Engineering topics
Publications and source records attributed to Peggs, Stephen.
Island tune spectra are introduced, beginning with curves of resonance trajectories plotted in the (Q x , Q I ) tune domain plane (where Q x is horizontal tune and Q I is island tune) for many resonance orders N. A slice in the (Q x , Q I ) plane delivers a Q I -spectrum at a fixed horizontal tune, with one spectral line for every appropriate value of N.
The Resonance Island Jump (RIJ) scheme for transition crossing in HSR (or the CERN SPS [1, 2]) is radically new, and untested. Beam experiments in RHIC will be necessary if the RIJ scheme is to be considered as a serious alternative to upgrading the first order linear jump scheme currently implemented in RHIC [3]. This note outlines the planning and the theoretical foundations that are necessary for such a beam experiment, perhaps to be performed in 2024.
Transition is crossed during acceleration in the Hadron Storage Ring (HSR) for all species other than protons. A first-order transition jump scheme manipulates the value of the optical quantity γ T and distorts the optics of the HSR around the time that the beam energy γ beam crosses the nominal transition value γ T 0 . The jump scheme in the Relativistic Heavy Ion Collider (RHIC) uses 48 jump quads, driven by 12 bi-polar power supplies that each drive 4 quads in series. Ten of those 48 are eliminated in the preliminary September 2022 HSR layout (EIC-HSR-220921a) that was developed from RHIC with no initial regard for transition crossing. This note analyzes the performance of the 38-quad HSR scheme, by comparison with the 48-quad RHIC scheme. It is concerned only with optics – the manipulation and response of Twiss functions and related quantities – and not with beams. This evaluation is a necessary first step before enhancing the transition jump scheme to restore RHIC performance in the HSR. Eventually full beam simulations of transition crossing need to be performed.
The effect of resistive wall instabilities on beam break-up is investigated for beam pipes with finite and infinite thickness walls. Analytic solutions for the electric field along the beam axis are determined for both finite and infinite thickness walls using complex analysis. The longitudinal and transverse impedances are also determined; variations from the infinite wall case are observed when the wall thickness is of the order of the skin depth of the beam pipe material.
Hysteresis was observed in the H1 dipole magnets in CBETA SX line. This study is the first step in order to understand and mitigate the effects this has on field reproducibility. Ansys Maxwell 19.2 was used to model the MS1DIP05 magnet with a hysteretic core. The parameter space for a degaussing procedure using the trim coils was explored. It was found that significant degaussing can be obtained with a decaying ringing waveform of magnitudes above 10 A, so long as the frequency ƒ and decay constant w were chosen such that ƒ >> w. The circuit properties of the magnet were simulated and explored using a sinusoidal voltage input. It was found that for low frequencies ƒ << 1 Hz, the magnet behaves as a series LR circuit with R = (28.3 ± 0.5) mΩ and L = (5.7 ± 0.1) mH. At higher frequencies, hysteretic effects become significant, and resistance becomes proportional to the frequency due to hysteretic losses. A drop and saturation in inductance for higher frequencies was observed in the simulation data. Numerical instabilities were found for excitations below small voltages V < 50 mV. Experimental results on degaussing and current corrections were collected with Hall probe measurements. Good qualitative agreement with simulation results was found, but due to B-H curve details quantitative comparisons were inaccurate. Two procedures, MaxOp and MaxNull, were designed to produce repeatable operating field values to within 0.1 G.
The initial shielding of the ERL dump was described in a note ”Notes on CBETA Dump Shield” dated March 5th, 2019. In that note the primary concern was to minimize the radiation downstream from the dump as the electron beam entered the FFAG arc on the East side of LOE. Accordingly, the entrance into the dump in LOE was not shielded. In retrospect this was a mistake as it was found that radiation emitted by the dump in the ”upstream” direction into LOE causes gamma radiation monitors to set off at electron currents > 10µA. Also, the actual shielding configuration in the previous note does not reflect the actual shielding as implemented and the calculations described below take this into account and include addition shielding at the entrance to the dump that was not present in the previous note.
This paper creates models of various components of CBETA and conducts simulations of electron beam irradiation to calculate dose rates on NdFeB permanent magnets. Using Geant4 simulation software, two models were created which red 42MeV beams and measured the resulting dosage of radiation on nearby permanent magnets. In the first model, the beam was incident on a viewscreen, and the maximum dose rate obtained on a nearby magnet was .97±.0007 Gy/μC which corresponds to 1910±3 hrs of continuous operation at a beam current of 150 pA, if the maximum allowed dose is 1 kGy. The second model is of a splitter section, and the results indicate that radiation dose rate increases as a function of the incident angle as the beam enters the 42MeV chamber line. In Model II, a half-quadrupole magnet experienced the highest dose rate at 0.0909±.0005 Gy/μC, corresponding to 20,400±140 hrs of operation at 150 pA until the magnet reaches a dose of 1 kGy. For this second model, it was also observed that vertical momenta proportionally decrease radiation dosage on the magnets.
This project develops a method of finding instrument settings that minimize the power load on accelerating radio-frequency (RF) cavities in an energy recovery linac (ERL). A general symmetry theory, which has been developed in detail and was previously tested in simple ERL models, will be used to optimize more detailed models of the Cor nell Brookhaven ERL Test Accelerator (CBETA). CBETA specifications are detailed in Fig. 1. Due to the high overlap with previous work, this report briefly summarizes the relevant outcomes of older ERL symmetry models; however, the reader is encouraged to consult for a more comprehensive coverage of these topics. This report will focus primarily on extensions of symmetry modeling toward more complex systems, including multi particle beams or CBETA-specific cavity tracking.
This work describes first commissioning results from the Cornell Brookhaven ERL Test Accelerator fractional arc test. These include the recommissioning of the Cornell photoinjector, the first full energy operation of the main linac with beam, as well as commissioning of the lowest energy matching beamline (splitter) and a partial section of the Fixed Field Alternating gradient (FFA) return loop featuring first production Halbach style permanent magnets. Achieving these tasks required characterization of the injection beam, calibration and phasing of the main linac cavities, demonstration of the required 36 MeV energy gain, and measurement of the splitter line horizontal dispersion and R 56 at the nominal 42 MeV. In addition, a procedure for determining the BPM offsets, as well as the tune per cell in the FFA section via scanning the linac energy and inducing betatron oscillations around the periodic orbit in the fractional arc was developed and tested. A detailed comparison of these measurements to simulation is discussed.
The dump is in a location that is inaccessible to the overhead crane so that it cannot be shielded by large shielding blocks. In this note, the radiation field produced by a 1 mA, 6 MeV electron beam collected by the dump is investigated and a possible shielding arrangement proposed. The radiation field was calculated using the code MCNP6.
The main goal of this study is to find the tolerance of the CBETA FFAG lattice subjected to various errors. Common error sources include non-ideal magnet positioning and alignment, undesired multipole field strengths, and BPM offsets. In order to fix the beam trajectories (orbits) affected by the errors, dipole kickers have been allocated throughout the FFAG beamline to serve as correctors. Since there are four coexisting orbits in the FFAG, the correction scheme must aim to correct all four of them simultaneously. The simulation is performed using Bmad, developed by Cornell University to model relativistic beam dynamics in customized accelerator lattices. The core of orbit correction is the singular value decomposition (SVD) optimization. As the error magnitudes increase, the beam size and emittance after correction could undesirably increase, and our goal is to find the "individual limit" of each error type beyond which the increases become unacceptable. Typically this limit is reached before the orbits become un-correctable. Similarly we can also find the "combined limit" of various errors. Both limits will be properly defined in the later section.
Geant4 (`Geometry and Tracking') is a C++ Monte- Carlo simulation toolkit developed by the Geant4 Collaboration under CERN and is used to describe inter- actions between particles and matter. It has been shown, through significant benchmarking against physical experiments and other software, to be remarkably accurate at simulating interactions between particles and matter. Currently, Geant4 is being used on world-wide projects such as ATLAS, CMS, Muon g-2, and ILC, largely due to being accurate, open-source, and written in a modern language. This openness and modernity is striking when compared to other Monte-Carlo simulation toolkits such as the Fortran toolkit MCNP (Monte-Carlo N-Particle code), which is restricted in acquisition and use by the US government. Despite barriers in the use of MCNP and having been written in an antiquated language, MCNP is still considered industry-standard for many fields, including radiation shielding. This paper details the development and validation of a radiation shielding simulation framework in Geant4 aimed at flexibility and ease of simulations with the goal of replacing MCNP scripts.
Our goal is to find which beam position (x; y) has predicted signals that are as close to the measured button signals $V_k$ as possible. To simplify the problem, exploiting the geometry of our buttons, the first thing we do is calculate the difference-over-sum of the horizontal and vertical buttons, and define these as the new target quantities.
Control of beam trajectory is essential in particle accelerators. The trajectory of a beam is sensitive to various types of imperfections that cause it to deviate from its desired path along the accelerator. Orbit correction is hence fundamental for the control of beam trajectory. In this work we aim demonstrate orbit correction using singular value decomposition techniques applied to distorted orbits generated in the CBETA-V Virtual Machine. In particular we present two concrete examples of such correction in different sections of the CBETA Fractional Arc Test layout. Our goal is to construct the backbone of an orbit correction software toolkit to be tested in the CBETA-V Virtual Machine and eventually implemented in the completed CBETA control system.
In May, 2018, the Cornell-University Brookhaven-National-Laboratory Energy-Recovery-Linac Test Accelerator (CBETA) fired its first electron beam tests using the first pass (S1) splitter line and first girder of the return loop (FA). At various locations on the S1 Line and FA section, the positions of the electron beam were recorded using ten Beam-Position-Monitors (BPMs) as the strength of the steering and focusing electromagnets were varied. Datasets from these tests contain measurements of beam position during magnet scans for the following beam momenta: 38, 42, 47, and 53MeV.
Halos in high energy particle beams require constant monitoring to minimise beam losses and damage to accelerator components. Four methods of measuring halos are described: wire scanner, laser wire scanner, diamond detector and halo masking monitors. Details of their effectiveness’ from tests at various accelerators are used to evaluate how useful they are. It is found that methods making use of wire and laser wire scanners are unable to achieve a high enough dynamic range to measure the outer halo, where the intensity of particles is very low. Diamond detectors can be used to determine the presence of a halo, but are unable to reconstruct their shape. The halo masking method, which makes use of a digital micromirror device to block out high intensity light, produces very good results, but it is possible that light is scattered inside the phosphor screen used to produce an image from the beam, skewing the measurements.
On April 18 2018, the CBETA project achieved technical milestone 6, “Fractional Arc Test: beam through the Main Linac Cryomodule and prototype girder ”, which is also a major funding milestone. For this test the electron beam was accelerated to a final energy of 42 MeV in the Main Linac Cryomodule (MLC), before being passed through the first splitter beam line S-1 and through the first permanent magnet arc girder (PMA). The MLC contains six superconducting accelerating cavities; for this test only five of the six were powered due to a temporary fault in cavity number 5.
The two common dipole magnets in each of the two CBETA splitter sections serve to direct the beams of 42, 78, 114 and 150 MeV from the linac or FFAG return loop exit into the separate splitter lines for path length adjustment and to direct the beams from the splitter into the linac or FFAG return loop entrance. Here we report the results of design and tracking studies based on the Opera V18R2 software. The choice of the H-magnet design included consideration of using the coil design developed for the internal splitter H-dipoles by Elytt Energy. Considerations of transverse space constraints and the bend angle of about 30 degrees for the 42-MeV beam led to the adoption of a T-shape for the pole. The XYZ magnet dimensions are 34 × 42 × 16 cm. The central field value is 3.6-3.9 kG, but the magnet can operate as high as 4.9 kG, allowing for tests of a beam energy as high as 54 MeV in the S1 splitter line. We include results of tracking studies showing optimal placement of the magnets based on bend angle uniformity for the four beam energies. The project now proceeds to the full engineering design and fabrication to be performed by Elytt Energy.