Aperture Considerations in the FEL Upgrade

1
Aperture Considerations in the FEL Upgrade

• Accepted design process
• generate design Þ s known
• set aperture Ns W
• N typically 4 to 6
• W is beam handling allowance
• example IR Demo has A 6s 4 cm
• Other restrictions may apply
• constraints imposed by FEL - optical mode size
• Here, programmatic considerations force deviation
  from accepted practice Þ risk escalates
• Can reduce risk by using all available
  information
• previous design studies
• experience with IR Demo

2
What Do We Know?

• No design Þ b unknown
• Injector not quantitatively understood Þ eN135 pC
  unknown
• s unknown
• FEL optical mode larger Þ 3 aperture needed
  unless we can compress e- beam transport

3
What Can We Reasonably Surmise?

• eN135 pC gt eN60 pC
• bupgrade gt bdemo
• larger machine Þ larger b and/or more quads
• more quads undesirable
• higher cost
• increased chromatic aberration (in turn a limit
  on larger required momentum acceptance)
• 1st iteration linac optics (actually, 2nd - 1st
  was UV Demo design study) has larger beam
  envelopes
• bs same in modules Þ 2 may be okay for
  modules provided emittance does not increase too
  much
• bs 2 x larger in warm regions
• triplet focussing needed to handle longer linac,
  higher RF focussing from increased module
  gradient
• Þ for same emittance, need bigger aperture

4
(No Transcript)
5
Geometric Emittance Comparison to Demo

• bupgrade gt bdemo with eupgradegeometric gt
  edemogeometric Þ larger spots
• eupgradegeometric gt edemogeometric with bupgrade
  bdemo Þ larger spots
• eN135 pC gt eN60 pC likely, bupgrade gt bdemo
  certain
• Injector setup required for high FEL gain
  (tapered wiggler tests) limited to 1.5 mA by BLM
  hits Þ 2 aperture inadequate even at 60 pC when
  high gain configuration required?

6
Conclusion 1

• Though 2 possibly (probably?) adequate in
  modules, peak bs in upgrade are in warm regions
  and will drive increase in aperture there
• Recommendation(s) 1
• Make effort to understand injector quantitatively
  - and run 5 mA CW at 135 pC
• helps define if 2 injector chamber allows
  reliable operation
• characterized normalized emittance at elevated
  charge
• 3 warm region in linac

7
Linac-to-FEL Transport at 100-200 MeV

• It is possible egeometricupgrade lt egeometricdemo
  in the module to FEL transport even with
  space-charge driven degradation (higher energy)
• Þ bupgrade gt bdemo is washed out in spot size in
  full energy transport
• note that at same energy (mid linac in upgrade,
  end of linac in demo) spot sizes are larger in
  upgrade
• at low end of energy range (100 MeV) spots may
  be same or larger in upgrade due to increased
  normalized emittance and larger beam envelopes

8
Conclusion 2

• 2 tube may be adequate for full energy beam from
  end of linac to start of FEL insertion
• Recommendation(s) 2

9
Component Reuse

• Larger aperture requirements limit component
  reuse to regions such as linac-to-FEL transport
• Diagnostics reusable without modification
• QB quads probably reusable without modification
• 48 MeV IR Demo QB maximum current 2 A
• QBs specd to 10 A with LCW
• Þ can get to 200 MeV with 20 headroom for
  matching
• Correctors may prove useful under similar
  analysis

10
FEL Insertion Region

• Optical mode significantly larger than in IR
  Demo
• either use 3 aperture (including dipoles)
• or restrict matching regions to 5 m length
• Current existence proof uses 10 m match
• manages aberrations at 5 momentum offsets by
  adjusting phase advances amongst telescopes/arc
  components
• causes destructive interference of chromatic
  effects
• y òds/b Þ if L reduced, b must reduce
• good for small apertures, but,
• b smaller Þ quads stronger
• stronger quads Þ aberrations larger
• higher order chromatics quadratic in quad
  strength, Þ halving lengths doubles quads,
  quadruples aberrations

11
Conclusion 3

• 10 m match meets specÞ5 m match 4 x out of
  spec
• - go with 3
• Recommendation(s) 3
• FEL insertion region
• basic optimization for matching telescope length
  must balance keeping b small - for good
  performance and acceptance while keeping L large
  - to limit quad strength
• Þ 10 m match in this machine

12

• Choose magnet families to keep construction
  simple
• fringe models developed for spectrometer magnets
  3 is not large so predictive capability likely
  okay
• match magnet gaps in similar families

• p-bends probably tolerate 2 because b, (and h)
  smaller
• power requirements dominated by p-bends (180o out
  of 300o bending per end loop, so draw most of
  power)
• IR Demo successful matching magnets within and
  across families should anticipate similar
  results in upgrade

13
Conclusion(s) 4

• To avoid undue risk must make FEL insertion 3
• Little additional cost in making all reverse
  bends 3
• moderate additional DC power (most in p-bends)
• no overhead in lost magnets
• no dipoles lost as none upgrade
• need new trim quads, 6-poles, 8-poles due to
  horizontal aperture increase necessary to
  accommodate 10 dp/p
• significant risk reduction, especially for lower
  energy operation at higher space charge (can
  tolerate 2x larger emittance)

14
Injection/Reinjection Region - 2 or 3?

• bupgrade 2 or 3 x bdemo at reinjection
• eNupgrade gt eNdemo (space charge)
• egeo.upgrade 1/2 to 1/3 egeo.demo (adiabatic
  damping)
• Þ it will not get better
• How good is it now?
• Cavity 8 tunes a fair bit (Þ losses)
• ILM0F062 hits have been limitation
• ILM0F06 hits are a limit when running injector
  for high wiggler gain

15
Conclusion 5

• 3 prudent risk reduction at modest incremental
  cost
• new injection/extraction dipoles needed to
  increase available dynamic range of
  injection/final energy
• small magnets (DU/DV) Þ minor power impact
• QJ quads/associated correctors support 3
• need additional quads for recirculator
• not enough QBs to populate reinjection region
• at very least, need to re-coil some QGs (4 for
  linac to FEL transport, this region would require
  an additional 6 or 7)
• could build an additional half-dozen 3 quads