| The LIGO
Special Report on VirgoVirgo Joining the GW Interferometer's Club
To: My LIGO Friends and Colleagues
From: Riccardo DeSalvo
At: Cascina, Italy
Date: July 3, 2001
I am here in Italy, visiting Virgo for the summer, and hopefully mine is just the first of many trips that several of you will have the opportunity to follow, thus strengthening the growing links between the two observatories. I have been here only a week, but there is already much to write home about.
At right: An aerial view of Virgo, the North arm pointing to Monte Serra and West arm on the left. Note the road bridges crossing over the pipes. The high seismic isolation of the superatteuators and the in-vacuum, suspended mode cleaner/injection bench make Virgo quite insensitive to ground perturbations.
Virgo is quickly joining the club of the working GW interferometers. The vacuum pipes are rapidly growing inside the tunnels, and the central interferometer--a Michelson operating between the inner test masses--just started up with a bang. Everything seems to be solidly on its way to operation within not much more than a year.
The elegant North tunnel is already finished and equipped with most of its 3-km vacuum pipe. By mid-July the pipe section welding will be finished in this tunnel and will halt, awaiting the finishing touches of civil engineering in the West tunnel. Then the welding of pipe sections will recommence in the West tunnel, and is scheduled for the second half of September. It's a task that should take about nine months. The first 300 meters of pipe have been tested already and found leak-free while a second section of 600 meters is presently under vacuum testing.
Above: At left, a view down the Virgo tunnel. Right, vacuum pipe sections ready for welding.
The terminal buildings are also growing rapidly and their tower vacuum chambers and seismic attenuation mechanics are waiting in the sun.
While civil and vacuum engineering are humming along, commissioning of the central interferometer is at the moment the most impressive part of this whole operation. The central interferometer presently includes the beam splitter, the two inner masses, the detection bench with its mode cleaner and the power recycling mirror. Originally, it was expected to include also the mode-cleaner/injection-bench unit. Owing to lingering problems with that unit's optics, the central interferometer is currently being fed by an auxiliary laser and injection telescope, which had been built to test the detection bench mode cleaner.
The entire inner interferometer is mounted on the passive superattenuator chains, the bigger brothers of the LIGO-designed TAMA-SAS chains ( recently reported on in this newsletter) which effectively isolates the mirrors from all external disturbances.
Above: At left, assemlby of the injection bench. Next, close-up view of the Virgo injection bench. The structure seen at center is a ULE reference cavity. All around the bench are the control coils of the actuators. Third, the top of a superattenuator chain, with filter and inertial damping instrumentation. Last, an eagle's eye view of a superattenuator tower.
The most satisfying thing to witness is the almost preposterous ease with which the interferometer catches lock. After the system is artificially excited, you just wait a little, watching the amplitude oscillations of the light in the dark port on a monitor. After a few minutes, only one to a few fringes per second pass by. At this point, you simply switch on the mirror and marionetta feedback, and the interferometer snaps into lock. The entire operation requires the pressure of a button. I am told that if you do not artificially excite the towers, or if you wait longer than a few minutes, the inertial damping of the superattenuator rigid body oscillations is so effective that the interferometer stays on a fringe for substantial times completely uncontrolled.
People keep milling around, doing their business, while the interferometer control people do their engineering runs. Nobody minds the eighteen-wheelers passing by the building outside--hauling pipe sections to the arms; nor the earthmovers digging the foundation of the second office building in the adjoining lot. But climbing on the tower metal structures does disturb the lock, and causes loud complaints from the operators. Using the overhead crane for more than a couple of minutes progressively feeds vibrations into the superattenuators, exciting its slow oscillations to growing amplitudes, eventually larger than the marionetta dynamic range. Once the mirror controls are overwhelmed, the lock is lost.
Above: First, at left, the control room during an inner interferometer development run. The middle TV screen shows the beam spot, while the centermost monitor shows the control actuator currents and the mirror residual motion. Pictured here are: (l-r) M.A. Bizouard, F.Bellachia, D.Garrot, M.Barsuglia, L.Holloway, and R.Passaquieti. Note the well-deserved champagne bottle on the shelf. Next, in the center photo, we see Matteo Barsuglia tuning the mode cleaner and injection bench. Then, at right, a view of half of the Virgo control electronics.
Below: Left, The 6-meter tall vacuum chambers housing the beam splitter and west inner mirror superattenuator towers. Next, pictured in the control room are: (from foreground to background) G.Losurdo, L.Holloway, M.A.Bizouard, F.Bellachia, M.Barsuglia, and F.Frasconi. Last, at right, the view screen images show residual mirror movements and mirror longitudinal actuation currents.
What is scientifically most impressive is that the little over 10-10 meter residual seismic motion measured at 1 Hz, even without optimization, already virtually ensures that Virgo will be able to achieve the control's "Holy Grail." I should be able to run its interferometer with uncontrolled mirrors while restricting all controls on the marionettas. Such performance is all the more impressive if one considers that the superattenuators are designed to reach the thermal noise floor starting at 5 Hz (in comparison with the future Advanced LIGO 10 Hz wall and the 40 Hz of the initial LIGO), and that the five seismic attenuation towers were fabricated with less than a million dollars in parts and components.
The output mode cleaner already works quite well on the Michelson's bright fringe, but presently there is insufficient power (microwatts) in the dark fringe for its operation. The Virgo team expects to add power recycling to the interferometer shortly, thus allowing the output mode cleaner to start its regular service.
The in-vacuum suspended input bench and 300 meter mode cleaner are by themselves a complex piece of optics. Being at the beginning of the chain, they were expected to have been ready by now and are somewhat behind schedule. Because of technical problems, they have been bypassed by the auxiliary input bench and have been developed in parallel so far. But they are starting to catch up. The input beam works reasonably well with in-lock times of tens of minutes, although with some excess noise. It still needs some fixing before it can be connected to feed power to the inner interferometer and thus complete the first stage of the Virgo commissioning. The complete inner interferometer is expected to be completely in operation by the end of the year.
Finally, it is worth mentioning that the substrates of the first large mirrors of the main interferometer optics have been polished and will be coated by the fall.
All the pieces are falling together and everybody is eagerly looking forward to getting the Virgo interferometer in working condition so as to complement the LIGO interferometers in a coordinated Gravitational Wave search. Given what I've seen so far, it's safe to say that I will have a very exciting summer here in Cascina, Italy.