The Evolution of Vacuum System Design for Accelerators over 30 years at Daresbury Lab

1
The Evolution of Vacuum System Design for
Accelerators over 30 years at Daresbury Lab

• Ron Reid
• Group Leader, Vacuum Science Group
• ASTeC
• CCLRC Daresbury Laboratory
• Warrington WA4 4AD, UK

2
Outline

• Our second accelerator (SRS)
• System Design
• Pressure calculations
• Materials
• Pumps
• Gauges
• Conditioning
• Our next accelerator

3
The SRS

• NINA
• SRS (1980)
• First 2nd Generation Light Source
• Medium Energy
• 2GeV, 1A (350mA), 100m, 3.88Å
• 10-9 Torr 12 hours
• lt 10-9 Torr, gt40 hrs _at_300mA
• Two major upgrades

4
The SRS
5
The SRS
6
Simple Vacuum Layout
7
Vacuum System Design - Then

• Highly simplified
• Analytical solution of gas flow equations
• No PCs mainframes with turnround of days
• Simple finite element models
• Fortran decks

8
Vacuum System Design - Now

• PCs ubiquitous
• Realistic calculations of complex systems
• Monte Carlo simulations
• Angular coefficients
• Quick optimisations of layouts

9
Materials

• Stainless steels as now
• Aluminium less common then than now
• Absorbers
• Copper
• Finite element calculations were not available
• Glidcop was not available
• Fabrication Techniques
• EB and plasma welding now more generally available

10
Pumps
11
Pumping schemes

• In SRS about 50 of circumference is dipole
  magnets vacuum vessels
• Main radiation absorbers are linear (distibuted)
  in dipole vessels
• Main gas load
• Most restricted pumping
• Solution DIPs

12
DIPs
13
Pumping schemes

• In SRS the DIPs give almost half the applied
  pumping speed during beaming
• Losing just one DIP gives a noticeable decrease
  in lifetime

14
Pumping schemes for our current designs

• Our new machine designs have much smaller
  apertures, dipoles occupy only a small fraction
  of the circumference
• DIPs no longer appropriate
• Machines use discrete photon stops
• Current preferred solution
• NEG (discrete and film)

15
Gauges

• Little change in type of gauge
• Except BAG often replaced by IMG
• Main difference is the electronics now compact,
  more reliable
• Biggest difference is in control - then home
  brewed, now uses standard interfaces

16
Conditioning

• Then - beam scrubbing effect was known
• But limited data and best measured values of ?
  were 10-6 mol ?-1
• SRS design needed 10-7
• Cleanup rate not well characterised
• Now much greater range of systematic data and
  clean up rates predictable as a function of beam
  dose

17
Our next accelerator 4GLS
18
4GLS

• Direct vacuum challenges seem to be few
• Most of the machine needs to be at modest UHV,
  maybe HV
• But
• Photocathode gun is XHV
• Superconducting linacs
• Particulate Cleanliness is a real challenge

19
Concluding Ramblings

• Thanks for your indulgence!