SCUs for the LCLS-II HXR FELP. Emma, et. al.

July 9, 2014

Hard X-Ray (HXR) FEL for LCLS-II must cover 1-5 keV (4-GeV)SASE to 5 keV is just possible (with no margin)Self-seeding with PMU not possible beyond 4 keVRadiation dose at 1 MHz threatens PMU fields (DK/K ~ 0.01%)SCU can extend photon range toward 8 keV (1 MHz), and beyond, can allow TW peak-power levels (120 Hz), and with much less sensitivity to radiation dose

L2-Linac L3-Linac

HXU

SXU

Sec. 21-30

LH BC1 BC2

BC3

D2

D10

m-wall

0.65 m0.93 m

2.50 mL1

“kicker”LTUH

LTUSLCLS-ILinac

Pe 120 kW

Pe 120 kW

Pe 250 kW Lu 145 m

PMUNbTi Nb3Sn

g = 5 mm

g = 4 mmg = 4 mm

E = 4.0 GeV (nominal)

145 m

gex,y = 0.40 mmIpk = 1 kAsE = 500 keVb = 16 m20% Lu margin3.4-m seg’s1.0-m breaks1 und. missing1.5 keV low-lim.

magnetic gap is 2.3 mm larger than vac. gap, g

7.6 keV SASE4.8 keV SASE6.8 keV

SASE

lu = 25.6 mm, 18.4 mm, 16.8 mmK = 0.6-2.4, 1.1-3.0, 1.3-3.2B = 0.2-1.0 T, 0.6-1.8 T, 0.8-2.0 T

PMUNbTi Nb3Sng = 5 mm

g = 4 mmg = 4 mm

E = 4.2 GeV (stretch)

8.2 keV SASE145 m

lu = 25.6 mm, 18.4 mm, 16.8 mm

~7 keV HXRSS limit

gex,y = 0.40 mmIpk = 1 kAsE = 500 keVb = 16 m20% Lu margin3.4-m seg’s1.0-m breaks1 und. missing

magnetic gap is 2.3 mm larger than vac. gap, g

E = 4.8 GeV (20% upgrade)

10 keVSASE

PMUNbTi

Nb3Sn

g = 5 mm

g = 4 mm

g = 4 mm

145 m

lu = 25.6 mm, 18.4 mm, 16.8 mm

8.3 keV HXRSS limit

magnetic gap is 2.3 mm larger than vac. gap, g

P (T

W)

z (m)

Add field taper to SCU

TeraWatt Peak Power Possible

C. Emma, C. Pellegrini, Z. Huang

1.2 TW

Nb3Sngm = 7.2 mmge = 0.4 umE = 7.8 GeVf = 120 HzIpk = 4 kA

Possible SCU Layout in LCLS-II (HXR)

Joel Fuerst, ANL

Joining Three 1.5-m Magnets into a Single 4.5-m Device

• Magnets conduction cooled through gap separation extrusion

• Gap separation ensures “seamless” core-to-core joint

• Core center channels may be omitted

End corrector End correctorPhase shifter dipole

Alignment verification quadsand Bx correction

Lb

Lb

Second Field Integral with phase shifter

+k +k

-2k

• Compact phase shifter uses one end corrector from each undulator and one extra dipole magnet in between

• Distance between the undulator cores ~13 cm for this layout (could be reduced if alignment quadrupoles are not necessary)

• Joint sections for Nb3Sn undulator are 4 cm long for each core

Joining Two 1.5-m Segments (13 cm extra)

Cold Break (quadrupole magnet)Conceptual design of a compact quadrupole magnet at breaks

– Directly attached to undulator cold mass– Integrated quadrupole strength of 4 T (LCLS-II) can be obtained– Independently powered coils can be used for x-field correction

End correctorQuadrupole Magnet

Vertical Alignment with Alignment Quadrupoles• Use reference quads at each end of ~3-m structure

– Tuning and calibration based on line between magnetic center of two quads

– Fiducialization performed with stretched wire measurement and referenced to fiducials on outside of cryostat

– Allows for beam based alignment by moving cryostat to find center of quads with electron beam

Small Alignment QuadFull Length Quadrupole

DC Resistive-Wall Wakefield (cold bore & warm)

cold bore

warm boreCu Al

Cu

Al

Parameter NbTi (ANL) Nb3Sn (LBNL) unitsPhoton Energy (4 GeV) 1.5 - 5.0+ 1.5 - 5.0+ keVUnd. Segment Length ~2 ? ~2 ? mUnd. Break Length (cold) 0.5 - 0.7 0.5 - 0.7 mUnd. System Length 145 145 mMissing Segments (HXRSS) 1-2 1-2 -Number of Segments 54 ? 54 ?Magnetic full gap 6.3 6.3 mmVacuum full gap 4.0 4.0 mmUndulator period 18.4 16.8 mmOn-axis field 0.6 - 1.8 0.8 - 2.0 TK parameter 1.1 - 3.0 1.3 - 3.2 -Cryo Power Required ~250 ~250 W

Undulator Parameters (4 GeV)

SCU Advantages

• Dramatically improved HXR-FEL performance (~7 keV with SC-Linac & 1-TW with Cu-Linac)

• Orders of magnitude less radiation dose sensitivity (1 MW!)• No mechanical motion (DC power supply drives each

segment)• Less tunnel space required – looks like LCLS-I und. (?)• Can easily be arranged as vertical polarizer (hor. fields)• Cryo-plant (4.5K, 280 W) might serve as injector stand-in at 2K

(7.5M$) – cryo-dist. system not incl. • Might drop 5 linac CM’s (3.4 GeV) and still get 5 keV (Ti or Sn)?• Or drop 10 linac CM’s (2.8 GeV) and still get 4 keV (Sn only)?

1. What decision criterion should be used to make this technology decision in 2015.

2. What is the impact on the LCLS-II project schedule and what additional

resources are needed to develop production SCUs and install them by the same April 2018 date presently planned for the PMUs?

3. What other subsystems would have to be developed / designed / specified

and become part of the baseline? How would they fit into the facility / tunnel?

4. What other reviews would we have to organize before this could become

baseline? 5. What other assurance would you want as the project director before you

can support the change?

Questions