Session 3P - Superconducting Magnets 1, Superconducting RF, Control Systems W, Radiation Monitoring and Safety.
POSTER session, Tuesday morning, May 13
Pacific Ballroom,

[3P.01] Novel Design of Superconducting Helical Dipole Magnet

Superconducting helical dipole magnets with a nominal field of 4~Tesla are needed for the spin physics program at the Relativistic Heavy Ion Collider (RHIC) at Brookhaven National Laboratory. The magnets are required to operate at a relatively low current of 400~A since many of these magnets have to be independently controlled. The Advanced Magnet Lab, Inc., in Palm Bay, FL has designed and built two prototype magnets using advanced computer controlled coil winding technology. The AML design is extremely cost effective since it avoids magnet specific tooling despite the required complex coil pattern and any precision machined inserts or spacers. It is the first time an accelerator magnet of this technology has reached a field above 4~Tesla. Results from the prototype testing at BNL are presented.

[3P.02] Removal of Axial Twist in RHIC Insertion Quadrupole Magnets

The focusing triplets located on either side of the six interaction points of RHIC each consist of three 13cm aperture quadrupoles with magnetic lengths of 1.44m (Q1), 3.40m (Q2), and 2.10m (Q3). The field quality and alignment of these magnets are most critical to the performance of the accelerator. The maximum allowable axial twist of the cold mass, defined as the standard deviation in the quadrupole roll angle, is 0.5 mrad. This requirement has occasionally exceeded the capabilities of the assembly fixturing and the procedures used to complete the axial welding of the shell halves around the cold mass yoke. A corrective shell welding technique has been successfully employed to remove excessive axial twist of the 13cm quadrupoles. This ``custom straightening" method will be described along with the before and after mechanical inspection data. The magnetic results which confirm the untwisting procedure shall also be discussed.

[3P.03] Parameterization and Measurements of Helical Magnetic Fields

Magnetics fields with helical symmetry can be parameterized using multipole coefficients (\tildea_n,\tildeb_n). We present a parameterization that gives the familiar multipole coefficients (a_n,b_n) for straight magnets when the helical wavelength tends to infinity. To measure helical fields all methods used for straight magnets can be employed. We show how to convert the results of those measurements to obtain the desired helical multipole coefficients (\tildea_n,\tildeb_n).

[3P.04] A Common Coil Design for High Field 2-in-1 Accelerator Magnets^*.

This paper presents a common coil design concept for 2-in-1 superconducting accelerator magnets. It practically eliminates the major problems in the ends of high field magnets built with either high temperature or conventional superconductors. Racetrack coils, consisting of rectangular blocks built with either superconducting cables or tapes, are common to both apertures with each aperture containing one half of each coil. The two apertures are in the same vertical plane in an over-under geometry. A set of common flat coils are placed vertically on left and right side of the two apertures producing field in the opposite directions. The ends are easy to wind with the conductors experiencing little strain. The ends can be fully supported by a simple 2-d geometry to contain the large Lorentz forces. The overall magnet design, construction and tooling are also expected to be simpler than in conventional cosine theta magnets. The block design for high field magnets uses more conductor than the cosine theta design but is preferred for dealing with the large Lorentz forces in the body of the magnet. The concept is also suitable for a variety of other high field superconducting, moderate field superferric, multi-aperture and combined function magnet designs.

\vskip 5mm ^*Work supported by the U.S. Department of Energy.

[3P.05] Change in Field Harmonics after Quench and Thermal Cycles in Superconducting Magnets^*.

A change in field harmonics after quench and thermal cycle has been observed in superconducting magnets for the Relativistic Heavy Ion Collider (RHIC). This change has been systematically examined in a number of RHIC dipole and quadrupole magnets. Limited data are also available on prototype dipole magnets built for the Superconducting Super Collider (SSC). These changes in field harmonics may limit the achievable field quality and its reproducibility in superconducting magnets. The harmonic change in RHIC magnets corresponds to small changes in magnet geometry. The magnitude of the harmonic change may, however, depend on the details of the magnet design and construction.

\vskip 10mm ^*Work supported by the U.S. Department of Energy under contract No. DE-AC02-76CH00016.

[3P.06] MAGNETIC DESIGN OF SUPERCONDUCTING QUAD\-RU\-POLES FOR A SC LINAC FOR APT.

This paper describes the magnetic design for superconducting quadrupoles designed for the superconducting option of the proton linac for the Accelerator Production of Tritium (APT) project. The quadrupole magnets provide the focusing for a singlet FODO lattice used in the linac. The magnets have a 16 cm (13 cm) aperture in the hi-\beta (medium-\beta) section of the linac. The proposed design is a coil dominated magnet with an iron yoke at a larger radius to shield the magnetic field from other parts of the accelerator. An alternate iron dominated superconducting design is also analyzed. A 3D calculation of the magnetic field using the finite element program TOSCA has been performed. The field performance, field quality, peak field and forces at the conductor are discussed.

[3P.07] Test Results from the Completed Production Run of Superconducting Corrector Magnets for RHIC

The production of all 420 superconducting 8cm arc corrector magnets required for the Relativistic Heavy Ion Collider (RHIC) has been completed at Brookhaven National Laboratory (BNL). All have been tested at 4.35K for quench performance and at room temperature for magnetic field quality. In addition, magnetic field measurements at 4.35K have been done for a 20% sample of these. Summaries of the harmonic and quench test results are presented. A comparison of the quench performance for magnets made with different coil preload pressures is also shown.

[3P.08] RHIC D0 Insertion Dipole Design Iterations During Production^*.

Iterations to the cross section of the Relativistic Heavy Ion Collider (RHIC) D0 Insertion Dipole were made during the production of the first six cold mass units. This was included as a part of the production plan because no Ramp;D or pre-production magnets were built prior to the start of production. The first magnet produced had the desired coil pre-stress and low field harmonics in the body of the magnet and is therefore being used in the RHIC Machine. On subsequent magnets, iterations were carried out to minimize the iron saturation and to compensate for the end harmonics. This paper will discuss the details of the iterations made, the obstacles encountered, and the results obtained. Also included will be a brief summary of the magnet design and performance. \vskip 5mm ^*Work supported by the U.S. Department of Energy under contract No. DE-AC02-76CH00016.

[3P.09] Helical Dipole Magnets for Polarized Protons in RHIC

The Brookhaven Relativistic Heavy Ion Collider (RHIC) will be able to support experiments using polarized proton beams. Siberian Snakes are used to maintain polarization in this high energy superconducting collider. To make efficient use of available space while taking advantage of high field superconducting magnets, 4~Tesla helical dipole magnets will be used. These magnets generate a central dipole field in which the field direction rotates through 360^\circ about the longitudinal axis over the length of the device. An arrangement of four such magnets can produce the desired change in the spin direction while keeping the proton orbit outside of the ``Snake'' unaltered. Similar magnet arrangements will be used to produce longitudinal polarization at the two major interaction points in RHIC. The basic requirements and layout of these magnets are described, as well as tolerances on field quality and integrated field strengths. First results of tests of prototype helical magnets will be discussed.

[3P.10] A Helical Magnet Design for RHIC^*.

Helical dipole magnets are required in a project for the Relativistic Heavy Ion Collider (RHIC) to control and preserve the beam polarization in order to allow the collision of polarized proton beams. The project requires superconducting magnets with a 100 mm coil aperture and a 4 Tesla field in which the field rotates 360 degrees over a distance of 2.4 meters. A design restraint is that the magnets operate at relatively low current (less than 500 amperes) in order to minimize the heat load from the current leads. A magnet has been developed that uses a small diameter superconducting cable wound into helical grooves machined into a thick-walled aluminum cylinder. The design and test results of this prototype magnet will be described. \vskip 10mm ^*Work supported by the U.S. Department of Energy.

[3P.11] Dynamic Field Quality of LHC/Saclay Arc Quadrupole Magnet Prototype

As part of the magnet Ramp;D program for the Large Hadron Collider (LHC), CEA Saclay has designed and built two 3~m long, 56~mm twin-aperture arc quadrupole magnet prototypes. Extensive field quality measurements were taken during the cold-test of the second prototype using a system based on a rotating coil array. The measurement sequence included measurements as a function of ramp rate. After describing the method used to analyze the rotating coil data and correct for the effects of time-varying field, a review of the field quality data as a function of ramp rate is presented. The observed field distorsions are explained in terms of inter-strand coupling currents arising from low and non-uniform resistances at the cross overs between the Rutherford-type cable strands. Last, the distribution of crossover resistances within the quadrupole magnet coil is determined.

[3P.12] State of the LHC Main Magnets.

The main features of the dipole magnet design have been frozen in 1996 and important steps for the preparation of their series production are being taken in the current year. To finalize the technical specifications of the superconducting cables and other components, a number of detail variants are being validated with the construction and test of short and long magnets. Thus, beside a number of 1~m long models, four 10~m long models of the main dipoles and two 14.2~m prototypes are being assembled in industry and at CERN. The fabrication of a further set of 3 full length dipoles is also starting in industry to verify the reproducibility of production and performance. The lifetime and fatigue test of the String Test Facility, consisting of three dipoles and one quadrupole and simulating the basic periodic cell of the LHC, has been successfully concluded. The String was repetitively cycled between the injection field of 0.6~T and the operational field of 8.4~T, 24 hours per day, and has accumulated more than 2100 cycles, corresponding to about ten years of machine operation. The final design of the main quadrupoles, based on the use of the dipole coil outer layer cable, is being worked out by CEA, Saclay, and the construction of two new prototypes has been launched.

[3P.13] Mechanical Design of the LHC Standard Half-Cell

The LHC Conceptual Design Report issued on 20th October 1995 (CERN/AC/95-05 (LHC) - nicknamed "Yellow Book") introduced significant changes to some fundamental features of the LHC standard half-cell, composed of one quadrupole, 3 dipoles and a set of corrector magnets. A separate cryogenic distribution line was introduced, which was previously inside the main cryostat. The dipole length has been increased from 10 to 15 m and independent powering of the focusing and defocusing quadrupole magnets was chosen. Individual quench protection diodes were introduced in magnets interconnects and many auxiliary bus bars were added to feed in series the various families of correcting superconducting magnets. The various highly intricate basic systems such as: cryostats and cryogenics feeders, superconducting magnets and their electrical feeding and protection, vacuum beam screen and its cooling, support and alignment devices have been redesigned, taking into account the very tight space available. These space constraints are given by the necessity to have maximum integral bending field strength for maximum LHC energy, and the existing LHC tunnel. Finally, cryogenic and vacuum sectorisation have been introduced to reduce downtimes and facilitate commissioning.

[3P.14] The Short Straight Sections for the LHC

During more than five years a close collaboration between CERN and CEA Saclay led to the development and construction of two prototype quadrupole magnets and the integration of one of them into the short straight section of the LHC half-cell test string at CERN. In the frame of the special host country contribution to the LHC project this collaboration has been extended to the CNRS laboratory in Orsay and covers besides the quadrupole magnets the complete cold mass assembly (CEA) and the integration into the short straight section cryostat (CNRS). The short straight sections include not only the main lattice quadrupoles with their protection diodes, they also house different corrector magnets and the beam position monitors. Further, they provide the cryogenic feed units for a half-cell with all the magnet interconnections and the jumper connection to the separate cryo-line. The paper will show the general lay-out of these complex units and elaborate the different aspects of their assembly.

[3P.15] Towards Series Measurements of the LHC Superconducting Magnets

The development phase of the superconducting dipole magnets for LHC is being completed. Extensive power tests of these magnets required the development of new techniques to study the quench and training behaviour. The related equipment will be adapted for the reception tests of the future series magnets. Extensive magnetic measurements of short and long model dipoles have allowed to understand and quantify the time dependent behaviour of the field quality during the current flat top needed during beam injection. The main dipole field is only 0.54 at injection T for a nominal field of 8.3 T at collision energy. The experience gained is employed for the design of the measuring benches and tests stations presently under construction for the series measurements of all types of LHC magnets. The economically important issue of how many magnets have to be measured in the superconducting state at 1.8 K will be addressed in view of the field quality required for the performance of the LHC.

[3P.16] Forces on Interaction Region Quadrupoles and Dipoles Due to a Detector Solenoid Magnet

Several interaction region accelerator magnets are being built partly or wholly inside a high energy physics detector solenoid field. The forces on the coils due the solenoid field can be very large. Forces due to misalignments, given the proximity of the detector steel, can also be large. In this paper we derive several generally applicable formulae to calculate forces due to the solenoid, discuss symmetries, and evaluate the results for the CESR Phase III interaction region magnets. Forces due to nearby steel were calculated using ANSYS and the results are presented.

[3P.17] The Superconducting Interaction Region Magnet System for the \urllinkCESRhttp://www.lns.cornell.edu Phase III Upgrade

Deep inside the CLEO III detector will be two pairs of new high gradient superconducting interaction region magnets. Their close proximity to the interaction point, short focal lengths and large warm bore diameter are all optimized for the highest possible luminosity using crossing angles and bunch trains. Superposed with the main quadrupole fields are fields from skew quadrupole coils for solenoid coupling compensation, and fields from horizontal dipole coils for correction of small vertical misalignments of the magnetic centers. The magnet cryostats are held by active positioning systems for beam-based mechanical alignment. The magnet system is being designed and manufactured by Tesla Engineering Ltd of England.

[3P.18] Choice of Configuration for the DIAMOND Superconducting Dipole Magnet.

The current design for the proposed UK third generation synchrotron source DIAMOND has thirty two dipole bending magnets. Whilst the majority of these will be of conventional design, the option of including a small number of superconducting magnets, to provide higher energy X-ray radiation down selected beam-lines, is currently under consideration. The required field strength of 4.35 T is modest compared to the demands of high energy proton machines, but the large horizontal to vertical aperture ratio, needed to allow the fan of radiation to exit the magnet, places unusual constraints on the design. To meet these requirements, a number of possible basic configurations for the magnet have been studied. These include a straight cylindrical arrangement, with cosine theta coil distribution and cold iron, and a quasi-rectangular Helmholtz pair configuration, with warm iron. The use of a split solenoid has also been considered. The paper details these feasibility studies, indicating the likely choice of configuration for the final engineering design.

[3P.19] Niobium Quarter-wave Cavity for the New Delhi Booster Linac

This paper reports the completion of development of a 97 MHz niobium coaxial quarter-wave cavity to be used in a booster linac for the New Delhi 16UD pellatron electrostatic accelerator. A prototype cavity, which incorporates a niobium-bellows tuning device, has been completed and operated at 4.2 K at accelerating gradients above 4 MV/m for extended periods of time. Operational characteristics will be discussed.

[3P.20] The Higher-order Mode Damping Concept of the LHC 400 MHz Superconducting Cavities

Each beam of the LHC has its own accelerating structure of eight single-cell 400 MHz cavities connected with large-aperture beam tubes. Higher-order modes have to be damped sufficiently in order to prevent coupled-bunch instabilities and to limit parasitic mode losses. The first two higher-order modes do not propagate into the beam tubes. However, strong damping can be obtained with a special coupler resonant at both modes. Because of restricted space, a compact design is used. The other higher-order modes propagate and form coupled modes with unequal field distributions. They are damped by wideband couplers positioned on either side of each cavity cell. The cavities have a mode at about three times the fundamental frequency which does not propagate although its frequency is above the cut-off frequency of the beam tubes. We discuss how the trapping effect depends on cavity cell length and beam tube radii and that by an appropriate choice the dangerous minimum damping situation can be avoided. We present the designs of the higher-order mode couplers together with external Q measurements on a real cavity model.

[3P.21] Development of Superconducting RF for CESR

After the successful CESR beam test of August 1994 the continued development of a superconducting RF system for the CESR Luminosity Upgrade is in progress at the Laboratory of Nuclear Studies, Cornell University. The system description as well as recent results are presented.

[3P.22] A Superconducting RF System for DIAMOND.

The proposed UK Light Source DIAMOND has an operating energy of 3 GeV, with a current of 300 mA. With the full complement of insertion devices the energy loss per turn is greater than 2.25 MeV, and the required RF energy acceptance is greater than 3percent, consequently an accelerating voltage of 5.1 MV is needed. Investigations using a warm cavity system for DIAMOND have been reported elsewhere (The DIAMOND RF System, D.M.Dykes, Proceedings of the 5th European Particle Accel. Conf., Sitges 1996, pp 1943-5.), this paper discusses the advantages of using a superconducting RF system, and outlines a possible system arrangement that fulfils the requirements.

[3P.23] Note on the SC Linear Collider TESLA Cavity Design

The experience we have gained over the last few years from experiments with superconducting cavities for the TESLA test facility justifies a revision of the design decided almost five years ago. The proposed new design takes advantage of the high quality factor Q_0 > 10^10 and the low electron emission as demonstrated by some tested cavities. The main aim of the new design is to simplify the production and preparation of sc cavities and thus to reduce the cost of the linear collider. The new cavity shape has an enlarged iris diameter with the following advantages: significant lower loss factors, simplified and less expensive scheme for the HOM damping, suitability of hydroforming and higher stability of the field profile.

[3P.24] The Fermilab Radiofrequency Power Coupler for the TESLA Test Facility

This paper summarizes the work at Fermilab on radiofrequency power couplers for the TESLA Test Facility in Hamburg, Germany. The coupler is required to transport 210 kW of 1.3 GHz rf power at a pulse length of 1.3 ms and a repetition rate of 10 Hz during normal accelerator operation. The coaxial coupler features two ceramic rf vacuum windows, bellows for coupling adjustment and to accommodate thermal shrinkage, and copper-plated stainless-steel construction. Couplers have been produced at Fermilab in quantities sufficient to complete the injector capture cavity and the first eight-cavity accelerating module. Results of first beam tests shall be reported if possible.

[3P.25] Superconducting Cavity for BTCF

The RF system design for BTCF (Beijing Tau-Charm Factory) is dominated by its relative high RF voltage and low beam power requirements. In this paper the merits of using superconducting cavity in BTCF are discussed by comparing with using normal conducting cavity. The existed B-Factory Superconducting Cavities are considered as candidates for BTCF, its corresponding characteristics and hardware capabilities are also analyzed.

[3P.26] Mechanical Stabilisation of Superconducting Quarter Wave Resonators

A mechanical device, which performs significant reduction of the quality factor of mechanical vibration modes which can prevent the phase stabilisation in superconducting quarter wave resonators, was designed, constructed and tested. This device can eliminate the necessity of fast tuners and related electronics, resulting in a significant simplification and cost saving both in the construction and in the operation of superconducting low beta cavities.

[3P.27] Beam Test of a Superconcucting Damped Caviy for KEKB

For the feasibility study of a superconducting damped cavity to KEK B-factory, a prototype module was constructed and tested in TRISTAN Accumulation Ring. This module stored the beam current of 500mA with a cavity voltage of 1 - 2MV, and 350mA with 2.5MV(10.3MV/m). These currents were limited by heating up of other ring components. The peak current of 573mA was achieved in 16 bunches with 1.2MV. A pair of HIPped ferrite dampers suppressed HOM modes sufficiently and absorbed a power of 4.2kW in maximum. A coaxial type input coupler supplied the RF power of 280kW to the cavity under perfect reflection. The maximum power transferred to the beam was 160kW.

[3P.28] Cryogenic High-Conductivity Cavity

A refrigerated cavity saves a considerable amount of driving power owing to the cryo enhancement in conductivity. An L-band copper cavity is expected to have about six times the room-temperature Q factor at 40 K, which is nowadays easily attainable by a compact refrigerator, and therefore reduce the required power to one sixth. The reduction in operation power enables the use of solid state amplifiers as a compact and handy power source. The combination of the cryogenic high-conductivity cavities and the solid state amplifiers, with each cavity fed separately by each amplifier, provides a flexible accelerating system applicable to a wide range of particle velocities. A post-accelerator for deuterons has been proposed as the first pilot system employing individually driven cryo cavities, which should serve to realize a simultaneous acceleration of protons and deuterons in the KEK PS injector. A 400 MHz copper cavity, equivalent to each of the unit cavities composing the post-accelerator, passed a test operation at liquid nitrogen temperature with a 20 Hz-repetition 0.3 ms-width pulsed power up to the peak value 5 kW giving rise to the specified voltage. It will be subjected to further operation at lower temperature.

[3P.29] Beam Test Results on HOM Absorber of Superconducting Cavity for KEKB

In 1996, three series of beam tests were performed on the superconducting cavity for KEKB in TRISTAN Accumulation Ring. The cavity had two higher-order-mode absorbers made of ferrite on the upstream and downstream sides. In the tests, the absorbed power went up to 4.2 kW in total without problems. There were no changes on absorbing characteristics during the beam tests. Loss factor of the system was measured at 1.5 to 9 cm bunches and was in good agreement with theoretical predictions. Details of the tests and results will be shown.

[3P.30] Higher Order Mode Analysis of the APT Superconducting Cavities

In another contribution to this conference \footnote ``Superconducting Cavities for the APT Accelerator'', by Frank L. Krawczyk et al, LANL the design of superconducting cavities for relatively low velocity proton beams has been reported. Besides an optimization of the rf properties of the accelerating \pi-mode, other modes, possibly excited by the traversing proton beam, need to be regarded. The full spectrum of modes in \beta=0.64 and \beta=0.82 5-cell cavities has been calculated up to a frequency of approximately 2.5 GHz. These have been evaluated for their potential to affect the beam. The presence of ``trapped'' modes has also been investigated. In addition to the specific mode spectrum, the total power deposited into the cavities by the beam has been determined from the wakefields. Due to the operation with beams below the velocity of light extreme care was required to prevent incorrect results by wave reflections from the boundaries of the calculation volume. The simulations indicate that a power deposition of up to 17W per cavity can be expected. This power has to be removed by higher order mode couplers, which is a technically simple task. Transporting this power out to a room temperature dump does not even noticably increase the requirements to the cryogenic system.

[3P.31] In-Situ Proton Irradiation and Measurement of Superconducting rf Cavities under Cryogenic Conditions

The Accelerator Production of Tritium (APT) Project is investigating using a superconducting linac for the high-energy portion of the accelerator. As this accelerator would be used to accelerate a high-current (100-mA) CW proton beam up to 1700 MeV, it is important to determine the effects of stray beam impingement on the superconducting properties of a 700-MHz niobium cavity. To accomplish this, two 3000-MHz elliptical niobium cavities were placed in a cryostat, cooled to 2 K in sub-atmospheric liquid helium, and irradiated with 798 MeV protons at up to 490 nA average current. The beam passed through the equatorial regions of both cavities in order to maximize sensitivity to any changes in the superconducting surface resistance. After irradiation with 6x10e16 protons, the cavities were warmed to 250 K, then recooled to investigate the effects of a room-temperature annealing cycle on the superconducting properties of the irradiated cavities. These results are important to employing superconducting RF technology to future high-intensity proton accelerators for use in research and transmutation technologies.

[3P.32] Superconducting Cavities for the Reduced-Beta Section of a Proton Linac.

A proton linac with energies beyond 1.0 GeV has been proposed in (C. Rubbia and J.A. Rubio, "A Tentative Program Towards a Full Scale Energy Amplifier", CERN/LHC/96-11 (EET), Geneva 15th July 1996). In this proposal the LEP2 superconducting cavities should be used, after decommissioning of LEP, in the high energy part of the linac. Energies ranging from 100 MeV to 1.0 GeV should be covered by superconducting cavities designed to accelerate protons with reduced value of beta. The basic idea is to keep the shape of the spherical LEP cavities for beta = 1.0 and to adjust their length, for beta < 1.0, to get optimal acceleration for "nominal energy" protons with beta = beta0. Three different values of beta0 allow to cover the required range with a reasonable acceleration efficiency. Hence the design of superconducting cavities for beta0 = 0.50, 0.625 and 0.80 is presented here. In particular the choice of an elliptic iris profile is discussed.

[3P.33] New Window Designs for CEBAF Energy Upgrade

As the Jefferson Laboratory upgrades the existing CEBAF SRF electron accelerator to operate at higher energies, the fundamental power coupler windows will be required to operate with lower RF dissipation and increased immunity to radiation from cavity field emission. New designs and modifications to existing designs are described which can achieve these goals.

[3P.34] Improvement of the Operational Performance of SRF Cavities via In Situ Helium Processing and Waveguide Vacuum Processing

The useful performance range of the superconducting rf (SRF) cavities in the CEBAF accelerator at Jefferson Lab is frequently limited by electron field emission and derived phenomena. In addition, numerous cavities cannot be used at their highest gradients due to instability of the input waveguide vacuum. Improvements are required to support future operation of the accelerator at higher than 5 GeV. We report our progress to date toward sustaining higher performance of the installed SRF cavities. Several installed and operational cavities have been successfully processed to higher useful operating gradients via rf-helium processing while on the beamline. Progress against field emission was evidenced by improved high-field Q, reduced x-ray production and greatly reduced incidence of arcing at the cold ceramic window. There was no difficulty reestablishing beamline vacuum following the processing. Cavities previously limited to 4-5 MV/m are now operating stably at 6-8 MV/m. We report details of the process and progress toward maximizing its efficiency. By applying a pulsed-rf processing technique, we have also improved the pressure stability of the thermal transition region of the input waveguide for several cavities. This has produced durable improvement of the stability of the waveguide vacuum.

[3P.35] Jefferson Lab IR FEL Cryomodule Modifications and Test Results

The Infrared Free Electron Laser being constructed at the Thomas Jefferson National Accelerator Facility will require a 42 MeV, 5 mA electron accelerator. The accelerator design requires a 10 MeV injector and a two pass 32 MeV linac, one pass for acceleration and one pass for energy recovery. In order to minimize the cost of the linac, standard CEBAF 1497 MHz Superconducting Radio Frequency cavities and cryomodules are being used with minimal changes. Two SRF cavities, housed in a quarter cryomodule, operate at a nominal 10 MeV/m to provide the injector energy. The linac is composed of one cryomodule, housing eight SRF cavities operating at an average gradient of 8 MeV/m. The modifications to the cryomodule are being made to handle the higher beam current, to improve RF control, and to increase machine reliability. The modifications to the Higher Order Mode loads, cavity tuners, cavity beam line, warm and cold RF windows, and cryogenic shield are described. Test results from the injector quarter cryomodule are also presented.

[3P.36] Innovative Aspects of the SDL Control System

The Source Development Lab at BNL consists of a 230 MeV electron linac and 10m long wiggler for short wavelength FEL development. The control system is based on that in use at the NSLS. Two new extensions of the control system using VXI equipment are described. The first extension is the replacement of patch panels and lab oscilloscopes to monitor RF equipment. Instead, the RF waveforms are fed through a multiplexor into VXI digitizers. The waveforms can then be monitored remotely on any control console. The second extension is the replacement of the analog RF hardware needed to process beam position monitor signals. A digital system based on very fast (sub-nanosecond) VXI waveform digitizers is under development. The difficult operations requiring precise time alignment are then done in software.

[3P.37] Control and Operational Models for Vacuum Equipment

Operational models which describe the behaviour and the physical values associated with the vacuum equipment as seen by an operator have been studied for some time at CERN. Recently, they have been completed by control models, which define in a formal way the data structures required to access the physical values described in the operational model. The control models also define the operations that an application program has to send to the vacuum equipment to modify its state. Object Modelling Techniques (OMT) have been used to formalise the description of the models. In order to test the validity of the concepts, we have made a working prototype in the LEP accelerator. This prototype is being built on top of the CERN SL-Equip equipment access package and uses the "cdev" C++ library, developed at TJNAF, for the interface to application programs. SL-Equip is used for data transmission between front-end computers and vacuum equipment. We use the "cdev" networking facilities to communicate between the workstation and the front-end computers, and the "cdev" generic server as the framework for implementing the vacuum controls software. These packages were used in order to minimise the required software investment, but also to prove that these models are hardware and software independent.

[3P.38] PC-Based Applications Programming on the SRS Control System.

The CERN PC-based ISOLDE control system has been installed at the SRS electron storage ring at Daresbury Laboratory. The use of Windows NT for the control consoles together with PC and VME front-end computers running under several operating systems has resulted in a flexible and reliable system for accelerator control. The implementation and philosophy of control application programs, based around a suite of Microsoft Visual Basic and Excel programs, is described. In particular, the use of Excel to provide adaptable programs online allows rapid generation of new control functions; orbit correction and servoing at the application level are described as examples of this.

[3P.39] The Usage of Transient Recorders in the Daily HERA Machine Operation

Many parameters of HERA machine components like RF systems or quench protection as well as important beam parameters are continuously measured by transient recorders. In general, these recorders are not synchronized among one another and sample the data with very different rates ranging from 200 Hz to 50 MHz. At present, work is going on to integrate the existing transient recorders into a global system. The arcticle reviews the transient recorder hardware in operation at HERA. In addition, the proposed trigger distribution based on the HERA Integrated Timing system as well as the software concept to archive, retrieve and display the data will be described.

[3P.40] A High Reliability Accelerator Control System

This paper describes the control system developed at IUCF for the Cooler Injector Synchrotron (CIS). The Hardware system is VME based and employs fiber optic data transmission for high noise rejection. The hardware includes several modules designed and manufactured at IUCF to meet specifications not attainable with commercial hardware. These modules feature active redundancy with automatic switch over for high reliability; built in test and self diagnosis with centralized failure and system health monitoring for rapid maintenance; and very low drift and self calibration for maximum repeatability. We describe several modules including high precision ramping and non-ramping DAC/ADCs and a deep memory timing sequencer. We give a brief overview of the software system, which is based on Vsystem from Vista, Inc.

[3P.41] Nuclotron Control System

The superconducting synchrotron named Nuclotron based on a miniature iron-shaped field SC-magnets was put into operation at the LHE JINR in 1993.The Nuclotron Control System (NCS) project,which is still under development,started in 1992 and has provided efficient support for the machine commissioning through all its phases.This paper presents the current status of the NCS. The control system architecture is hierarc- hical in nature and consists of two physical levels. High performance workstations,together with a general purpose server computers, are used at the top level.Workstations act as an operator consoles,while the servers provide massive disk data storage,printing utilities,a common database, program library and data exchange between Nuclotron and its experiments. The front-end level comprises as industrial com- puters equipped with I/O boards and data acquisition modules, as in- telligent CAMAC crate-controllers with embedded micro-PCs. NCS is distributed system,in which subsytems geographically separated by as much as 500 m.The total number of computers presently installed is 25. An Ethernet Local Area Network,which runs IPX/SPX and TCP/IP communi- cation protocols ,connects the console computers to the front-end le- vel and physicists workstations.

[3P.42] Use of a Virtual Accelerato for a Development of an Accelerator Control System

Concept of Virtual accelerator is introduced to develop beam control application programs in the KEKB accelerator control system. A virtual accelerato is a computer process which simulates behaviour of a beam in an accelerator and responds to the accelerator control program under development in a same way as an actual accelerator do. Users of the virtual accelerator can test their control algorithm and user inferface without affecting the operation of the accelerator.

EPICS(Experimental Physics and Industrial Control System) jointly developed by LANL and APS/ANL will be used as a base of KEKB accelerator control system. In the EPICS based control system, a device is represented as a collection of records in a EPICS runtime database. A control program on a host computer can access devices in the system only through names of record fields, called `channel'. This abstraction allows us to construct a virtual accelerator control system.

Channel access interfaces was implemented into the modeling program SAD to realize this idea. Sample application of the method will be also presented.

[3P.43] Operation Logging System using Database for Synchrotron Radiation Beam Lines at the Photon Factory

The operation logging system has been designed and built using the Oracle database for twenty-two synchrotron radiation beam lines at the 2.5 GeV positron storage ring at the Photon Factory where X-ray/VUV synchrotron radiation experiments are simultaneously carried out. The operation logging system has a real-time capability to automatically store the database with all possible operational events of all vacuum valves/shutters and interlock signals, and all static operational data, including pressures of the beam lines and the storage ring, and related operational data which represent physical behaviors of the beam lines. By retrieving any combination of operational information, the system allows to virtually reproduce physical behaviors that have occurred in the beam lines.

[3P.44] Integrating Commercial and Legacy Control Systems With EPICS

The Experimental Physics and Industrial Control System (EPICS) is a software toolkit, developed by a worldwide collaboration, which significantly reduces the level of effort required to implement a new control system. Recent developments now also significantly reduce the level of effort required to integrate commercial, legacy and/or site-authored control systems with EPICS. This paper will illustrate with examples both the level and type of effort required to use EPICS with other control system components as well as the benefits which may arise.

[3P.45] Completion of the LANSCE Proton Storage Ring Control System Upgrade -- A Successful Integration of EPICS Into a Running Control System

The Los Alamos Neutron Scattering Center (LANSCE) Proton Storage Ring (PSR) control system upgrade was completed in 1996. In previous work, much of a PDP-11-based control system was replaced with Experimental Physics and Industrial Control System (EPICS) controls. Several parts of the old control system which used a VAX for operator displays and direct access to a CAMAC serial highway still remained. The old system was preserved as a "fallback" if the new EPICS-based system had problems. The control system upgrade completion included conversion of several application programs to EPICS-based operator interfaces, moving some data acquisition hardware to EPICS Input-Output Controllers (IOCs), and the implementation of new gateway software to complete the overall control system interoperability. Many operator interface (OPI) screens, written by LANSCE operators, have been incorporated in the new system. The old PSR control system hardware was removed. The robustness and reliability of the new controls obviated the need for a fallback capability.

[3P.46] Automizing ELETTRA Operation with "One Button Machine"

ELETTRA has reached good operational stability. Repetitive tasks, like the injection procedure, can be automated in order to maximize the beam time delivered to users. One Button Machine is a task spawner that executes different tasks concurrently, following a pre-defined control flow logic. It uses conventional UNIX routines for process execution and communication, and a Motif interface. A description of its various features is presented, covering operator intervention during flow execution, error recovery and step-by-step help when manual operation is selected. Operational results are given.

[3P.47] Operator Interface for the PEP-II Low Level RF Control System

We describe the operator interface for the Low Level RF Control System being built for the PEP-II storage rings at SLAC. The operator interface consists of displays for monitor and control from UNIX workstations, slow feedback loops and control sequences residing on microprocessors running the VxWorks real time operating system, and an interface to the existing SLAC Linear Collider control system for archiving, alarming, message logging, configuration, correlations, and multiknobs. The displays, database, and slower controls and monitoring are implemented using the EPICS control system package. Control sequences include the cavity tuner loop, automatic processing, klystron voltage loop, and RF station state control. Monitored subsystems include the machine protection system and low level RF. Challenges encountered in the operator interface design consist of the interface to a legacy control system, ability to change hardware and databases with minimal software impact and quick turnaround, intuitive user displays and tools, and reliable, bumpless, 24 hour a day operation.

[3P.48] Software Architecture of the PEP-II Low Level RF Control System

We describe the software architecture of the Low Level RF Control System being built for the PEP-II storage rings at SLAC. This VXI based system utilizes a commercially available embedded CPU controller board running the VxWorks real time operating system for each RF station. The operator interface and slower data processing are implemented using the EPICS control system package. The embedded processors are used to load, calibrate, monitor and diagnose various components of the system. The system consists of custom hardware modules containing fixed algorithm DSPs with programmable coefficients to implement FIR and DDF filters, and software programmable DSPs to implement the line supply ripple cancelation feedback loop and the ion clearing gap adaptation algorithm. Machine protection is implemented using a PLC based platform.

[3P.49] Network Upgrade for the SLC: PEP-II Network

We shall describe the PEP-II network which consists of a central host connected to an FDDI ring located around the PEP-II ring. FDDI-to-ethernet bridges connect the networked nodes to the FDDI ring. This network is separated from the rest of SLAC networks by the multihomed OpenVMS central host and two multihomed UNIX application gateways. The UNIX hosts act as EPICS control system OPIs providing data and displays to exterior networks. The network is addressed using a Private IP network address to further improve security. TCP/IP is required for EPICS and other applications to control the PEP-II equipment. This was a driving force to convert the SLC control system micro clusters to ethernet-based Multibus Single Board Computers along with TCP/IP software. This obsoletes the current SLAC-developed SLCNET communications system in areas accessible by the new PEP-II FDDI ring. We shall describe the details behind some of our design choices and the issues we faced while implementing this network.

[3P.50] Network Upgrade for the SLC: Control System Modifications

Current communications between the SLAC Linear Collider control system central host and the remote micros is built upon the SLAC developed SLCNET communication hardware and protocols. We will describe how the Internet Suite of protocols (TCP/IP) are used to replace the SLCNET protocol interface. The major communication pathways and their individual requirements are described. This includes modifications to applications on the central VMS based host and also modifications to the micro cluster software. A proxy server is used to reduce the number of total system TCP/IP connections and is an intergral part of the implementation. A third party TPC/IP stack was integrated with the micro cluster operating system to provide a fast and maintainable communications interface. Other software modifications were required to move to a new embedded micro computer which contained an intergrated ethernet controller. We shall explain some of the design decisions along with our experiences adding a new communications system to a legacy control system.

[3P.51] A client/server system for acquisition of distributed timing synchronous data

The Remote Data Gathering Service is an RPC service intended for communication with remote nodes on which data is available through BSD sockets. It was designed to interface the SLC and NCLTA control systems with those developed by acceperator users on their own hardware and using such programs as VeeTest.

[3P.52] Using WWW in SRRC Control System

The SRRC (Synchrotron Radiation Research Center - in Taiwan) control system was originally implemented in DEC VMS and DEC OSF/1 UNIX systems. Other computer system in SRRC (e.g. SUN, SGI, and IBM PC ...) were not able to connect to the control system. On sequently a hardware and OS independent system is required. The SRRC light source WWW site (http://www-icg.srrc.gov.tw) is built for this propose. With a IBM comptable PC (Pentium-133) runing Microsoft WindowsNT 3.51 and Microsoft IIS 1.0, it supports an easy way to build a WWW server. The program connected to SRRC control system is written with JAVA language, using JAVA applets to support runtime interactive ablitity.Using WWW browser(such as Netscape Navagator 3.0 or Microsoft Explorer 3.0), it is possible to monitor and control every storage ring device. This paper will describe the detail of the hardware and softwarestructures.

[3P.53] The Design of the Control System for ANKA.

ANKA is a low budget light source. It has to be built with minimal cost. This includes a short construction period to keep the personnel cost low. The use of ANKA as an industrial source on the other hand places tight constraints on the reliability of its components and its operation. In the creation of the control system one has to minimise in-house development and maximise the use of commercial-quality components. Here, the term commercial- quality applies equally to off-the-shelf-products and to products in the public domain or from other accelerator centres that are field proven and are documented at a level equal to commercial products. It has been our goal to search for the simplest model and thus the most reliable one. Our extensive investigations have shown that none of the actions and responses of the control system need a real-time operating system. Therefore, we have disposed of the whole middle layer of the control system including VME crates in favour of a fieldbus network which connects the individual devices directly to the console computers. The consoles will be PCs running Windows NT. The fieldbus candidate is LonWorks(R) from Echelon which provides plug-and-play and requires almost no additional development. On the software side we have opted for TACO, an object oriented control system environment developed at the ESRF. The architecture of the control system and results of tests with LonWorks(R) will be presented in this paper.

[3P.54] Automated Startup of the CEBAF 45 MeV Injector

In order to improve the speed and reproducibility of restoring the beam in the CEBAF 45 MeV injector after a full or partial shutdown of the accelerator, a program has been written using the Tcl/Tk scripting language to automate most of the required steps. The procedure is separated into three main parts. The first consists of preliminary checks that verify that the hardware is set correctly and that systems are ready to be activated. The second part turns on the main interlocked systems including high power magnets and RF. The final step turns on the beam and verifies that the beam quality is satisfactory by measuring the transmission, orbit, bunch length, transverse emittance and match, energy spread, and dispersion. Minor corrections for phasing are also performed in the program. In order to identify inefficiencies in the startup, each step is timed and parameter changes are logged so that system drifts can be tracked. This paper describes the software implementation, the logic to achieve a successful startup, and efficiency results.

[3P.55] Access Control and Interlock System at the Advanced Photon Source

The Advanced Photon Source (APS) consists of a linac, positron accumulator ring (PAR), booster synchrotron, storage ring, and up to 70 experimental beamlines. The Access Control and Interlock System (ACIS) utilizes redundant programmable logic controllers (PLCs) and a third ``hard-wired'' chain to protect personnel from prompt radiation generated by the linac, PAR, synchrotron, and storage ring. This paper describes the ACIS's design philosophy, configuration, hardware, functionality, validation requirements, display interfaces, beamline interfaces, and operational experience.

[3P.56] Radiation Levels Around the Fermilab Main Injector Extraction Septa

The Fermilab Main Injector extraction system will be capable of delivering a uniform 120\,GeV beam of \sim\,3\,\times\,10^13 protons per spill to the fixed target experiments. Up to 2% of the beam is expected to be lost at the extraction septum wires and the Lambertson septum. As a result, one expects increased radiation levels around the septa compared to other parts of the Main Injector. Realistic Monte-Carlo simulations have been performed to estimate the instantaneous and residual radiation levels in the beam extraction region. The results of these studies are presented and implications are discussed.

[3P.57] Activation of Fermilab Main Injector Extraction Septum

The Fermilab Main Injector extraction system will be capable of delivering a uniform 120\,GeV beam of \sim\,3\,\times\,10^13 protons per spill to the fixed target experiments. Up to 2% of the beam is expected to be lost at the extraction septum wires and the Lambertson septum. As a result, one expects the septa to be much more exposed to the high energy radiation than many other parts of the Main Injector. Realistic Monte-Carlo simulations have been performed to estimate the instantaneous and residual radiation levels in the beam extraction region. The results of these studies are presented and implications are discussed.

[3P.58] Polymerconcrete for Radiation Background Shielding of Detectors at Hadron Colliders

New shielding material -- polymerconcrete with density 1.2\dots 3.6 g\cdotcm^-3 and H, Li or B contents (4\dots6.6)\cdot10^22 H\cdotcm^-3, 1.3\cdot10^22 Li\cdotcm^-3 or 3.3\cdot10^21 B\cdotcm^-3, respectively, is developed. Granular polythene, lead powder, lithium salts or boron carbide are cemented by the special cement, dissolved in the liquid polymer. Material samples have lower limits of the resistance to compression 42 MPa, the resistance to rupture 6 MPa and the dynamic coefficient of elasticity 10^3 MPa. Radiation resistance, tested at the reactor IBR--2 and accelerators at JINR, IHEP and ITEP is 10^3 Mrad at 10 krad\cdots^-1 and 1 Mrad at 10 rad\cdots^-1. It is enough for a future colliders. Our experimental results demonstrate, that the induced radioactivity levels for the polymerconcrete are lower than ones for a ordinary concretes. It is shown that density 3.2\dots 3.6 g\cdotcm^-3 of a steel-concrete composition is optimum to get the minimum shielding thickness and cost. Polymerconcrete can be used as bricks or for the filling in any forms. It is suggesting to use this material instead CH_2 and Pb for shielding of the setups D0 and CDF at the Tevatron and CMS, ATLAS, ALICE at the LHC.

[3P.59] LIONs at the Stanford Linear Accelerator Center

The term LION is an acronym for Long Ion chamber. This is a distributed ion chamber which is used to monitor secondary ionization along the shield walls of a beam line resulting from missteered charged particle beams in lieu of the use of many discrete ion chambers. A cone of ionizing radiation emanating from a point source as a result of missteering intercepts a portion of 1 5/8" Heliax cable (about 100 meters in length) filled with Argon gas @ 20 psi and induces a pulsed current which is proportional to the ionizing charge. This signal is transmitted via the cable to an integrator circuit whose output is directed to an electronic comparator, which in turn is used to turn off the accelerated primary beam when preset limits are exceeded. This device is used to prevent potentially hazardous ionizing radiation resulting from missteered beams in areas which might be occupied by people. This paper describes the desigh parameters and use experience in the Final Focus Test Beam area of the Stanford Linear Accelerator.

[3P.60] Radiation Protection in the NLC Test Accelerator at SLAC

This paper describes the elements of the design of the NLC Test Accelerator pertaining to ionizing radiation protection and safety. The NLC Test Accelerator is an accelerator physics research facility at SLAC designed to validate 2.6-cm microwave linear accelerator technology for a future high-energy linear collider (the "Next Linear Collider"). The NLC Test Accelerator is designed for average beam power levels up to 1.5 kW, at energies up to 1 GeV (roughly equivalent to 1/500 of an NLC linac). The design for radiation protection incorporates shielding, configuration controls, safety interlock systems for personnel protection and beam containment, and operations procedures. The design was guided by the DOE Accelerator Safety Order, internal Laboratory policy, and the general principle of keeping radiation doses as low as reasonably achievable.

[3P.61] Modular Reliability Modeling of the CEBAF Personnel Safety System

A reliability model for the Thomas Jefferson National Accelerator Facility (Jefferson Lab) personnel safety system has been developed. The model, which was implemented using an Excel spreadsheet, allows simulation of all or parts of the system. Modularity of the model's implementation allows rapid ``what if'' case studies to simulate change in safety system parameters such as redundancy, diversity, and failure rates. Particular emphasis is given to the prediction of failure modes which would result in the failure of both of the redundant safety interlock systems. In addition to the calculation of the predicted reliability of the safety system, the model also calculates availability of the same system. Such calculations allow the user to make tradeoff studies between reliability and availability and to target resources to improving those parts of the system which would most benefit from redesign or upgrade. The model includes calculated, manufacturer's data, and Jefferson Lab field data. This paper describes the model, methods used, and comparison of calculated to actual data for the Jefferson Lab personnel safety system. Examples are given to illustrate the model's utility and ease of use.

[3P.62] Jefferson Lab Personnel Safety Fast Beam Kicker System

The CEBAF accelerator at Thomas Jefferson National Accelerator Facility (Jefferson Lab) uses a continuous electron beam with up to 800 kilowatts of average beam power. The laboratory beam containment policy requires that in the event of an errant beam striking a beam blocking device, the beam must be shut off by three methods in less than 1 millisecond. One method implemented is to shut off the beam at the gun. Two additional methods have been developed which use fast beam kickers to deflect the injector beam on to a water cooled aperture. The kickers designed and implemented at Jefferson Lab are able to deflect the injector beam in less than 200 microseconds. The kicker system includes self-test and monitoring capabilities that enable the system to be used for personnel safety. This paper will describe the requirements and performance of the fast beam kicker system.

[3P.63] Emerging Standards with Application to Accelerator Safety System Design

This paper addresses international standards which can be applied to the requirements for accelerator personnel safety systems. Particular emphasis is given to standards which specify requirements for safety interlock systems which employ programmable electronic subsystems. The work draws on methodologies currently under development for the medical, process control, and aerospace industries. The paper then goes on to show how these methods may be applied to accelerator safety system design. Detailed examples are drawn from the recently approved standard ``ISA-S84'' and the draft standard ``IEC1508''.

Part 3 of program listing