Seismic Isolation Retrofit Plan for the LIGO Livingston Observatory

1
Seismic Isolation Retrofit Plan for the LIGO
Livingston Observatory

• Dennis Coyne
• 29 Nov 2001

2
Issues

• Spiky seismic noise 1 - 3 Hz band
• related to human activity, especially cutting
  lumber
• worst at y end, appears as wide band noise
  through narrow band filter
• coincident with transmission peak in test mass
  isolation stack and seismic amplification in the
  soil under the Y-end station
• precludes interferometer lock during work day
• Rayleigh surface waves vertical/ horizontal
  1.5/ 1, 200 - 300 meters/ sec
• Most likely growing with epoch, tracking
  population
• 1988 LSU survey not evident
• 1995 Rohay study about 1/ 2 as large and 1/ 2
  rate of today
• need to include margin in fix
• Strategy to deal with the noise
• higher peak current controller short term, adds
  noise
• active external isolation short study, feedback,
  feed forward.
• need to be installed before reaching design
  sensitivity - 1 year

3
Improved Seismic Isolation

• Improved isolation is essential for Test Masses
  BSC Chambers
• Need factor of 20 reduction in the rms motion
  between 1 and 5 Hz (Firm requirements are being
  developed)
• Improved isolation for the Mode Cleaner optics
  may be necessary HAM chambers
• Simulation is being developed to study the effect
  of the resulting frequency noise modest means
  of reducing the rms

4
Improved Seismic Isolation

• Industry
• No current commercial off-the-shelf system is
  directly applicable
• Have contacted some isolation specialty firms
  not promising
• Reluctant to place the responsibility
  out-of-house as the problem is clearly unique
  challenging for industry
• Will continue to study adaptation of TMCs STACIS
  system others possibly
• Fine Actuators
• Existing single degree of freedom actuation
  system (piezoelectric) used for tidal correction
  and microseismic feedforward compensation could
  be employed for isolation (feedback or
  feedforward)
• Considerable cross-coupling in the stack modes
  makes this approach unlikely to succeed
• Experiments at Livingston are planned

5
Improved Seismic Isolation

• Tuned Mass Damping
• Passive method to increase modal damping
  requires (perhaps brief) entry into the vacuum
  system
• Unlikely to get more than a reduction of a factor
  of 8
• May suffice for the HAM stacks, or may help an
  active system on the BSC
• Analysis is underway
• Active 6-dof pre-isolator
• Preferred approach
• retrofit fine/coarse actuation with an active
  pre-isolation system under the existing passive
  isolation stacks hope to preserve alignment
• Leading candidate is the hydraulic actuator that
  the LSC (Stanford) is developing for advanced
  LIGO
• Alternate actuators (e.g. piezo, EM) are under
  study in parallel

6
Hydraulic system
7
Hydraulic prototype test results

• 2-DOF system
• Feedforward andfeedback
• Potential to reduceambient seismicnoise also
  forinitial LIGO if indicated

8
Implementation

• Actuator prototypes component test stands are
  underway at Stanford for the actuator and
  sensors at MIT for the hydraulic distribution
  system
• Full scale prototype system test including
  retrofit installation is planned on the LASTI BSC
• BSC stack has been installed
• Long lead items procured in parallel with
  prototype testing
• Team
• Accelerated development by the existing LSC adv.
  LIGO team, augmented by LIGO Lab staff
• Effort transitions out of LSC RD into Lab
  implementation through the LASTI prototype
  development

9
Schedule Cost

• Preliminary Design Review in Jan
• System modeling results
• Sensors selected characterized
• Hydraulic actuator prototype tested (bolted
  version)
• Distribution system designed modeled
• Prototype installation test _at_ LASTI May Sep
• Installation at Livingston starts Oct
• Rough estimate is 750K
• for 4 BSCs, 2 HAMs
• beyond the RD costs associated with the
  prototype development (which is mostly part of
  the existing RD plan)