8/14/2019 dodd 1991 0002

1/3

SATURNUS: The UCLA Infrared Free-Electron Laser ProjectJ.W. Dodd, S.C. Hartman, S. Park, C. Pellegrini, J.B. Rosenzweig, J.A. Smolin, G. Hairapetian,J. Kolonko, W.A. Barletta, D.B. ClineCenter for Advanced Accelerators, Department of Physics, U.C.L.A.. Los Angeles,A, 90024-1547

J.G. Davis, CJ. Joshi, N.C. Luhmann Jr.Department of Electrical Engineering, U.C.L.A., Los Angeles, CA 90024-1594

S.N. Ivanchenkov, A.S. Khlebnikov. Yu.Yu. Lachin, A.A. VarfolomeevI.V. Kurchatov Institute of Atomic Energy, Moscow, 123182, U.S.S.RABSTRACT: A compact 20 MeV linac with an RF laser-driven electron gun will be used to drive a high-gain (10cmgain length). 10.6pm wavelength FEL amplifier, operating inthe SASE mode. Satumus will mainly study FEL physics inthe high-gain regime, including start-up from noise, opticalguiding, sidebands, saturation, and superradiance, withemphasis on the effects important for future short wavelengthoperation of FELs. The hybrid undulator was designed andbuilt at the Kurchatov Institute of Atomic Energy in theU.S.S.R. The primary magnetic flux is provided by C-shapediron yokes, where between the poles thin blocks ofneodymium-iron-boron magnets are placed to provideadditional magnetic flux along the undulator axis, The fieldstrength is adjusted by moving the thin Nd-Fe-B blocks on aset-screw mount. The initial assembly will have forty periods,each l&m long. The gap distance between the yoke pole-pieces s fixed at 5mm. The undulator field has been measured,yielding an on axis peak value of 6.6kGauss, which closelymatchescomputer simulations.

Table 1. Satumus design parameters.CTRON BEAMEnergy, nominal 16SMeVEnergy spread, .m.s. 0.2%

Peak current 200AKlystron frequency 2.865GHzPulse repetition rate 5HzMacropulseduration 3.5pecmicropuise duration, r.m.s. =1.6psecChargeper bunch =InCNormalized emittance. r.m.s. 8mm-mradDUUTORDrift tube I.D. within undulator 4mmElectron beamdiameter n undulator 0.4mmPeriod 1.5cmLength 6OcmFixed gap betweenpole pieces 5mmField on axis 6.6kGauss

Wavelength 10.6pmGain length 1OcmSaturation length 13OcmPeaksaturatedpower 5oMwACKNOWLEDGEMENTS: Our group appreciates hehelp received from H. Kirk, K. Batchelor, J. Xie. J. Sheehan,G. Bennett, J. Wurtele, M. Allen, G. Loew, M. Baltay, and H.Hoag. This work is supported by DOE Grant No. DOE-D&A&3-90ER-40583.

I. INTRODUCTIONA compact 16.5 MeV linac with an RF laser-drivenelectron gun[l] is being built at UCLA (See S.C. Hartman, etal.. this Conf.). This linac will be used to study theproduction o f high-brightness electron beams, and to drive ahigh gain, 10.6pm wavelength FEL amplifier, capable ofoperating in the SASE mode. Satumus will use a 1.5cmperiod, 6.6kG peak field undulator developed and built at theI.V. Kurchatov Institute of Atomic Energy. Table 1 showsbeam parameters.Satumus will mainly study FEL physics in the high-gain regime. including start-up from noise, optical guiding,sidebands,saturation, and superradiance, with emphasison theeffects important for future short wavelength operation ofFELs.[2] For the electron beam we will study ways toimprove the beam brightness and peak current. We will havetwo beam lines, one leading straight into the undulator, theother designed for beam diagnostic and longitudinal bunchcompression. This second beam line will also be used forother particle beam physics experiments, including plasmawakefield acceleration.[3]

II. FEL OUTPUT:SIMULATIONS AND DIAGNOSISThe FEL performance n SASE s evaluated using theTDA code developed by Tran and Wurtele at MIT.[4] TDApredicts a gain length of lOcm, with a saturation power of50MW. This assumes that a full 200A peak current isachievable. Fig. l(a) graphs the power versus distance alongthe undulator as projected by TDA. Fig. 1 uses an inputpower of 40mW , corresponding to the spontaneous radiationemitted in one gain length, within an angle 0 = l/(?*dN) and aline width 1/(2N). This radiation is assumed o be fmused lo a

spot radius (XLUNDUI.) If2 141~:t the centre of the first gainlength. Fig. l(b) plots the calculated r.m.s. optical beamradius versus the distance along the undulator. Notice thatonce growth occurs the optical beam collapses and remains inthe vicinity of the electron beam as it propagates down theentire undulator. It is interesting to note that the optical beamradius and the gain length depend on lhe extent of detuning theelectron beamenergy from the resonantenergy.The FELs output wavelength will closely match the10.6pm radiation which CO2 lasers produce, simplifying thediagnosis o f the output radiation by allowing the possibility toinject the FJZL with an external CO2 laser. For cryogenic0-7803.01358/91$01.00 @IEEE 275 1

1991 IEEE. Personal use of this material is permitted. However, permission to reprint/republish this material

for advertising or promotional purposes or for creating new collective works for resale or redistribution to servers

or lists, or to reuse any copyrighted component of this work in other works must be obtained from the IEEE.

PAC 1991

8/14/2019 dodd 1991 0002

2/3

infrared detectorssensitive to lwrn wavelengths, the expectedpower output from spontaneous emission in the FEL(excluding gain) is marginally detectable. For 1nC charge perbunch, 1.2*107 photons per micropulse are expected fromspontaneous emission within the angle and linewidth givenabove over a 1Ocmdistance. For smaller charges one mayneed to integrate the signal over many pulses. As seen romfig. l(a, at the end of the undulator section, roughly 3kW ofpower occurs from exponential gain. This power levelcorresponds to 101* photons per micropulse and is easilydetected.

III. HYBRID UNDULATORThe hybrid undulator was designedand constructed atthe Kurchatov Institute of Atomic Energy in the U.S.S.R.The primary magnetic flux is provided by SmCo5 magnetsplaced n the middle of C-shapedvanadium-permenduryokes.Thin blocks of neodymium-iron-boron magnets are placedbetween the poles to provide additional magnetic flux. Thefield strength is adjusted by moving the thin Nd-Fe-B blockstowards or away from the axis on a set-screwmount, Fig. 2 isa schematic diagram of the Kurchatov undulator. The initialassembly has forty periods, each period being 1.5 cm long.The gap distance between the yoke pole-pieces is fixed at

E+09EM6

z i307c:

E+O62 lz+o52 004

EE+03

0 E+O?LiI E*olLA. E+OO

E-o 9 .4 .6 .a 1.0 11 1.4

UNDULATOR LENGTH(m)

I.2 .4 .6 .6 1.0 1.2 1.4UNDULATOR LENGTH (m)

(a)

(b)

Figure 1. TDA simulations of SASE using the parametersin table 1: (a) power growth through the undulator and (b)radius of the optical beam within the undulator. Opticalguiding is evident since the optical beam remains confinednear he electron beam, once power growth predominates.

Figure 2. The Kurchatov hybrid undulator (g = 5mm).5mm. The complete undulator was measuredand yielded anaverage peak field on axis of 6.6kGauss. A computersimulation using the TOSCAtS] code gave similar results.The deviation o f the peak field within the undulator has beencontrolled within a tolerance of 0.5%.

Fig. 3 shows he measured alues of the perpendicularmagnetic field as measured along the undulators axis. The600mm end includes a compensating taper by tilting theendmost Nd-Fe-B magnets 45. The zero millimeter endshows he field distribution without this compensation.The first phase of construction consists of this single60cm undulator section. The second phase of constructionwill have two separate 60cm undulator sections with adispersive array installed in between the other sections. Thedispersive section provides a shorter distance for the velocitymodulation induced by the first section to enhance thebunching of the electron stream n the ponderomotive we11.[6]IV. ELECTRON BEAM DIAGNOSTICS

A. Current and Posirion:The electron beam position d iagnostic is basedupon adesign used on ATF at BNL. Each monitor consists of fourstripline electrodes placed symmetrically around the electronbeam, nside and electrically isolated from the vacuum vesselvia ceramic, vacuum feed through% The total signal inducedon the four striplines is proportional to the total charge in abunch. The difference of the signals between two diametricallyopposed striplines determines the relative position of theh * t I I,, ti 6 L 7E ,fIIl

!:,!,!!!!.~!;:~~!I,! II 1: 4 - zg2k T:31:i/ 3-

Y3iI~liilliliil~iiiiii~i~~lli~i~i 1-8 -100 0 100 200 300 400 500 600 700Z-AXIS (mm)

Figure 3. Measured B-field configuration along the axis ofthe Kurchotov undulator.2152

PAC 1991

8/14/2019 dodd 1991 0002

3/3

electron bunch from the geometric center of the four striplines.The sum and difference signals are subsequently sent to aheterodyne receiver with local oscillator at 2856GHz. in orderto rectify the signal sent to the preamplifier. Ejch microbunchgeneratesa positive and negative signal of equal magnitude onthe electrode, separated by twice the transit time of theelectrodes ength (i.e., 5OOpsec).[7]The heterodyning processreverses the polarity of the second pulse, therefore the netintegratedcharge s doubled insteadof cancelled n the preamp.B. Emittance, Energy Spread, und Overail Eneqy.Accurate measurements of the electron beamproperties are important to understand he FEL behavior. Thebeam emittance will be merlsuredusing a technique similar tothat developed at ATF.[8] A phosphor screen is placed tointercept the electron beam. The spot size is reflected into anoptical camera. The emittance can be unfolded from thecomparison of the spot sizes at two different settings of thebeam transport system.]91In the dispersive region between two dipole magnets,the energy and energy spread of the electron beam can bemeasured by placing the phosphour screen at the the beamfocus. Since the beam emittance is expected to be low, anangular spread in the spot size at the focus (as seen on thecharge-coupled camera viewing the phosphour screen)can byand large be attributed entirely to the energy spread in thebeam. The averageenergy of the beamcan be determined romlocating the centroid of the electron spot as seen on thephosphour screenand comparing IO he calculated trajectory.

Dienes, AIP Conf. Proc. No. 184 (Am. Inst. of Physics.New York, 1989). pp. 472-524; especially 94.3 on beamposition monitors.[8]D.P. Russell. and K.T. McDonald. A beam-profilemonitor for the BNL Accelerator Test Facility (ATF), Proc.1989 IEEE Particle Accel. Conf., Vol. 3, eds., F. Bennettand J. Kopta. Chicago. IL., pp. 1510-1512.[9]K.T. McDonald and D.P. Russell, Methods of EmittanceMeasurement, in the Pror. of the Jont US-CERN Schoolon Observation, Diaqnosis and Correction in Part. Beums,Capri, Italy, October 20-26, 1988.

C. Temporal Pulse Duration.The time duration of the electron pulse will bemeasuredby passing he electron beam hrough a thin plate ofdiclcctric material. then sweeping the Ccrenkov radiationthereby produced, on the picosecond time scale using a streakClllW-3.

V. REFERENCES[I] K. Batchclor. H. Kirk. J. Sheehan. M. Woodle, and K.McDonald, Proc. Europeon Particle 4ccel. Conf., Rome.Italy, June 7-12. 1988: also, K. Batchelor. et al.. Proc. I989Particle Accel. Conf., Chicago, IL. p. 273.[2] C. Pellegrmi. Nucl. Instr. & Meth. Phys. Res. A272

(1988) 364.[3] J. Smolin, T. Katsouleas. C. Joshi, P. Davis. C.Pellegrini. Design of a plasma wakefield acceleratorexperiment at UCLA, Bui. Am. Phys. Ser. to be publ.Nov. 1990. Cf.. T. Katsouleas. J.J. Su. W.B. blori. C.Joshi. J.M. Dawson, A compact 1OOMeV cceleratorbasedon plasma wakefields, Microwuve and Partlcie BeumSources nnd Directed Energy Conrepts, H.E. Brandt. cd..Proc. SPIE 1061. pp. 428-433 (1989).[4]T.M. Tran and J.S. Wurte le. Free-electron laser simulationtechniques. Physics Reports 195 (1990) l-21.[S]IEEE Proc. Vol. 127. Pt. B. No. 6 (1980).[b]J. Gallardo and C. Pellegrini. Optics Comm.. 77 (1990)35.[7]K.-Y. Ng. Fields. Impedances.and Structures. in Phys. ofPart. Accelerators. Vol. One, eds., M. Month and >I. 2753

PAC 1991