With the CHESS-U upgrade in 2018, the south arc region of Cornell Electron Storage Ring (CESR) has been replaced with new magnets and vacuum chambers, and CESR will operate with higher beam energy and beam current at lower beam emittance. A new crotch absorber has been designed for the upgrade, based on previous CHESS crotch absorbers. The main body of the crotch absorber is a water-cooled hollow cylinder made of oxygen-free high-conductivity copper, and it has a 5-mm thick beryllium band brazed on the outside in the middle section to reduce the power density it absorbs on the surface. In this paper, we describe the mechanical design, present finite element thermal and structural analysis results with a focus on the region with high temperature and high thermal stress, and discuss the fatigue lifetime of the absorber. For typical operation with 6 GeV beam energy and a 200 mA beam current, the highest temperature on the crotch is approximately 300 °C based on the thermal analysis. Correspondingly, the fatigue lifetime is conservatively estimated to be around 4 × 104 cycles. We also show some details in the fabrication of the crotch assembly, especially about the vacuum furnace brazing joint between the beryllium band and the copper main body, including X-ray fluorescence inspection of the brazing joint. In addition, in order to have a relevant comparison in an effort to estimate the fatigue lifetime, we carry out finite element analyses (FEA) for the low cycle fatigue experiment with beryllium tiles using electron-beam heating of Watson et al. [Fusion Eng. Des. 37, 553 (1997)] in 1990s and provide new understanding for some of their experimental observations. These analyses assure the effectiveness of the design.

During the summer and fall of 2018 the Cornell High Energy Synchrotron Source (CHESS) is undergoing an upgrade to increase high-energy flux for x-ray users. The upgrade requires replacing one-sixth of the Cornell Electron Storage Ring (CESR), inverting the polarity of half of the CHESS beam lines, and switching to single-beam on-axis operation. The new sextant is comprised of six double-bend achromats (DBAs) with combined-function dipole-quadrupoles. Although the DBA design is widely utilized and well understood, the constraints for the CESR modifications make the CHESS-U lattice unique. This paper describes the design objectives, constraints, and implementation for the CESR accelerator upgrade for CHESS-U.

During the summer and fall of 2018 the Cornell High Energy Synchrotron Source (CHESS) is undergoing an upgrade to increase high-energy flux for x-ray users. The upgrade requires replacing one-sixth of the Cornell Electron Storage Ring (CESR), inverting the polarity of half of the CHESS beam lines, and switching to single-beam on-axis operation. The new sextant is comprised of six double-bend achromats (DBAs) with combined-function dipole-quadrupoles. Although the DBA design is widely utilized and well understood, the constraints for the CESR modifications make the CHESS-U lattice unique. This paper describes the design objectives, constraints, and implementation for the CESR accelerator upgrade for CHESS-U.

CHESS has developed and successfully deployed a novel Dual Array Valence Emission Spectrometer (DAVES) for high energy resolution, hard x-ray spectroscopy. DAVES employs the simplest method for scanning multiple spherical crystals along a Rowland Circle. The new design achieves unique 2-color collection capability and is built to take special advantage of pixel array detectors. Our initial results show why these detectors greatly improve data quality. The presentation emphasizes flexibility of experimental design offered by DAVES. Prospects and benefits of 2-color spectroscopy are illustrated and discussed.

In November 2014 two in-air 1.5m CHESS Compact Undulator (CCU) magnets built by KYMA S.R.l. were installed in Cornell Electron Storage Ring (CESR) in canted arrangement and after few days of commissioning their regular operation started. CCU magnets are compact, lightweight and cost efficient devices. They have very stable magnetic field integrals independent of deflection parameter value. This feature greatly simplifies the storage ring operation. The CCU concept was developed at Cornell in 2011 and the first 1m in-vacuum CCU magnet was beam-tested in 2012. The article presents CCU concept and some details of the design. It describes also the layout of CCUs installation in CESR, their performance and characteristics. The current status of operation and future plans are discussed as well. Presently, at CHESS two CCU magnets provide radiation for 5 out of 11 experimental stations.

We describe operation of the CESR-TA vertical beam size monitor (xBSM) with er. beams with E-b=4 GeV. The xBSM measures vertical beam size by imaging synchrotron radiation x-rays through an optical element onto a detector array of 32 InGaAs photodiodes with 50 pm pitch. The device has previously been successfully used to measure vertical beam sizes of 10-100 mu m on a bunch-by-bunch, turn-by-turn basis at e(+/-) beam energies of similar to 2 GeV and source magnetic fields below 2.8 kG, for which the detector required calibration for incident x-rays of 1-5 keV. At E-b =4.0 GeV and B=4.5 kG, however, the incident synchrotron radiation spectrum extends to similar to 20 keV, requiring calibration of detector response in that regime. Such a calibration is described and then used to analyze data taken with several different thicknesses of fillers in front of the detector. We obtain a relative precision of better than 4% on beam size measurement from 15 to 100 pm over several different ranges of x-ray energy, including both 1-12 keV and 6-17 keV. The response of an identical detector, but Lilted vertically by 60 in order to increase magnification without a longer beamline, is measured and shown to improve x-ray detection above 4 keV without compromising sensitivity to beam size. We also investigate operation of a coded aperture using gold masking backed by synthetic diamond. (C) 2015 Elsevier B.V. All rights reserved.

Novel methods & materials have been used to produce dynamically bent Bragg diffraction analyzer crystals in a modified von Hamos geometry.They are used to collect a wide energy range X-ray emission spectrum over a large solid angle with electron volt resolution. Crystals fabricated from silicon-on-insulator wafers by photolithography and deep reactive ion etching can bend to 10 cm radius without increased lattice strain. The design permits adjustment of energy dispersion for individual analyzers in an array. A multilayer mono, mono-capillary focusing, and multi-crystal spectrometer together collect signals at a bend magnet beamline comparable to those from an undulator. Preliminary measurements validate this new energy dispersive spectrometer.

We describe the design and performance of optical elements for an x-ray beam size monitor (xBSM), a device measuring e+ and e− beam sizes in the CESR-TA storage ring. The device can measure vertical beam sizes of 10–100μm on a turn-by-turn, bunch-by-bunch basis at e± beam energies of ~2–5GeV. x-rays produced by a hard-bend magnet pass through a single- or multiple-slit (coded aperture) optical element onto a detector. The coded aperture slit pattern and thickness of masking material forming that pattern can both be tuned for optimal resolving power. We describe several such optical elements and show how well predictions of simple models track measured performances.

We describe the construction and operation of an X-ray beam size monitor (xBSM), a device measuring e+ and e− beam sizes in the CESR-TA storage ring using synchrotron radiation. The device can measure vertical beam sizes of 10–100μm on a turn-by-turn, bunch-by-bunch basis at e± beam energies of ~2GeV. At such beam energies the xBSM images X-rays of ϵ≈1–10keV (λ≈0.1–1nm) that emerge from a hard-bend magnet through a single- or multiple-slit (coded aperture) optical element onto an array of 32 InGaAs photodiodes with 50μm pitch. Beamlines and detectors are entirely in-vacuum, enabling single-shot beam size measurement down to below 0.1mA (2.5×109 particles) per bunch and inter-bunch spacing of as little as 4ns. At Eb=2.1GeV, systematic precision of ~1μm is achieved for a beam size of ~12μm; this is expected to scale as ∝1/σb and ∝1/Eb. Achieving this precision requires comprehensive alignment and calibration of the detector, optical elements, and X-ray beam. Data from the xBSM have been used to extract characteristics of beam oscillations on long and short timescales, and to make detailed studies of low-emittance tuning, intra-beam scattering, electron cloud effects, and multi-bunch instabilities.

We developed, built and beam tested a novel, compact, in-vacuum undulator magnet based on an adjustable phase (AP) scheme. The undulator is 1 m long with a 5mm gap. It has a pure permanent magnet structure with 24.4mm period and 1.1 Tesla maximum peak field. The device consists of two planar magnet arrays mounted on rails inside of a rectangular box-like frame with 156 mm × 146 mm dimensions. The undulator magnet is enclosed in a 273 mm (10.75") diameter cylindrical vacuum vessel with a driver mechanism placed outside. In May 2012 the CHESS Compact Undulator (CCU) was installed in Cornell Electron Storage Ring and beam tested. During four weeks of dedicated run we evaluated undulator radiation properties as well as magnetic, mechanical and vacuum properties of the undulator magnet. We also studied the effect of the CCU on storage ring beam. The spectral characteristics and intensity of radiation were found to be in very good agreement with expected. The magnet demonstrated reproducibility of undulator parameter K at 1.4 × 10−4 level. It was also found that the undulator K. parameter change does not affect electron beam orbit and betatron tunes.

We developed, built and beam tested a novel, compact, in-vacuum undulator magnet based on an adjustable phase (AP) scheme. The undulator is 1 m long with a 5mm gap. It has a pure permanent magnet structure with 24.4mm period and 1.1 Tesla maximum peak field. The device consists of two planar magnet arrays mounted on rails inside of a rectangular box-like frame with 156 mm x 146 mm dimensions. The undulator magnet is enclosed in a 273 mm (10.75”) diameter cylindrical vacuum vessel with a driver mechanism placed outside. In May 2012 the CHESS Compact Undulator (CCU) was installed in Cornell Electron Storage Ring and beam tested. During four weeks of dedicated run we evaluated undulator radiation properties as well as magnetic, mechanical and vacuum properties of the undulator magnet. We also studied the effect of the CCU on storage ring beam. The spectral characteristics and intensity of radiation were found to be in very good agreement with expected. The magnet demonstrated reproducibility of undulator parameter K at 1.4 x10 -4 level. It was also found that the undulator K parameter change does not affect electron beam orbit and betatron tunes.

X-ray beam size monitors at SuperKEKB must withstand high, sustained incident power loads. Two prototype optics elements were fabricated and tested at CesrTA, using incident X-ray power densities comparable to those expected at the SuperKEKB LER. One element was based on a silicon substrate, the other a CVD diamond substrate, with each substrate supporting a coded aperture mask pattern in gold on its surface. The diamond substrate mask showed superior performance to the silicon substrate mask, with the the mask pattern on the silicon substrate melting at the highest incident power level tested, where the diamond-substrate mask survived. We will present here the high-power test results, along with a simulation of X-ray power absorption and heat transfer in the two prototype elements, and the resulting implications for the design of the optics, beam line and heat sink for SuperKEKB. INTRODUCTION X-ray beam size monitors will be used at SuperKEKB to measure low-emittance bunch profiles in both the Low Energy Ring (LER) and High Energy Ring (HER).[1][2] The total beam currents in the LER and HER will be 3.6 A and 2.6 A respectively, which creates a high average SR power load on the optics elements. Two types of prototype optics elements (URA masks[3]) have been fabricated and are being used for both imaging tests and high-power burn tests at CesrTA. The latter tests are described here. Table 1: Mask Parameters Parameter Silicon mask Diamond mask Substrate material 625 m monocrystalline silicon 350 m polycrystalline CVD diamond Buffer material 5 nm Cr 100 nm Cr Mask material 18.75 m Au 8.7 m Au Mask pattern 31 x 5 m URA 59 x 10 m URA The two types of mask tested are described in Table 1. The first type tested, made by NTT-AT Nanofabrication, was made of 18.75 m-thick gold on a 625 m-thick silicon substrate. To prevent the gold layer from peeling away from the silicon due to differences in thermal expansion coefficients, there is a 5 nm buffer layer of chromium between the gold and silicon. There is no chromium in those parts of the mask pattern where there is no gold. The second type of mask, made by Cornes Technology, is made of 8.7 m gold on a 350 m CVD diamond substrate, with a 100 nm Cr buffer layer. BURN TESTS The burn tests were carried out at the CesrTA D Line, using 5.3 GeV beam. In the first test carried out, the silicon-substrate mask was exposed to 200 mA of beam (corresponding to the heat load expected at the LER) for approximately 5 minutes, after which it was removed from the beam line and examined, as shown in Fig. 1; at this stage, it appears completely unharmed. Figure 1: Au+Si mask after exposure to 200 mA beam. Figure 2: Au+Si mask after exposure to 243 mA beam. To determine what margin of safety there may be, the same mask was re-installed in the beam line, and exposed to ~20% higher beam current, 243 mA. Figure 2 shows the result of this second burn test. The gold coating in the WEPF15 Proceedings of IBIC2013, Oxford, UK ISBN 978-3-95450-127-4 C op yr ig ht c ○ 20 13 by JA C oW — cc C re at iv e C om m on sA tt ri bu tio n 3. 0 (C C -B Y3. 0) 844 Beam Profile Monitors center of the chip has melted and slumped down. The Si substrate appears to be intact, and the Cr bonding layer between the gold and the silicon appears to be still attached to the Si. The center of the mask evidently reached at least the melting temperature of gold (1064°C), but not that of silicon (1414°C), or chromium (1907°C). Following this somewhat alarming result, a similar test was conducted on a prototype diamond-substrate mask. A factory-reject sample was used, with scratches and displaced traces. Photos of the Au+CVD diamond mask before and after exposure to beam currents up to 243 mA are shown in Fig. 3. No obvious damage resulting from the beam exposure is seen. Figure 3: Au+CVD diamond mask (factory-reject) before (left) and after (right) exposure to 243 mA beam. The 5.3 beam current was raised in steps, with beam image scans taken through the mask after each step at 4 GeV, as shown in Fig. 4. No change in the beam image through the mask is seen between the scans taken after exposure to 60 mA (left) and 243 mA (right) 5.3 GeV beam. Figure 4: Beam image scans taken through Au+CVD diamond mask (factory-reject ) after exposure to 60 mA (left) and 243 mA (right) beam. From the photographs and beam image scans, the Au+CVD diamond mask would appear to have survived the exposure to the maximum beam heat load possible at CesrTA. The superiority in high-power applications of the diamond-substrate mask over the silicon-substrate, due to the higher thermal conductivity of diamond than that of silicon, was clearly demonstrated. SIMULATIONS ANSYS simulations were carried out to replicate the burn test results, and to evaluate how much margin is gained by using a diamond-substrate mask. The mask is mounted flush with a copper heat sink plate with a 2.38 mm diameter hole in it, through which synchrotron radiation is incident on the mask. A stainless steel mounting bracket is used to hold the mask in close thermal contact with the copper heat sink, and the bracket is mounted with stainless bolts that pass through the copper plate. The assembly is water cooled at the center of the top surface. A drawing of the holder assembly is shown in Fig. 5; the simplified model of the assembly used in simulation is shown in Fig. 6. The total incident power on the mask and its holder was calculated according to P[W/mr/A]=13.8Ee[GeV]/[m] (Ref. [5], Eq. 3.39, e.g.), and modeled as a heat source on the surface of the mask and holder in the form of a horizontal ribbon with a vertical width corresponding to the projection of the critical angle at the mask’s distance from the source. Figure 5: Heatsink mount for coded aperture at CesrTA. Figure 6: ANSYS simulation of temperature distribution over simplified model of heatsink mount. For the simulation, we used the temperature-dependent thermal conductivities of the mask and holder materials from the AIST material properties database [6]. The thermal conductivity data for diamond are only available up to 400°C, so we fit the data (for the sample in the database with the lowest thermal conductivity) to an inverse power law in degrees Kelvin, and extrapolated the Proceedings of IBIC2013, Oxford, UK WEPF15 Beam Profile Monitors ISBN 978-3-95450-127-4 845 C op yr ig ht c ○ 20 13 by JA C oW — cc C re at iv e C om m on sA tt ri bu tio n 3. 0 (C C -B Y3. 0) thermal conductivity data out to 1500°C, following the model of phonon-phonon scattering dominance for thermal conductivity dependence in that regime [7]. Figure 7: ANSYS simulation of temperature distribution of gold layer on silicon substrate. Figure 8: ANSYS simulation of temperature distribution of silicon substrate behind gold layer. Figure 9: ANSYS simulation of temperature of stainless steel retaining bracket behind substrate, and bolts. Figure 7 shows the temperature distribution of the gold layer of the Au+Si mask for an incident SR power equivalent to 243 mA at 5.3 GeV at CesrTA. Figure 8 shows the temperature of the underlying Si substrate, and Fig. 9 shows the temperature distribution over the stainless steel retainer bracket and mounting bolts. Finally, Fig. 10 shows the temperature of the gold layer of the Au+CVD diamond mask under the same incident heat load. The peak temperature of Au on the Si substrate reaches 863°C, close to melting temperature of Au of 1064°C. The peak temperature of the Au on the CVD diamond substrate, on the other hand, is 327 degrees lower at 536°C. As mentioned previously, the actual Au+Si mask must have reached a peak temperature between 1064°C and 1414°C. If the actual Au+Si mask temperature was near the upper bound of 1414°C, then from the difference in peak temperatures the Au+CVD diamond mask peak temperature might have come quite close to the melting temperature of gold. Figure 10: ANSYS simulation of temperature distribution of gold layer on diamond substrate. EFFECT OF HEAT-SINK SHIELDING Based on the above results, we simulated the effect of placing a slotted screen in front of the mount, to prevent SR power from being absorbed outside the mask aperture and raising the base temperature of the copper heat sink mount. The results are shown in Table 3. Placing a slotted shield in front of the holder so that the SR fan hits only the mask and not the copper heat sink lowers the peak temperature a further ~200-300 degrees. Table 3: Maximum Au Temperature Reached in Simulation Design Silicon mask Diamond mask Unshielded heat sink 863C 536C Shielded heat sink 586C 352C In addition to lowering the temperature of the mask itself, such shielding also keeps the stainless steel bolts, which otherwise reach peak temperatures of 1100 degrees under direct exposure to SR, from excessive temperature build-up. WEPF15 Proceedings of IBIC2013, Oxford, UK ISBN 978-3-95450-127-4 C op yr ig ht c ○ 20 13 by JA C oW — cc C re at iv e C om m on sA tt ri bu tio n 3. 0 (C C -B Y3. 0) 846 Beam Profile Monitors LESSONS AND PLANS FOR SUPERKEKB Table 3 shows a summary of SR heat loads for the burn tests conducted at CesrTA, and those expected for the SuperKEKB LER and HER. Two types of heat load are shown: the on-axis solid-angle power density in W/mr/A (calculated as per Ref. [5], Eq. 3.13 and following), and the line power density as described above. Their projections onto the mask in terms of peak W/mm, and integrated W of linear power, respectively, are also shown. For the LER, while the maximum beam current is much higher than that at CesrTA, the lower beam energy and greater distance to the optics mount gives an area power density the same as that seen at CesrTA at 200 mA, 23 W/mm. Based on the burn test, this should be an acceptable power density for the LER. Table 3: X-ray Beam Line Heat Loads at CesrTA and the SuperKEKB LER and HER Parameter CesrTA SuperKEKB

Motivated by the needs of Cornell High Energy Synchrotron Source (CHESS), we developed and are now constructing a novel type of planar undulator magnet called CHESS Compact Undulator. The device consists of two pure permanent magnet (PPM) arrays mounted inside of a rectangular box-like frame on linear bearings providing magnet array motion along the beam axis. The outside frame dimensions are 156 mm 148 mm. Undulator is 1 meter long. It has 5 mm constant gap. The magnetic field strength is controlled by the array longitudinal motion (adjustable phase scheme). Magnetic structure has 24.4 mm period and maximum peak field 1.1 Tesla. The PM material is Nd-Fe-B of 40UH grade. The undulator magnet will be enclosed in a 273 (10.75") mm diameter cylindrical vacuum vessel while the driver mechanism responsible for moving the magnet array will be placed outside. The following sections describe mechanical and magnetic design of undulator and present results of the magnetic field measurement and evaluation of vacuum properties.

beam dynamics in the presence of electron cloud effects. These new instruments are also required to develop low emittance beam conditions which are key to the success of Differential Position Accuracy Channel-to-Channel Sampling Time Accuracy BPM Tilt Errors (after correction) 10 μm 10 ps 10 mrad the damping ring design for the International Linear Collider. This poster will detail some of the architecture and tools which have been developed to support these efforts. Experiment Specific Instruments •Server/client code •Multi Language Functions •Lab wide networking •Hardware interface designs