LCLS-II project and beamline instruments: mechanical ... · 10/18/2016 · LCLS-II project and beamline instruments: mechanical engineering challenges NSLS-II Engineering Seminar

October 18, 2016NSLS-II Engineering Seminar Series

LCLS-II project and beamline instruments:mechanical engineering challenges

L. Zhang

LCLS, SLAC National Accelerator Laboratory2575 Sand Hill Road, Menlo Park, CA, 94025, United States

[email protected]

2

Co-workers at LCLS/SLAC

LCLS-II project and beamline instruments: mechanical engineering challengesNSLS-II Engineering Seminar Series, Oct 18, 2016, L. Zhang

• L. Amores• D. Cocco• C. Hardin• J. James• N. Kelez• J. Krzywinsk• D. Morton• D. Schafer• V. Srinivasan now India• P. Stefan• R. Whitney

3

Outline of the presentation


Ø Introduction§ LCLS, LCLS-II project§ X-ray Transport and Experimental Systems (XTES)§ Beamline instruments

Ø Mechanical engineering challenges in XTESØ Attenuators, beam stoppers and beam collimatorsØ X-ray optics and KB mirror systems§ Optics cooling§ Flat mirror§ Bendable KB mirror systems

4

LCLS vs. LCLS-II


NEH

FEH

LCLS office

5

LCLS vs. LCLS-II


4 GeV SC Linac

LCLS LCLS-II

Accelerator (Linac) Copper Linac (3-15 GeV)Superconducting Linac (4 GeV)+ Copper Linac (3-15 GeV)

Undulator Fixed/Tapered gap SXU + HXU: variable gap

Repetition rate 120 Hz ~ 1 MHz

6

LCLS vs. LCLS-II


LCLS-I HXU - Cu HXU - SC SXU - SC SXU – Cu

Photon Energy Range (keV) 0.25 - 12.8 1 - 25 1 - 5 0.25 - 1.3 0.25 - 6

Repetition Rate (Hz) 120 120 929,000 929,000 120

Per Pulse Energy (mJ) ~ 4 ~ 4 ~ 0.2 ~ 1 ~ 7

Max. av. power (W) 0.48 0.48 200 600 0.8

Photons/Second ~ 1014 ~ 1014 ~ 1016 ~ 1017 ~ 1014

LCLS-II-HE: à 13 keV (≥100 kHz, SC Linac)

7

X-ray Transport & Experimental Systems (XTES)


Ø X-ray beam transport system§ Diagnostics and components

Ø X-ray experimental system§ X-ray instruments

Ø X-ray Optics

XTES Schematic Layout

LCLS-II DOE Review, October 12-14, 2016

SXR BranchSXR Branch

HXR BranchHXR Branch

• X-ray optics- Offset (flat) mirrors

• X-ray diagnostics- Beam imagers- (Gas) energy monitors- HXR spectrometer- HXR K-monochromator

• X-ray components- (Gas and solid) attenuators- Stoppers, collimators,

apertures/slits• New SXR endstation

- KB focusing optics- (Gas) photon dump

Existing LCLS

Upgraded by LCLS-II

Development (not in scope)

New LCLS-II

FEE

NEH

EBD

Courtesy Y. Feng

9

X-ray instrument plans for LCLS-II


• 7 instruments fed by a single undulator at present• 9 instruments available for LCLS-II

NEH 1.1: Atomic, Molecular and OpticalNEH 2.1: Resonant Inelastic X-ray ScatteringNEH 2.2: Soft X-ray ResearchNEH 1.2: Tender X-ray InstrumentXPP: X-ray Pump ProbeXCS: X-ray Correlation SpectroscopyMFX: Macromolecular Femtosecond CrystallographyCXI: Coherent X-ray ImagingMEC: Matter in Extreme Conditions

3 Soft X-ray

5 Hard X-ray

1 “tender” x-ray

SXUSXU

HXUHXU

FarHall

XCS MFX CXI MEC

NearHall

N1.1 N1.2 XPP

N2.1

N2.2

~ 50 m ~ 70 m

10

Optics Configuration in Front End


11

KB mirror systems for Soft and Tender X-ray


à 6 pairs of KB mirror systems

NEH 1.1• TMO• Bendable K-B Pair

─ 1 μm• Fixed Figure K-B Pair

─ 300nm• 250-1300 eV

NEH 1.2• Tender X-ray Instrument• SXR Bendable K-B Pair

─ 1 μm• HXR Bendable K-B Pair

─ 1 μm• 400-6000 eV

NEH 2.1• RIXS• Bendable K-B Pair

─ 2x10μm• 250-1350 eV

NEH 2.2• Spectroscopy• Bendable K-B Pair

─ 1x4 μm• 250-1350 eV

12




ØMechanical engineering challenges in XTESØ Attenuators, beam stoppers and beam collimatorsØ X-ray optics and KB mirror systems§ Optics cooling§ Flat mirror§ Bendable KB mirror systems

13

Mechanical engineering challenges in XTES


Ø Extremely high peak power, mJ energy per pulseà TW§ Single shot damage

Ø High rep-rate FELà High average power: 200 ~ 600 W§ Effective cooling necessary§ Multiple shots fatigue and damage issue

Ø Nearly monochromatic beam (especially with self-seeding)§ Offset mirror removes only 10% of beam power§ (~)All the optics, including final focusing mirrors to be actively cooled

Ø Fully coherent photon beam à Wavefront preservation

14

Design consideration for heat load components


Peak power versus average powerØ Peak powerà single shot damage threshold§ Instantaneous absorbed dose per atom

§ Datom << 1 eV/atomà Large ρatom d material: B4C, C (Diamond, graphite),…

Ø Average powerà stress, strain, thermal fatigue§ High thermal conductivity,

lower thermal expansion,high strength materials

Ø Multiple-shot fatigue and damage

Datom =aF sinqratomd

F: fluence (energy/cm2)α: absorption coefficientθ: incidence angleρatom: number of atoms per unit of volumed: photon beam extinction length

15

Mechanical engineering challenges: optics requirements


Ø Fully coherent photon beam à Wavefront preservation§ 2*FWHM beam size needed

§ Shape error requirement (SR ≥ 0.97)

§ Sub-nm shape requirement

¥ acceptance 2 FWHM accept. 1 FWHM accept.

Unfocussed beam Unfocussd beam Unfocussed beam 0

20

40

60

80

100

120

0.00.20.40.60.81.01.2

0 500 1000 1500

Div

erge

nce

[µra

d]

Bea

mFo

otpr

int[

m]

Photon Energy [eV]

FootprintDivergence

θ=14 mrad

SR=0.97

SR=0.80

Unfocussed beamFocused beam

16




Ø Mechanical engineering challenges in XTES

ØAttenuators, beam stoppers and beam collimatorsØ X-ray optics and KB mirror systems§ Optics cooling§ Flat mirror§ Bendable KB mirror systems

17

Photon beam stoppers and collimators


Ø Design principle§ 750-µm CVD Diamond absorbs most of photon beam power§ 2-µm Graphite coating to prevent diamond from graphitization at Carbon

k-edge§ 10-mm SiC + 10-mm Heavy metal absorb high energy photons§ Lateral cooled by contact (+ eutectic GaIn interface)

CVD Diamond: 750µmGraphite: 2 µm coating

SiC: 10mm

Heavy Met: 10mm

Heat sink

18

Beam Collimators, Apertures/Slits


Courtesy of S. Forcat Oller

HeavyMet (WHA)

SiC

graphite coatedCVD Diamond

Core assembly

CVD DiamondOptical grade polycrystalline CVD diamond50±0.1 OD x (8, 8.5, 9, 12, 18 ID) mm750µm thickness with 2 µm graphite coating

SiC discSiC Diamond composite, grade Thermadite 100-60

60 vol.% of 100um diamond

70±0.1 OD x (8, 8.5, 9, 12, 18 ID) mm

10mm thickness

HeavyMet discAlloy HD18DV (95% tungsten + 3.5% nickel + 1.5% iron)

ASTM-B-777-15 - Class 3, for vacuum application

Density: 18 g/cm3

70 OD x (8, 8.5, 9, 12, 18 ID) mm

10mm thickness

Copper holder

19

Photon stoppers


Courtesy of HZ Wang, YP Feng

Beam

83x83x101mm3, W

Burn through monitor

OFHC heat sink

SiC plate

CVD diamond plate

SiC plate

CVD diamond plate

Unfocused fullbeam

750 mm Diamondw/ 2 mm graphitecoating+ 10-mm SiC plate+ heavy metal

redundant stopperspair

20

SXR Gas Attenuator


Ø Similar concept to LCLS gas attenuator, but§ 15 m long, N2 à Ar, higher pressure (up to 10 Torr)§ Differential pumping w/ variable size apertures

Ø Design figures§ Windowless, gas flow (no cooling)§ High differential pumping stages (9~10 order of magnitude)

Ø Requirements§ SXR (200 ~ 1300 eV)

10-5 Attenuation§ HXR (1000 ~ 2500 eV)

10-3 Attenuation

15 m

21

SXR Gas Attenuator: High rep-rate issues


Ø Gas power absorption à temperature increase à densitydecrease à less power absorption

Gas attenuator test at ESRF, 2009

Primary Slits26m

Xenon Attenuator30.9 m

Be window (0.3mm) Be window (0.3mm)

Beam

Diamond window (0.3mm)31.7 m

Calorimeter31.5 m

Diamond window (0.3mm)30.5 m

BeamSpectrumMeasurement

32.5 m

air

Diamond window (0.3mm)

Source

U42 or W70

2-mm Alspectral ratio (400mBar Xe / noGas)

0

0.2

0.4

0.6

0.8

1

0 20 40 60 80 100

eph (keV)

spec

tralr

atio

calc_P_0.40Measurementcalc_P_0.26

22

SXR Gas Attenuator: High rep-rate issues


Ø High Repetition Rate DensityDepletion Experiments at SLAC

Courtesy: YP Feng,D. Schafer

0μs 1μs 2μs 3μs

4μs 5μs 10μs 20μs

50μs 100μs

20 Torr

Courtesy: Dr. Eric Galtier(SLAC-MEC)

800[nm]“Pump”

10 images averagedat each time delayand divided by theaveraged referenceimages

hot core

pressurewave

Optical Test in SLACResearch Laser Lab

Optical Test Images

23




Ø Mechanical engineering challenges in XTESØ Attenuators, beam stoppers and beam collimators

ØX-ray optics and KB mirror systems§Optics cooling§ Flat mirror§ Bendable KB mirror systems

24

Thermal deformation of the optics

Optimizing X-ray mirror thermal performance using matched profile cooling,SRI 2015, NYC, L ZHANG et al.

Thermal Bending• Due to the temperature gradient -

variation in the thickness• Spherical shape deformation

• Dominant when Lfootprint ~ Loptics :white beam mirror

• Can be suppressed by optimizingmirror geometry and water cooling(see next slide)

Thermal Bump• Due to the temperature gradient -

variation along x-axisà Variable thickness along x-axis

bump shape deformation

• Dominant when Lfootprint < Loptics :monochromator crystal

• How to reduce this deformation ?Liquid Nitrogen cooling or else ?

Thermal deformation of the mirror - bending

Ø Top-up-side cooling + notches§ Fully illuminated§ Top-up-side cooling§ Optimized cross sectionà Minimize temperature

gradient in the thicknessand bending deformation§ Possible for multiple undulator settings

(photon energy tuning)§ Possible for multiple coating stripes

§ 1st application at ESRF ID26 (2007)

§ Applied to all ESRF beamlines using white beam mirror, and/or manymultilayer optics since then

~ 0.01 µrad

Zhang L. et al., SRI2012, Phys. Conf. Ser. 425, 052029

26

Variable footprintà mirror cooling challenging


LCLS-II power distribution on the offset mirror

θ=14 mrad

27

Mirror cooling design – 3 schemes

Ø Top-up-side water cooling

Preliminary Design Review of the KB mirrors for LCLS-II SXRAugust 27, 2015, L. Zhang & D. Morton

back face view

2. Variable-length cooling

3. Electric heater + Single-length cooling

optical face view

1. Single-length cooling

• L. Zhang et al. J. Syn. Rad. (2015). 22,1170–1181• L. Zhang et al. , SRI 2015 Conference

RMSthermal : = f (Lheater, Paheater, x)

28

Optimal heater + single-length cooling

Optimizing X-ray mirror thermal performance using matched profile cooling,SRI 2015, NYC, L ZHANG et al.

Ø At least, two parameters can beoptimized: heater length andpower (or power density)

Ø There is an optimal heater powerdensity for a given heater length

Ø Optimal heater parameters (lengthand power density) for a givenpower load distribution (or eph)

( )optcoolmiroptheater LLL -- -»21

Wheater = 5 mm

800eV, 2*FWHM=304 mmXFEL power: 20 W

29

Mirror cooling design – performance

Ø LCLS-II SXR K-B mirrors§ For 20 W of XFEL beam power, full-length (top-up-side) cooling is

sufficient§ For 200 W of XFEL beam power, optimal, variable-length cooling is

needed


Resistive Element Adjustable Length

REAL Cooled Optics(DoE funded R&D project – WavefrontPreserving Mirrors,2017-2018 FY, SLAC, BNL, ANL, LBNL)

LCLS-II FAC Review, July 19-21, 201630

Mirrors characteristics/requirements - Baseline

1.4 m 2 m 2.5 mz

Undulator flat mirror H-KB V-KB FP1 variable to FP2

90m (from undulator exit) 15 m

Flat mirrorØ 1 m long (950X25 mm2 useful)Ø Angle of incidence 12 mradØ Coating B4C and NiØ Shape errors

< 0.3 nm rms on 300 mm< 0.6 nm rms with beam< 1 nm rms on 950 mm< 2 nm rms with beam

KB mirrorsØ 1 m long (950X25 mm2 useful)Ø Angle of incidence 14 mradØ Coating B4C and NiØ Shape errors

< 0.3 nm rms on 300 mm< 0.6 nm rms with beam after bending< 1 nm rms on 950 mm< 2 nm rms with beam after bending

31

Offset mirror (flat)


Ø Sub-nm shape errorà Jtec (EEM-Elastic Emission Machining)Ø First 3 mirrors received in May 2016 at SLACØ Performance: < 0.2 nm rms in the central 300 mm, and < 0.6 overallà the best ever manufactured mirror in the world!

Measured shape errors at parallel linesat the vendor premise

32

Offset mirror (flat)


5 cooling circuits, 3 lengths(1st step towards variable-length cooling)

mirrorsurface

33




Ø Mechanical engineering challenges in XTESØ Attenuators, beam stoppers and beam collimators

ØX-ray optics and KB mirror systems§ Optics cooling§ Flat mirror§ Bendable KB mirror systems

34

KB mirror system, technical challenges


Ø Kirkpatrick-Baez (K-B) mirror configuration

Ø Ellipsoidal shape

Ø Technical challenges§ Large Acceptance à Long mirror§ Variable Source & Focal Pointsà Bendable Mirror§ Sub Nanometer Shape Errorà Limited Suppliers§ High Demagnificationà Tight Bending (stress issues,…)§ Few tenth nrad residual bending errorà Variable Mirror Width§ High Thermal Loads & Variable Footprint à Innovated Cooling§ Minimize the coupling between the mirror Bending & Cooling

35

Design Solution

Vertical Focusing Mirror

Horizontal Focusing Mirror

Ø LCLS-II K-B mirror system§ Dynamically bendable§ Water cooled

• Invar 36 Strongback• Titanium Flexure Benderw/ Height Correction

• Titanium Flexure Benderw/ Twist Correction

36

Bender

• Flexure Lock Plates• Titanium Leaf SpringLever Arms

• Titanium Push PullFlexures

• UHV Linear Actuators• High StrengthAerospace Epoxy

• Variable Width Profile SiMirror

http://www.janssenprecisionengineering.com/precisionpoint/Hart-Smith, L.J. (1983), “Designing to Minimize Peel Stresses inAdhesive-Bonded Joints”

• Cooling Pads

37

Cooling

• Invar Supports• Copper Cooling Rails• Clamps• Gallium Indium

38

Technical challenges


§ Large Acceptance à Long mirror§ Variable Source & Focal Pointsà Bendable Mirror§ Sub Nanometer Shape Errorà Limited Suppliers§ High Demagnificationà Tight Bending (stress issues,…)

§ Sub-µrad residual bending errorà Variable Mirror Width§ High Thermal Loads & Variable Footprint à Innovated Cooling§ Minimize the coupling between the mirror Bending & Cooling

39

Mirror profile optimization


Ø Width profile defined by Bending Equation (BE)

Ø Limitation of the analytical formula(Beam theory approximation)

)()(12)( 3 xREt

xMxw =

F1 = F2 = 60 N

F1 = 62.92 NF2 = 63.58 N

Residual Slope Error (RSE) :Δslope = slope – slopeellipse

ANSYS Release 16.0 16.0AUG 4 201508:31:00ELEMENTS/EXPANDEDPowerGraphicsEFACET=1

1

FE model with bending forces (VFM)

F

X Y

Z

LCLS-II KB mirror: VFM, Fin=60, Fout=60 N, Ndxc=8, i=5

R(x)

40

Mirror profile optimization


ANSYS Release 16.0 16.0AUG 4 201508:31:00ELEMENTS/EXPANDEDPowerGraphicsEFACET=1

1

FE model with bending forces (VFM)

F

X Y

Z

LCLS-II KB mirror: VFM, Fin=60, Fout=60 N, Ndxc=8, i=5

Silicon crystal orientation(low stress & bending force)

• Mirror optical surface //Si (110) plan

• Tangential-axis // [001]

Optimized Mirror Profile (VFM, q=2m)

V201505

• L. Zhang, SMEXOS (2009), Grenoble, France• L. Zhang et al., AIP Conference Proceedings

1234, 801 (2010); doi: 10.1063/1.3463335

41

Optimized Mirror Profile – bending performance


iter=1 iter=2 iter=3 iter=4 iter=5

RMSΔslope (reduction factor : ~ 104)§ 43.7 μrad (with the profile defined by BE)§ 0.005 μrad (with the optimized profile by FEA)

RMSΔslope-opt /slopePV-ellipse ~ 2 10-6

à Following effectsto be taken into account

§ Bender stiffness (not negligible)§ Anticlastic-bending effects§ Anisotropy of the Si crystal§ Geometrical non-linear effects in

the simulation

42

Technical challenges


§ Large Acceptance à Long mirror§ Variable Source & Focal Pointsà Bendable Mirror§ Sub Nanometer Shape Errorà Limited Suppliers§ High Demagnificationà Tight Bending (stress issues,…)§ Sub-µrad residual bending errorà Variable Mirror Width§ High Thermal Loads & Variable Footprint à Innovated Cooling

§ Minimize the coupling between the mirror Bending & Cooling Minimization of mechanical constraint effects of Eutectic GaIn

as thermal interface

Bend cooling blocks (design optimization practice)

43Minimization of mechanical constraint effects of Eutectic GaIn as thermal interfaceMEDSI 2016, Sept. 11-19, 2016, L. ZHANG

GaIn interfaces (pink lines) , gap = 50 μm

Mirror bendingà mirror motion (up-down)relative to fixed cooling blocks

MN

MX X Y

Z

What’s the mechanical constraint effectsduring the mirror bending?

Eutectic GaIn as thermal contact interface

44

Eutectic Gallium-Indium (eGaIn)

Minimization of mechanical constraint effects of Eutectic GaIn as thermal interfaceMEDSI 2016, Sept. 11-19, 2016, L. ZHANG

Ø eGaIn: 75.5% Gallium (29.76ºC)24.5% Indium (156.6ºC)

Ø Thermal interface§ Thermal conductance > 105 W/m2.K, ~ 10 times better than Indium foil§ Interface or trough / bath for indirect cooling of the X-ray mirrors

Ø Mechanical properties§ Bulk viscosity ~ 2 mPa*s (2µwater)§ Modulus ??§ Thin skin (oxides of Ga)

Dickey et al., Adv.Funct.Mater.2008,18,1097–1104

Dickey, ACSAppl.Mater.Interfaces2014,6,18369−18379

Tmelting=15.7ºC

45

GaIn test Setup


Ø Si-wafer (D=4”, t = 4 mm, both side polished)

Ø GaIn gap: (51, 102, 152, 203, 254 μm)

Ø Measurements§ Displacement of Si-wafer dwafer§ Forces of Si-pads on

Left side and Right side: F_L, F_R§ Relative displacement between wafer and pads: dpad = dwafer – (F – F0)/ kFG§ Force gauge stiffness: kFG = 10.57 N/mm

Displ. Sensor A

Displ. Sensor B

Silicon wafermoving part

Silicon padfixed on aforce sensor

46

GaIn Test results – Data fitting, modulus


47

GaIn test results – gap dependant; Modulus G1 and G2

Minimization of mechanical constraint effects of Eutectic GaIn as thermal interfaceMay 2nd, 2016

G2 = 9 Pa

Shear modulus G of GaIn interfacegap G2_av G2_min G1

μm Pa Pa Pa50 43.7 20.0 1421

100 16.2 8.6 534

150 10.3 3.4 232

200 8.2 4.7 285

250 8.3 4.8 94

G1

G2

# cycles(motionpatterns)

48

Bending Operation Optimization with GaIn Interface


BE

q=4 m

q=2 m

~2 mN

49

FEA results with cooling blocks and GaIn interface


ANSYS Release 16.0 16.0MAR 30 201617:20:31PLOT NO. 199NODAL SOLUTIONSTEP=1SUB =1TIME=1/EXPANDEDUZ (AVG)RSYS=0PowerGraphicsEFACET=1AVRES=MatDMX =.412588SMN =-.175605SMX =.025594

1

MN

MX

Vertical displacement of the mirror Uz (mm)

F

X Y

Z

-.175605-.15325-.130894-.108539-.086183-.063828-.041473-.019117.003238.025594

VFM,FP1, GaIn(200um, G=9 Pa), Fin=90.014,Fout=90.011N, Uz_BC=0um

G2=9 Pa

50

FEA results with cooling blocks and GaIn interface


Residual Slope Error (RSE) :Δslope = slope – slopeellipse

51

GaIn gaps


gGaIn = 500 μm

gGaIn = 50 μm

gGaIn = 200 μm

Before assemblingfor test at gGaIn=152 μm

After test atgGaIn= 152 μm

Wafer side

Wafer side

pad

pad

52

Intensity distribution around focus (1300 eV)


SRW Sirepo(Oleg Tchoubar +Radiasoft)

Single lengthcooling(no correction)

Single lengthcooling(focus correctionby translation)

REAL cooling(focus correctionby translation)

10 mm in front at the focus 10 mm behind

53

Intensity distribution around focus (500 eV)


SRW Sirepo(Oleg Tchoubar +Radiasoft)

Ø Intensity distribution at focus

54

Collaboration with NSLS-II


Wavefront preserving mirrors(DoE funded R&D project, 2017-2018 FY, SLAC, BNL, ANL, LBNL)

NSLS-II:Mourad Idir: wavefront sensorOleg Tchoubar: wavefront propagation simulation…...

55

Acknowledgement


• E. Anderssen LBNL• R. Baker ESRF• J.C. Castagna SLAC/LCLS• M. Church SLAC/SSRL• R. Duarte LBNL• Y.P. Feng SLAC/LCLS-II• S. Forcat Oller SLAC/LCLS-II• D. Harrington SLAC/SSRL• T. Rabedeau SLAC/SSRL• A. Ringwall SLAC/SSRL• E. Ortiz SLAC/LCLS-II• B. Schlotter SLAC/LCLS-II• O. Tchoubar BNL/NSLS-II• H.Z. Wang SLAC/LCLS-II

Many SLAC colleagues

These works performed under DOE Contract DE-AC02-76SF00515.

Documents

LCLS-II project and beamline instruments: mechanical ... · 10/18/2016 · LCLS-II project and beamline instruments: mechanical engineering challenges NSLS-II Engineering Seminar