LCLS Undulators Present Status and Future Upgrades HeinzDieter

LCLS Undulators – Present Status and Future Upgrades Heinz-Dieter Nuhn – LCLS Undulator Group Leader March 1, 2010 LCLS Undulator Status March 1, 2010 Heinz-Dieter Nuhn nuhn@slac. stanford. edu

Linac Coherent Light Source INJECTOR LINAC BEAM TRANSPORT UNDULATOR HALL LCLS Undulator Status March 1, 2010 2 2 Heinz-Dieter Nuhn nuhn@slac. stanford. edu

Undulator Hall nd Undulator L Room for 2 ine 33 Undulator Segments Installed LCLS Undulator Status March 1, 2010 3 3 Heinz-Dieter Nuhn nuhn@slac. stanford. edu

Short Break Section Undulator Segment BFW Part of WPM Support RF Cavity BPM Quadrupole and horz/vert Correctors Segment Slider Girder LCLS Undulator Status March 1, 2010 HLS Sensor 4 4 Girder Mover (cam) Heinz-Dieter Nuhn nuhn@slac. stanford. edu

Fully Assembled Girder (seen from downstream end) Quadrupole Undulator Segment Vacuum Chamber RF Cavity BPM Girder LCLS Undulator Status March 1, 2010 5 5 Heinz-Dieter Nuhn nuhn@slac. stanford. edu

Girder Precision Alignment on CMM Quadrupole Undulator Segment with mu-Metal Shield Coordinate Measurement Machine Position Sensor RF Cavity BPM LCLS Undulator Status March 1, 2010 6 6 Heinz-Dieter Nuhn nuhn@slac. stanford. edu

LCLS Undulator Components Vacuum Chamber and Support BFW Segment 3. 400 m Quadrupole Cam Shaft Movers Short Break 47. 0 cm WPM Manual Adjustments BPM Horizontal Slides Not visible Sand-Filled, Thermally Isolated Fixed Supports LCLS Undulator Status March 1, 2010 Long Break 89. 8 cm 7 HLS 7 Heinz-Dieter Nuhn nuhn@slac. stanford. edu

Vacuum Chamber Inserted into Gap Undulator Segment Magnet Block Horizontal Trajectory Shim Holder Vacuum Chamber Pole Piece LCLS Undulator Status March 1, 2010 8 8 Heinz-Dieter Nuhn nuhn@slac. stanford. edu

LCLS Undulator Module Pole Canting Pole canting enables remote K adjustment for fixed gap undulators. • Canting comes from wedged spacers • 4. 5 mrad cant angle • Gap can be adjusted by lateral displacement of wedges • 1 mm shift means 4. 5 µm in gap, or 8. 2 G • Keff can be adjusted to desired value LCLS Undulator Status March 1, 2010 9 9 Heinz-Dieter Nuhn nuhn@slac. stanford. edu

Undulator Roll-Away and K Adjustment Pole Center Line Neutral; K=3. 4881; Dx= 0. 0 mm First; K=3. 5000; Dx=-4. 0 mm Vacuum Chamber Neutral; K=3. 4881; Dx= 0. 0 mm Roll-Away; K=0. 0000; Dx=+80. 0 mm Neutral; K=3. 4881; Dx= 0. 0 mm Horizontal Slide LCLS Undulator Status March 1, 2010 10 10 Heinz-Dieter Nuhn nuhn@slac. stanford. edu

LCLS-I Undulator Parameters units Nominal Undulator Parameter K 3. 5 Undulator Period lu 30 mm Undulator peak Field, Bpk 1. 249 T g 6. 8 mm Full Gap Height (fixed) Undulator Type Planar Hybrid Permanent Magnet Material Nd 2 Fe. B 14 Pole Material Vanadium Permendur Magnet Block Dimensions h×t×w 66× 9× 56. 5 mm 3 Pole Dimensions h×t×w 44× 6× 48 mm 3 Periods per Segment Gap Cant Angle 113 a 4. 5 Number of Installed Segments LCLS Undulator Status March 1, 2010 mrad 33 11 11 Heinz-Dieter Nuhn nuhn@slac. stanford. edu

Taper Design Considerations 1. Compensation of spontaneous radiation (linear tapering over 132 m) 2. Compensation of vacuum chamber wakefields (linear tapering over 132 m, for 0. 25 n. C) From Wakefield budget based on S 2 E Simulations 3. Gain enhancement (linear tapering before saturation) [Z. Huang] 4. Enhanced energy extraction (quadratic tapering after saturation) [W. Fawley] The ratio between changes in E and K to maintain the resonance condition at a given wavelength is LCLS Undulator Status March 1, 2010 12 12 Heinz-Dieter Nuhn nuhn@slac. stanford. edu

K Tapering Requirements K for segment 1 0. 3 % 1. 5 Å K for segment 33 spont wake gain post sat 15 Å 0. 3 % spont gain wake post sat LCLS Undulator Status March 1, 2010 13 13 Heinz-Dieter Nuhn nuhn@slac. stanford. edu

Figure 3: K Tapering Scenarios (Continuous) Avoid Reliance on Good Field Region at 1. 5 Å LCLS Undulator Status March 1, 2010 14 14 Heinz-Dieter Nuhn nuhn@slac. stanford. edu

Measured Field Integrals on SN 25 y : +200 µm +0 µm -200 µm LCLS Undulator Status March 1, 2010 15 15 Heinz-Dieter Nuhn nuhn@slac. stanford. edu

Beam Based Measurement: 1 st Field Integral SN 14 Horizontal (I 1 X) and vertical (I 1 Y) first field integrals measured by fitting a kick to the difference trajectory as function of undulator displacement SN 14 Requires 20 nm BPM resolution Reference Point LCLS Undulator Status March 1, 2010 Beam Based Measurements SN 14 MMF Measurement 16 16 Heinz-Dieter Nuhn nuhn@slac. stanford. edu

Segmented Undulator Pre-Taper LCLS Undulator Status March 1, 2010 17 17 Heinz-Dieter Nuhn nuhn@slac. stanford. edu

CMM Keff Measurements for U 33/SN 20 K=3. 497 K=3. 468 LCLS Undulator Status March 1, 2010 18 18 Heinz-Dieter Nuhn nuhn@slac. stanford. edu

Segmented Undulator K Control TEMPERATURE CORRECTED KACT K ADJUSTMENT RANGE (MEASURED) TAPER REQUEST K ADJUSTMENT RANGE (MEASURED) LCLS Undulator Status March 1, 2010 19 19 Heinz-Dieter Nuhn nuhn@slac. stanford. edu

Tolerance Budget Analysis based on time dependent SASE simulations with GENESIS Eight individual error sources considered: Beta-Function Mismatch, Launch Position Error, Segment Detuning, Segment Offset in x, Segment Offset in y, Quadrupole Gradient Error, Transverse Quadrupole Offset, Break Length Error. The ‘observed’ parameter is the average of the FEL power at 90 m (around saturation) and 130 m (undulator exit) The Results are combined into the Error Budget LCLS Undulator Status March 1, 2010 20 20 Heinz-Dieter Nuhn nuhn@slac. stanford. edu

Segment K Errors Simulation and fit results of Module Detuning analysis. The larger amplitude data occur at the 130 -mpoint, the smaller amplitude data at the 90 -m-point. 130 m Module Detuning (Gauss Fit) 90 m Location Fit rms Unit 090 m 0. 042 % 130 m 0. 060 % Average 0. 051 % Budget Tolerance LCLS Undulator Status March 1, 2010 21 21 Heinz-Dieter Nuhn nuhn@slac. stanford. edu

Individual Studies (Example K) Choose a set of DKm/K values to be tested, e. g. { 0. 000%, 0. 045%, 0. 100%, 0. 200%} For each DKm/K choose 33 DKs values from a random flattop distribution with maximum DKm. Apply these errors, DKs, to the respective segment Ks values and perform a GENESIS FEL simulation. Evaluate the simulation result to extract power levels at the 90 m and 130 m points, P 90, m and P 130, m, respectively. Loop Plot these results, P 90, m and P 130, m, versus the rms of the distribution, i. e. Apply Gaussian fit to obtain rms-dependence. LCLS Undulator Status March 1, 2010 22 22 Heinz-Dieter Nuhn nuhn@slac. stanford. edu

Horizontal Segment Misalignment Simulation and fit results of Horizontal Module Offset analysis. The larger amplitude data occur at the 130 -mpoint, the smaller amplitude data at the 90 -m-point. 130 m Horizontal Model Offset (Gauss Fit) 90 m Location Fit rms Unit 090 m 0782 µm 130 m 1121 µm Average 0952 µm Budget Tolerance LCLS Undulator Status March 1, 2010 23 23 Heinz-Dieter Nuhn nuhn@slac. stanford. edu

Vertical Segment Misalignment Simulation and fit results of Vertical Module Offset analysis. The larger amplitude data occur at the 130 -mpoint, the smaller amplitude data at the 90 -m-point. 130 m Vertical Model Offset (Gauss Fit) 90 m Location Fit rms Unit 090 m 268 µm 130 m 268 µm Average 268 µm Budget Tolerance LCLS Undulator Status March 1, 2010 24 24 Heinz-Dieter Nuhn nuhn@slac. stanford. edu

Tolerance Budget Gaussian fit yields functional dependence of power reduction on error amplitude: Assuming that each error is independent on the others other, i. e. each error source causes a given fraction power reduction independent of the presence of the other sources: tolerance fitted rms fi=qi/si LCLS Undulator Status March 1, 2010 25 25 Heinz-Dieter Nuhn nuhn@slac. stanford. edu

LCLS Tolerance Budget si f i si fi @ 130 m (24. 2% red. ) Hor/Ver Optics Mismatch (z-1)0. 5 0. 71 0. 452 0. 32 Hor/Ver Transverse Beam Offset 30 0. 176 3. 7 µm Module Detuning DK/K 0. 060 0. 400 0. 024 % Module Offset in x 1121 0. 125 140 µm Module Offset in y 268 0. 298 80 µm Quadrupole Gradient Error 8. 8 0. 029 0. 25 % Transverse Quadrupole Offset 4. 7 0. 214 1. 0 µm Break Length Error 20. 3 0. 049 1. 0 mm Error Source LCLS Undulator Status March 1, 2010 26 26 Units z < 1. 1 0. 64<b/b 0<1. 56 Heinz-Dieter Nuhn nuhn@slac. stanford. edu

Model Detuning Sub-Budget Typical Value rms dev. dpi 3. 5 0. 0003 a. K -0. 0019 °C-1 0. 0001 °C-1 DT 0 °C 0. 32 °C b. K 0. 0023 mm-1 0. 00004 mm-1 Dx 1. 5 mm 0. 05 mm Parameter pi KMMF LCLS Undulator Status 27 March 1, 2010 Note ± 0. 015 % uniform Thermal Coefficient ± 0. 56 °C uniform without compensation Canting Coefficient Horizontal Positioning 27 27 Heinz-Dieter Nuhn nuhn@slac. stanford. edu

Beam Based Alignment Tolerance Verification Beam Based Measurements Random misalignment with flat distribution of widh ±a => rms distribution a/sqrt(3) LCLS Undulator Status March 1, 2010 28 28 Heinz-Dieter Nuhn nuhn@slac. stanford. edu

Beam Based K Tolerance Verification Beam Based Measurements LCLS Undulator Status March 1, 2010 29 29 Heinz-Dieter Nuhn nuhn@slac. stanford. edu

LCLS Undulator Tolerance Budget BB Verification Tolerance Budget Components si f i si fi @ 130 m (24. 2% red. ) Hor/Ver Optics Mismatch (z-1)0. 5 0. 71 0. 452 0. 32 Hor/Ver Transverse Beam Offset 30 0. 176 3. 7 µm 0. 400 0. 024 % 0. 125 140 µm 0. 298 80 µm 0. 029 0. 25 % Error Source 0. 06 Units Module Detuning DK/K 0. 060 Module Offset in x 1121 Module Offset in y 268 Quadrupole Gradient Error 8. 8 Transverse Quadrupole Offset 4. 7 0. 214 1. 0 µm Break Length Error 20. 3 0. 049 1. 0 mm Module Offset in x @ z. SAT 780 1200 8. 8 770 µm MEASUREMENTS LCLS Undulator Status March 1, 2010 30 30 Heinz-Dieter Nuhn nuhn@slac. stanford. edu

LCLS-II An initial rough evaluation of LCLS-II undulator parameters will be presented. Priority is given to the Soft-Xray line, which is likely to be based on short variable gap undulators. Shortness is required to enable the low beta -functions needed for optimum FEL performance. LCLS Undulator Status March 1, 2010 31 31 Heinz-Dieter Nuhn nuhn@slac. stanford. edu

Phased Enhancement Plan for LCLS-II full polarization 4 -Ge. V SXR and control 14 -Ge. V HXR 6 -60 Å 2 -pulse 2 6 -60 Å simultaneous op’s -color adjust. gap with bypass line EEHG*? adjust. gap 4 -14 Ge. V SXR 2 (45 m) self-seeding option 5 m Phase-3 Phase-2 Phase-1 Existing Phase-0 FEE-2 full polarization control Large Gap. SHAB Shortened Large Gap Larger Gap Undulator Shortened 74 -m Undulator Existing 112 -m Undulator (1. 5 -15 Å) (0. 75 -7. 5 Å) (0. 5 -5 Å) 30 m (1. 5 -15 Å) (0. 5 -5 Å) self-seeding HXR option 0. 75 -15 Å (2 bunches) 5 m 240 nm 6 nm SXR 1 (45 m) 5 m 3 -7 -Ge. V bypass FEE-1 0. 75 Å No civil construction. Uses existing beam energy and quality. * G. Stupakov, Phys. Rev. Lett. 102, 074801 (2009) LCLS Undulator Status March 1, 2010 32 32 Heinz-Dieter Nuhn nuhn@slac. stanford. edu

sg = 2. 8 Ipk = 3000 A, gexy= 0. 6 µm LCLS-I U 1 Enhancement LCLS Undulator Status March 1, 2010 33 33 Heinz-Dieter Nuhn nuhn@slac. stanford. edu

helical linear <b> = 5 m, sg = 2. 8 Ipk = 2000 A, gexy= 0. 6 µm LCLS-II U 2 FEL Performance Estimate LCLS Undulator Status March 1, 2010 34 34 Heinz-Dieter Nuhn nuhn@slac. stanford. edu

helical linear <b> = 5 m, sg = 2. 8 Ipk = 2000 A, gexy= 0. 6 µm LCLS-II U 2 FEL Performance Estimate LCLS Undulator Status March 1, 2010 35 35 Heinz-Dieter Nuhn nuhn@slac. stanford. edu

Beta-Function at 6 nm Optimum LG~0. 69 m for bx, y = 5 m LG~0. 65 m for bx, y = 4 m Smallest practical beta function 4 -5 m is above optimum. LCLS Undulator Status March 1, 2010 36 36 Heinz-Dieter Nuhn nuhn@slac. stanford. edu

‘Optimum’ Beta-Function at 6 nm LG~0. 27 m for bx, y ~ 0. 1 m Optimum beta function would reduce undulator length by more than factor 2 but is not accessible. LCLS Undulator Status March 1, 2010 37 37 Heinz-Dieter Nuhn nuhn@slac. stanford. edu

Optimum Beta-Function at 0. 6 nm Considered Value Optimum Value At 0. 6 nm beta function of 4 -5 m is close to optimum. LCLS Undulator Status March 1, 2010 38 38 Heinz-Dieter Nuhn nuhn@slac. stanford. edu

Beta Function and Undulator Length The smallest average beta-function achievable with a FODO lattice is The FODO length is determined by segment length and break length Breaks between segments need to be sufficiently wide to allow space for essential components, such as quadrupole, BPM, Chicane. Smallest practical quadrupole separation is 2. 5 m, corresponding to a FODO length of 5 m. Bellows EXAMPLE: Break 0. 70 m Undulator: 1. 80 m Break 0. 70 m Half FODO Length: 2. 50 m Minimum <bx, y> = 5 m LCLS Undulator Status March 1, 2010 39 39 Chicane RF Cavity BPM Quadrupole Heinz-Dieter Nuhn nuhn@slac. stanford. edu

Example Chicane Dimensions Multi-Segment variable gap undulators require phase shifters between segments to adjust gap dependent inter-segment phase slippage. An example for such a chicane is shown here. Field levels have been kept low to reduce in-tunnel power release. L = 9 cm xmax L = 4. 5 cm L =4. 5 cm 3 cm L = 24 cm LCLS Undulator Status March 1, 2010 E 7. 0 3. 0 Ge. V lr 1. 2 6. 0 nm B 203 195 G x’ 78 175 µrad xmax 7. 6 17 µm Df 360 deg. Xray hx -5. 9 -13. 2 µm R 56 0. 74 3. 7 nm 40 40 Heinz-Dieter Nuhn nuhn@slac. stanford. edu

Undulator Types A number of different variable field undulator types are under consideration Parallel-Pole Variable Gap Fixed Linear Polarization Hybrid or Pure Permanent Magnet Apple Type Variable Gap Variable Linear/Circular Polarization Hybrid or Pure Permanent Magnet Delta Type Variable Phase Variable Linear/Circular Polarization and Intensity Pure Permanent Magnet Superconducting Helical Variable Excitation current Fixed Circular Polarization [Substantial R&D required] New Designs … Key issues are Precision Hall probe measurements K stability and settability Compact design to mount on movable girders. Gap > 7 mm LCLS Undulator Status March 1, 2010 41 41 Heinz-Dieter Nuhn nuhn@slac. stanford. edu

Summary The LCLS-I undulators have performed very well during commissioning and first user operation. Initial parameter development for the LCLS-II undulators has started, giving priority to the new soft x-ray line. The goal is a compact variable gap design to cover wavelengths between 6 nm and <0. 6 nm at electron energies in the range 3 -7 Ge. V. The low emittance and lower electron energy require beta functions of order 5 m or smaller for best utilization. Low beta-functions require a short FODO length, i. e. , short undulator segments of length 1. 8 m and compact break sections. The total length of each of the 2 soft x-ray undulator lines is expected to be about 50 m. LCLS Undulator Status March 1, 2010 42 42 Heinz-Dieter Nuhn nuhn@slac. stanford. edu

End of Presentation LCLS Undulator Status March 1, 2010 43 43 Heinz-Dieter Nuhn nuhn@slac. stanford. edu