Download A HIGH-BRIGHTNESS THERMIONIC MICROWAVE ... and more Study notes Microwave Engineering and Acoustics in PDF only on Docsity!
c
SLAC- UC-41 4 (A)
A HIGH-BRIGHTNESS THERMIONIC
MICROWAVE ELECTRON GUN*
Michael Borland Stanford Linear Accelerator Center Stanford University, Stanford, CA 94309
February 1991
Prepared for the Department of Energy under contract number DE-AC03-76SF
Printed in the United States of America. Available from the National Technical
- Information Service, U.S. Department of Commerce, 5285 Port Royal Road, Springfield, Virginia 22161 -mm------- *Ph.D. Thesis
Abstract
In a collaborative effort, by SSRL, AET Associates, and Varian Associa,tes, a high- bright.ness microwave electron gun using a thermionic cathode has been designed, built,, t,ested, and installed for use with the SSRL 150 MeV linear accelerat,or. This thesis discusses the physics behind the design and operat,ion of the gun and associat,ed syst,ems, presenting predictions and experimental tests of the gun’s performance. _- The microwave gun concept, is of increasing interest, due t,o its promise of providing higher-current, lower-emittance electron beams than possible from convent.ional, DC gun t,echnology. In a DC guns, accelerat,ing gradients are less than 8 MV/m, while those in a microwave gun can exceed 100 MV/m, providing much more rapid initial accelerat,ion, dhereby reducing the delet,erious effects of space-charge. Microwave guns produce higher moment,um beams than DC guns, thus lessening space-charge effect.s during subsequent beam transport. Typical DC guns produce kinetic energies of 80-400 KeV, compared to 2-3 MeV for the SSRL microwave gun. “Stat,e-of-the-art,” microwave gun designs employ laser-driven photocat,hodes, pro- viding excellent, performance but, with great,er complexity and monetary costs. A thermionic microwave gun with a magnetic bunching system is comparable in cost and comp1exit.y to a conventional system, but, provides performance that, is orders of magnitude bet,ter. Simulations of the SSRL microwave gun predict a normalized RMS emittance at, the gun exit of < 10 T. m,c. pm for a, beam consisting of approximat.ely 50% of the particles emitted from the gun, and having a momentum spread of 110 %. These emit,tances are for up to 5 x loge- per bunch. Chromatic aberrat,ions in the t.ransport’ line between the gun and linear accelerat,or (GTL) increase this t,o t’ypically < 30 7r - m,. pm. (^)... 111
Acknowledgements
As with any project. of the complexit,y of the SSRL microwave gun and its associat,ed syst,ems, the task of going from the concept,ual stage t.o the point where the syst,em is turned over for operational use involves a great, many people. It. is my pleasure to acknowledge the contributions of those who participated in bhe project, that, I am reporting on in this thesis. Wit.h so many people deserving of thanks, there is t.he .danger.^ that.^ 1 have^ overlooked^ someone^ or failed^ to fully^ recognize^ his^ contributions. I have done my best, to avoid this, and apologize in advance if I fail in this regard. First and foremost, I wish t,o thank my thesis advisor Helmut Wiedemann for introducing me t,o t.he exciting field of accelerat,or physics, for providing me with guidance and instruct,ion as I learned about, t,he field, and for giving me the rare opport,unit*y to take on the microwave gun project. I am particularly grat,eful for Helmut’s patience, support, and confidence in me when ot.hers thought. that more conventional t,echnology and a more experienced person were needed. Virt.ually all of the work that. I report on here has benefit,ed from Helmut,‘s suggestions, ideas, scrutiny, and criticism. Jim Weaver of SSRL participated in the design of t.he gun and many components of the gun-to-linac (GTL) syst,em, in addition to designing the RF syst,em for the gun and preinjector. I am convinced that without, Jim’s tireless attention to det.ail and his sound design ideas, the gun, the alpha-magnet chamber, and the chopper would not, have been as successful as they were. I am also grateful to Jim for convertming inches t,o millimet,ers when speaking t,o me and for laughing at most of my jokes. Michael Baltay of SSRL did most of the mechanical engineering for the GTL and directed t,he assembly and alignment as well. All of the GTL magnet.s were designed in collaboration bet.ween Mike and myself (with many helpful ideas from Helmut, John V
Voss, and others). Mike also helped me to assemble the necessary hardware for the magnetic measurement,s and helped with alignment, of the measurement, equipment. Other SSRL and Injector people made significant, cont,ributions to the project: John Voss was the project manager for the Injector, and as such helped to organize the effort, and push us toward completion. Clarence Chavis designed and directed the building of the modulators, and he and Jim Ha.ydon frequently came to the res- cue when something broke. Harry Morales oversaw the vacuum system work in 6he GTL and contribut.ed to the design of many components. Paul Golceff’s meticulous- ness when it. came to the gun vacuum syst,em and his prominent, participation in
- the assembly of the GTL line, alpha-magnet, chamber, and chopper are very much appreciated. Defa Wang assembled much of the equipment, used for running the RF system, did cold test, measurements on RF system components, and calibrated the gun RF diodes; he also proved tireless in the RF processing of the gun and linac, -and in the commissioning of the chopper and linac. Jim Sebek did the TDR of the chopper, ‘tested and refined the toroid design, stayed all night several times during t.he commissioning of the chopper, and, with Bob Hettel, helped to track down noise problems in many of the diagnostic signals. Ra.y Ortiz found and packaged the coaxial swit.ches that, proved so useful during the commissioning and during experiments, and also assembled the necessary circuits for running the gun filament. Bill Lavender’s “test.” cont,rol program for the GTL magnets ended up being the principle means of controlling the GTL during the commissioning and experiments. Brad Youngman, Tom Sanchez, and Ken Ruble participated in t,he design of t,he alpha-magnet chamber, chopper, and other GTL components. Several of Helmut”s ot#her graduate st,udents participated in t.he gun and GTL commissioning as well. Louis Emery in part,icular was indispensible during the crucial early gun experiments at’ SSRL, when many remained to be convinced that, the gun could function reliably. In addition, it, was very valuable to be able to refer to the UTEX input for his thesis in preparing mine. James Safranek was responsible for t,he phosphor screens, TV cameras, and relat.ed hardware, and assumed his share of those seemingly endless RF processing shifts. Subst,antial credit for t.he success of t,he microwave gun goes to Eiji Tanabe of AET Associates and Luther Nelson of Varian Associates. While I performed all of
_
Contents
Abstract
Acknowledgements V
1 Introduction
..^ 1.
Revi.ew of Fundamental Concepts .................... 1.1.1 Phase-Space and Liouville’s Theorem .............. 1.1.2 Beam Emitt,ance ......................... 1.1.3 Beam Brightness ......................... Applications of High-Brightness Beams ................. 1.2.1 Synchrotron Radiation ...................... 1.2.2 Coherent Radiation ....................... 1.2.3 Free Elect,ron Lasers ....................... RF Guns ................................. 1.3.1 Varieties of RF Guns ....................... 1.3.2 A Brief Hist,ory of RF Guns ................... 1.3.3 Fact.ors That, Degrade Electron Beam Brightness ........ 1.3.4 Advant,ages and Disadvantages of RF Guns ........... Overview of Thesis ............................
2 Gun Design and Simulations 39 2.1 Gun Design Overview. ..........................^41 2.1.1 Design Charact,eristics ......................^41 2.1.2 Gun Operating Cycle .......................^44
111^.^.^.
ix
X
5.1 Quadrupole Strengths for GTL Optics Solutions for Various Values of &............................... 265
xv
--
I
List of Figures
1.1 Young’s^ Two-Slit^ Experiment,^.^.^.^.^.^.^.^.^.^.^.^.^.^.^.^.^.^.^.^.^.^.^20 1.2 Schemat,ic Layout, of the Gun-to-Linac Region of the SSRL Preinjector (^36)
.- 2.
Cross Sect,ional View of the SSRL RF Gun............... Pyrometric Measurements of t,he Gun Cathode Temperature..... Gun On-Axis Longidudinal and Radial Electric Field Profiles..... SUPERFISH Field Line Plots for RF Gun Cells............ Profile Used in MASK for the First. Cell................ Profile Used in MASK for the Second Cell............... Longitudinal Field Profiles from SUPERFISH, MASK, and Bead-Drop Measurement ............................... RF Current Waveform for Exciting Cells in MASK .......... Nonlinear Field Terms in the RF Gun ..................
2.10 Gun Longitudinal Phase Space for Various Values of cy........ 77 2.11 Dependence of Beam Properties on o and E,z............. 79 2.12 Dependence of Back-Bombardment Properties on o and E,z..... 80 2.13 Dependence of Beam Properties on RF Frequency........... 82 2.14 Dependence of Longit.udinal Phase-Space on RF Frequency...... 83 2.15 rfgun Predictions for cr = 2.9 and Various Initial Phase Intervals.. 85 2.16 rfgun Transverse Phase-Space Results for Ep2 = 75 MV/m and Q = 2.9 87 2.17 rfgun Longitudinal Phase-Space Results for En2 = 75 MV/m and cr = 2.9 88 2.18 rfgun: Transverse Phase-Space Evolution in the First. Cell, for EpZ = 75MV/mandcr=?.9.......................... 90 xvi --
Beam-sizes for GTL Optics Solution for ii = llcm, from First- and Xix
- 2.1.3 Matching t,o the RF Source - 2.1.4 On-Axis Field Profiles - 2.1.5 Design Goals - 2.2 Simulat,ion Codes and Methodology - 2.2.1 Tuning and Boundary Conditions - 2.2.2 Gun Cavit,y Parameters - 2.2.3 Methodology of MASK Simulations - 2.2.4 Cathode Simulation - 2.2.5 Compensation of Ceil Frequency Mismat.ch - 2.2.6 The rfgun Program - 2.2.7 Off-Axis Field Expansion - 2.2.8 Non-Linear Field Terms - 2.2.9 Boundary Conditions for rfgun
- _. 2.3 Simulation Results and Predictions - 2.3.1 Effects of the Cell Field Ratio - 2.3.2, Effects of bhe RF Frequency - 2.3.3 Effects of Non-Linear Field Terms - 2.3.4 Effects of Parameter Errors in MASK Runs - 2.3.5 rfgun Predictions versus Momentum Spread - 2.3.6 MASK Beam Snapshots - 2.3.7 Calculating Emittance from Cylindrical Coordinates - 2.3.8 Tests of the Independent, Bunch Assumption - 2.3.9 Transverse Beam Evolution - 2.3.10 Accuracy of MASK Field CaIculat,ions - 2.3.11 Adequacy of Off-Axis Expansion - 2.3.12 Space-Charge Effects - 2.3.13 MASK Predictions of Gun Performance
- 3 The Alpha-Magnet - 3.1 Magnetic Characteristics and Design of the Alpha-Magnet - 3.1.1 Asymmetric Quadrupole Design - 3.1.2 Panofsky Quadrupole Design - 3.1.3 Comparison of the Two Designs :
- 3.2 Particle Motion in the Alpha-Magnet - 3.2.1 Scaled Equation of Motion - 3.2.2 Ideal Trajectory - 3.2.3 Numerical Solution of the Equations - 3.2.4 Dispersion and Achromatic Pat,h-Length
- 3.3 Alpha-Magnet, Transport, Matrix Scaling - 3.3.1 Curvilinear Coordinates and Matrix Notation - 3.3.2 Relationships Between Curvilinear and Fixed Coordinates - 3.3.3 Coordinat,e Scaling - 3.3.4 Scaled Equation of Motion with Dispersive Terms - 3.3.5 Scaling of the Transport, Matrices - 3.3.6 Alternative Treatment, of Dispersive Terms
- _- 3.4 Transport Matrices from Numerical Integration - 3.4.1 One-Variable Terms - 3.4.2 Two-Variable Terms - 3.4.3 Three-Variable Terms - 3.4.4 Initial-Vector Ensemble - 3.4.5 Accuracy Considerations and Limits - 3.5 Transport, Makices for the Alpha Magnet - 3.5.1 Program Tests and Choice of Initial Amplitudes Ii - 3.5.2 Final Results Ii - 3.6 Effects of Field Errors - 3.6.1 Multipole Errors - 3.6.2 Entrance-Hole-Induced Errors - 3.7 Alpha-Magnet Beam-Optics Experiments - 3.7.1 Charact,erization of the Steering Magnet - 3.7.2 Comparison of Experimental Results and Theory
- 4 Longitudinal Dynamics - 4.1 Longitudinal Dynamics in Linear Accelerators - 4.1.1 Approximat,e Treat,ment for Highly-RelativisGc Particles - 4.1.2 Numerical Solution and the Contour Approach
- 4.2 Magnetic Bunch Compression - 4.2.1 First, Order Solution for Bunch Compression - 4.2.2 Achieving Momentum-Dependent, Pat,h Lengt,h - 4.2.3 Options for Implementing Magnetic Compression
- 4.3 Optimized Bunch Compression for the RF Gun - 4.3.1 ‘Cise of alpha-opt to Optimize Bunch Compression - 4.3.2 Optimization of the Injection Phase - 4.3.3 Opt.imizations for Various Current, Densities - 4.3.4 Effects of Transport Aberrations - 4.3.5 Comparison wit,h Other Inject,ors
- 5 Gun Experimental Characterization - 5.1 Experimental Configuration
- 5.1.i Gun-to-Linac Components - 5.1.2 Gun-t,o-Linac Instrumentat.ion - 5.1.3 Beamline Control and Data Acquisition - 5.1.4 The Preinjector Linear Accelerator - 5.2 Gun-to-Linac Optics - 5.2.1 Modeling of the Quadrupoles - .5.2.2 Optical Makhing - 5.2.3 Higher-Order Effects - 5.2.4 Experimental Tests of t,he Quadrupole Model - 5.3 Momentum Spectra - 5.3.1 Principle of Spectrum Measurements - 5.3.2 Practical Considerations and Simplifications - 5.3.3 Scraper Calibrat,ion and Sources of Error - 5.3.4 Measurement of Beam Power - 5.3.5 Experimental Results - 5.4 Emittance - 5.4.1 Principle of the Emittance Measurements - 5.4.2 Inclusion of Experimental Errors - 5.4.3 Thin Lens Treatment. - 5.4.4 Emit,tance Measurement Lattice and Procedure - 5.4.5 Analysis of Digitized Beam-Spot Images - 5.4.6 Measures of the Beam Size - 5.4.7 Imaging and Phosphor Resolution - 5.4.8 Choice of Screen Mat.eriaI and Experimental Limitations - 5.4.9 Overview of the Experiments - 5.4.10 Estimation of Uncertainties - 5.4.11 Experiment al Results - 5.4.12 Comparison of Experiments and Simulations - 5.4.13 Possible Sources of Discrepancies
- 5.5 Bunch Length Measurement
- 5.5.1 Principle of Bunch-Length Measurements
- 5.5.2 Practical Considerations
- 5.5.3 Experimental Results
- 5.5.4 Comparison with Simulat#ions
- A Computational Issues and Techniques
- A.1 The mpl Scientific Toolkit
- A.2 The awe Self-Describing Dat,a Format
- Bibliography - xl - 2. - 2. - 2. - 2. - 2. - 2. - 2.
- 75MV/manda=2.9 rfgun: Transverse Phase-Space Evolution in the Second Cell, for En2 = - rfgun: Effects of Different, Non-Linear Fields - rfgun: Effects of (1 and Frequency Errors - Emittance and Brightness for a = 2.9,f = 2856 MHz, and Q = 3, f = - MHz - rfgun Predictions for o = 2.9 and Various Final Momentum Intervals - MASK Beam Snapshots at, Various Phases-First Cell - MASK Beam Snapshots at. Various Phases-Second Cell - Exit,-Time and Momentum Hisdograms as Functions of Bunch Number
- tions of Bunch Number Normalized Average Momentum and Normalized Emittance as Func-
- MASK Transverse Phase-Space Evolution in the First Cell, for J --+
- Cell as Calculat,ed by MASK and SUPERFISH Comparision of Derivatives of On-Axis Longit,udinal Fields in the First
- URMEL Longitudinal Fields in the First Cell as Calculated by MASK and
- Off-Axis Expansions of Various Orders Comparison of E,(z) for r=2.87mm, as Calculated by MASK and Using
- Int.ervals, for MASK- and SUPERFISH-Calculated Fields Comparison of rfgun Emitstance Predictions for Various Initial Phase-
- Calculations for J t Comparison of rfgun results for MASK-Calculated Fields with MASK
- Calculations for J -+ Comparison of rfgun results for MASK-Calculated Fields with MASK
- for J = 80A/cm2 MASK-Calculated Transverse Phase-Space Evolution in the First Cell,
- MASK-Calculat,ed Transverse Beam Distribution at z = X/12, for J -+
- and J = 80A/cm2
- ond Cell MASK Longitudinal Phase-Space Distributions at the End of the Sec-
- Based Compression 2.38 MASK Longit.udinal Phase-Space Distributions aft.er Alpha-Magnet.-
- 2.39 MASK Results for Normalized Charge Per Bunch
- 2.40 MASK Results for Normalized RMS Emit.tance
- 2.41 MASK Results for Normalized Brightness
- 2.42 MASK Results for Transverse Brightness
- 2.43 MASK Results for Transverse Phase-Space Distribution, for Ep2 = - MV/m and AP/P = lo%-Parb
- 2.44 MASK Result,s for Transverse Phase-Space Dist.ribution, for Ep2 = - MV/m and AP/P = lO%-Part.
- 2.45 MASK Result,s for Transverse Phase-Space Dist,ribution, for Ep2 =
- .- 3.
- MV/m and AP/P = lO%-Part,
- Simplified Cross-sectional view of the SSRL alpha-magnet.
- Comput,ed and Measured Gradient, of the SSRL Alpha-Magnet’
- Measured Excitabion Curve of the SSRL Alpha-Magnet
- Panofsky quadrupole
- Alpha-magnet coordinat*e system
- Ideal Trajectory in the Alpha-Magnet’
- Reference plane and coordinates at, the entrance
- Reference plane and coordinat,es at the vertical midplane
- Reference plane and coordinat,es at, the exit
- 3.10 Effects of Sext.upole Errors-Part,
- 3.11 Effects of Sext.upole Errors-Part
- 3.12 Effects of Sextupole Errors-Part,
- 3.13 Hole-Induced Gradient Errors vs q1
- 3.14 Hole-Induced Gradient Errors vs q2
- 3.15 Effect,s of Hole-Induced Gradient, Errors-Part
- 3.16 Effects of Hole-Induced Gradient Errors-Part
- 3.17 Effects of Hole-Induced Gradient. Errors-Part,
- 3.18 Magnetic Measurements for GTL,CORR2
- 3.19 Measured and Theoretical Alpha.-Magnet r12’s - xv - 3.20 Measured and Theoretical Alpha-Magnet, rs4’s - 4. - 4. - 4. - 4.
- I 4.
- Constant Final Momentum Contours - Constant Final Phase Contours - Explanation of Slopes of Constant Final-Momentum Cont,ours - Expanded View of Constant Final Momentum Contours - Expanded View of Constant, Final Phase Contours - Particle Motion in a Wedge Bending Magnet - Result. for Compression Optimized for a Short Bunch at Linac Entrance - Results for Compression Optimized for a Short Bunch at, Linac Exit - Optimization for a Short Bunch at the Linac Exit for Various &, - sities, for (AP/P)i = AlO % Results of Optimized Compression for Various Cathode Current Den- - sities, for (AP/P)i = 15 % Results of Optimized Compression for Various Cat,hode Current. Den- - sities, for (AP/P)i = l t2.5 % Resu1t.s of Optimized Compression for Various Cathode Current Den- - Cathode Current, Densities and Ep2 = 75MV/m, for (AP/P)i = f10 % elegant/MASK Results after First Accelerator Section, for Various - Cathode Current Densities and Ep2 = 75MV/m, for (AP/P)i = 15 % elegant/MASK Results after First Accelerator Sect,ion, for Various - results for E,-,2= 75MV/ m, J = 10A/cm2, and (AP/P)i = 110 %). Longit.udinal Phase-Space at Various Points in the GTL (elegant /MASK - Bright.ness and Peak Current, for Various Injectors - Gun-to-Linac Region of the SSRL Preinjector - GTL Quadrupole Gradient. vs Longitudinal Position - GTL Optics Solution for & = llcm - ticle Dist,ribution, AP/P = 10%. Second-Order Tracking with elegant, for MASK-generated Initial Par-
5.22 Normalized Forward and R.eflected RF Power Waveforms for High and Low Forward Power............................ 5.23 Signal Analysis for a Video Scan..................... 5.24 Resolution Test for Horizontal Imaging................. 5.25 Integrated Video-Signal Int,ensity for Several Q5 Settings, for Various Valuesoff................................. 5.26 Integrated Video-Signal 1ntensit.y for Several Q4 Settings, for Various Values of f................................. 5.27 lo%-Contour Graphs of for Several Q4 Settings, from Experiments wit,h f=0.08................................ 330 5.28 Collapsed Horizontal Beam-Int,ensit,y Profiles for Several Q4 Settings, from Experiments with f=0.08...................... 5.29 Collapsed Vertical Beam-Intensit,y Profiles for Several Q4 Sett,ings,
.. from^ Experiments^ with^ f=0.08.^.^.^.^.^.^.^.^.^..^.^.^.^.^.^.^.^.^.^.^. 5.30 Horizontal Beam-Size versus Q4 Strength, for f = 0.08. (Points are experimental data, lines are fits.).................... 5.31 Vertical Beam-Size versus Q4 Strength, for f = 0.08. (Points are experimental data, lines are fits.).................... 5.32 Horizont,al Emittance versus f (Momentum Spread) as Inferred from Variation of Q4............................... 5.33 Vertical Emittance versus f ( Momentsum Spread) as Inferred from Vari- ation of Q4................................. 5.34 Variation of the RMS Geometric Emit.tance Along the Beamline for the Emit.tance Measurement, Lat,tice with Q4 at 90 mm2, from MASK/elegant Simulations......... .’....................... 5.35 Emittance at the Chopper Screen as Altered by Q4, for f = 0.08, from MASK/elegant Simulations........................ 5.36 Beam-Size as a Function of Q4 Strength, from MASK/elegant simula- tions wit,h f = 0............................. 5.37 Emitstance from w vs. Quadrupole Strength, as a Function of Half- Momentum Spread, f, from Simulations and Experiment. (See t#ext for explanation.)............................... xxi
I
:
5.38 Emitt.ance from r vs. Quadrupole Strengt,h, as a Function of Half- Momentum Spread, f, from Simulations and Experiment. (See text for explanation.)............................... 347 5.39 Data for Bunch-Length Measurements for Various Momentum Frac- tions Allowed Through the Alpha-Magnet............... 354 5.40 elegant/MASK-Simulated Beam Parameters After First, Linac Sec- tion vs Alpha-Magnet, Gradient for f=O.O3, (E,z = 65MV/m, J = 10 A/cm2)................................... 356 xxii --
