Thomas Jefferson National Accelerator Facility 1 of 20 Distribution State A “Direct” Injection...

Thomas Jefferson National Accelerator Facility

1 of 20

Distribution State A

“Direct” Injection

D. Douglas, C. Tennant, P. Evtushenko

JLab


2 of 20


Acknowledgements

• Initial funding provided by ONR• Recent work supported by AES under JTO

funding• Initial simulations (sanity check!), useful feedback

provided by John Lewellen, discussions with Steve Benson, operational help from Kevin Jordan


3 of 20


“Direct” (off-axis) Injection

• Rather than merge beams using DC magnetic fields, inject beam into linac at large amplitude and use RF focusing & adiabatic damping to bring orbit into line

• Can use reverse process for extraction of energy-recovered beam

0.15 m

current sheet or field clamp

linac centerline

0.075 m

injected beam

recirculated beam, reinjected for energy recovery

accelerated and recovered beams in linac


4 of 20


Direct Injection/Extraction

start of linac

-0.075

-0.050

-0.025

0.000

0.025

0.050

0.075

-0.075 -0.050 -0.025 0.000 0.025 0.050 0.075

x (m)

y (m

)

1st pass

2nd pass

end of 1st cavity

-0.075

-0.050

-0.025

0.000

0.025

0.050

0.075

-0.075 -0.050 -0.025 0.000 0.025 0.050 0.075

x (m)

y (m

)

1st pass

2nd pass

end of 2nd cavity

-0.075

-0.050

-0.025

0.000

0.025

0.050

0.075

-0.075 -0.050 -0.025 0.000 0.025 0.050 0.075

x (m)

y (m

)

1st pass

2nd pass

end of 3rd cavity

-0.075

-0.050

-0.025

0.000

0.025

0.050

0.075

-0.075 -0.050 -0.025 0.000 0.025 0.050 0.075

x (m)

y (m

)

1st pass

2nd pass

end of 4th cavity

-0.075

-0.050

-0.025

0.000

0.025

0.050

0.075

-0.075 -0.050 -0.025 0.000 0.025 0.050 0.075

x (m)

y (m

)

1st pass

2nd pass

end of 5th cavity

-0.075

-0.050

-0.025

0.000

0.025

0.050

0.075

-0.075 -0.050 -0.025 0.000 0.025 0.050 0.075

x (m)

y (m

)

1st pass

2nd pass

end of 6th cavity

-0.075

-0.050

-0.025

0.000

0.025

0.050

0.075

-0.075 -0.050 -0.025 0.000 0.025 0.050 0.075

x (m)

y (m

)

1st pass

2nd pass

arc/dump split

-0.075

-0.050

-0.025

0.000

0.025

0.050

0.075

-0.075 -0.050 -0.025 0.000 0.025 0.050 0.075

x (m)

y (m

)

1st pass

2nd pass

cross-sectional view of both passes of beam (first = blue, second = pink) looking down linac from injection to dump


5 of 20


Issues & Solutions

• Concerns• Possible emittance dilution from finite phase extent of bunch in RF fields

(thanks to Steve Benson for pointing this out…)• Potential for HOM excitation/BBU instability

• Approach• Estimates & analysis (emittance, BBU)• Simulation (PARMELA, GPT)• Beam studies on JLab Upgrade Driver


6 of 20


Head-Tail RF-Driven Emittance Dilution

• Reviewed head-tail issue• assumed beam was 8 degrees long (6head to tail

(~Jlab injected length)

• Simulated RF steering of injected beam with simple cavity matrix model

Results:• Propagated beam envelopes vary only slightly• Differential steering not dramatic


7 of 20


beam envelop for head, centroid, tail

0

2

4

6

8

10

12

0 2 4 6 8 10

s (m)

bet

a x

(m)

beta x -24 deg

beta x for centroid

beta x -16 deg

central orbit for tail, centroid, and head

-0.04

-0.03

-0.02

-0.01

0

0.01

0.02

0.03

0.04

0.05

0 2 4 6 8 10

s (m)

x (m

) 4 degrees late

centroid

4 degrees early

central orbit for tail, centroid, and head

0.04

0.041

0.042

0.043

0.044

0 2 4 6 8 10

s (m)

x (m

) 4 degrees late

centroid

4 degrees early


8 of 20


• head/tail (orbit) centroid move ~ ±0.2 mm in position, ±30 microrad in angle.

• compare to the beam size – for 5 mm-mrad normalized emittance at 100 MeV, with 10 m beta:

• x ~ sqrt()=sqrt(10*5e-6/(100/0.51099906)) ~0.5 mm• x’ ~sqrt(/)=(5e-6/10/(100/0.51099906)) ~50 rad

• with stated assumptions about the bunch length get ~ ± ½ sigma motion – over the full (6) bunch length

Conclusion: emittance dilution may not be too bad; look at more carefully…


9 of 20


Detailed Study

• Performed as part of JTO-funded AES merger study• Three part investigation

• More careful analytic estimates• Simulations with space charge• Beam study on Jlab IR Upgrade

Conclusions:

emittance growth very modest; tolerable for IR systems

BBU thresholds unaffected; additional power goes into HOM loads

Several cm pass-to-pass possible


10 of 20


Results – Theory/Simulation

0 1 2 3 4 5 6 7 8 9 10 11 12

GPT z

-0.04

-0.03

-0.02

-0.01

0.00

0.01

0.02

0.03

x

GPT simulation of beam size in single-module linac (C. Tennant)

• Estimates emittance growth negligible for IR FELs

• Emittance growth negligible in simulation• Beam quality not degraded

• Analysis BBU threshold independent of injection offset• C. Tennant, JLAB-TN-07-011

• Power into HOMs depends on injection offset


11 of 20


Bunches Traveling Through Linac: Animation

Injected on-axis

Injected 10 mm off-axis

C. Tennant and D. Douglas | July 24, 2008


12 of 20


Machine Study

• Measured impact of injection offsets on beam quality in JLab IR Upgrade• Aperture limited to ~1 cm offsets• Able to run CW @ 1 cm BBU tests possible• Tested at nominal (9 MeV) and low (5 MeV) injection energy

Conclusion: No observable impact on beam quality; BBU-related measurements underway


13 of 20


Machine Study: Method

• Measure injected emittance (multislit)• Quad scan emittance measurement after linac

• On axis & several displacements• Tomography in recirculator

• BBU – look at power into HOMs in 7-cell module


14 of 20


Steering

• “off-axis” emittance tests: steer off into 1st module, grab at end of module where RF focusing bring (nearly ) to node (no offset downstream)

• “BBU” tests: steer off into linac, resteer in recirculator to maintain 2nd pass transmission

note path-length/phase/energy effects in arc…


15 of 20


Machine Study: Results

• Transversal beam sizes and profiles largely independent of injected orbit over ±1 cm offsets in H and V

• Machine drift much higher impact than orbit offset• Initial data analysis of emittance data emittance unaffected by steering (to

resolution of measurement)• Working through error propagation

• BBU: • set up CW configuration, acquired initial signals, whereupon machine

crashed (refrigerator trip); • lost rest of run to LCW line break before follow-on shifts • will schedule more study time over the summer


16 of 20


Beam Profile At End of Linac

x=-10 mm x=0 mm x=+10 mm

y=-10 mm y=0 mm y=+10 mm (some scraping) profile measurement by P. Evtushenko & K. Jordan


17 of 20


Transverse Emittance (5 MeV injection)

• Measured with 3 methods:

1. “multislit” in injector

2. quad scan at end of linac

3. tomography in recirculator backleg • Results generally consistent and roughly match values w/ full energy injection

Location Method Result (mm-mrad)

Injector Multislit ~ 13

End of linac Quad Scan ~10-15

Backleg of recirculator Tomography ~10 (horizontal)


18 of 20


Emittance Data @ 5 MeV Injection

Multislit: ~ 13 mm-mrad

Tomography: ~ 10 mm-mrad

beam spot

reconstructed phase space

Quad scan: ~ 12-15 mm-mrad

tomography courtesy C. Tennant

Multislit courtesy P. Evtushenko


19 of 20


“Direct” Injection @ 5 MeV

• Test of “merger-less” merger• Low-loss operation with large (~ cm) injection offsets• Beam behavior ~independent of injection orbit


20 of 20


Conclusions

• Direct injection provides possible alternative to traditional merger• Beam quality requirements are key

• likely appropriate for IR systems, • may not be quantitatively appropriate for, e.g. shorter wavelength

applications• Lower frequency better (i.e. “easier”, more available aperture!)• Few-several cm separations possible• Still need to evaluate emittance data (error analysis) and measure HOM power

deposition

Date post:	20-Jan-2016
Category:	Documents
Upload:	camron-charles
View:	213 times
Download:	0 times

Thomas Jefferson National Accelerator Facility 1 of 20 Distribution State A “Direct” Injection...

Documents