The High Power Electrodynamics Group at Los Alamos National Laboratory

Transcription

1 The High Power Electrodynamics Group at Los Alamos National Laboratory Steven J. Russell Los Alamos National Laboratory Slide 1

2 Los Alamos National Laboratory Valles Caldera Los Alamos Slide 2

3 Los Alamos National Laboratory Los Alamos (2.2 km elevation) Ski Area Los Alamos National Laboratory (100 km2) Slide 3

4 Los Alamos National Laboratory (May 2000) Slide 4

5 Los Alamos has Two Large Accelerator Facilities LANSCE DARHT Slide 5

6 Dual-Axis Radiographic Hydrodynamic Test (DARHT) Facility Two, 20 MeV induction linacs. Axis 1: A single 60 ns, 2000 A beam pulse. Axis 2: A single 2 μs, 2000 A beam is chopped into 4, 80 ns pulses. Linacs are at a 90 angle with respect to each other. Bremsstrahlung x-rays are generated at the intersection point for radiography experiments. Slide 6

7 Los Alamos Neutron Science Center (LANSCE), Formerly the Los Alamos Meson Physics Facility (LAMPF) 800 MeV, 1 ma Proton Accelerator Our Experimental Building Me 0.75 km Slide 7

8 LAMPF is the Reason LANL has an Accelerator Program Started by Louis Rosen, LAMPF began operation in the 1970s and for many years was the center of basic research at Los Alamos. Today LAMPF, now LANSCE, plays a more limited role. The main areas of focus are delivering beam for our spallation neutron source, proton radiography and isotope production. LAMPF led to the development of strong R&D programs in accelerators, pulsed power and microwave sources. Starting in the mid 1980s Los Alamos began research in free electron lasers. The most successful of these experiments lased from the infrared to the UV and led to the development of photoinjectors (Sheffield and Shafer). Our group was formed from the former free electron laser group and a pulsed power and microwave source R&D group. Slide 8

9 The High Power Electrodynamics Group at Los Alamos National Laboratory: What do we do? Compact pulsed power. Computer modeling. Accelerators and microwave sources Electromagnetic simulations Plasma physics. Plasma sources Plasma thrusters Coherent light sources. Microwave tubes (1 GHz 100 GHz, high power) Free electron lasers (infrared, high-power) Accelerators. High-brightness electron accelerators (photoinjectors) Plasma wakefield accelerator experiments Beam dynamics Slide 9

10 Compact pulsed power Slide 10

11 48 Stage Solid State Marx modulator 48 kv, 160 A Optically triggered Fault tolerant 11 kg (24 lbs) 2 ft. x 1 ft. x 0.5 ft. Patented Winner LANL Distinguished Performance Award 2005 Nominated for R&D 100 (2005) Slide 11

12 Typical Operating Waveforms output voltage (kv) µs rectangular pulses at 50 Hz into a 300-Ω load. Waveform taken after 10 minutes of operation (a) time (μs) MW peak power output current (A) kw average power Charge voltage = 1000 V per stage (b) time (μs) Slide 12

13 Pulse width variability output current (A) output voltage (kv) time (μs) time (μs) Six pulse burst into a 300 Ω load consisting of one 4-µs pulse, two 2-µs, and three 1- µs pulses. Relative pulse positions maintain a 50% duty factor. Note that the last 3 pulses are modulated at 500 khz. Charge voltage = 1000 V per stage Slide 13

14 Wave Shape Variability output voltage (kv) (a) time (μs) 50 output voltage (kv) Output voltage pulse for (a) ramp, (b) triangle, and (c) inverted triangle. Initial charge voltage is 1000 V for all three cases. Each shape consists of 12 control increments. For the triangle, the turn-on commands are separated by ~ 200 ns, and the turn-off commands are separated by ~ 400 ns. output voltage (kv) (b) time (μs) (c) time (μs) Slide 14

15 High frequency, planar vacuum tubes Slide 15

16 Cross Section of Conceptual Schematic of Planar 94 GHz TWT Planar Focusing Slow-Wave Copper Structure Slide 16

17 Planar Geometries are a Natural Approach to High-Power, High- Frequency Microwave Tubes RMS Envelope Equations d ε K x + k x = 0 dz x 2 x y 2 2 x 2 RMS x RMS 3 RMS RMS RMS ( + ) d ε y + k y = 0 dz y 2 x y 2 2 y K 2 RMS y RMS 3 RMS RMS RMS ( + ) Planar RF structures lend themselves well to established micro-fabrication techniques. By spreading the electron beam in one dimension, we can transport a high net beam current, enabling very high power devices. Slide 17

18 Wide Planar 94 GHz RF Structure Concept Ridged Waveguide Full Structure Wide RF Structure Sheet-beam Slide 18

19 Two Converters are Combined by 45 Waveguide Sweeps Using two converters allows us to launch the proper mode in the highly overmoded wide RF structure. Converter Beam Pipe Combiner Function The two converters are combined by 45 degree waveguide sweeps in which the TE mode is converted to a TM mode. Slide 19

20 Full Model of Coupler Matched into Wide RF Structure with Tapered Vanes Full view of coupler and wide RF structure Cut view with calculated fields Calculation of the S parameter shows a bandwidth of approximately 6% at 94 GHz. Bandwidth ~6% Slide 20

21 We Built a Narrow Planar Structure for a Proof-of-Principle Experiment with a Pencil Electron Beam Full Narrow Structure Combiners Combiners with Tapered Matching Section 10 GHz aluminum cold model matched simulations very well. Slide 21

22 Network Analyzer Measurements of a 94 GHz Copper Structure Agreed well with Simulations Agreement between measurement and simulations using HFSS and Microwave Studio were very good. 94 GHz copper narrow planar structure built by Maroney Company using wire EDM being tested with network analyzer. Slide 22

23 We Tested the 94 GHz Copper Structure with a 120 kv, 2A Pencil Beam and Showed RF Gain Slide 23

24 Our Current Approach to Generating a Sheet Electron Beam is to use an Elliptical Pole Solenoid The solenoid is placed downstream from a standard Pierce geometry 120 kv, 20 A electron gun, transforming an initially round beam to form an elliptical sheet beam. Our elliptical pole solenoid is 2.75 inches long and has an elliptical aperture with a horizontal semi-axis radius of inches and a vertical semi-axis radius of inches. Slide 24

25 Electron Gun Simulation TRAK Simulation of Electron Gun Beam Parameters Current: Energy: Radius: Emittance: A kev 0.6 cm 3.26 mm-mrad Emittance is RMS normalized and includes thermal effects. Slide 25

26 Demonstration of Sheet Beam Formation with Elliptical Solenoid at 20 kv, 1.42 A Measurement Simulation (OmniTrak) Slide 26

27 We Would Like to Make our Sheet Electron Beams using Emittance Cooling With proper placements and strengths of quadrupoles (rotated 45 ), we can transfer emittance from one transverse plane to the other. 2 εlarger 2L εsmaller εintrinsic L = ebcathπr 16m c e 2 cath The intrinsic emittance is the thermal emittance of the beam (assuming gun is well designed). R. Brinkman, Ya. Derbenev, and K. Floettmann, Phys. Rev. ST Accel. and Beams, 4, (2001). Slide 27

28 After the First Quadrupole, the Beam Shears Vertically on Either Side of the x = 0 Axis y (m) y (m) x (m) x (m) The beam exits the electron gun with angular momentum due to the solenoid field on the cathode. The first skewed quadrupole stops much of the horizontal motion of the beam electrons, leading to vertical beam shear about the x = 0 axis. Slide 28

29 Quadrupoles Two and Three Complete the Process of Removing Horizontal Beam Motion y (m) y (m) x (m) x (m) After the second skewed quadrupole, the beam is again spinning but now has a highly elliptically shape. The final skewed quadrupole, like the first, stops the horizontal motion of the beam. This completes the transfer of horizontal motion to the vertical direction. The result is a sheet beam that is very cold in the horizontal plane. Slide 29

30 Once Formed, the Sheet Beam is Transported in a PCM Array or Wiggler Array Periodic Cusped Magnetic (PCM) Array N S N S N S N S N N S N S N S N S N N S N S N S N S Wiggler Array N S N S N S N S N S N S N S N S N S N S N S S N S N Slide 30

31 PCMs and Wigglers Provide the Same Focusing Force, but have Different Stability Requirements B Required Field Magnitude mv γ e b K ε y mag = + 4 e 2yRMS ( xrms + yrms ) y RMS λ Stable Transport Places Constraints on the Array Period Wiggler Period is limited by magnet length needed to turn electrons around = ε 2πβγc max 2 ny576frf 2 PCM Period is governed by Mathieu s equation (assuming emittance dominated beam) 0.662πβγc λ = ε max 2 ny576frf 2 B. E. Carlsten, L. M. Earley, F. L. Krawczyk, S. J. Russell, J. M. Potter, P. Ferguson and S. Humphries, Stable two-plane focusing for emittance-dominated sheet-beam transport, PR-STAB, 8, (2005). Slide 31

32 Original Sheet Beam Experiment with Single Plane Focusing Wiggler 3D OmniTrak Simulation Measured Beam 6.4 cm Inside Wiggler (30 kv, 2.6 A) Horizontal and Vertical Projections of OmniTrak Simulation Slide 32

33 Notched Wiggler Dual Plane Focusing 3D OmniTrak Simulation Notched magnets leads to linear vertical focusing and nonlinear horizontal focusing (notch is exaggerated). Beam Horizontal and Vertical Projections of OmniTrak Simulation Slide 33

34 PCM/Quad Hybrid Dual Plane Focusing Experiment 3D OmniTrak Simulation The PCM/Quad hybrid dual plane focusing experiment superimposes a quadrupole field on a planar PCM structure to achieve focusing in both planes. This experiment has been designed and is currently being fabricated. Horizontal and Vertical Projections of OmniTrak Simulation Slide 34

35 PCM/Quad Hybrid was Tested in September 2006 with a 120 kv, 20 A Beam 95% beam transmission to current monitor. Flat beam measured at center using foil (to right) 7 mm 0.3 mm Slide 35

36 At 300 GHz, Emittance Cooling Allows for Practical Transport Magnet Parameters The following table gives the important parameters for a PCM or Wiggler array used to focus a 120 kv, 2A, 1 cm wide sheet electron beam in a 300 GHz TWT. The calculations assume different initial sources for the electron beam and assume no contribution to the beam emittance due to aberrations in focusing magnets upstream from the sheet beam transport array. All cathodes emit with a 4A/cm 2 current density. Cathode Shape Cathode Width Cathode Height Required Beam Final Height Normalized Emittance Lower Bound in Narrow Plane (Theory) Required Magnitude of PCM or Wiggler Field Maximum Stable PCM Period Maximum Stable Wiggler Period Conventional Gun 0.80 cm 0.80 cm 0.17 mm 0.95 μm (thermal emittance) 1.32 T 0.68 cm 0.59 cm Elliptical Gun 1.00 cm 0.64 cm 0.17 mm 0.76 μm (thermal emittance) 1.06 T 0.85 cm 0.73 cm Sheet Beam Gun with Rectangular Cathode 1.00 cm 0.50 cm 0.17 mm 0.69 μm (thermal emittance) 0.96 T 0.93 cm 0.81 cm Conventional Gun with Emittance Converter (433 Gauss at cathode) 0.80 cm 0.80 cm 0.17 mm μm 0.10 T 113 cm 8.2 cm Slide 36

37 Electron accelerators and free electron lasers Slide 37

38 The Advanced Free Electron Laser (AFEL) Commissioned in early 1990s Operates in the infrared 20 MeV electron photoinjector 1 nc in 10 ps typical from photoinjector 2 μm normalized emittance First demonstration of a regenerative amplifier FEL D. C. Nguyen et. al, Nucl. Instrum. Phys. Res. A, 429, p. 125 (1999). Slide 38

39 A Second 8 MeV Photoinjector We Use for Beam Compression and Plasma Wakefield Accelerator Experiments 5 ½ cell, 1300 MHz standing wave structure Output energy 7 8 MeV Charge per bunch 10 nc Normalized emittance (10 ps beam): 2.5 μm at 1 ps. Bunch length: 20 ps to < 1 ps (with magnetic compression). Slide 39

40 A Second 8 MeV Photoinjector We Use for Beam Compression and Plasma Wakefield Accelerator Experiments Slide 40

41 We Use a Magnetic Chicane to do Longitudinal Beam Shape Manipulation Photoinjector Chicane Slide 41

42 We Use a Magnetic Chicane to do Longitudinal Beam Shape Manipulation Slide 42

43 Depending on Chicane Bend Angle and Beam Injection Phase, we can Manipulate the Beam Shape Density (arb. units) Initial Beam Bunch Density (arb. units) Time (ps) Time (ps) B. E. Carlsten and S. J. Russell, Phys. Rev. E., 53, p (1996). Density (arb. units) Density (Arb. Units) Longitudinal Position in Time (ps) Time (ps) Slide 43

44 The Wedge Shaped Beam is Especially Interesting for Plasma Wakefield Accelerator (PWFA) Experiments Numerical simulations of 1D wake fields produced by various bunch shapes: a) Triangular bunch. b) Gaussian rise, σ r = 7.2c/ω p. Gaussian fall, σ p = 0.1c/ω p. c) σ r = 7.2c/ω p, σ f = 1c/ω p. d) σ f = 3c/ω p. T. Katsouleas, Physical mechanisms in the plasma wakefield accelerator, Nucl. Instrum. Phys. Res. A, 33, p (1986). Wedge shaped beam Large Transformer Ratio Slide 44

45 Using a Xenon Gas Jet, We Have Performed a Preliminary PWFA Experiment with a Drive Beam Only Xenon Gas Plume Bunch #4 + e e - e e - + e e e e - e + - e- + Bunch #2 Bunch #1 To Spectrometer The first two beam bunches ionize the gas jet. Subsequent bunches experience strong wakefields. The ionization process is driven mainly by the secondary electrons produced by collisions of neutrals with bunches #1 and #2. Gas Jet Slide 45

46 We Measure an Obvious De-Acceleration of our Wedge Shaped Drive Beam Gas Jet Off Gas Jet On Spectrometer screen for a 5 bunch electron beam. Slide 46

47 The Results are Intriguing, but Leave a Lot of Questions Unanswered Intensity vs. Energy, Gas Jet On 3.0E E-02 Intensity (arb. units) 2.0E E E-02 1 Bunch 2 Bunches 3 Bunches 4 Bunches 5 Bunches 5.0E E Energy (kev) Gas plume width: 3 mm Peak Energy Los: De-Acceleration Gradient: 180 kv 60 MV/m Transformer Ratio: 7? Accelerating Gradient: 420 MV/m? Slide 47

48 In June 2007 We Take Delivery of a 100 ma Average Current Photoinjector Cavity Design 700 MHz, π-mode, 2½ cell On-axis electrically coupled OFE copper on Glidcop Beam Parameters Bunch charge Beam energy Normalized emittance Bunch length (rms) Energy spread (rms) 1 3 nc 2.0 MeV 7 μm 9 ps 40 kev Photocathode: Cs 2 Te, Cs:GaN, K 2 CsSb Drive laser: 4X, 3X or 2X Nd laser Klystrons: CPI MW 700 MHz (X2) Slide 48

49 2D Design Solenoids Magnetic field Resonant cavities Electric field Vacuum plenum Cathode 3-cell Injector Beam Aperture Gradient (E 0 ) 7 MV/m Cathode field 9.92 MV/m Temperature 20 C Ohmmic losses 807 kw Maximum heat flux 117 W/cm 2 Cavity Q 30,000 Shunt impedance Z 31.5 MΩ/m Solenoid Magnetic Field B (Gauss) On-axis RF Electric Field Distance (cm) On-axis DC Magnetic Field Slide 49

50 3D Cavity Design The The cavity cavity design design of of pillboxes pillboxes with with large large apertures apertures is is a trade-off trade-off between between maximizing maximizing the the accelerating accelerating gradient gradient and and RF RF surface surface thermal thermal management. management. 105 W/cm ka/m 3D Electric Field Distribution 3D Surface Current Distribution Slide 50

51 Over the Next Several Years we Hope to Build a 100 kw FEL Demonstration Experiment Tentative Time Line June 2007: Begin thermal testing of injector with no beam. December 2007: Begin testing of photoinjector with beam. Now 2009?: Design energy recovery linac/fel 2009??: Build and test 100 kw FEL Slide 51

52 The End Slide 52