In present scenarios of a Heavy Ion Inertial Fusion (HIIF) facility a combination of linacs and storage rings has been proposed as a driver. After some funneling steps the main linac has to accelerate and focus an intense heavy ion beam (e.g. Bi 1+, 400 mA) to a final energy of 10 GeV. Using well known analytical formulae a study has been performed to find a range of beam and structure parameters (e.g. frequency, shunt impedance, beam current, emittances, focusing scheme), in which the requirements on a DTL can be fulfilled well.

* Work supported by BMBF under contract 06OF841I.

MO4016
Luminosity Monitor Studies for TESLA

O. NAPOLY, CEA/Saclay; D. SCHULTE, CERN

To optimise the beam parameters at the collision point in a linear collider a fast relative measurement of the luminosity with a precision on the percent level is required. We study 3 options for such a luminosity monitor in the context of the TESLA-500 linear collider: 1) calorimetry of low energy e⁺e^- pairs in the masking system, 2) detection of the bremsstrahlung leptons around 50 GeV outside of the last doublet, 3) calorimetry of the beamstrahlung photons emitted forward in a narrow cone. The large spacing between bunches may allow to scan the beam parameters within a bunch train.

MO4017  (oral poster MO3002)
Demonstration of Two Beam Acceleration in CTF II

R. BOSSART, H.H. BRAUN, G. CARRON, F. CHAUTARD, R. CORSINI, J.P. DELAHAYE, J.C. GODOT, S. HUTCHINS, G. SUBERLUCQ, P. TENENBAUM*, L. THORNDAHL, M. VALENTINI, I. WILSON, W. WUENSCH, CERN

The 2nd phase of the Compact LInear Collider (CLIC) Test Facility (CTF II) at CERN aims to demonstrate the feasibility of a linear electron/positron collider based on 30 GHz technology using a high-charge drive beam, parallel to the main beam, as the RF power source. In CTF II the two beams are generated in 3 GHz RF photo-injectors and are accelerated in 3 GHz linacs before injection into the 30 GHz modules. The 3 GHz drive beam linac has to accelerate a 16 ns long train of 48 bunches of 13.4 nC each. To cope with the very substantial beam-loading, special accelerating structures are used running slightly off the bunch repetition frequency. A magnetic chicane compresses the bunches after acceleration to less than 5 ps (FWHM), which is needed for efficient 30 GHz power generation. The 30 GHz modules are fully engineered building blocks of CLIC, including transfer structures to extract the 30 GHz power from drive beam, 30 GHz accelerating structures to accelerate the main beam, high resolution BPMs and a wire-based active alignment system. The performance achieved so far as well as the operational experience with the first accelerator of this type are reported.

* Present address: Stanford Linear Accelerator Center, Stanford, California.

MO4018
The CLIC 30 GHz Two-Beam Test Accelerator

I. WILSON, W. WUENSCH, W. COOSEMANS, C. ACHARD, CERN

As part of the experimental effort to demonstrate the feasibility of the CLIC scheme, a 3 m long representative section of the CLIC accelerator has been constructed, installed in the CLIC Test Facility (CTF) and successfully commissioned with beam. Prototype 30 GHz components developed during the course of the CLIC study including accelerating structures, power generating transfer structures, and high power RF loads, have been successfully integrated with the micron-precision active alignment system, the vacuum system, the water cooling system, and the high power RF distribution system to produce a very compact two-beam test accelerator. This paper describes the layout and gives details of the important subsystems.

MO4019
Beam Loading Compensation using Phase to Amplitude Modulation method in ATF

S. KASHIWAGI, Graduate Univ. for Advanced Studies; H. HAYANO, K. KUBO, T. NAITO, K. OIDE, S. TAKEDA, N. TERUNUMA, J. URAKAWA, KEK; T. KORHONEN, PSI; S. NAKAMURA, YNU

For future linear colliders, one of the essential techniques to get a sufficient luminosity is the ability to accelerate multi-bunch beam with small bunch spacing. Beam loading voltage generates a large energy spread along the bunch train. This energy spread is critical for the lattice design and, if not properly compensated, induces emittance growth and in turn lowers the luminosity. A method to compensate for beam loading effects in a multi-bunch beam is under development at Accelerator Test Facility (ATF) in KEK. We will report the beam tests for early injection and phase modulation to amplitude modulation method using multi-bunch beam with 2.8 ns bunch spacing. This energy compensation method compresses energy spread of multi-bunch beam by changing the input rf waveform into accelerating structures.

MO4020  (oral poster MO3003)
Results from Hardware R&D on C-band RF-System for e⁺e^- Linear Collider

T. SHINTAKE, N. AKASAKA, H. MATSUMOTO, KEK; J.S. OH, PAL-POSTECH; M. YOSHIDA, Univ. of Tokyo; K. WATANABE, Tohoku Univ.; Y. OHKUBO, H. YONEZAWA, Toshiba Co.; H. BABA, NKH Co. Ltd.

Hardware R&D on the C-band (5712 MHz) rf-system for the electron and positron linear collider (0.5 - 1.5 TeV) started in 1996 at KEK [1]. During two years R&D, we have developed most hardware components: a 50 MW class C-band klystron, its modulator, the high-power waveguide system and the accelerating structure. Concerning the rf-pulse compressor, using a cold model, a flat pulse generation using a three-cell coupled cavity was demonstrated. This paper summarizes the C-band RF-system development and discusses future applications in the industrial field.

Classification Category: A03

MO4021
RK-TBA Studies in KA-Band

S.M. LIDIA, S.S YU, LBNL; J. GARDELLE, J.L. RULLIER, CEA/Cesta; J.T. DONOHUE, CENBG; G.A. WESTENSKOW, LLNL

It is well-established that operating frequencies in the 30GHz range and higher are required to obtain the high accelerating gradients needed by linear collider systems that wish to probe center-of-mass energies significantly higher than 1 TeV. As an rf power source for high-energy linear colliders, relativistic klystron two-beam accelerators (RK-TBAs) have been shown theoretically to scale favorably to frequencies higher than X-band. To complement our studies of RK-TBA systems at 11.424 GHz, we are undertaking experimental tests of components at both 35 and 30 GHz. These studies will be conducted at the CEA/Cesta facility using the LELIA and PIVAIR beamlines, respectively, and will concentrate on the interaction of the intense, modulated electron beam with rf cavity structures, compact induction modules, and permanent magnet quadrupole transport lattices. Details of the rf cavity design and the proposed experiments are presented.

MO4022

Contribution withdrawn.

MO4023

Contribution withdrawn.

MO4024
High Charge Short Electron Bunches for Wakefield Accelerator Structures Development*

M.E. CONDE, W. GAI, R. KONECNY, J. POWER, P. SCHOESSOW, ANL

The Argonne Wakefield Accelerator Group develops accelerating structures based on dielectric loaded waveguides. To study these structures we have a facility comprised of two photocathode based RF guns and two standing-wave linac modules (L-band). The first electron gun generates high charge (10-100 nC) short electron bunches that are used to excite wakefields in the dielectric loaded structures. The second gun generates a low charge probe bunch that is accelerated by the wakefields left behind by the drive beam. We report measurements of the parameters of these two beams (charge, bunch length, emittance) and also initial results of the dielectric loaded accelerating structures. We have studied acceleration of the probe beam in these structures and we have also made measurements on the generated RF pulses. Single drive bunches, as well as multiple bunches separated by an integer number of RF periods have been used to generate the accelerating wakefields.

*Work supported by the U.S. Department of Energy, Division of High Energy Physics, under contract No. W-31-109-ENG-38.

MO4025
Laser Driven Subpicosecond Electron Linac

R. A. CROWELL, C. D. JONAH, ANL

Time-resolved investigations of fast chemical processes have been part of radiation chemical studies almost from the beginning. Recent advances in accelerator technology are making it possible generate subpicosecond electron pulses. Additionally, solid state laser development is making it more routine to generate subpicosecond optical pulses. A merging of these two technologies will provide an invaluable tool for probing ultrafast radiation induced processes. At the Argonne Chemistry Division efforts are underway to develop a sub-picosecond electron beam pulse radiolysis facility for chemical studies. The target output of the accelerator is to be generate electron pulses that can be adjusted from 3nC in .6ps to 100nC in 45ps. The following key problems will be addressed:

1. Development of an intense (>3mJ) 5ps UV laser pulse that is needed to drive the photocathode.


2. A .1ps probe pulse must be derived from the same laser and be independently tunable from the UV to the IR.


3. The UV pump and the probe laser pulses must be synchronized to the RF of the accelerator to within .5ps.


4. Design of an accelerator that will deliver the target output.


5. Development of novel, sensitive and specific detection techniques to make the fullest use of the potential information.

Classification Category: A07

MO4026
Feasibilty Study of a 2 GEV Superconducting H^- Linac as Injector for the CERN PS

R. GAROBY, H. HASEROTH, C.E. HILL, A. LOMBARDI, J.M. TESSIER, M. VRETENAR, CERN; P.N. OSTROUMOV, INR RAS

This preliminary feasibility study is based on the availability of the CERN LEP2 superconducting RF system after LEP de-commissioning. The option that is explored is to use this system as part of a high energy H^- linac injecting at 2 GeV in the CERN PS, with the aim of reliably providing at its output twice the presently foreseen transverse beam brightness at the ultimate intensity envisaged for LHC. This requires the linac to be pulsed at the PS repetition rate of 0.8 Hz with a mean beam current of 10 mA which is sufficient for filling the PS in 240 microseconds (i.e. about 100 turns) with the ultimate intensity foreseen for injection for the LHC. The linac is composed of two RFQ's with a chopping section, a room temperature DTL, a superconducting section with reduced beta cavities up to 600 MeV, and a section of LEP2 cavities up to 2 GeV. This study deals, in particular, with the problems inherent in H^- acceleration up to high energy and in the pulsed operation of SC cavities. Means for compensating microphonic vibrations in the SC cavities are considered, with the aim of reducing the final overall energy spread to the tight requirements of injection into a synchrotron. Other possible applications of such a machine are also briefly reviewed, making use of its potential for working at a higher duty cycle than required by LHC alone.

MO4027
Recent Progress in the Development of a Circular Ion Induction Accelerator for Space Charge Dominated Beams at LLNL*

L. AHLE, T. C. SANGSTER, D. AUTREY, G. CRAIG, A. FRIEDMAN, D. P. GROTE, E. HALAXA, R. L. HANKS, M. HERNANDEZ, H. C. KIRBIE, B. G. LOGAN, S. M. LUND, G. MANT, A. W. MOLVIK, W. M. SHARP, LLNL; D. BERNERS**, S. EYLON, D. L. JUDD, L. REGINATO, LBNL; A. DEBELING, W. FRITZ, Bechtel Nevada Corp.

The Heavy Ion Fusion Group at Lawrence Livermore National Laboratory has for several years been developing the world's first circular ion induction accelerator. This machine has recently been extended to 90 degrees, or 10 half-lattice periods (HLP) with full beam transport. In addition, induction cores have been installed on five of the HLP's, each with an independent arbitrary waveform pulser. An arbitrary waveform pulser for the bending electrostatic dipoles has also been enabled. Together, they have allowed the first attempts at coordinated bending and acceleration of the beam. The results of these attempts will be reported on in this paper by examining the output of various diagnostice devices, such as the Capacitive Beam Probes (C-probes), slit scanners and the Gated Beam Imager (GBI). The C-probes, which measure the charge centroid of the beam in each HLP, demonstrate control of the beam, while the slit scanners and the GBI, which measure emittance, demonstrate controlled emittance growth through the 90 degree bend.

* This work has been performed under the auspices of the US DOE by LLNL under the contract W-7405-ENG-48.
** Present address: Stanford University

MO4028
The LINAC of the Munich Fission Fragment Accelerator

O. KESTER, D. HABS, T. SIEBER, R. RAO, M. GROSS, A. KOLBE, P. THIROLF, J. OTT, LMU München; U. KÖSTER, TU-München; A. SCHEMPP, IAP-Univ. of Frankfurt; U. RATZINGER, GSI Darmstadt

An linear accelerator for the new Munich high flux reactor FRMII is under design [1,2]. This LINAC will be able to deliver intense beams of very neutron rich fission fragments from a target ion source located inside a through going reactor tube for the production of super heavy elements [3]. In order to obtain an efficient acceleration in the LINAC, charge breeding of the 1+ ion beam from the reactor to a q/A > 0.16 is required. The LINAC will work with 10% duty cycle with a final energy between 3.7 and 5.9 MeV/u. The LINAC of the fission fragment accelerator will consist of an Radio Frequency Quadrupole (RFQ) accelerator, three interdigital H-type (IH) structures similar to the lead LINAC at CERN and two seven-gap IH-resonators for the variation of the final energy of the ions. In the poster the concept, particle dynamics calculations and first cavity design calculations with MAFIA will be presented.

Classification Category: A06

MO4029
Field Description in an RFQ and its Effect on Beam Dynamics

R. DUPERRIER, R. FERDINAND, J.-M. LAGNIEL, P. MATTEI, CEA/Saclay; S. NATH, LANL

In a Radio Frequency Quadrupole (RFQ) linac, the electric field distribution is generated by four poles arranged symmetrically around a central z-axis. It's a customary practice to use a formulation based on cylindrical harmonics. That leads to inaccurate description of the fields near the boundary walls. Here, we present an analysis of errors based on an alternate numerical method to map the field correctly in the entire space between the vanes. This is especially important for high currents when the beam tends to fill nearly all available cavity-space. The difference in the field-description and its effect on beam dynamics in a high current RFQ with the same input-beam distribution will be presented.

MO4030

Contribution withdrawn.

MO4031  (oral poster MO3004)
The Drive Beam Decelerator of CLIC

A. RICHE, D. SCHULTE, CERN

In the Compact Linear Collider (CLIC) a high current low energy beam will be decelerated in a chain of transfer-structures to produce the RF-power necessary to accelerate a low charge high energy beam in the main linac. The transverse dynamics of the decelerated beam are discussed. The very large energy spread and strong transverse wakefields as well as the high group velocity of these fields and the considerable length of the bunch train are important factors. Static and dynamic imperfections are considered including ground motion. The choice of parameters for the structures is investigated. A promising beam-based alignment technique is presented that makes use of a low emittance beam.

MO4032
Wake Field Effects Analysis in APT Linac*

S.S. KURENNOY, LANL

The 1.7-GeV 100-mA CW proton linac is now under design for the Accelerator Production of Tritium (APT) Project. The APT linac comprises both the normal conducting (below 211 MeV) and superconducting (SC) sections. The high current leads to severe restrictions on allowable beam losses (<1 nA/m), that requires analyzing carefully all possible loss sources. While wake-field effects are usually considered negligible in proton linacs, we study these effects for the APT linac to exclude potential problems at such a high current. Loss factors and resonance frequency spectra of various discontinuities of the vacuum chamber are investigated, both analytically and using 2-D and 3-D simulation codes with a single bunch as well as with many bunches. Our main conclusion is that the only noticeable effect is the HOM heating of the 5-cell SC cavities. The analysis helps to identify dangerous modes for designing HOM couplers properly.

*Work supported by the U.S. Department of Energy.

MO4033
Design Studies of the DARHT Phase II Injector with the GYMNOS PIC Code*

W.M. FAWLEY, E. HENESTROZA, LBNL; Y.-J. CHEN, D.W. HEWETT, LLNL

We have used the GYMNOS 2D (r-z) electro-/magneto-static PIC simulation code to help design a high current, high brightness, 3.5-MV injector for the DARHT Phase II accelerator. GYMNOS is particularly noteworthy for its use of piece-wise linear approximations to curved boundaries within a regular orthogonal mesh, which aids in modeling complex surfaces. We present a number of comparisons between GYMNOS and EGUN results and discuss the numerical limitations of each code. We also illustrate the various practical tradeoffs that must be considered in the injector design, such as the dependence of output beam brightness upon maximum allowed electric field stress upon different surfaces.

* This work was supported by the U.S. DOE under Contracts No. DE-AC03-76SF00098 (LBNL) and W-7405-ENG-48 (LLNL).

MO4034
Emittance Growth from Bend/Straight Transitions for Beams Approaching Thermal Equilibrium*

J. J. BARNARD, B. LOSIC, LLNL

In certain applications such as heavy ion fusion, intense beams with large space charge tune depressions will be transferred from linear transport sections into bent transport sections. In some designs, such as recirculating induction accelerators, transport through bends will occur over thousands of betatron periods and in some driver designs the final transport through a bend will occur over tens of betatron periods. Over such distances, non-linear space charge forces should be expected to produce particle phase space distributions which in some sense are close to thermal equilibrium, especially with respect to lower order moments. Here we calculate the properties of thermal equilibrium beams in bends assuming uniform focusing, as a function of beam parameters, such as tune depression and energy spread. We also calculate the change in emittance for a beam that is initially in thermal equilibrium in a straight transport section, and that finally reaches thermal equilibrium in a bent system, using an energy conservation constraint to connect the two states. Implications for heavy-ion fusion will be discussed.

*Work performed under the auspices of the U.S. DOE by LLNL under contract W-7405-ENG-48.

MO4035

Contribution withdrawn.

MO4036
Coupling Slot Measurements Against Simulations for TRISPAL Accelerating Cavities

P. BALLEYGUIER, CEA

At the LINAC 96 conference, we presented a new coupling scheme for the Trispal CCL accelerating cavities: the "4-petal" slots. It resulted from a MAFIA optimization in which we tried to maximize the shunt impedance. Since that time, we designed and built a mock-up for an accurate measurement of the coupling-slots induced Q-drop. Indeed, we achieved a good accuracy and reliability in Q-drop measurements, but the value was rather disappointing: -22.5% (± 0.5), instead of -5% as predicted. As a consequence, the "4-petal" coupling scheme was abandoned, and we learned that local power losses computed by cavity simulation codes can be widely underestimated. Further simulation showed that improving the mesh resolution could give better results, though only a more subtle meshing method (like triangular cells or partially filled cells) could lead to realistic values.

MO4037
External Q Studies for APT Superconducting Cavity Couplers*

P. BALLEYGUIER, CEA; F. KRAWCZYK, LANL

Coupling coefficients of the APT superconducting cavity couplers have been predicted using an improvement on the method previously developed for the French Trispal project. We here give the method and a proof of the formula used to compute the external Q. Measurements on a single-cell copper cold model exhibited a very good agreement against simulation. With our method, we established that the original coupler design lead to an insufficient coupling in the beta=0.64 cavities. Different solutions were proposed to fix this problem, such as combining impedance discontinuities in the line and an off-centered disc end tip. Finally, it was decided to increase the beam tube diameter, though this has some influence on the cavity end-cell performance.

* Work supported by the U.S. Department of Energy.

MO4038
2.5D Cavity Code with High Accuracy

Y. IWASHITA, ICR-Kyoto Univ.

PISCES II calculates eigensolutions of any axisymmetric cavity with 2.5D finite element method and can handle periodic boundary conditions. Dipole and higher multipole solutions are obtained by edge elements. The accuracy of the frequency in a solution obtained from PISCES II is improved by use of higher order elements. In order to reduce the computation time, the eigenvalue solver is improved.

MO4039
Modeling Linear Collider RF Components in the Time Domain with Unstructured Grids*

C.-K. NG, V. SRINIVAS, Z. LI, B. MCCANDLESS, W. MI, M. SAPOROV, Y. SUN, M. WOLF, K. KO, SLAC

A number of simulation packages, including MAFIA and HFSS, have been in effective use for designing and analyzing accelerator RF components. Nevertheless, there is strong interest in new modeling capabilities, especially for future accelerators like the NLC, that will result in improvement in accuracy as well as increase in problem size. As part of the DOE Grant Challenge effort, the Numerical Modeling Group at SLAC has undertaken the development of a time-domain electromagnetic field solver, Tau3, that is based on an unstructured grid for better geometry modeling, and is designed to be portable to MPP platforms for large scale simulation. The solver has a broadband waveguide termination so that the transmission properties of an RF structure can be calculated over a wide frequency range in a single run. Initial results from applying Tau3 to determine the match of some typical NLC components, such as power splitters, combiners, and input couplers, will be reported and compared with simulations using MAFIA and HFSS.

*Work supported by the U.S. Department of Energy, contract DE-AC0376SF00515.

Classification Category: D03

MO4040

Contribution withdrawn.

MO4041
Computing Eigenmodes in Highly Lossy Accelerating Structures

S. SETZER, T. WEILAND, TU-Darmstadt

The numerical calculation of eigenvalues in structures containing high loss dielectric and permeable materials is of importance in the field of accelerators as well as in many other high frequency applications. While satisfying algorithms exist for loss free and small-loss problems, the numerical problem of highly lossy material insertions is still a big challenge. We examine the the Jacobi-Davidson method that proves to be a rather suitable method for calculating a set of eigenvalues even for the case of highly absorbing materials. Furthermore, unlike the commonly used sub-space methods, this algorithm is not limited to calculate extreme eigenvalues only. It is also capable of finding the eigenfrequencies located around any user specified frequency. Another practical advantage of this method is the absence of parameters such as the upper limit for an eigenvalue spectrum. Besides the basic theory we present results for typical accelerator components such as a preliminary model of a higher order mode damped accelerating cavity built at Fermilab.

MO4042
Simulation Results with an Alternate 3D Space Charge Routine, PICNIC

N. PICHOFF, J.-M. LAGNIEL, S. NATH*, CEA/DSM/DAPNIA/SEA

One of the major problems in the beam dynamics calculation dealing with high current linacs is the treatment of space charge effects. The widely used SCHEFF routine, originated and developed at Los Alamos, is often critiqued as being simplistic because of its inherent assumption of transverse symmetry. Here we report preliminary work on an alternate fully 3D space-charge routine for a bunched beam. It is a particle-in-cell approach based on numerical calculation of the interaction between cubes (PICNIC). The principle underlying the method and the comparative results of simulation with SCHEFF and PICNIC are reported.

* On sabbatical from Los Alamos National Laboratory, Los Alamos, NM, USA.

MO4043
MUSTAFA -- A Tool for Numerical Simulations of the Beam Behaviour in a Linac

G. GUIGNARD, J. HAGEL, CERN

MUSTAFA stands for MUltibunch Simulation and Tracking Algorithm for Future linear Accelerators and provides an environment for tracking a multibunch beam accelerated in the linac of a collider. It is made of three parts: i) an interactive utility MBUNCH which allows to run the simulations, edit the input data, systematically vary some parameters, launch statistics on misalignment realisations and select the outputs, ii) the tracking program itself MBTR that includes magnetic focusing, acceleration, wakefields, misalignments and trajectory correction, and iii) a graphics facility called MOVIE that makes available an animated representation of the transverse motion within each bunch of a train, either in the geometrical space or in the phase space. While MBTR is written in FORTRAN, the other two codes are written in QUICKBASIC (MS V4.5). They are run on PC pentium computers in a stand-alone mode, independently of any network. Hence, MUSTAFA is easily transportable and offers a unique graphics facility. It has been extensively used for multibunch simulations in the Compact LInear Collider scheme (CLIC) as well as for demonstrations of the beam behaviour in different stability or beam break-up conditions.

MO4044
Simulation of Halo Formation in Breathing Round Beams in Periodic Focusing Channel*

Z. HUANG, Y. CHEN, CIAE

Halo formation in high-intensity axisymmetric beams in periodic focusing channel is analyzed using particle-in-cell simulations. In order to explore self-consistently the fundamental properties of breathing round beams propagating in periodic focusing channel, the initial phase-space distribution of a beam injected into a linac is adopted to be a sufficiently realistic distribution such as Gaussian, waterbag and parabolic. Numerical results such as halo intensity and emittance growth are obtained by means of multiparticle simulations.

*Work supported by the National Natural Science Foundation of China and the Science Foundation of Chinese Nuclear Industry.

MO4045  (oral poster MO3005)
Optics Elements for Modeling Electrostatic Lenses and Accelerator Components IV. Electrostatic Quadrupoles and Space Charge Modeling

G. H. GILLESPIE, G. H. Gillespie Associates, Inc.

A set of optics models for a variety of electrostatic lenses and accelerator columns has been developed for the computer code TRACE 3-D. TRACE 3-D is an envelope (matrix) code that includes a linear space charge model primarily used to model bunched beams in magnetic transport systems and radiofrequency (RF) accelerators. Several new matrix models have been developed that allow the code to be used for modeling beamlines and accelerators with electrostatic components. The new models include a number of options for: (1) einzel lenses, (2) accelerator columns, (3) electrostatic deflectors (prisms), and (4) electrostatic quadrupoles. A prescription for setting up the initial beam appropriate to modeling 2-D (continuous) beams has also been developed. The models for (4) are described in this paper and examples of their use are illustrated. The relationship between the 3-D (bunched beam) and 2-D (DC beam) space charge modeling is discussed and comparisons of numerical results to other calculations are presented.

MO4046
Beam Dynamics in a High Current SC Proton Linac for Waste Transmutation and Energy Production

G. BELLOMO, P. PIERINI, INFN-Milano-LASA

INFN is studying a 100 MeV to 1.6 GeV high current (25 mA) proton linac for waste transmutation and energy production. The beam dynamics in the three beta sections of the linac have been studied with linear and nonlinear transport and tracking codes. The results of this activity will be presented here, with a description and preliminary results of a simulation codes that we are developing for space charge dominated beam dynamics with nonaxisymmetric beams.

MO4047  (oral poster MO3006)
Recent Developments in the Accelerator Design Code PARMILA*

H. TAKEDA, J. H. BILLEN, LANL

The name PARMILA is an acronym for "Phase And Radial Motion in Ion Linear Accelerators." The PARMILA code, which originated in the 1960s for designing drift-tube linacs, now designs and simulates the performance of many types of rf linear accelerators. The structure types include the drift-tube linac, the coupled-cavity drift-tube linac, the coupled-cavity linac, and several types of superconducting linacs. The new code can deal with multiple types of linac structures in a single run. This multi-structure run can include the effects of errors such as quadrupole-magnet misalignments, cavity field amplitude errors, etc. The latest version of PARMILA features a more logically organized input sequence for the different linac structures and their properties. A PARMILA run can include sequences of beam-transport elements before, after, and between linac sections. Alternatively, the run may consist only of transport elements with no linac structure at all. In this paper, we describe the new user interface highlighting the implementation of multiple rf structures. We also discuss the design algorithm used for superconducting linac structures. A comprehensive manual for the code is in preparation.

* This research is sponsored by the Division of Materials Sciences, DOE, under contract DE-AC05-96OR22464 with Lockheed Martin Energy Research Corp.

MO4048

Contribution withdrawn.

MO4049
Simulations of the Nonlinear Transverse RF Fields on the Beam Dynamics in the Low Energy X-band SW Linacs*

X. SUN, Y. LIN, Tsinghua University

The detailed beam dynamics simulations in the low energy X-band SW linacs are presented. The codes CAV, CAV9 and TRACK are developed. They are used to investigate radial beam dynamics including the nonlinear transverse RF fields. Examples are given of the study of transverse effects in different initial phases of the beam, injection voltages of the electron gun and accelerating gradients in the first cavity.

*Work supported by the National Science Foundation of China.

MO4050
A Novel Structure of Multi-Purpose RF Gun

E. TANABE, AET Associates, Inc.; F. ODA, Kawasaki Heavy Industries

RF guns have recently been very commonly used in high-energy research accelerators, as well as in various applications of accelerators such as FEL, because it offers significant advantages over the conventional techniques, which use electron guns and bunchers. In this paper, the novel structure of the RF gun, utilizing the coupled cavity, pi/2 mode, standing wave structure is described. The new structure offers much higher group velocity and lower emittance over the structure that is commonly used. As a result, the new structure has a more stable operation for higher beam current without sacrificing beam emittance. Moreover, the structure offers unique characteristics in reduction of back bombardment for the application of thermionic cathode in RF guns. The cavity structure and beam parameters were optimized by using EMSYS(2D) and MAFIA(3D). An overview of design detail and prototype structure, which can be used with either thermionic cathode or photocathode, will be presented.

MO4051
The High Voltage System for the High Intensity CERN Proton Source

C.E. HILL, M. O'NEIL, CERN

For the RFQ injector mounted on the CERN 50 MeV proton linac, the source needs to provide about 300 mA of 100 keV beam in pulses of 20 - 150 µs at 1 Hz. Although the High Voltage supply is fairly conventional, a number of measures had to be taken to ensure not only the reliability of the source electronics but also other equipment installed in the near vicinity. In view of the high current demanded from the source a new, and very simple, form of beam load compensation was developed to stabilise the preinjector voltage to values acceptable to the RFQ.

MO4052
First Experience of Works with a Compact Injectors for Trials and Drills of RF Linac Structures

V.V. KUSHIN, N.A. NESTEROV, S.V. PLOTNIKOV, D.N. SELEZNEV, A.S. SUVOROV, V.P. ZUBOVSKY, ITEP; E.P. BOGOLYUBOV, V.T. BOBYLEV, V.A. PRESNYAKOV, All-Russian Research Institute of Automatics

There is a well-known problem of gas loading on vacuum in RF linac structures from the side of ion injectors. At the starting stage of high power linac trials it often leads to necessity of significant augmentation of pumping means. This problem is mostly vital at adjustment of multiaperture linac structures. To simplify the problem the new test injector based on spark ion source is developed. It is designed for generation of ion beam and its preliminary acceleration up to 50-150 keV. Beam generation is based on spark discharge desorption of deuterium, stored in some metals (Ti, Zr, Sc) due to occlusion and its ionization within vacuum arc in consequence of discharge spark. An industrial vacuum tube of TNT-147 usually used in neutron generators is a main element of the injector accelerating block. The injector is rather compact due to the possibility of location both ion source and high voltage block within the same package of 680 mm long and 130 mm in diameter. This configuration allows to get protons, deuterons or heavy ions formed in pulses of ns and mcs ranges at repetition rate up to tens and hundred pps. Such injector is fixed at the input flange of RF linac structure at once due to the property of discharge initiation within the spark ion source with no initial gas. It makes possible to merge injector and structure vacuum systems. The results of works with experimental samples of injectors designed for protons, deuterons and heavy ions are considered.

MO4053

Contribution withdrawn.

MO4054
The Cathode Test Stand for the DAHRT Second-Axis

C. FORTGANG, B. CARLSTEN, D. PRONO, LANL; C. HUDSON, K. MOY, R. STURGIS, Bechtel; C. PETERS, S. WU, LBNL

The DAHRT second-axis requires a 3.5 MV, 4 kA, 2 microsecond injector. The injector will use a thermionic dispenser-cathode with an 8" diameter. Because such a cathode has requirements beyond existing cathode assemblies, we are building a modest facility for measuring its most important engineering features. The cathode will be tested in situ as part of an operational electron-gun which has most of the design features envisioned for the DAHRT-2 injector. The cathode requires 5 kW of heater power. We will measure: temperature uniformity over the cathode surface, differential thermal expansion between the cathode and the focusing electrodes, and outgassing and vacuum properties. We will also test field-emission from the focusing electrodes at an electric-field strength of 160 kV/cm. We will extract a 500 kV, 500 A beam using a high-voltage modulator. This modulator uses a PFN and a step-up transformer to generate a 1.2 microsecond pulse. After a modest compression of the beam and a short transport section, we will measure the beam profile to assess electron-current emission-uniformity, paying particular attention to beam-halo in preparation for further studies of emittance growth due to non-linear focusing forces.

MO4055  (oral poster MO3007)
Comparison of Beam Simulations With Measurements for a 1.25-MeV, CW RFQ*

H. V. SMITH, JR., G. O. BOLME, J. D. SHERMAN, R. R. STEVENS, JR., L. M. YOUNG, T. J. ZAUGG, LANL

A Los Alamos experiment to test the injector for the Low-Energy Demonstration Accelerator (LEDA) using the Chalk River Injector Test Stand (CRITS) radio-frequency quadrupole RFQ [1] as a diagnostic instrument is reported. The LEDA injector [2] produces 50-keV, DC proton beams for injection into the 1.25-MeV RFQ. The RFQ transmission and spatial profiles are measured as a function of injected current and low-energy beam transport (LEBT) solenoid excitations [3]. At lower RFQ currents (50 mA), the RFQ beam transmission is 90%. For RFQ currents near the 75-mA design value, the measured RFQ transmission decreases to 80-85%. The expected system performance is calculated using the computer codes TRACE and SCHAR to model the LEDA LEBT and the computer code PARMTEQM to model the acceleration of those beams in the CRITS RFQ [4]. The predictions of the computer simulations are compared with the beam measurements.

* Work supported by the U.S. DOE, Defense Programs.

Classification Category: T01

MO4056
The SNS Front End Accelerator Systems

J. STAPLES, D. CHENG, M. FONG, M. HOFF, R. KELLER, K. KENNEDY, M. LEITNER, A. RATTI, S. VIROSTEK, LBNL

The Spallation Neutron Source front end comprises a 35-70 mA volume H^- source, a multi-element electrostatic LEBT including chopping and steering, a 402.5 MHz RFQ with low output emittance and a 2.5 MeV MEBT also including chopping. The beam duty factor is 6%, with possible extension to 12%. This system, along with a LANL-supplied 1 GeV linac and a BNL-supplied storage ring, provides an average beam power of 1-2 MW to an ORNL and ANL-supplied neutron target and beam instruments facility. The current status of the front end design is described, along with results of several R&D projects leading to the final design.

MO4057
A Ferroelectric Ceramic: A Novel Efficient and Robust Photocathode

I. BOSCOLO, R. PARAFIORITI, A. PORCARI, A. SCURATI, University and INFN-Milano; M. CASTELLANO, L. CATANI, M. FERRARIO, P. PATTERI, F. TAZZIOLI, INFN-LNF; A. DORIA, G.P. GALLERANO, L. GIANNESSI, E. GIOVENALE, ENEA-CRE; J. HANDEREK, University of Katowice

Ferroelectric ceramics of the PLZT type, i.e. lead zirconate titanate lanthanum doped, in form of thin disks have been tested as photocathodes in a new geometrical configuration. The disk of material, with the back surface only metallized and held in place by a metallic ring, has been set in front of a solid anode. The applied accelerating field ranged from zero up to 20 kV/cm, the light pulse exciting the sample was 25 ps long, 532 nm wavelength and its energy arrived up to 4 mJ on an area of 10 mm². The vacuum was very poor. The maximum output charge was 1 nC, but it was clearly limited by space charge effect. We have observed different replies with samples having different compositions. A theoretical explanation of the results is only hinted because the material surface structure is very complicated and various emission mechanisms may concur. Emphasis is given to measurements performed to characterize the cathode as photoemitter. The high current, the green exciting light and the insensitivity to poor vacuum make this photocathode an interesting candidate for linac injectors.

Classification Category: T01

MO4058
Fields Induced by Chopped Beams in the Tank Cavity

E. TAKASAKI, Z. IGARASHI, F. NAITO, K. NANMO, T. TAKENAKA, KEK

In order to accelerate high-intensity beams at the 12-GeV proton synchrotron complex at KEK, the 40-MeV proton linac has successfully operated as an injector. Recently, acceleration of chopped beams will be planned for the longitudinal beam matching from the 500-MeV Booster to the 12-GeV main ring. Fields induced in the second tank, which has post couplers, by chopped beams have measured. The results measured will be described in this report. The effects of adjustment of the phase between two tanks on the transversal motion will also be described.

MO4059
Design and Development of the LEDA Slow Wire Scanner Profile Measurement*

J. F. O'HARA, AlliedSignal FM&T**; J. F. POWER, J. LEDFORD, J. D. GILPATRICK, M. STETTLER, LANL

The Low Energy Demonstration Accelerator (LEDA) is being developed at Los Alamos as part of the Accelerator Production of Tritium (APT) project. One of the diagnostics being developed to commission LEDA is a slow wire scanner, which will be used to measure beam profiles. Initial measurements of the H⁺ beam will be made at 6.7 MeV beam energy and 100 mA beam current. The wire scanner is an interceptive device which will move two silicon carbide fibers (positioned along the x axis and y axis) through the beam to obtain the profile. As the protons interact with the fiber, freed secondary electrons from the silicon carbide fibers are detected as an integrated charge depletion at each wire position throughout a scan. The actuation of the scanner is provided by a modified, commercially available, motorized, linear, feedthrough driven by a step motor. Some of the design considerations discussed are wire materials, scanner operating parameters, signal detection, signal processing, wire position determination, on-line calibration, timing, and controls.

*This work is supported by the Department of Energy.
**Operated for the U.S. Department of Energy under Contract No. DE-ACO4-76-DP00613.
© Copyright AlliedSignal Inc., 1998.

MO4060
Improvements on the Accuracy of Beam Bugs*

Y. J. CHEN, T. FESSENDEN, LLNL

At LLNL resistive wall monitors are used to measure the current and position of intense electron beams in electron induction linacs and beam transport lines. These, known locally as "beam bugs," have been used throughout linear induction accelerators as essential diagnostics of beam current and location. Recently, the development of a fast beam kicker has required improvement in the accuracy of measuring the position of beams. By picking off signals at more than the usual four positions around the monitor, beam position measurement error can be greatly reduced. A second significant source of error is the overlap of the foil that forms the resistor. In addition, in-situ bugs used on ETA-II show a drop in signal due to a fast redistribution time constant of the signals. This paper presents the analysis and experimental tests of the improved beam bugs used for beam current and position measurements in and after the fast kicker.

*This work was performed under the auspices of the U.S. Department of Energy under Contract No. W-7405-Eng-48.

MO4061
A Coaxial Cable Beam Loss Monitor Ion Chamber System for High Power Multi-Bunch Beams*

M. C. ROSS, D. MCCORMICK, SLAC

Gas filled coaxial cable beam loss monitors are a proven diagnostic in short pulse linacs and transport lines. At the SLAC linear collider (SLC), where the bunch length is ~ 1 mm, monitor cables with lengths ranging from 100 m to 3 km are used to locate beam losses of 5 E 8 particles (1.5% of nominal) with a resolution of ± 1 m. The monitor is effective because of the simplicity of its installation and signal interpretation. Future linear colliders will have high power beams made up of many closely spaced bunches and will therefore require more careful signal processing in order to locate losses. Typical collider operation will involve the use of pilot pulses, made up of only one bunch, to test subsystem performance prior to full power operation. A simple signal processor will be able to locate losses by comparing the evolution of the loss monitor signal as the number of bunches is increased. The monitor must have 10 x greater sensitivity than the SLC monitors in order to provide a prediction of the expected beam loss at full power using only the signal from the pilot pulse. This paper describes the proposed linear collider loss monitor system.

* Work supported by the Department of Energy, contract DE-AC03-76SF00515.

MO4062
Time Resolved, 2-D Hard X-ray Imaging of Relativistic Electron-Beam Target Interactions on ETA-II*

C. CRIST, SNL; S. SAMPAYAN, LLNL; M. KROGH AlliedSignal FM&T; G. WESTENSKOW, G. CAPORASO, T. HOUCK, J. WEIR, D. TRIMBLE, LLNL

Advanced radiographic applications require a constant source size under 1 mm. To study the time history of a relativistic electron beam as it interacts with a bremsstrahlung converter, one of the diagnostics we use is a multi-frame time-resolved hard x-ray camera. We are performing experiments on the ETA-II accelerator at Lawrence Livermore National Laboratory to investigate details of the electron beam/converter interactions. The camera we are using contains 6 time-resolved images, each image is a 5 ns frame. By starting each successive frame 10 ns after the previous frame we create a 6 frame movie from the hard x-rays produced from the interaction of the 50 ns electron beam pulse.

*Work performed jointly for the U.S. DOE by Sandia National Laboratories under contract DE-AC04-94AL85000, Lawrence Livermore National Laboratory under contract W-7405-ENG-48 and by AlliedSignal FM&T under contract DE-ACO4-76-DP00613.

MO4063  (oral poster MO3008)
Study of the Compensation Process of Ion Beams*

A. JAKOB, H. KLEIN, A. LAKATOS, J. POZIMSKI, IAP-University of Frankfurt

For investigation of the space charge compensation process due to residual gas ionization and the experimental study of the rise of compensation, a LEBT system consisting of a pulsed ion source, two solenoids and a diagnostic section was set up. The potentials at the beam axis and the beam edge were ascertained from time-resolved measurements by a residual gas ion energy analyzer. A numerical simulation of self-consistent equilibrium states of the beam plasma has been developed to determine plasma parameters which are difficult to measure directly. The temporal development of the kinetic and potential energy of the compensation electrons has been analyzed by using the numerically gained results of the simulation. To investigate the compensation process the distribution and the losses of the compensation electrons were studied as a function of time. The acquired data show that the theoretical estimated rise time of space charge compensation neglecting electron losses is shorter than the build up time determined experimentally. To describe the process of space charge compensation an interpretation of the achieved results is given.

* Work supported by BMFT under contract No. 06 OF 841.

MO4064
Absolute Beam Position Monitoring Using HOM-Damper Signals

C. PESCHKE, G. SCHREIBER, P. HÜLSMANN, H. KLEIN, IAP-University of Frankfurt

To preserve the required beam quality in an e⁺/e^- collider it is necessary to have a very precise beam position control at each accelerating cavity. An elegant method to avoid additional length and beam disturbance is the usage of signals from existing HOM-dampers. The magnitude of the displacement is derived from the amplitude of a dipole mode whereas the sign follows from the phase comparison of a dipole and a monopole HOM. To check the performance of the system, a measurement setup has been built with an antenna which can be moved with micrometer resolution to simulate the beam. Furthermore we have developed a signal processing to measure the absolute beam displacement. Measurements on the HOM-damper cell can be done in the frequency domain using a network analyser. Final measurements with the nonlinear time-dependent signal processing circuit has to be done with very short electric pulses simulating electron bunches. Thus, we have designed a subnanosecond pulse generator using a clipping line and the step recovery effect of a diode. The measurement seems to be possible with a resolution of about 10 micrometers. Measurements and numerical calculations of the measurement setup and the pulse generator are presented.

MO4065
Design of the Phase Reference System for the TESLA Linear Collider

T. PLAWSKI, S.N. SIMROCK, J. SONNENBERG, DESY

The frequency distribution system for TESLA must deliver a highly phase stable signal to the 716 rf stations over a length of 33 km. At the operating frequency of 1300 MHz a short term and long term stability of the order of 1 degree with respect to the accelerated beam is required. The design of the phase distribution system combines three different systems. A low loss coaxial power distribution system is used to distribute a low frequency subharmonic of the operating frequency to the individual rf stations where multipliers generate the klystron drive signal. The local oscillator signal for the downconverters is also generated from this signal using a table driven vector modulator. At selected locations the 1300 MHz signal is compared to an ultrastable fiber optic distribution system to detect long term phase drifts with an accuracy of a few degrees. A third system measures the beam phase with respect to the cavity field at each rf system and is used for precise phase calibration of the reference line. This system is transparent to beam operation and will continually monitor and correct slow phase drifts.

MO4066
Frequency Source for the ISAC RFQ

S. FANG, K. FONG, M. LAVERTY, TRIUMF

The ISAC RFQ nominally operates at a frequency of 35 MHz. It can operate in either self-excited or driven mode. In the self-excited mode, the RFQ cavity itself is one of the frequency determining elements. In the driven mode, the frequency is slaved to an external stabilized frequency source. In order to achieve glitchless transitions between the two modes, the RFQ frequency source is a phase-locked loop. RF switches in the loop configure the loop to lock to the external frequency source when in driven mode. Reconfiguration of the rf switches in self-excited mode turns the phase-locked loop into an oscillator whose frequency is determined partly by the phase shift of the RFQ cavity. The damping in the phase-locked loop prevents sudden shifts in either frequency or phase during switching between the two operating modes.

MO4067
A Development and Integration Analysis of Commercial and In-house Control Subsystems

M. MOORE, WSRC; R. DALESIO, LANL

The acquisition and integration of commercial automation and control subsystems in physics research is becoming more common. It is presumed that these systems present lower risk and less cost. This paper will study four subsystems used in the Accelerator Production of Tritium (APT) Low Energy Demonstration Accelerator (LEDA) at the Los Alamos National Laboratory (LANL). The Radio Frequency Quadrupole (RFQ) Cooling System, the High-Power RF System and the RFQ Vacuum System were outsourced; the Low-Level RF (LLRF) System was developed in-house. Based upon our experience, a careful evaluation of the costs and risks in acquisition, implementation, integration, and maintenance associated with these two approaches is given.

MO4068
Performance of the Klystron HV Pulse Modulators at the S-Band Test Facility at DESY

S. CHOROBA, J. HAMEISTER, M. KUHN, DESY

The klystrons at the S-band test facility at DESY require high voltage pulses of 550kV at a flat top pulse duration of 3microsec., a current of 700A and a repetition frequency of 50Hz. Two HV pulse modulators with a nominal power of 375MW were built. The design and results of the commissioning of the first system were described in [1]. The second system came into operation and was operated at an output power up to 475MW (610kV, 780A). In addition it was modified to investigate capacitor charging power supplies, which might be used in a linear collider tunnel. This type of power supply promises a better efficiency than the conventional resonant charging method and also saves space in the linear collider tunnel. This paper describes the operation experience with the HV modulators at the test facility especially with the second system. It reports the modifications which were required to operate the modulator with the new type of HV power supply and presents results of the operation.

Classification Category: T04

MO4069
Advanced Buck Converter Power Supply "ABCPS" for APT*

R. W. STREET, T. OVERETT, General Atomics

A DC power supply is required for each of the approximately 250 1MW CW klystrons in the RF power system. The requirements are that the power supply meet output performance specifications, provide fault protection for the klystron, have high efficiency, high reliability, good maintainability, and be readily producible. As the power supplies are one of the largest cost items in the accelerator, a technology review was made to determine the most economical approach to satisfy the requirements. A switch-mode power supply employing a buck-regulator was identified as being potentially the lowest cost approach. As the switch is a high risk development a small scale prototype has been constructed for feasibility evaluation. A description of the hardware will be presented and the results will be discussed

*Work supported by DOE contract: DE-AC04-96AL89607

MO4070
Status of the 36 MHz RF-System for the High Current Injector at GSI

W. VINZENZ, G. HUTTER, W. GUTOWSKI, GSI Darmstadt; B. ROSSA, Thomcast AG Turgi Swiss

In 1999 the 1.4 MeV/u prestripper section will be replaced by the High Current Injector HSI with an accepted mass over charge ratio for heavy ions up to 65. A 36 MHz RFQ and two IH drift tube tanks will replace the four tank Widerve section. Each of this three cavities will be fed by 2MW final amplifier stages designed at GSI. The duty factor is up to 2% at maximum rf level, the specified rf power fraction into the beam is ranging up to 40%. GSI specified the 200kW amplifier stages, while the major design part as well as the construction was performed by THOMCAST AG. In total six 200kW stages will be under operation, because there is a need for three rebuncher cavities along the UNILAC and the injection line into the synchrotron SIS. The described UNILAC upgrade causes also large modifications within the low level rf equipment, including the master oscillator system and the low level rf distribution. To allow acceptable beam operation, also at unstable beam intensities, the bandwidth of the amplitude and phase control loops has been raised to around 500kHz. The main design steps as well as the status of the rf project will be described.

MO4071

Contribution withdrawn.

MO4072
Klystron RF Stabilization Using Feedforward Circuit

H. HAYANO, M. AKEMOTO, T. NAITO, S. TAKEDA, KEK; M. HIGUCHI, T. SAKAMOTO, D. AIZAWA, Tohoku-Gakuin University

Klystron RF stability in the linac of KEK ATF is important to make beam injection into the damping ring stable. The current pulse-to-pulse performance of the ouput RF is 0.2% peak-to-peak in the amplitude and 1.5 degree in the phase. The linac beam energy is correlated with RF amplitude jitter directly. In order to make more stable beam current storage and extraction, pulse-to-pulse jitter of beam energy should be half of it. By the analysis of RF jitter, the amplitude and the phase jitter are correlated with charging voltage of the modulator. We developed a feedforward circuit which refined the deQ circuit performance with about a half of it. The analysis of RF jitter and feedforward circuit test are described.

MO4073
Klystron Modulator for the KEKB Injector Linac

H. HONMA, T. SHIDARA, S. ANAMI, K. NAKAHARA, KEK

An upgrade of the klystron modulator for the KEKB injector linac has been completed until March, 1997. SLED (SLAC Energy Doubler) along with 46-MW klystron is utilized to accelerate electron and positron beams up to 8 and 3.5 GeV, respectively. It is desirable to increase the pulse flat top of the modulator output pulse voltage for SLED operation. Improvements to the output waveshape had been carried out and a flat-top width of 3.7 micro-seconds was finally obtained. In addition to these, the time jitter of the modulator output pulse voltage was reduced by modifying the driver circuit for a thyratron. This paper describes the details of modulator upgrade and improvements mentioned above.

MO4074
New High Power 200 MHz RF System for the LANSCE Drift Tube Linac*

J. LYLES, C. FRIEDRICHS, M. LYNCH, LANL

The LANSCE linac provides an 800 MeV direct H⁺ proton beam, and injects H^- to the upgraded proton storage ring for charge accumulation for the Short Pulse Spallation Source. Accelerating these interlaced beams requires high average power from the 201.25 MHz DTL RF system. Three power amplifiers have operated at up to three Megawatts with 12% duty factor. Recent modifications to the 201.25 MHz RF plant have significantly improved the system availability. The total number of electron power tubes in the RF amplifiers and their modulators has been reduced from fifty-two to twenty-four. The plant continues to utilize the original design of a tetrode driving a super power triode. Further increases in the linac duty factor are limited, in part, by the maximum dissipation ratings of the triodes. A description of the system modifications proposed to overcome these limitations includes new power amplifiers, using low-level RF modulation for tank field control. The first high power Diacrode® was obtained and a new amplifier is being designed. With only eight power tubes, the new system will deliver both peak power and high duty factor, with lower mains power and cooling requirements. The remaining RF components of the new system will be discussed.

*Work supported by the U.S. Department of Energy.

MO4075
Accelerator Production of Tritium 700 MHz and 350 MHz Klystron Test Results*

D. REES, M. LYNCH, P. TALLERICO, LANL

The Accelerator Production of Tritium project (APT) utilizes a 1700 MeV, 100 mA proton Linac. The radio frequency (RF) power is provided by 244 continuous wave (CW) klystron amplifiers at 350 MHz and 700 MHz. All but three of the klystrons operate at a frequency of 700 MHz. The 350 MHz klystrons have a nominal output power of 1.2 MW at a DC-to-RF conversion efficiency of 65%. They are modulating-anode klystrons and operate at a beam voltage and current of 95 kV and 20 A. The design is based on the CERN klystron. The 700 MHz klystron is a new development for APT. Three 700 MHz klystrons are currently under development. Two vendors are each developing our baseline klystron that has a nominal output power of 1.0 MW at a DC-to-RF conversion efficiency of 65%. A 700 MHz klystron is also under development that promises to provide an efficiency in excess of 70%. The 700 MHz klystrons operate at a maximum beam voltage of 95 kV and a maximum beam current of 17 A. The test results of these klystrons will be presented and the design features will be discussed.

*Work supported by the U.S. Department of Energy.

MO4076
Electromagnetic Cold-Test Characterization of the Quad-driven Beam Deflection Kicker

S. D. NELSON, LLNL

The first kicker concept design for beam deflection was constructed to allow stripline plates to be driven thus directing, or ``kicking'', the electron beam into two subsequent beam lines. This quad-driven kicker is an eight port electromagnetic network and consists of two actively driven plates and two terminated plates. Electromagnetic measurements performed on the bi-kicker and quad-kicker were designed to determine: 1) the quality of the fabrication of the kicker including the component alignments; 2) quantification of the input feed transition regions from the input coax to the driven kicker plates; 3) identification of properties of the kicker itself without involving the effects of the electron beam; 4) coupling between the line current source and the plates of the kicker; and 5) effects on a driven current to simulate an electron beam through the body of the kicker. Included in this are the angular variations inside the kicker to examine modal distributions. The goal of the simulated beam was to allow curved path and changing radius studies to be performed electromagnetically. The cold test results produced were then incorporated into beam models [1].

Classification Category: T04

MO4077
Solid-State Switch Modulator Deck for the MIT-Bates S Band Transmitter

C. WOLCOTT, A. ZOLFAGHARI, R. CAMPBELL, A. HAWKINS, W. NORTH, L. SOLHEIM, R. TREPSAS, MIT-Bates; M. GAUDREAU, Diversified Tech., Inc.

This paper describes how to modernize and simplify the design of a pulse amplitude modulator for klystron power amplifiers. The exisiting modulator design uses two parallel-connected Litton Injectron Beam Switch Tubes (BSTs) in series with the cathode of an RF amplifier klystron. The vacuum-tube circuitry that produces the high-voltage modulating-anode pulses for the BSTs was state-of-the-art thirty years ago, but in recent years has proven difficult and expensive to maintain. The new design replaces this circuitry with a single switch, comprising multiple series-connected, high-voltage, high-current Insulated-Gate-Bipolar Transistors (IGBTs). The switch is connected in series with the BSTs in a cathode-switching configuration. The modulating anodes of the BSTs are connected to an adjustable highly regulated high-voltage DC power supply. The prototype of this system along with its control and feedback circuitry has been built at the Bates Linear Accelerator Center and has been successfully tested in one of the Bates accelerator RF transmitters. The system has produced klystron beam current pulses of 80 Amps with a width of 25 microseconds at 600 pulses per second, and has injected RF power output from the klystron into the accelerator beam line. The new cathode switching design will not only replace obsolete and failing technology but will also significantly improve the reliability and performance of the accelerator and RF transmitters.

MO4078  (oral poster MO3009)
Design, Analysis and Testing of a High Thermal Conductivity Waveguide Window

T. SCHULTHEISS, V. CHRISTINA, M. COLE, J. RATHKE, Q. SHU, Northrop Grumman Corp.; T. ELLIOT, V. NGUYEN, L. PHILLIPS, J. PREBLE, TJNAF

Design, analysis, and testing of a waveguide window capable of propagating greater than 100 kW average power operating at 1500 MHz has been performed. This is made possible by the high thermal conductivity of the BeO window. Brazing the window to a soft copper frame and then brazing the frame to a flange provides the vacuum seal. RF analysis combined with thermal and structural analysis shows the benefits of thermally conductive materials. RF testing has shown that this system provides high levels of RF power with a small temperature rise within the window. The basics of this design can be expanded to applications with lower frequencies and higher average power.

MO4079
Design Considerations for Very High Power RF Windows at X-Band

W. R. FOWKES, R. S. CALLIN, E. N. JONGEWAARD, D. W. SPREHN, S. G. TANTAWI, A. E. VLIEKS, SLAC

RF window designs of the Pillbox type are capable of transmitting peak rf power up to about 10 MW. The power levels now being produced by advanced high frequency power sources are beyond the level that can safely be transmitted through a single rf window of conventional design. New approaches are required to keep the rf electric fields at a manageable level in the vicinity of the rf window. This paper describes some of the recent rf window designs at SLAC with some of the test results. Windows that operate in larger waveguide in higher order modes such as TE01 and in a mix of modes such as TE11, TM11 and TE12 are described. RF chokes and corona shields, circular polarization and forced electric field reduction are also discussed.

MO4080
Strategies for Waveguide Coupling for SRF Cavities*

L. DOOLITTLE, TJNAF

Despite widespread use of coaxial couplers in SRF cavities, a single, simple waveguide coupling can be used both to transmit generator power to a cavity, and to remove a large class of higher order modes produced by the beam. There are balances and tradeoffs to be made, such as the coupling strength of the various frequencies, and the transverse component of the coupler fields on the beam axis, and the magnitude of the surface fields and currents. This paper describes those design constraints, categories of solutions, and examples from the CEBAF Energy Upgrade studies.

* Work supported by the U.S. DOE Contract # DE-AC05-84ER40150.

MO4081
Systematic Design of an S-band Pillbox-Type RF Window

A. JOESTINGMEIER, M. DOHLUS, N. HOLTKAMP, DESY

The scattering matrix technique and the MAFIA computer code are employed for the systematic design of a 75MW S-band pillbox-type window. It is shown that with the standard pillbox-type design the window cannot be matched using ceramic disks with a thickness in the range from 4mm to 8mm. Nevertheless such disks are mechanically more robust and easier to manufacture than the usual 3mm disks, especially if high purity alumina is considered. Therefore a new design of the pillbox-type window with an inductive iris approximately one quarter wavelength in front of the transition from the rectangular to the circular waveguide is presented so that ceramic disks of arbitrary thickness may be used. Furthermore, it is demonstrated how the bandwidth of the window can be optimized by fine tuning the thickness of the ceramic disk. The edges at the waveguide transition give rise to field singularities. The electrostatic module of the MAFIA computer code is applied to investigate the local field enhancement at a smoothed edge. The final design of the window is validated by the application of the MAFIA time domain module.

MO4082
Design of a HOM Broadband Absorber for TESLA

A. JOESTINGMEIER, M. DOHLUS, N. HOLTKAMP, H. HARTWIG, DESY

For the TESLA FEL operation very short intense bunches of electrons have to be accelerated. These bunches excite a broad spectrum of HOM up to frequencies of some THz. Two HOM couplers per cavity are foreseen in the present design proposal of TESLA in order to extract some of these HOMs from the accelerating structure. In this contribution an additional HOM broadband absorber is suggested which is to be installed between two cryogenic modules at 70K. Its task is to prevent that the really high frequency HOM are absorbed somewhere in the accelerator at the 2K level where the heat capacity of helium is very low. The performance of various structures is investigated using MAFIA. It is found that the additional contribution to the beam impedance which corresponds to the proposed absorber is small. The best attenuation of HOM is obtained for a structure consisting of a stack of metallized SiC washers. Nevertheless other structures seem to be more favourable from the manufacturing point of view which is also discussed.

MO4083
Biperiodic Disk-and-Washer Cavity for Electron Acceleration

H. AO, Y. IWASHITA, T. SHIRAI, A. NODA, M. INOUE, ICR-Kyoto Univ.; T. KAWAKITA, M. MATSUOKA, Mitsubishi Heavy Industries, Ltd.

The biperiodic Disk-and-Washer (DAW) cavity for electron acceleration has been studied. The DAW structure has features in high stability, good vacuum properties and high shunt impedance. A high power model of DAW consists of two 1.2m long accelerating tubes connected by a coaxial bridge coupler. The coaxial bridge coupler has an RF coupler, vacuum port, and three frequency tuners. The operating frequency is designed at 2857MHz so that it can replace one of three existing disc-loaded waveguide accelerating tubes. Dimensions of DAW were optimized by 3-D calculations and aluminum model measurements. All of the parts are made of OFC for the high power model. Fabrication techniques and details will be described.

MO4084
The Estimations for Mechanical Vibrations of Stem-Like Elements in RF Cavities

A.S. LEVCHENKO, V.V. PARAMONOV, R.S. TER-ANTONYAN, INR RAS

In such elements of accelerating cavities as stems, posts, spirals, split rings, low frequency (several tens Hz) mechanical oscillations may exite. Analytical expressions for resonant frequencies are presented. If the channel for cooling liquid is placed inside the element, depending on flow parameters the source of noises may exist from turbulence. Estimations for flow parameters and possible spectrum of noises are given.

MO4085
A Fabrication of the C-band (5712 MHz) Choke-Mode Type Damped Accelerator Structure

H. MATSUMOTO, T. SHINTAKE, N. AKASAKA, KEK

In 1992, T. Shintake proposed an original Choke-mode type damped accelerating structure. In 1994 at KEK, the first high-power model at S-band frequency was successfully operated and accelerated a beam at 50 MV/m with 120 MW of peak rf power. In 1996, hardware R&D on C-band rf system for linear collider started at KEK. The C-band accelerating structure of this type is under development at Mitsubishi Heavy Industries, Inc. This will be a complete model of 1.8 m-long, which is equipped with the HOM absorbers made by SiC, the double feed couplers and integrated BPMs. The wake field damping performance will be investigated at SLAC-ASSET in this year.

Classification Category: T05

MO4086
An Electro-plating Fabrication Method of Electron Accelerating Structures

H. MATSUMOTO, T. SHINTAKE, KEK; Y. IINO, Z. KABEYA, Mitsubishi Heavy Industries, Inc.

The electro-plating fabrication method is the original technique developed by Mitsubishi Heavy Industries, Inc. for the accelerating structure, collaborated with INS Tokyo University in 1970. It has been currently applied for the KEK-2.5-GeV injector linac for 15 years without problem. It is now applied to the new project at KEK; the KEKB injector linac, which extends the beam energy 2.5 to 8 GeV. The electro-plating technology, plates a thick copper layer (5 mm) at the outside of accelerator structure to assemble disks and cylinders. The vacuum seal is kept by the plated copper without brazing process. Advantage of this method is that the negligible frequency during this process, therefore no tuning is required. An integrated phase error for 51 cavities (1.8 m-long) is typically within 2.5 degree at the S-band frequency. Additionally this process does not produce excess deformations on the structure, resulting in a good straightness, which will be an important feature in the next generation scientific applications.

MO4087
Status of Engineering Development of CCDTL for Accelerator Production of Tritium*

R. L. WOOD, J.H. BILLEN, P.O. LESLIE, R.J. ROYBAL, F.E. SIGLER, LANL

The Coupled-Cavity Drift Tube Linac (CCDTL) is a relatively new rf accelerator structure which plays a major role in the APT Low-Energy Linac design. Engineering development is pushing ahead on several fronts, including thermal management, fabrication procedures, cavity and coupling slot tuning, high-power prototype fabrication and testing, supports and alignment, vacuum, and provisions for beam diagnostics. Fabrication of the CCDTL Low-Beta Hot Model is nearly complete, and high-power rf tests will commence soon. In 1999, we will begin the fabrication of 11 meters of CCDTL to be added to the Low-Energy Demonstration Accelerator. In 2001, it will take the 100mA beam from 6.7MeV to 10.05MeV, producing the world's most powerful proton beam. We are also starting the design of a CCDTL High-Beta Hot Model to demonstrate cooling of a 96MeV version of the structure. (The 14cm-long 9cm diameter drift tube has roughly 5kW dissipated on it.) This all leads to the final mechanical design of the 113m long CCDTL for the APT plant linac.

*Work supported by the US DOE under the APT program.

MO4088
Tuning the LEDA RFQ 6.7-MeV Accelerator*

L. M. YOUNG, L. J. RYBARCYK, LANL

This paper presents the results of tuning the 8 meter long Radio Frequency Quadrupole (RFQ) built for the Low Energy Demonstration Accelerator (LEDA). This RFQ resonating at 350 MHz is split into four 2 meter-long-RFQ's. Then they are joined with resonant coupling to form an 8 meter-long-RFQ. This improves both the longitudinal stability and the transverse stability of this long RFQ. The frequencies of the modes near the RFQ mode are measured. We show the effect on the RF fields of an error in the temperature of each one of the 2 meter-long-RFQ's. Slug tuners distributed along the outer walls tune the RFQ. The program RFQTUNE [1] is used to determine the length of the tuners. The tuners are machined to length when the final tuning is complete.

*Work supported by the U. S. Department of Energy.

Classification Category: T05

MO4089
Temperature Distribution Calculations on Beryllium Windows in RF Cavities for Muon Colliders

D. LI, J. CORLETT, W. TURNER, LBNL

We report the temperature distribution calculations on beryllium windows in RF cavities for muon colliders. The cavity resembles a closed pillbox cavity with the conventional beam iris apertures covered by thin beryllium foils to enhance the electric fields on the beam axis. The cavity may be operated at liquid nitrogen (LN) temperature to reduce the rf power losses. Heating caused by the RF power dissipations on the beryllium windows is a concern for both the room and LN temperature operations. Based on MAFIA simulations, considering the heat conduction in the beryllium windows only, a pillbox and an approximated "linear" models are used to solve the temperature distribution analytically. Formulae are given for the temperature distributions on flat and tapered windows. Preliminary calculations indicate that no special cooling designs are needed for proposed 0.125 mm thickness windows.

MO4090
Conceptual Design of the SNS RFQ*

A. RATTI, C. FONG, M. FONG, R. MCGILL, R. GOUGH, J. STAPLES, M. HOFF, R. KELLER, S. VIROSTEK, R. YOURD, LBNL

The conceptual design of the RFQ included in the front end injector of the Spallation Neutron Source is described. The RFQ operates at 402.5 MHz, with a maximum H^- input current of 70 mA, 6% duty factor. It is 3.72 m long and made out of four equally long sections. A brazed copper structure has been chosen due to the high power, high duty factor operation. The 1.25 MW peak rf power is coupled into the structure via eight ports, two per section. A set of tuners is provided for final frequency adjustment and local field perturbation correction. Quadrupole mode stabilization is obtained with a set of P-mode stabilizing loops. The conceptual design, assembly processes and status report are presented.

* This work is supported by the U.S. Department of Energy under contract no. DEAC0376SF00098.

MO4091
Test Results for a Cold Model of a CCDTL Two Gap to Three Gap Transition Region to be used in the Accelerator Region*

M. COLE, H. BLUEM, Northrop Grumman Corp., J. BILLEN, P. LESLIE, LANL

As part of the APT project we have tested cold models of a CCDTL 2 to 3 gap transition region. This is the region where cavities change from having 2 gaps and 1 drift tube per cavity to having 3 gaps and 2 drift tubes per cavity. We have established that frequency effects of coupling slots can be accurately predicted allowing calculation of the cavity frequencies within the allowed tuning range. The base cavity frequency is very accurately predicted by Superfish and correction for stems and tuners is well established; the major uncertainty prior to these tests was predicting the frequency effect of the coupling slots. These tests have also shown that the cell to cell coupling for a given slot geometry can be predicted accurately enough to allow a minimum of slot tuning range to be left in the structure. We have also found that a large desired field tilt between the two coupled accelerating cavities can be set up by varying the cell to cell couplings between the accelerating cells and the coupling cells. Further the tilt can be predicted accurately enough to allow a minimum of slot tuning range to be left in the structure.

*Work supported by the U.S. DOE under the APT LEDA program

MO4092  (oral poster MO3010)
The Dipole Wake Field for a Rounded Damped Detuned Linear Accelerator Structure with Optimised Cell-to-Manifold Coupling

R.M. JONES*, SLAC; N.M. KROLL* **, SLAC & UCSD; R.H. MILLER*, Z. LI*, J. WANG*, SLAC; T. HIGO, KEK; J. IRWIN*, SLAC

A redesign of the basic cell configuration of the Damped Detuned Structure has been briefly reported in [1] where the cells are referred to as ellipsoidal cavities, and accelerator structures incorporating them are designated DDS 5 and 6. This new structure type has been renamed RDDS, and the first of this series, RDDS 1, is presently under design and fabrication. The carefully sculpted cell profile (fabricated on computer controlled lathes incorporating diamond point machining) provides a 20% increase in shunt impedance which, when combined with other parameter changes, allows for a dramatic reduction in the RF power required for NLC. The detuning profile, the damping manifold taper, and the cell to manifold coupling constant profile have all been carefully optimised so as to permit decoupling the cells at the ends of the structure from the manifolds while still adequately minimizing the transverse wake. The decoupling is required in order to fit adequately matched terminations for the manifolds into the structure. The single structure analysis has been supplemented with studies of wake degradation arising from random fabrication errors and wake improvement obtained by combining manifold damping with structure interleaving.

* Supported by U.S. DOE grant number DE-FG03-93ER40759.
** Supported by U.S. DOE grant number DE-AC03-76SF00515.

Classification Category: T05

MO4093
The Effect of Periodic Misalignments of Cells on the Wake Field and on Beam Based Structure Alignment for Advanced Damped Detuned Linear Accelerator Structures

R.M. JONES*, SLAC; N.M. KROLL**, SLAC & UCSD; R.H. MILLER*, SLAC; T. HIGO, KEK; Z. LI*, J.W. WANG*, SLAC

The damped detuned structure consists of 206 cells and four manifolds running along the outer edge of the cells. In order that the beam be centered precisely (and hence transverse wake field effects be minimized) it is required that the cells be well-aligned. It is shown that should the structure possess a periodic misalignment in the cells (i.e., alternating cell to cell displacement), the phase advance per cell shifts by Pi and hence the interaction shifts towards the the 0 mode. This gives rise to a large kick factor contribution from the upper dipole band and this in itself will cloud the interpretation of the signals received from the manifold. Furthermore, the structure has been designed to damp out the lower dipole mode and as the wake will in these circumstances be largely in the upper dipole band it will be inefficiently damped. Here, we present calculations of the wake field and an interpretation of the energy radiated to the manifold in this oscillating structure case. The need to carefully align structures to ameliorate these effects is pointed out.

* Supported by U.S. DOE grant number DE-FG03-93ER40759.
** Supported by U.S. DOE grant number DE-AC03-76SF00515.

MO4094
Application of a Mapping Function Technique to the Rapid Calculation of Dipole Wake Fields in Damped Detuned Structures

R.M. JONES*, SLAC; N.M. KROLL** * , SLAC & UCSD; R.H. MILLER*, SLAC

In order to calculate the wake field in the damped detuned structure (DDS) we require the circuit parameters of each of the 206 cells which constitute the 1.8 m structure of the present design. Initial calculations proceed using the MAFIA code to obtain the dispersion curves for the 11 fiducial cells. However, this is a time consuming process (and also memory intensive). Here we parameterize all quantities so that they are functions of the synchronous frequency of the cells and we then utilise a mapping technique for successive re-designs of the structure. The new cell dimensions and circuit parameters are readily obtained, and the phase of the fundamental accelerating mode is preserved under the mapping.

* Supported by U.S. DOE grant number DE-FG03-93ER40759
** Supported by U.S. DOE grant number DE-AC03-76SF00515.

MO4095
Optimum Operating Temperature of Superconducting Cavities

H. SAFA*, LANL and CEA/Saclay

Superconducting radiofrequency (SCRF) cavities require cryogenic temperature to operate in the superconducting state where the RF power losses are extremely low. But this power has to be removed at very low temperatures using a refrigerator. While reducing the operating temperature results in lower losses for the cavities, it may severely impact the cryogenic plant, drastically reducing its efficiency. On the other hand, the exponential dependence of the BCS surface resistance with temperature will ask for much higher refrigeration power at higher temperatures. Therefore, an optimum working temperature results, depending on frequency, accelerating field and cavity performance. This temperature is determined using a cost optimization including the capital cost of the refrigerator and its operation cost integrated over the expected life time of the accelerator.

* Work performed while on assignment at Los Alamos National Laboratory.

MO4096  (oral poster MO3011)
Basic Research on Horizontal Assembly Method for SC Cavities with High Q and High Gradient

K. SAITO, E. KAKO, T. SHISHIDO, S. NOGUCHI, M. ONO, H. INOUE, T. FUJINO, Y. YAMAZAKI, KEK; M. SHIRATAKE, NP

In the application of superconducting rf cavities for the linear collider (e.g., TESLA) or proton linacs for the nuclear waste transmutation (NWT), high Q and high gradient performance is demanded: Q=10¹⁰, Eacc=30MV/m (TESLA), 10MV/m (NWT). Recently such excellent intrinsic cavity performances: Q>10¹⁰, Eacc > 30 MV/m have been reliably realized in the vertical cold test. To establish the performances in the accelerator, horizontal cavity assembly into the cryo-module is a very important issue. In such a high Q performance, particulate contamination problem from opening the cavity after the cold test or from accessories; input couplers, HOM couplers which are put on the cavities in the assembly, is very serious even assembled in the clean room. We have started a basic research to take cures for the problem in the horizontal cavity assembly. In this paper, the relationship between the leak rate of the introduced nitrogen gas to open cavities and the field emission or effects of the high pressure water rinsing after putting the accessories on the cavity, and so on will be presented.

Classification Category: T06

MO4097  (oral poster MO3012)
Simulation of High-Average Power Windows for Accelerator Production of Tritium*

L. D. DAILY, C. C. SHANG, D. J. MAYHALL, S. D. NELSON, LLNL; K. A. CUMMINGS, LANL; J. SALEM, NASA Lewis Center

Development of a robust, high-average-power (250 kW, CW) microwave transmission line system for the Accelerator Production of Tritium (APT) facility is a stringent engineering and operational requirement. One key component in this RF transmission system is the vacuum barrier window. The requirement of high-power handling capability coupled to the desirability of good mean time to failure characteristics can be treated substantially with a set of microwave, thermal-structural, and Weibull analysis codes. In this paper, we examine realistic 3-D engineering models of the ceramic windows. We model the detailed cooling circuit and make use of accurate heat deposition models for the RF. This input and simulation detail is used to analyze the thermal-structural induced stresses in baseline coaxial window configurations. We also use a Weibull-distribution failure prediction code (CARES), using experimentally obtained ceramic material failure data and structural analysis calculations, to infer mean-time-to-failure.

*The work was performed under the auspices of the U.S. Department of Energy by LLNL under contract W-7405-ENG-48.

MO4098
An Accelerator-assisted Nuclear Fuel Assembly for a Future Project at KURRI

Y. KAWASE, S. SHIROYA, RRI-Kyoto Univ.; M. INOUE, ICR-Kyoto Univ.

A hybrid system of an accelerator and a nuclear fuel assembly has been proposed as a future neutron source at Kyoto University Research Reactor Institute (KURRI). The first step of the project composed of a 20MeV deuteron linac and an existing small critical assembly is considered as a teststand for the final goal which is composed of a 300MeV proton linac and a 5MW subcritical nuclear fuel assembly. The deuteron linac can accelerate H₂⁺ ions. Therefore it may be used as an injector of the 300MeV proton linac. The linac of each step can be also used independently for many purposes. The subcritical nuclear fuel assembly multiplies more than ten times neutrons generated by the linac. The hybrid system promotes the basic study of a nuclear engineering system to solve a future nuclear fuel problem such as thorium cycle and nuclear waste incineration. On the other hand, intense neutrons controlled in energy and pulse width by this system are expected to bring new advanced research fields such as neutron scattering experiments by pulsed neutrons, elementary particle physics by ultra cold neutrons and the boron neutron capture therapy (BNCT) by epithermal neutrons.

MO4099
Design of 1 GeV, 30 mA Proton Linac with Superconducting Cavities

V.M. BELUGIN, B.I. BONDAREV, A.P. DURKIN, A.P. FEDOTOV, Y.D. IVANOV, B.P. MURIN, I.V. SHUMAKOV, N.I. UKSUSOV, MRTI

Main parameters and principal decisions on CW 1 GeV, 30 mA proton linear accelerator are presented. First part of accelerator (up to the energy of 50 MeV) operates at room temperatures on a frequency of 325 MHz. Second part of accelerator (1056 MHz operating frequency) is based on 9-cells superconducting cavities; accelerating rate along the cavity is 10 MeV/m. Accelerator efficiency is estimated. Accelerator design is developed for electronuclear power installation with sub-critical blanket intended for weapon plutonium conversion or for "energy amplifier" in thorium cycle.

MO4100
Bremsstrahlung Pair-Production of Slow Positrons with Low Neutron Background*

E. LESSNER, M. WHITE, ANL

Minimization of component activation is very desirable at linac-based positron sources. Electrons in the 8- to 14-MeV energy range impinging on a target produce photons energetic enough to create electron-positron pairs; however very few of the photons are energetic enough to produce photoneutrons. Slow positron production by low-energy electrons impinging on a multilayer tungsten target with electromagnetic extraction between the layers was studied by simulation. The neutron background from 14-MeV electrons is expected to be significantly lower than that encountered with higher-energy electron beams. Numerical results are presented and ideas for a low-activation slow-positron source are discussed.

* Work supported by the U.S. Department of Energy, Office of Basic Energy Sciences, under Contract No. W-31-109-ENG-38.

MO4101
The KEK-PF Slow-Positron Facility at a New Site

T. SHIDARA, T. KURIHARA, A. SHIRAKAWA, A. ENOMOTO, H. KOBAYASHI, K. NAKAHARA, KEK

The KEK-PF slow-positron facility was relocated to the 1.5-GeV point of the upgraded electron/positron injector linac relevant to the KEKB project. A dedicated linac for slow-positron use is also installed utilizing the remnants of the injector linac reformation. We expect a slow-positron intensity of more than 100 million positrons/sec with a maximum primary beam power of 1 kW for their application in various fields of solid-state physics.

MO4102
Faraday Cup Measurements of the Plasma Plume Produced at an X-ray Converter*

T.L. HOUCK, M. GARCIA, S. SAMPAYAN, LLNL

The next generation of radiographic machines based on induction accelerators is expected to generate multiple, small diameter x-ray spots of high intensity. Experiments to study the interaction of the electron beam with the x-ray converter are being performed at the Lawrence Livermore National Laboratory using the 6-MeV, 2-kA Experimental Test Accelerator electron beam. The physics issues of greatest concern can be separated into two categories. The multiple pulse issue involves the interaction of subsequent beam pulses with the expanding plasma plume generated by earlier pulses striking the x-ray converter. The plume expands at several centimeters per microsecond and defines the minimum transverse spacing of the pulses. The single pulse issue is more subtle and involves the extraction of light ions by the head of the beam pulse. These light ions could propagate at velocities of several millimeters per nanosecond through the body of the pulse resulting in a moving focus prior to the converter. In this paper we describe Faraday cup measurements performed to quantify the plasma plume expansion and velocities of light ions.

* The work was performed under the auspices of the U.S. Department of Energy under contract W-7405-ENG-48.

MO4103
Experimental Investigation of Beam Optics Issues at the Bremsstrahlung Converters for Radiographic Applications*

G. A. WESTENSKOW, G. J. CAPORASO, Y.-J. CHEN, L. HOUCK, A. C. PAUL, S. SAMPAYAN, R. RICHARDSON, J. T. WEIR, LLNL; C. E. CRIST, SNL

To obtain the desired x-ray source for advanced radiographic applications the beam optics for linear induction accelerators will need to maintain the electron beam size at the bremsstrahlung converter at a constant spot size under 1 mm. Light ions emitted from the converter's front surface by the field from the electron beam may pose a problem even for the existing one shot machines. We are performing experiments on the ETAII accelerator at LLNL to investigate issues associated with this problem. We will be studying whether we can bias the converter to stop ion emission from the converter or use plasma focusing under these conditions. We are also studying issues raised when several pulses per view will be required in future radiographic machines.

*The work was performed under the auspices of the U.S. Department of Energy by LLNL under contract W-7405-ENG-48.

MO4104
RF Power Distribution and Phasing at SSRL Injector Linac

S. PARK, J.N. WEAVER, SLAC/SSRL

At the Stanford Synchrotron Radiation Laboratory injector linac, one thermionic RF gun and all three linac sections were powered by three XK-5 klystrons in 1990. After five years of operation, one SLAC type 5045 klystron replaced two XK-5's. To ensure stability and reliability, the RF power from the 5045 was branched out to the last linac section in 1997, thus having one klystron powering all the loads. During the process, a 5 dB waveguide directional coupler was developed, and RF phasing was done by cold tests and by beam based tuning at full power. The remaining XK-5's are currently being utilized at the Gun Test Facility, located in the SSRL linac vault.

Classification Category: A01

MO4105

Contribution withdrawn.

MO4106
Induction Core Performance*

A.W. MOLVIK, W.R. MEIER, LLNL; A. FALTENS, L.L. REGINATO, LBNL; C.H. SMITH, Nonvolatile Electronics, Inc.

Large masses of magnetic core material are required for many of the induction accelerator-based projects currently under study; the quantities required exceed 1E7 kg for a linear heavy-ion fusion driver, so core performance and cost are critical issues. We have evaluated cores of amorphous alloys from AlliedSignal and MRTI (Moscow RadioTechnical Institute) and nanocrystalline alloys from Hitachi and Vacuumschmelze. The cores were of moderate size, between 1 and 11 kg. We characterized the materials in terms of the flux swing dB from -B(remanent) to +B(saturation), and the energy loss versus dB/dt. We found sources for each material that could coat, wind, and then anneal the cores. This required the development of thin coatings that withstand 350-550 deg. C anneal temperatures. The result is core performance near the ultimate small sample performance of each material, with higher dB and lower losses than the earlier approaches of using as-cast material or rewinding after anneal, in both cases usually cowinding with thin mylar (~4µm thick). We are beginning system code studies of tradeoffs between dB and losses.

*This work performed under the auspices of the U.S. DOE by LLNL under Contract W-7405-ENG-48, and by LBNL under Contract DE-AC03-76F00098.

Classification Category: T08

MO4107
High Resolution BPM for Future Colliders

M. JUILLARD, Y. LUSSIGNOL, C. MAGNE, A. MOSNIER, B. PHUNG, CEA/Saclay

A beam position monitor using a coaxial reentrant RF cavity has been designed and tested following the principle initially proposed by R.Bossart at CERN. Advantages are : the axial symmetry of the cavity, high precision, excellent linearity. The off-resonance operation allows high resolution especially in the case of short intense electron bunches. The BPM has been tested on the injector #1 of the TTF linac at DESY (low charge, 216 MHz bunches). The resolution is 10 microns. On the injector #2, which will have 8 nC, 1 MHz bunches, the expected resolution is 0.7 micron.

Classification Category: T02

MO4108
Heat Transfer Coefficient in Serpentine Coolant Passage for CCDTL

P. LESLIE, R. WOOD, F. SIGLER, A. SHAPIRO, A. RENDON, LANL

A series of heat transfer experiments were conducted to refine the cooling passage design in the drift tubes of a coupled cavity drift tube linac (CCDTL). The experimental data was then compared to numerical simulation data to determine differences between analytical solutions for long, straight, circular tubes and realistic curved, short, rectangular channels. Different flow rates were used to provide a larger spectrum of cooling conditions in the experimental drift tube and the coolant inlet and delta (outlet minus inlet) temperatures were recorded. Also recorded were two drift tube body temperatures and power input from cartridge heaters placed around the drift tube. The numerical simulation consisted of iterations of an axisymmetric model where the boundary conditions were modified to match the experimental data. The results were then used to determine the coefficients of the Nusselt number equation which is then used in the determination of the heat transfer coefficient. The determined coefficients are valid only for this specific geometry. The entire thermal analysis will be presented.

Classification Category: T08

MO4109
Complete Dynamic Focusing for Linear Collider

J. IRWIN, SLAC

Dynamic focusing refers to the use of secondary (lens) beams to form the final focus lenses for the primary (main) beams. Complete dynamic focusing refers to the situation where an initial lens-lens beam collision focuses these beams for their collision with the primary beam. This paper describes the techniques for the formation of a uniform lens shape from an initial Gaussian shape, the necesary main- and lens-beam parameters and their scaling, hardware requirements for a 1 TeV c.m. application, operations and tuning issues, and a proposal for a demonstration experiment at SLAC. Advantages of this scheme include the complete elimination of the conventional final focus and collimation systems, elimination of any beamline elements within the detector, and the loosening of linac alignment and ground motion tolerances by more than a factor of 10.

Classification Category: A03