Slides

Transcription

Slides
Seismic isolation system and room
temperature payload for KAGRA
NAOJ, ICRRA, ERIB, Univ. SannioC, INFN RomeD, NIKHEFE
Fabián Erasmo Peña Arellano, Takanori SekiguchiA, Yoshinori Fujii, Ryutaro Takahashi, Mark Barton,
Naoatsu Hirata, Ayaka Shoda, Hideharu Ishizaki, Naoko Ohishi, Kazuhiro YamamotoA, Takashi UchiyamaA,
Tomotada Akutsu, Yoichi Aso, Osamu MiyakawaA, Masahiro KamiizumiA, Akiteru TakamoriB, Riccardo
DeSalvoC,Ettore MajoranaD,Eric HennesE,Jo van den BrandE,Alessandro BertoliniE,Kazuhiro AgatsumaE,
J. van HeijningenE
Outline
• Description of the Type B prototype.
• Vibration isolation chain.
• Payload.
• The prototype in TAMA300
• Damping the resonances of the suspension.
• Alignment range of the payload.
• Performance.
• Work at Kamioka.
• Conclusion
Elites Meeting 2015
2
Requirements
• Vibration isolation requirement: 1 × 10−17 m
Hz.
• RMS velocity of the mirror: less than 0.5 𝜇m s (depends on the
bandwidth of the control system.)
• RMS angular displacement: less than 1 𝜇rad.
• Increase of observation time: short recovery time after an
unexpected excitation of the system (e.g. an earthquake).
Elites Meeting 2015
3
Type B prototype
Top GAS filter
Pre-isolator
Inverted
pendulum
Standard GAS filter
Bottom GAS filter
Intermediate
mass assembly
Payload
Optic assembly
Elites Meeting 2015
4
Pre-isolator
GAS blade springs
Geophones and
Seismometers on
top (not pictured)
LVDT and coil-magnet
Actuator.
• IP: three flexures.
• Geophones: output proportional to velocity
above 0.3 Hz.
• LVDT: used below 0.3 Hz.
• Signal blending of both signals.
• Actuator IP: coil-magnet actuator.
• Resonant frequency IP: less than 0.1 Hz.
• Resonant frequency TF: 0.37 Hz.
Inverted
pendulum
flexure
Elites Meeting 2015
5
Magnetic damper
Standard Filter
Blade springs
• Sensor: calibrated LVDT.
• Actuator: coil-magnet actuator.
• Resonant frequency: 0.38 Hz.
• Damper: For yaw and pendulum
LVDT and coilMagnet actuator
modes of the chain.
Elites Meeting 2015
6
Bottom Filter
Fine tilt
adjustment
Blade springs
• Sensor: calibrated LVDT.
• Actuator: coil-magnet actuator.
• Resonant frequency: 0.44 Hz.
• This filter can be tilted by moving
massive bodies on top: for fine
LVDT and coilmagnet actuator
alignment of the payload only.
Elites Meeting 2015
7
The payload prototype (1)
• Test mass
• Recoil mass
• Intermediate mass (marionette)
• Intermediate recoil mass
• 10 OSEMs
Test mass: 200 μm steel wire.
Recoil mass: 600 μm tungsten wire.
Elites Meeting 2015
8
The payload prototype (2)
• Test mass
• Recoil mass
• Intermediate mass (marionette)
• Intermediate recoil mass
• 10 OSEMs
Test mass: 200 μm steel wire.
Recoil mass: 600 μm tungsten wire.
Elites Meeting 2015
9
The payload prototype (3)
• Test mass
• Recoil mass
• Intermediate mass (marionette)
• Intermediate recoil mass
• 10 OSEMs
Test mass: 200 μm steel wire.
Recoil mass: 600 μm tungsten wire.
Elites Meeting 2015
10
Intermediate mass (1)
Elites Meeting 2015
It was calculated that this system
provides roll and pitch adjustment of
approximately ±2.5 degrees.
11
OSEM (1)
• OSEM: Optical sensor and
electromagnetic actuator.
• shadow sensor.
• Coil-magnet actuator.
Elites Meeting 2015
12
OSEM (2)
• OSEM: Optical sensor and
electromagnetic actuator.
• shadow sensor.
• Coil-magnet actuator.
Elites Meeting 2015
13
OSEM calibration
Measurement range (linear regime along Y): 1 mm.
Alignment tolerance (along X): ± 400 μm for a ±5% error.
Sensitivity: 𝟎. 𝟓𝟓 𝐧𝐦
𝐇𝐳 at 1 Hz.
Elites Meeting 2015
14
Alignment range
Elites Meeting 2015
15
Three stage usage of the system in KAGRA
• Calm-down phase: the resonances are damped with decay times of
60 s and shorter.
• Lock acquisition: 0.5 𝜇m s is needed.
• Observation: keeping the lock.
Elites Meeting 2015
16
Calm-down phase
Coil-magnet actuators
at IP, TF, SF and BF.
Damping control
Seismometer at IP
Control system
Displacement sensors.
(LVDT at IP, TF, SF and BF.)
Actuators (OSEM
at RM and IRM)
Displacement sensor
(OSEMs at RM and IRM)
Optical lever at the optic
Elites Meeting 2015
Damping control
17
Damping the suspension resonances (1)
We used the traditional method:
1.
The modes were excited separately using the OSEMs.
2.
An exponentially damped sinusoidal function was fitted to the ringdown.
3.
The quality factor was calculated as 𝑄 = 𝜋𝑓0 𝜏.
Elites Meeting 2015
18
Damping the suspension resonances (2)
• When the control system was on the
decay times were less than 60
seconds for all the modes considered.
• The coil driver amplification after the
DAC after the control system was too
large: we could not acquire data when
the amplitudes were small.
Elites Meeting 2015
19
Lock acquisition
Coil-magnet
Actuators at IP
Damping
control
Actuators (OSEM
at the IRM)
DC and damping control
Seismometer at IP
Control system
Displacement
sensors at IP.
Displacement sensor
(damping control except
in pitch and yaw)
Optical lever at the
optic (pitch and yaw)
Elites Meeting 2015
DC and damping control
20
Lock acquisition: on top of the IP
1 𝜇m s
• RMS values are in 𝜇m s.
• Along the optic axis the velocity
0.5 𝜇m s
was reduced:
1 𝜇m s ⇒ 0.5 𝜇m s
Elites Meeting 2015
21
𝐻𝑧)
Pitch and yaw RMS is above
the requirement:
pitch: ~2 𝜇rad
(𝜇rad
(𝜇rad 𝐻𝑧)
(um/rtHz)
Lock acquisition: control of the optic
yaw: almost 3 𝜇rad
The micro-seismic motion
may have been particularly
bad when the measurement
was taken.
Elites Meeting 2015
22
Observation mode
DC and damping control
Coil-magnet
actuators at IP
Damping
control
Actuators
at the IRM
Actuators
at the RM
Seismometer at IP
Control system
Displacement
sensors at IP.
OSEM sensors not used
Optical lever at the
optic (pitch and yaw)
Main interferometer
Along the optic axis
Elites Meeting 2015
DC and damping control
23
Observation mode
Ground vibrations: the sensor
is supported by the EQ stop
• The lock was
Damping control off
maintained for 5 days.
• After an earthquake
the lock was recovered
in 5 minutes.
Elites Meeting 2015
24
For iKAGRA
Support
• The Type Bp does not have a damper to damp the
pendulum mode at 0.45 Hz
• The translational velocity is larger than the
requirement for Type Bp: 1 𝜇m s.
• A temporary solution is to remove the SF in order
to come close to the requirement: 1.3 𝜇m s (at
the price of a smaller vertical vibration isolation
and a larger bandwidth).
Cite Shoda-san.
Elites Meeting 2015
25
Assembly test using the spare mirror
RM
RM PR3
Elites Meeting 2015
26
Assembly test using the spare mirror
Elites Meeting 2015
27
Conclusions
• The payload prototype was assembled and tested at NAOJ.
• Values of quality factors were measured: most of the decay times are
lower then 60 seconds.*
• Alignment range of the payload was given.
• Velocity at the top of the IP was reduced to 𝟎. 𝟓 𝝁𝒎 𝒔.
• RMS pitch and yaw of the optic are ~𝟐 𝐚𝐧𝐝 𝟑 𝝁𝐫𝐚𝐝 respectively.*
• The lock was achieved and maintained for 5 days. The recovery time
after an earthquake was 5 minutes.
• The procedure for hanging the optics for iKAGRA was tested in
Kamioka with the spare mirror.
Elites Meeting 2015
28