[Motivations]
Normal-conducting accelerating structures, usually made of high-purity oxygen-free copper, are at the hearts of many modern particle accelerators. A vacuum arc in such a structure can lead to vacuum breakdown, which can limit the high-gradient accelerator performance; however, the actual trigger mechanism of breakdown is still a mystery despite many breakdown studies having been performed by many researchers around the world.

[Solutions]
I proposed a method based on direct in-situ observation of inner surfaces of accelerating structures using cameras at the moment of breakdowns, which does not depend on indirect or secondary signals arising from vacuum breakdown, such as RF reflection, current flash, X-ray, and acoustic emission. This method has been inspired by the assumption that "seeds" of breakdown are so minuscule that the information they contain can be mostly destroyed by the much larger phenomenon of vacuum arc.

[Results and related papers, articles, reports, and presentations in time series]

[Motivations]
A common normal-conducting high-gradient accelerating structure of particle accelerators is the "disk type", which has dozens of machined, stacked, and bonded disks. In the disk type, huge surface currents associated with a magnetic field of an accelerating mode (~10⁸ A/cm²) flow across disk-to-disk junctions, which might trigger RF breakdowns and/or cause yield aggravation in structure fabrication. Another structure is the "quadrant type" ("half type" for undamped structures), which is orthogonal to the above type (disk v.s. quad(half)). In the quadrant type, bonding planes are parallel to a beam axis, and therefore, this type has the remarkable feature that no such surface current flows across any junction. Furthermore, we can expect significant cost reduction and yield improvement in structure fabrication. However, the naive quadrant-type X-band (11.4 GHz) accelerating structure with 18 cells failed in high-gradient tests performed by KEK and SLAC in 2009. We could not find any experimental evidence of the failure.

[Solutions]
In 2012, I proposed and designed an improved quadrant-type by performing simulation studies to overcome all disadvantages of the naive quadrant-type design. This new design ("improved quadrant type") has the following two key characteristics: where the chamfer of 0.4 mm and the gap of 0.1 mm are the results of optimization by simultaneously minimizing the field enhancements and deterioration of the shunt impedance. We fabricated and tested an improved quadrant-type X-band (11.4 GHz) accelerating structure in the form of a single-cell test cavity, which is easier to make and test than multi-cell structures.

[Results]
Performing a high-gradient test of the single-cell test cavity in 2017, we experimentally demonstrated its efficacy and the potential for the improved quadrant-type structure to be successful for gradients of 100 MV/m or higher at sufficiently low breakdown rates required for future linear colliders.

[Motivations]
A positron damping ring (DR) has been constructed to fulfill the requirement for low-emittance positron-beam injection into the main ring (MR) of the SuperKEKB accelerator complex at KEK. Previously, the required total accelerating voltage in the initial DR design was 0.26 MV, meaning that one cavity was enough for DR operation. However, it was shown theoretically that single-bunch instabilities caused by coherent synchrotron radiation would significantly affect the operation if such a total accelerating voltage was used. Now, a total voltage of 1.4 MV is required, but the large theoretical uncertainty is still a concern. It should be noted that 1.4 MV is required per the DR design specification, and the hardware (cavity) performance should exceed this requirement.

[Solutions]
To supply an accelerating voltage higher than 1.4 MV to the DR in the limited space of a sole RF section, which was originally designed for one cavity, I proposed and designed a 509-MHz accelerating structure that can accommodate up to three cavities with unique space-saving features of:
In this accelerating structure, all of the related HOMs are damped so that the electromagnetic field in each cavity has high independence. Furthermore, the number of cavities used in this structure is variable, and cavities are replaceable, yet it can still be considered a single mechanical structure with solid connections between the components. This "multi single cell" design is the most significant characteristic of this structure.

[Results]
We successfully fabricated two production-version DR cavities based on development results of a prototype cavity, and demonstrated the high-power performance at a test stand up to a cavity voltage of 0.95 MV (= an accelerating gradient of 3.7 MV/m, continuous wave). In 2016, we successfully installed the two cavities into the DR tunnel with an alignment accuracy of ~0.3 mm in the transverse directions, based on a method I devised that includes selecting the best permutation of the components (cavities and beam pipes) out of 2048 permutations for the assembly in the DR tunnel. These RF cavities have been successfully and stably accelerating positrons in the DR operations.

[Motivations]
Coaxial-line high-power input couplers have been used for normal-conducting 509-MHz accelerating cavities (ARES) at the (Super)KEKB main rings. Regular coaxial lines have more multipactoring (also called multipacting), which is repeated discharge driven by RF fields, than rectangular waveguides because of more uniformity of the electromagnetic field. In 2003 and 2004, we had a serious problem caused by multipactoring in the coaxial lines of the input couplers for some of the ARES cavities used in KEKB accelerator operations. We could not solve it even replacing the input couplers by new ones.

[Solutions]
I adopted fine grooving of the conductor surfaces of the coaxial lines because it is expected to be a promising method for suppressing multipactoring under any conditions with heavy beam loading.

[Results]
Developing a dedicated simulation program, I performed simulation studies on the multipactoring to determine the best design of grooves. Then, we fabricated grooved input couplers based on the design, and demonstrated the completely-suppressive power of the grooving both in high-power tests at a test stand and KEKB accelerator operations. Now this groove design has been included in the standard drawing of the input coupler for the ARES cavities, and contributed to stable (Super)KEKB accelerator operations.

[Motivations]
Some high-energy physics experiments using particle accelerators have suffered from bitter experience on backgrounds of synchrotron radiation coming from the accelerators, where particle detectors could be severely damaged.

[Solutions]
We have developed a new calculation method of synchrotron radiation based on a real beam orbit in an accelerator, aiming at quantitative estimations of synchrotron-radiation backgrounds, and the construction of a possible alarm system for the background.

[Results]
We have successfully developed a new method to calculate synchrotron radiation based on a real beam orbit with reasonable offset corrections, where the orbit is obtained by global fitting of several sets of measurements of beam-position monitors. We have also successfully reproduced the Belle/SVD gain-drop accident quantitatively, so that the practicability of our method has been established.

[Motivations]
Research of an intermediate region between soft and hard processes in strong interaction, that was not yet well studied

[Solutions]
Measuring production cross sections of J/psi mesons (-->l+l-) in high-energy electron-proton collisions, the results were compared with theoretical predictions from the Regge phenomenology (soft physics) and from the perturbative QCD (hard physics), providing a new and unique object for the QCD study with a clean signal of the leptonic final state.

[Results]
Measured cross sections were well described by theoretical models based on the perturbative QCD, which indicates that the virtuality of the exchanged photon and/or the mass of the J/psi can be used in perturbative calculations as a hard scale although the quark partons are confined in a shell of the hadron state.

[Motivations]
Dilepton production in electron-proton collisions (e p → e l⁺ l^- X ) can be a significant background in many experimental data analyses. However, there was no event generator of dilepton production which contained all the physics processes, including accurate proton destruction processes.

[Solutions]
I developed a complete event generator by using an automatic calculation system GRACE, including all the tree-level Feynman diagrams in the standard model and accurate proton destruction processes.