system

ABSTRACT In a proton linear accelerator (linac), the proton beam is sometimes unexpectedly interrupted because of the electrical discharge caused by radio frequency, device/equipment failure, and other factors. Are these beam trips random? Traditionally, beam trips have been implicitly assumed to occur randomly. In this study, we investigated whether beam trips occur randomly in the Japan Proton Accelerator Research Complex (J-PARC) linac. The aim was to estimate the beam trip frequency in a superconducting proton linac for an accelerator-driven nuclear transmutation system. First, the J-PARC linac was classified into five subsystems. Then, the reliability function for the operation time in each subsystem was obtained by the Kaplan – Meier estimation, a reliability engineering method. This reliability function was used to investigate the randomness of beam trips. Analysis of operational data for five subsystems of the J-PARC linac for five years showed that beam trips occurred randomly in some subsystems. However, the beam trips did not occur randomly in many subsystems of the proton linac, including the ion source and the acceleration cavity, which are the primary subsystems of the proton linac.

Measurement of H0 particles generated by residual gas stripping in the Japan Proton Accelerator Research Complex linac

The Japan Proton Accelerator Research Complex (J-PARC) linac is a high-intensity accelerator in which beam loss is a critical issue. In the J-PARC linac, H− beams are accelerated to 191 MeV by a separated drift tube linac (SDTL) and subsequently to 400 MeV by an annular-ring coupled structure (ACS). Because there are more beam loss mechanisms in H− linacs than in proton linacs, it is imperative to investigate the beam loss circumstances for beam loss mitigation. Electron-stripping phenomena, which generate uncontrollable H0 particles, are characteristic beam loss factors of H− linacs. To clarify the beam loss causes in the J-PARC linac, a new diagnostic line was installed in the beam transport between the SDTL and ACS. In this diagnostic line, H0 particles were separated from the H− beam, and the intensity profiles of the H0 particles were successfully measured by horizontally scanning a graphite plate in the range where H0 particles were distributed. By examining the intensity variation of the H0 particles with different residual pressure levels, we proved that half of the H0 particles in the SDTL section are generated by the residual gas stripping in the nominal beam operation of the J-PARC linac.

Compact and efficient radio frequency digital feedback control system for accelerator applications

The digital feedback (DFB) setup employed in a low-level radio frequency (LLRF) control system is crucial to ensure RF field stability in accelerating cavities, thus guaranteeing successful operation. To this end, a novel in-house and multi-purpose prototype DFB setup was developed for multiple applications at the Japan Proton Accelerator Research Complex (J-PARC). Being efficient, low-cost, and compact is key to achieving a cost-effective system that fulfills the performance specifications. It also reduces the production and maintenance costs of future DFB setups. A field-programmable gate array (FPGA)-based design with a digital signal processing (DSP) function uses an analog in-phase/quadrature-phase (I/Q) demodulator for RF-to-baseband signal conversion and an I/Q modulator to generate RF power. This arrangement compensates for RF phase fluctuations and amplitude modulations by employing a proportional and integral (PI) feedback controller. In addition, systematic errors are minimized by applying a hardware-error-compensation process. The system was utilized to conduct high-power tests on the interdigital H-mode drift-tube linac (IH-DTL) cavity of a muon linac for short RF pulses. The setup was also tested on a buncher cavity (Buncher-2) in the J-PARC linac, achieving efficient performance for longer RF pulses. The stability of the RF accelerating field in the IH-DTL was achieved at ±0.25% peak-to-peak (pp) in amplitude and ±0.36 degree pp in phase. For the Buncher-2, the amplitude stability of ±0.18% pp and phase stability of ±0.13 degree pp were obtained. Therefore, this DFB setup can be used to conduct high-power tests of RF cavities and klystrons. This study discusses the design aspects of a cost-effective DFB system and reports high-power measurements.

Development of the novel and superior adaptive beam-loading compensation system calculated in the frequency domain for the J-PARC Linac

The Japan Proton Accelerator Research Complex (J-PARC) is one of the highest intensity proton accelerator facilities in the world. Its accelerator consists of a 400-MeV linac (LINAC), a 3-GeV rapid-cycling synchrotron (RCS), and a 30-GeV main ring (MR) and its beam is provided for wide-ranging applications. The peak current of the beam at LINAC had been increased from 30 mA to the design value of 50 mA in 2018. Therefore, the performance of the beam-loading compensation implementing in the low-level radio frequency (LLRF) system has been more important to satisfy the stability of the accelerating gradient. Although the iterative learning control (ILC) scheme with adaptive ability was demonstrated, the practical use of ILC was not convenient for large accelerators such as the J-PARC LINAC with forty-nine RF stations. Therefore, we developed at novel and superior adaptive beam-loading compensation system that was calculated in the frequency domain. Compared to the scheme of the previous ILC, the newly developed system is useful because it does not require parameter adjustment using the beam. In the experimental results utilizing this system, the optimum tables of FF_BEAM were able to be obtained as a smaller number of iterations. The peak-to-peak stability of amplitude was achieved to ± 0.23% and the peak-to-peak stability of phase was improved to ± 0.036°. In addition, errors such as white noise did not accumulate for the table of FF_BEAM and this system solved the problem of the divergence when the number of iterations was increased. This implementation of the adaptive beam-loading compensation system calculated in the frequency domain will potentially contribute to the improvement of the J-PARC LINAC LLRF, and this system may lead to a stable user operation. This system is so simple in principle and effective that it could be applied to other linear accelerators with the pulsed beam.

Low-power proton beam extraction by the bright continuous laser using the 3-MeV negative-hydrogen linac in Japan Proton Accelerator Research Complex

ABSTRACT The Japan Atomic Energy Agency (JAEA) has designed a Transmutation Physics Experimental Facility (TEF-P) as an experimental facility in the Japan Proton Accelerator Research Complex (J-PARC). The TEF-P is a critical assembly driven by a low-power proton beam, a maximum of 10 W, which is extracted from a high-power beam source, such as 250 kW of 400 MeV proton beam of the J-PARC linac. To extract such a low-power proton beam from the high-power proton beam, we developed a laser charge exchange (LCE) device and employed its technique, which is one of the non-contact beam extraction techniques. To confirm the performance of the LCE device, a low-power proton beam was extracted using a negative-hydrogen linac having an energy of 3 MeV, and a bright continuous laser. A 0.57 mW proton beam was successfully extracted with a laser stripping efficiency of . These values obtained are in good agreement with the estimated values.

VSWR Adjustment for ACS Cavity in J-PARC Linac

In the Japan Proton Accelerator Research Complex (J-PARC) linac, negative hydrogen beams are accelerated from 190 MeV to 400 MeV by Annular-ring Coupled Structure (ACS) cavities. The RF input coupler of the ACS21 cavity, which is the twenty-first (the last) accelerating cavity in the order of beam acceleration, had a comparatively larger Voltage Standing Wave Ratio (VSWR) value than the other ACS cavities. Therefore, we designed and manufactured a rectangular waveguide which have a capacitive iris to adjust the coupling factor of the ACS21 cavity. By making use of the summer maintenance period in 2018, we installed the newly manufactured waveguide to the cavity. Consequently, the VSWR of the ACS21 was successfully decreased to the target value which leads to the critical coupling under the nominal accelerating condition with 50-mA peak beam current.

Longitudinal Bunch Size Measurements with an RF Deflector at J-PARC LINAC

We have measured the longitudinal bunch size of beams from the radio-frequency quadrupole linac (RFQ) at the Japan Proton Accelerator Research Complex (J-PARC). The radio-frequency deflector (RFD) is installed in the medium-energy beam transport section (MEBT1) between the RFQ and the drift-tube linac (DTL). The RFD, which is usually used to produce a macro pulse configuration for the 3-GeV rapid-cycling synchrotron, is utilized to produce a different deflecting angle depending on the bunch arrival phase. Transverse beam profiles were measured after the RFD with and without deflection to extract the longitudinal bunch size. The Courant-Snyder parameters in the longitudinal direction were estimated by measurements for different strength of the upstream buncher. These measurements are one of the important steps to understand the longitudinal beam properties at a high beam current.

A beam position monitor for the diagnostic line in MEBT2 of J-PARC linac

In the linac of the Japan Proton Accelerator Research Complex (J-PARC), the neutral hydrogen (H0) beam from the negative hydrogen ion (H-) beam is one of key issues in mitigating beam losses. To diagnose H0 particles, we installed a set of beam-bump magnets to generate a chicane orbit of the H- beam. The beam position monitors (BPMs) in the beam line are used for orbit correction to maintain the beam displacement within 2.0 mm from the duct center. To measure the beam displacement under different drive currents of the beam-bump magnets, a new wide-range BPM was designed and manufactured to evaluate the horizontal beam position by using a correction function to compensate for non-linearity. We also employed the beam profile monitor (WSM: wire scanner monitor) to measure the H- beam profile, which helped us to compare the beam position measurements. In this paper, the design and the performance of the wide-range BPM are described. In addition, we present a comparison of the beam position measured by the BPM and the WSM.

Path to 1-MW at J-PARC Rapid Cycling Synchrotron

The accelerator system at Japan Proton Accelerator Research Complex (J-PARC) has been operational since May 2008 and has mainly been used to perform physics experiments. The accelerator system consists of a Linac, a Rapid Cycling Synchrotron (RCS), and a Main Ring Synchrotron. The originally designed RCS injection energy is 400MeV, but the first operation started at 181MeV. New acceleration cavities were installed in J-PARC Linac during the summer shutdown of 2013, and user operation by the Material and Life science Facility (MLF) at the injection energy of 400MeV was started from February 2014. Post beam commissioning of 400MeV injection energy, beam loss was small enough, and we established 300kW continuous operation. Subsequently, the peak current of the Linac was increased from 30mA to 50mA. This upgrade enabled us to try 1-MW beam acceleration. Finally, after some additional improvements, we successfully accelerated 1-MW equivalent protons.

Design and delivery of beam monitors for the energy-upgraded linac in J-PARC

In the J-PARC (Japan Proton Accelerator Research Complex) linac, an energy-upgrade project has started to achieve a design beam power of 1 MW at the exit of the downstream synchrotron. To account for the significant beam parameter upgrades, we will use the newly-fabricated beam monitors for the beam commissioning. This paper discusses the design and assembly of the beam position monitor, phase monitor, current monitor, transverse profile monitor, and beam loss monitor for the energy-upgraded linac. We periodically installed the newly-fabricated monitors for the upgraded beam line, as well as for longitudinal matching, because of the frequency jump between the original RF cavity and the newly-developed cavity. We employed two debunchers to correct for momentum spread and jitter. To account for the new debunchers, we fabricated and installed additional pairs of phase monitors in order to tune the debunchers to the adequate RF set point. Finally, we propose commissioning plans to support the beam monitor check. We will begin to establish the 181-MeV operation to confirm the proper functioning of beam monitors. Herein, we will examine the response to changes of the knobs that control the quadrupole magnets after the energy upgrade. After proper functioning of the beam monitors is confirmed, we will use the new beam monitors to establish the 400-MeV acceleration operation.

High-power test of annular-ring coupled structures for the J-PARC linac energy upgrade

Annular-ring coupled structures (ACSs) will increase the beam energy of the Japan proton accelerator research complex (J-PARC) linac from 181 to 400 MeV to achieve a beam power of 1 MW for a materials and life science experimental facility. The mass production of the ACS cavities commenced in March 2009. Before the installation, all cavities require power testing. High-power testing is essential not only for confirming the cavity’s design performance but also for preventing delays in cavity conditioning schedule. However, the 2011 Tohoku earthquake damaged J-PARC facilities, including the ACS power-test area, and cavity conditioning was interrupted for two years. After the facility’s restoration, two ACS cavities (M01 and M11) were conditioned. They performed 15–20% above the designed accelerating field of 4.2 MV/m. As M01 was initially conditioned six years ago, the most recent conditioning time required for M01 was drastically reduced. From this result, we confirmed that long-term stored ACS cavities purged with nitrogen gas do not produce critical cavity performance issues. During high-power operation of M11, which is a unique cavity equipped with a capacitive iris in a waveguide, no significant increases in the temperature and the discharge rate around the capacitive iris were observed. Even considering beam loss due to residual gas scattering, the vacuum pressure was sufficiently low (4 × 10−6 Pa). More stable operation can be expected following a month-long conditioning process before the beam is commissioned. M11’s conditioning successfully demonstrated an auto-conditioning program, and we established the conditioning scheme using this auto-conditioning program for all ACS cavities in a limited time and with limited manpower.

Progress of beam commissioning and beam loss mitigation in the J-PARC linac after the Tohoku earthquake

The Tohoku earthquake in March 2011 caused significant damage to the Japan Proton Accelerator Research Complex (J-PARC) linac and forced us to shutdown the accelerator for nearly nine months. After a significant effort for its restoration, we resumed beam operation of the J-PARC linac in December 2011. Since the resumption of beam operation, we have been suffering from beam-losses which were not observed before the earthquake. Tackling the beam-loss issues, we have finally reached the same beam power for user operation as before the earthquake. In this paper, we present our experience with the beam start-up tuning after the earthquake, with emphasis on beam-loss mitigation efforts.

Recovery from the earthquake and operation of the J-PARC linac

The Japan Proton Accelerator Research Complex (J-PARC) facilities were extensively damaged by the Great East Japan Earthquake on March 2011. At the linac facility, the buildings, especially those above the ground, were severely damaged. Owing to the damage, most of the conventional facilities were unavailable for several months. There were many cracks in the accelerator tunnel, and a large amount of groundwater (approximately 10 tons per day) leaked into the tunnel. The RF cavities and magnets themselves were not seriously damaged; however, vacuum leaks occurred in the ion source, the medium-energy beam transport 1 (MEBT1), and the separated-type drift tube linac (SDTL) sections from the destruction of the vacuum port, beam monitors, and bellows. Therefore, the accelerator components were exposed to highly humid air for several weeks. A maximum misalignment of 0.2 mm of the drift-tube (DT) bore center in the drift tube linac (DTL) was observed. A subsidence of more than 40 mm and a tilt of about 3 mrad were also observed in the accelerator tunnel. After significant recovery efforts, we resumed beam operation in December 2011 and user operation in January 2012. The beam power from the linac reached 13.3 kW in March 2012, which is equal to that just before the earthquake. In this report, we will discuss the recovery from the earthquake and the present operating status of the J-PARC linac.

Suppression of Beam Loss at the First Arc Section in the J-PARC Linac

J-PARC (Japan Proton Accelerator Research Complex) linac has recently started its continuous operation to provide a 181-MeV H <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">-</sup> beam to the downstream 3-GeV synchrotron and a neutron target. Immediately after the commencement of the continuous operation, significant residual radiation and beam loss were observed at the first bending magnet in the first arc section. The measured radiation and beam loss were localized in the side opposite to the deflection of the H <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">-</sup> beam. A possible beam loss source is the H <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">+</sup> produced by double electron-stripping of H <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">-</sup> in a residual gas scattering at the front end part of the linac. To confirm this supposition, an H <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">+</sup> signal was measured with a wire scanner monitor at the beam energy of 3 MeV by separating H <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">+</sup> from the H <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">-</sup> beam with horizontal dipole magnets. Then the separated H <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">+</sup> is successfully removed by a beam scraper originally for beam chopping. With this H <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">+</sup> suppression, the beam loss at the first bend was suppressed by factor of 24-31 without sacrificing the beam chopping performance. Consequently, the residual radiations were reduced to a comfortable level for the continuous operation.

The slow-ground-motion monitoring based on the hydrostatic leveling system in J-PARC linac

A slow-ground-motion monitoring system has been developed for J-PARC (Japan Proton Accelerator Research Complex) linac to improve its operational stability. The monitoring system utilizes a hydrostatic leveling system with non-contact water-level sensing, and it has remote water-level control capability. Online calibration has been performed with a remote water-level control and an accuracy of 0.02 mm over the range of 1 mm has been obtained. Characteristic time response of this monitoring system is evaluated based on a hydraulic model. This simple modeling can simulate dynamic fluctuation of a multi-vessel system with various conduit network topologies. We have performed a forced water-level variation test, and confirmed that the modeling shows a reasonable agreement with experiments. The time resolution of this system is confirmed to be around 540 s with a filling tank for remote water feeding. To demonstrate the system performance, the slow ground motion of J-PARC linac has been measured with this monitoring system. A periodic ground movement is observed with this system, which is supposed to be induced by a tide effect. The stability of the accelerator tunnel has some local variation possibly due to the difference of the geological circumstances and the inequality in the foundation constructions.

Tuning of RF amplitude and phase for the separate-type drift tube linac in J-PARC

It is important to accurately adjust the amplitude and the phase of RF power sources for a high intensity proton linac. J-PARC (Japan proton accelerator research complex) linac is one of those high-intensity linacs. J-PARC linac has 30 SDTL (separate-type drift tube linac) tanks to accelerate the negative hydrogen ions from 50 to 181 MeV. Two neighboring SDTL tanks are driven by one klystron, where the phase and the amplitude of these two tanks are controlled in terms of the vector sum. The target-value of the vector sum control should be determined with a beam-based tuning for each klystron. During the beam commissioning, the RF tuning has been performed with a phase scan method introducing a concept of phase signature matching. In the tuning, the output beam energy from the SDTL module is monitored while scanning the RF phase. Comparing the obtained phase dependence of the output beam energy with those from a numerical model, the target-values for the low-level RF control system has been tuned within the required accuracy of 1° in phase and 1% in amplitude. The same tuning procedure has successfully been applied to the RF tuning of buncher and debuncher cavities.

Macro-particle simulation study on transverse halo collimator for J-PARC linac

To realize the transverse emittance required for the ring injection, a transverse halo collimation system of the charge-exchange type has been developed for J-PARC (Japan Proton Accelerator Research Complex) linac. In this type of collimation system, negative hydrogen ions constituting a tail portion are charge-exchanged to protons with thin foil collimators. Then, the generated protons are separated by a bending magnet to be delivered to a dedicated beam dump. Because a two-species beam is transported along the collimation system, a special care is to be taken to avoid the uncontrolled beam loss along the beam transport. To this end, it is of practical importance to develop an accurate numerical model that has the capability of both particle tracking with space-charge effects and modeling of scattering at the charge-exchange foil. In this paper, a numerical model is developed for the collimation system using a three-dimensional particle-in-cell code, IMPACT, and a Monte-Carlo simulation tool kit, GEANT. Uncontrolled beam loss along the collimation system has been evaluated with this model, and the optimum foil thickness has been determined for the collimation system.

RF reference distribution system for J-PARC linac

The linac used by the Japan Proton Accelerator Research Complex (J-PARC) is a high-intensity proton linac with a peak current of 50 mA. The error in the accelerating field needs to be maintained within ±1% in amplitude and ±1° in phase. Thus, high phase stability is required for the RF reference distribution system. Our objective is to maintain the phase stability of the reference to less than ±0.3°. The beam quality or intensity of a proton linac strongly depends on the stability of the accelerating fields of each cavity; therefore, the stability of the RF reference distribution system is essential for the performance of the J-PARC linac. A highly stable and unique RF reference distribution system was developed for the J-PARC linac. A 12-MHz RF reference signal is converted into an optical signal and amplified by an optical amplifier. Then it is distributed through optical fiber links to 60 low-level RF control systems comprising klystron and solid-state amplifier driving systems. Phase-stabilized optical fiber (PSOF) is employed in the optical transfer line. The thermal coefficient of the PSOF transmission time was measured: its value was 0.4 ppm/°C. This property is insufficient for the required stability; accordingly, the temperature change in the PSOF should be controlled to be within ±0.5 °C by means of a cooling water system. New stable and low-jitter optical-to-electric (O/E) and electric-to-optical (E/O) converters were developed for this linac. The O/E converter is so compact that is can be mounted on a compact PCI board: its temperature is maintained constant by a Peltie device. A pulse driver (limiting amplifier) is introduced in the E/O and O/E converters for signal modulation to reduce the detection jitters. It transforms the 12-MHz signal into a rectangular wave: its the rise time is shorter than 200 ps. Then, a low-transfer jitter of 0.8 ps (rms) was obtained in the optical link of the E/O and O/E converters. The installation of the RF reference distribution system was completed. The phase stability of the distributed signal was evaluated, and a phase stability of the ±0.2° was obtained for a frequency of 927 MHz; consequently, the required system stability was achieved. The beam acceleration to a design energy of 181 MeV for the first phase (Phase I) was successfully performed in February 2007 and the centroid momentum fluctuation of the beam pulse was within ±0.015%, while the required stability was within ±0.025%. Now the beam commissioning has been steadily continued.

The development and beam test of an RF chopper system for J-PARC

The linac of the Japan Proton Accelerator Research Complex (J-PARC) is now under construction. It is a high-intensity linac with a peak H − current of 50 mA and a 1.25% duty factor. As a key component of the medium-energy beam-transport line of the linac, a chopper is necessary to obtain the required beam structure for the following 3-GeV rapid-cycling synchrotron, and also plays an important role in beam-loss control in such a high-intensity accelerator. Clean chopping requires a short rise time and a high deflecting field. An radio frequency deflector (RFD) was proposed as a chopper for the J-PARC linac, because of its merits of high deflecting field, compact structure, ease to manufacture and both reliability and flexibility of the total chopper operation. An RF chopper, consisting of two RFDs, has been developed. In a low-level RF test, the fundamental RF properties show good agreement with those of a simulation in the design. In a high-power test, the chopper works well without any discharge under 36 kW peak driving power. In a test of the coupled RFD system, the longer rise time of the second RFD and the delay of the second RFD excitation were observed. A beam test of the RF chopper was also successfully performed up to a peak current of 25 mA. The chopper system worked very stable during the beam test, and the sufficient separation between chopped and normal beam was obtained, with a short rise/fall time (10 ns) of the chopped beam.