Status of Virgo experiment Matteo Barsuglia, LAL/CNRS Orsay

1. Status of Virgo experiment Matteo Barsuglia, LAL/CNRS Orsay On behalf of the Virgo Collaboration TAMA symposium, Feb 18th 2005

2. The VIRGO Collaboration VIRGO is an Italian-French collaboration • ITALY - INFN • Firenze-Urbino • Frascati • Napoli • Perugia • Pisa • Roma • FRANCE - CNRS • ESPCI – Paris • IPN – Lyon • LAL – Orsay • LAPP – Annecy • OCA - Nice

3. Virgo site (20km from Pisa, Italy)at European Gravitational Obseervatory (EGO)

4. Expected Sensitivity Seismic wall at ~ 4 Hz

5. The VIRGO Interferometer Input Mode Cleaner 144 m long Michelson Interferometer with3 km long Fabry-Perot cavitiesin the arms and Power Recycling Output Mode Cleaner 4 cm long • High quality optics are: • located in vacuum • suspended from multi-stage pendulums Laser 20 W

6. Vacuum System • Two tubes: 3 km long, 1.2 m in diameter,in vacuum since June 2003, 400 modules • 6 long and 3 short superattenuators towers Central area Towers Tubes

7. The Suspension System The Superattenuator (SA) is designed toisolate the optical components from seismic activities (local disturbances). Working principle  multistage pendulum Expected attenuation @10 Hz:1014 Residual mirror motion (rms) rotation <1 mrad longitudinal <1 mm

8. The Top Stage Top of an inverted pendulum: - Inertial damping (70 mHz to 5 Hz) - Possibility to move the suspension point with small forces

9. Passive Filters • Five seismic filters: • Suspended by steel wires • Vertical isolation by a combination of cantilever springs and magnetic anti-springs

10. The Local Controls Marionette control:CCD camera, optical levers and fourcoil-magnet actuators: <2 Hz

11. Fast actuators Reference mass and mirror,four coil-magnet actuators

12. Injection System IMC RFC

13. Laser Master Slave • Located on an optical table outside the vacuum • Nd:YAG master commercial CW single mode (700 mW) @1064 nm • Phase locked to a Nd:YVO4 slave (monolithic ring cavity) • Pumped by two laser diodes at 806 nm (40 W power) • Output power: 20 W

14. Input Mode Cleaner • Triangular cavity, 144 m long, Finesse=1000 • Input optics and two flat mirrors are located on a suspended optical bench • End mirror suspended with a reference massfor actuation • Transmission 50% Injection Bench Mode Cleaner Mirror

15. Detection System Output Mode-Cleaner • Suspended bench in vacuum with optics for beam adjustments and the output mode cleaner (OMC) • Detection, amplification and demodulation on external bench Suspended bench External bench

16. Output Mode Cleaner Output Mode-Cleaner • 4 cm long ring cavity • Contrast improvement ~ 10 • Length control via temperature (Peltier element) • Lock acquisition takes 10 min Detection Bench

17. Photodiodes Output Mode-Cleaner • 16 InGaAs diodes for the main beam (dark port), in air and not suspended • 6 additional photo diodes for control purposes Main beam External bench

18. Digital Controls • Fully digital control, local and global • Feedback is send with 20-bit DACs @ 10kHz to thesuspensions • The suspension control is performed by dedicated DSPs (one per suspension) • Interferometer signals are acquired with 16-bit ADCs @ 20 kHz. The data is transferred via optical linksto Global Control (dedicated hardware and software that computes correction signals and sends them to the mirror DSPs)

19. Sensing and Control • Modulation-demodulation • scheme with only one modulation frequency • (6 MHz) to control: • 4 lengths • 10 angles

20. Data Acquisition and Storage • 16-bit ADCs, up to 20 kHz sampling frequency • Data in Frame format:- full signal (20 kHz)- down-sampled to 50 Hz- down-sampled to 1 Hz (trend data) • Frames available for data monitoringwith ~few sec delay • Current rate: 7 Mbytes/s (compressed) • Raw data buffer ~ 2.5 s

21. Control room

22. Current Status Central Interferometer (CITF) • 2001-2002 Commissioning of the central interferometer and the injection system • 2002-2003 Tubes commissioning and final mirror installation • Since September 2003 commissioning of the full interferometer

23. Commissioning of VIRGO The commissioning of the full detector has been divided into three phases: Phase A (sept 2003 - feb2004): the 3 km long arm cavities separatly Phase B (feb 2004 – summer 2004): recombined Michelson interferometer Phase C (from summer 2004) : Michelson interferometer with Power Recycling (full detector)

24. Commissioning of the arms • Test the cavity locking, and digital control chain • Test the automatic alignment (with the Anderson technique) • Test the frequency stabilization • Test the locking of the output mode-cleaner

25. Phase A: the two arm cavities are used separatly, starting with the north arm; North Arm Cavity commissioning West arm misaligned PR misaligned

26. In October 2003 the North arm cavity was locked on first trial using a control algorithm that was tested before with SIESTA, a time domain interferometer simulation The West arm cavity was locked in December 2003 North Arm Cavity

27. Recombined Interferometer • Phase B: • Recombined Interferometer • B2 (P) used to control common mode (L1+L2) • B2 (Q) used to control beam splitter • - B1/B1’ used to control differential mode (L1-L2) Recombined locked in February 2004 PR misaligned

28. Automatic Alignment • Anderson technique: • - Modulation frequency coincident with cavity TEM01 mode • - Two split photo diodes in transmission of the cavity (at two different Guoy phases) • - Four signals to control the 2x2 mirror angular positions (NI, NE)

29. Automatic Alignment • Alignment control allows to switch off local controls • Power inside the cavities becomes more stable • Installed and tested for the recombined interferometer • Bandwidth ~3 Hz • Residual fluctuations ~0.5 urad rms (1 nrad @ 10 Hz)

30. 1. The laser frequency is stabilised to the common lengths of thearm cavities (bandwidth ~17 kHz) 2. The arm cavities are stabilised to the reference cavity (bandwidth ~2 Hz) Frequency stabilization The `second stage´ of frequency stabilization

31. C4 run (june 2004) • Both cavities automatically aligned • BS alignment with local control • Michelson (l1-l2) controlled with ref_quad • Laser frequency stabilized to cavities common mode • Cavities common mode locked to reference cavity • Output mode-cleaner locked to dark fringe • Arms differential mode controlled with OMC transmission • Tide control on both arms

32. C4-june 2004 (recombined) • Configuration: recombined ITF with 90% complete control system: • - automatic alignment of input beam and beam splitter missing • Duration: 5 days • Test periods at the beginning and at the end of the run (~ 0.5 day) • 9 losses of lock during quiet periods (all understood, one due to an earthquake • in Alaska !) • Longest locked period: ~ 28 h, relatively stable noise level

33. C4 (recombined) sensitivity 2.5 · 10-20

34. DAC noise 103

35. Suspension hierarchical control - I Transfer of the low frequency component of the locking feedback force to the marionette DC-0.01 Hz 0.01-1.5 Hz 1.5-50 Hz • Main difficulty: • Driving of tilt modes when pushing on the marionette •  need for a good diagonalization of the driving

36. Suspension hierarchical control - II • Transfer of locking force to marionette tested • Crossing frequency ~ 0.5 Hz  1.5 Hz  8 Hz ! • Force applied to the mirror (via reference mass coils) decreased by ~10 Force applied to marionette (a.u.) Force applied to the mirror (a.u.)

37. After C4, recombined ITF • Almost all the controls running • Noise quite understood • C4/C5 data used for data analysis purposes  in july 2004 lock acquisition trial for the recycled ITF started

38. Light backscattered by the mode-cleaner Laser Frequency (Hz) Input laser beam ITF reflected beam Power divided by 10 • Solutions: • - Short term: insert attenuator between the IMC and the ITF • - Mean term: insert Faraday isolator (input bench upgrade)

39. Lock acquisition of the recycled ITF The variable finesse lock acquisition “A recycled ITF with a low recycling factor is similar to recombined interferometer “ • Lock the 4 degrees of freedom of the ITF on the half or white fringe • Bring the interferometer slowly on the dark fringe

40. Step 1: lock on the half fringe ASY DC POWER MICH ERROR SIGNAL With PR misaligned of 10 urad some light goes in the reflected beam -> Used to lock PR - Lock of the recombined on the HALF FRINGE - Lock of the long arms indipendently with the end photodiodes West arm North arm

41. Step 2: align the power recycling mirror West transmission_phase Coming from TAMA experience North transmission_phase Ref_3f phase Asy_ DC

42. Step 3: common mode laser frequency West transmission_phase  diff arm mode LASER Ref_3f phase Pick_off_Phase Asy_ DC

43. Step 4: reducing the offset Offset : 0.5  0.2 ASY DC POWER MICH ERROR SIGNAL

44. LASER Step 5: change the error signal for the michelson control DCAC West transmission_phase Ref_3f phase Pick_off_Phase Pick_off_Quad

45. Step 6: going to the dark fringe Recycling gain big increase: intermediate steps to arrive to the dark fringe ITF on the operating point

46. Power on beam splitter during lock acquisition Recycling gain ~ 25-30 300 times Recombined interferometer

47. LASER Switch to a “detection” mode Ref_3f phase prcl Pick_off_Quad mich Pick_off_Phase common arm mode Asy_Quad  diff arm mode

48. 2.5 hours lock Stored power

49. C5 run – december 2005 • Recycled ITF • Second stage of frequency stabilization (common mode servo) • output mode-cleaner

50. Recyled sensitivity Control noises BS and PR Oscillator phase noise (?) ?