TAMA Workshop — Opening Session

0:00 Good morning, friends and colleagues. It is my honor to welcome you all to this summer workshop by the name of the committee and a member of the summer. So, as the governor said, this is the second action by TAMA, as the funding failure to construct data into phenomena for detection of static waves will be over in March next year. But still, at the moment, we made these observations already, and you can have a chance to see our system tomorrow at Mitaka. And also, besides paper by many other projects, it is scheduled for cover 300 collaborators to present several papers here. Of course, there are still many problems to reduce noises in the telemetre, and we expect a lot of discussions about that, and also, it passed also to have been several cooperation collaboration project among us. Still, I hope that this meeting will stimulate to initiate such collaboration project in future, and I hope that this meeting will stimulate each together to have new ideas about the phenomena. Thank you very much.

2:30 Before we start the session, the program might be a little confusing, so I will clarify this. So the first talk by Albert Rattarini, according to the timetable, it's between 9.30 and 9.50. And the next talk is between 9.55 and 10.15. So there's a gap between 9.50 and 9.55. This is the discussion time. So five minutes of discussion time for each talk. And also, the chair will post five minutes, two minutes, and zero minutes at 9.45, 9.48, 9.50. So please follow the time schedule. So let's start the session. All right, it is my honor to open this time on workshop here, and... Do you think I need it? Opening this second TAMA workshop, and at the beginning, a few words on gratitude might be in place. I don't know how to focus this, but this one. And particularly in the name of us foreign visitors, I would like to thank the organization. First, of course, the host of this workshop which is becoming, let me say, a meeting place

5:00 deals with this wonderful turnout that we have here. But certainly, thanks should also begin to Seiki Kawamura who has done such a tremendous work and has taken over all the world of organizing such a teaching. But furthermore, a word of congratulation is due to the place, here you see it, that we are going to see tomorrow and we are all inspired, let me say, and definitely admire the speed with which the concept, the design, the construction and even preliminary operation of this Kama 300 detector has been done. Now, to start with, here's this morning's program, and after the talk of the design, we will start off with Alvaro Pazzarini, then is going to give an overview of the BOGO project. Benio Wilcoff on the DO600 detector, and somehow Seiji and RedWise have conspired to let me give a few minutes of talk on the plant new detector, DOGO. Perhaps the copy rate will be somewhat later. I wrote 10.49 instead of 10.45. I'll go.

7:30 Thank you. Is this any better? I'll use this one. First of all, I wanted to echo Albrecht's thanks to the town group for organizing this conference. It's a pleasure to be here to share with you the progress that we've been experiencing on the LIGO project. At the very top level, I would like to say that the construction project, the physical building of LIGO, is on the detector is complete with the exception of the completion of the P2 bakeout and that's progressing at Livingston right now. The detector installation is in progress, vigorously in progress at both sides. We're close to schedule. I would say we're looking at probably a three to four month delay from what we had projected five years ago. But we expect to be completed with the integration in 2000 and then go into the shakedown period and then still looking at a The bake-out, to remind you, uses only heating of V2 using a thin stainless steel wall as a resisted element. All four modules, eight kilometers in total, have been baked out. We need or exceed our goals designed to use advanced interferometers in the future, and we were able to learn if we went from Hampton to actually accelerate the baking process the original design, 150C to 168C. In three months, we were able to move all the equipment associated with the bakeout from Hanford to Livingston, Louisiana. The first of the four modules was completed last month, and at the longest will be another year before the remaining beam tubes in Livingston are baked. The results are as follows. The only hard measurement that we were able to make is the residual hydrogen outgassing which meets the required specifications that are listed here in blue. These are specifications such that pumping only at the 2A to the V tube, it will be possible to meet 10 or minus 9 total residual pressure for hydrogen 2, which is what we want for LIGO 2 and advanced LIGO interperometers. All the other more massive, the ASK residuals are all upper limits,

10:00 limited basically by the sensitivity of the residual gas analyzer that was being used. Oh, and they can note that the first of the modules of Hanford, the data of the recent, and they come in with no great surprise. In fact, the first plant that we were able to bake in Livingston was as good as, better than three of the four B-tubes in Hanford. So the learning curve is there. Going to the detector, seismic isolation systems, At the two-kilometer at Hanford, the two-kilometer system at Hanford, all the chambers except for the end of the two-kilometer of the X arm are complete. This includes the course actuation, which is used to move in three axes of translation and three degrees of tilt the entire stack I'll show that in the next slide. We've begun working on the four-kilometer interferometer at Yandrick, and in the process of integrating the psychic system, we did encounter a problem that took more than a month to resolve, and we're still in the process of resolving it. When the design for the psychic acts went to a quote-unquote all-mental system, it turned out that it was still necessary to have two small components of florel, that are used to interface between the metal-to-metal interface would otherwise take place between the masses and the strength. We concentrated on making sure that there was no noxious outgassing processes coming out of the viton, but it turns out that one thing that did get us is that there's a substantial amount of water that is outgassed from these fellows, and it turns out that it's too large a gas load for our L2 traps the way the current system is designed. do is risk wetting the beam tube again we have a very clean beam tube as I showed you and so we took a step we paused we took a look at what to do and we decided that we would need to pre-bake the seats for 48 hours at a low temperature 120 c to dry them out and we will have to go back retroactively and do that which involves the second step of realignment that will be required in fine optics the work is currently in progress at Hanford it's going as fast as we could have

12:30 because it will be finished before Thanksgiving, the end of November. This is a close-up, a very grainy close-up of the system that then defines in tip-tilt and translation the large seismic isolation system for LIGO. We have a scissors jack that controls the Z, which is used at each of the four points to define the plane of the seismic isolation system. There are two KTIL-driven XY translation stages that will be used actually in open loop but in real time to actually take care of tidal earth tides over a 12-hour period will be accommodated for in this serval system. And then we have a spherical knuckle which is used to provide for a torque-free ability to raise and lower the stack without introducing residual stresses in the system. This system was validated at Hanford. It works, and we're putting it in also at Livingston in the servo. They want it to be installed next month. We've also used Livingston to take advantage of expertise that's now at Louisiana State University with Joe Gianni's presence there. And he, with a student, went through and did a validation of in-vacual seismic isolation, which was something we had to get done with our stack. And I'll talk about that in a minute. And of course, we've got to go through and address quite a few of the chambers have already been integrated and those have to be modified. But one thing we've learned is that what we've learned at Hanford is being efficiently transferred to the group in Louisiana, things are generally progressing faster than we experienced the first time around, so that's reason for optimism. These are data that are more than a year old that was performed at the vendor of the company HITECH for in-air on the site of our stack and even though it's a very green picture, there's superimposed on here of a model of the stack performance superimposed upon data and because of the in-air performance, we're able to follow the transfer function down by three orders of magnitude. Since then, Joe, as I mentioned, has continued the measurements at Hanford, at Livingston, and we're able to follow the transfer function of the stack down another three orders of magnitude down by ten to minus six. The structure that one sees above about 20 hertz can be explained as a combination of the accelerometer and noise floor of the system that was being used and the shaker system had internal resonances that then manifest themselves in the transfer function because of the way that it was being normalized.

15:00 So there are no surprises in terms of the performance of our stacks down at least by six orders of magnitude. We have suspension systems that are part of the detector. At hand, for the two-kilometer parameter, all the large optics, except the end test mass on one arm, which is the XMID station for the two-kilometer, has been installed. And I say aligned once, since I mentioned we don't have to go back and do that once the stacks are replaced, are retrofitted with the dry bioton. There were several issues that were identified and resolved in the process of integrating the optics in a two-kilometer. I'll show you another picture. We have sensor actuator heads for use for both critically dampening certain degrees of freedom and also for monitoring for local alignment and control. It turned out that the windings on these coils had internal shorts that were caused by abrasion of the very fine wire on the Maycorp spool on which it's wound at the edge, so that had to be reworked. We had relatively few continued problems with regard to the magnet standoff assembly which had plagued us earlier in the integration phase. The ones that failed were reworked and we have seen no more problems since that point. And as I mentioned again, we'll have to realign the system once we've introduced baked vitarn of late 4LCs. At Livingston, the assembly has begun. We have a test rack in the electronics lab that's validating all the electronics integration that goes into controlling these large optic systems. The controllers have been installed for the input optics at Livingston. One of the large mirrors, it's the first large mirror in our optics system, which is the telescope expanding mirror that brings the beam out of the laser mode cleaner, and then conditions the beam to feed into the recycling mirror, the first power recycling cabinet, and the recycling mirror assembly has begun at least. The input optics, we have successfully installed and integrated the first of the three long mode cleaners, one per interferometer that we have. Its characterization is currently underway. It's a 15-meter mode cleaner. It routinely locks for many, many hours at a time without any real effort at the moment.

17:30 In fact, one thing we discovered was that you'd have to agitate the system to bring it into lock because the residual motion of the mirrors with our seismic oscillation is such that you don't go through a full fringe if it's pliescent. You have to actually push it to get it to a position where you can lock. The cabinet length is consistent with the specification. is 15 and a quarter meters, the line of the cabinet finesse as expected, and the scatter absorption transmission are a few parts, a few tens of parts per milli and also as expected. The frequency and length control servos are working as designed, and currently there is work being done on the suspension servos associated with angle control of the elements in the motor leader. I show here recent data that Nergis Mabalva provided me, which is a power residual angular fluctuations when the motor cleaner is in operation being fed by one watt out of our pre-stabilized laser and the end of the curve is approximately 1.4 micro ratings which is not consistent with the specification of what we need at this stage of the infant optics chain. One problem we did encounter and it was discovered in the two phenomena and we're currently thinking through how to fix it is that the optical design that we have attitude of the suspended optics as well as these little magnets that are bonded to the mirror plus voice coils that are situated arrayed around the optic and control the tip tilt and axial control. The original design, which was something that was verified and adopted from the 40 meter experience, which unfortunately had a 514 nanometer green beam, is such that these are infrared diodes and this is a DC coupled sensor system and what we find is that there's too much spectral overlap between the laser 104 micron pre-stabilized laser and these laser diodes that are locally used so that when you build up power in the cabins you get scatter from the resonant build up. Scatter is not anomalous, it's at the level that one expects from a few parts familiar but nonetheless it confuses the sensors because they're DC sensors And what we find is that you have alignment stability effects that cause it to go unstable if you build up the power too greatly. And that the local damping that's required from the side-to-side motion is also effective.

20:00 We're exploring short-term solutions which will work for us, which is basically looking at masks and strategically locating miniature baffles. But the longer-term solution is to consider going back and identify an acceptable sensor weight length filter so that we can spectrally isolate the high-power laser from the local sensing diodes. That's something that we can do retroactively once we've got it working in a short term with masks. At Livingston Observatory, installation has begun, although the mode cleaner and the input optics are not yet at all complete. Going on to the core optics, the main optics that constitute the resonant paddies for the fabric for all arms in a thermometer, all the optics have been polished and coated. The micro-optics have been speck with few 10 parts per million scattered, few 10 parts per million scattered. The radius of curvature are all within speck. Individually all the optics are within 5% of their specification. It turns out that because the beam splitter is slightly thinner than the input test masses, we've experienced that the diuretric coating that's used for the beam splitter causes a residual stress which produces a slight amount of defocus in the beam splitter, curvature in the beam splitter which went viewed at 45 degrees from after the stigmatism. And it's something that we're in the process of evaluating of the degree of degradation of the modem acting that we will get between the two arms but except for the bean splitter we're it is possible to enhance select the various core optics that are used in the 300 parameters and we get better than three percent radius of curvature matching for you to get the parameters the coatings are quite good in quality with point defects typically unless two parts of a million tested out 10 of the 40 optics which were our point defect presence and two of the ten were indeed rejected through we have to decide whether, if we have sufficient spare so that we can live, just put them aside or whether we have to actually reprocess them to reuse them in the initial installation of the interferometer. The absorption is also well within spec. All the objects with two-kilometer interferometer have been delivered and are at the site awaiting installation. The Livingston Observatory objects are currently being put through the metrology sequences at Caltech

22:30 for phase maps, and principally for phase maps to reduce the temperature validation. Although the recycling mirror has already been delivered to hand, it's been for installation. This is a rather old picture, but it showed for one of the flats that we have, measurements that were originally made in Australia at CSIRO during the polishing phase, and then we set up a one-micron interferometer. It was a modified WICOL system at Caltech, with a fraction of a nanometer RMS, the shape and the RMS flatness of the surface. of the pre-stabilized laser at this point. Excuse me. To remind you that we require at least eight watts in the fundamental mode at one micron. The pre-stabilized laser as an integrated unit needs to deliver about 10 millihertz per reverence and a part 10 to the sixth stable, and intensity stabilization out into the input optics and move into the 15-meter mode cleaner. Of the ten such laser systems that we have on Nordify have been delivered. Of those, one each has been installed at both Hanford and Livingston. At Hanford, the frequency and intensity control circles have been implemented. The system routinely blocks for days and time. I believe we'll say something about that in this talk later in the week. The integration of the 15-mode cleaner was successfully demonstrated. I heard that earlier. At Livingston, it's near completion and we've been able to incorporate important lessons at the first cycle of insulating at Hanford. We ended up modifying the intensity servo system. Originally, the plan was to use an acoustic optic cell, similar to what had been done in 40 meters. It turned out that a more elegant and simpler way, not including an extra active optic in the system, involved modulating the current power amplifier, and that seems to be working well. And certain details about the frequency characteristics of the cerebral loop and the frequency control have also been modified from experience gained at hand. I show here, these are older slides that demonstrate the intensity stabilization of the pre-stabilized laser is adequate. And this is using a one-meter reference cavity that's part of the pre-stabilized system to show the in-loop residual frequency noise

25:00 through some of the first measurements that were made at Hancred. And then once we had the 15-meter road cleaner working, we were able to use it as a much longer, more sensitive cavity. And so this is what we call the out-of-loop measurement of the same quantity. I transferred on here the same requirement curve. And the first time through, we have about a factor of 10 increased noise between the out-of-loop measurement and what the servo-loop was telling us. This is for 1W of operational power, and work is proceeding right now to understand in detail the discrepancy between these two. No surprise or expected. It was the first time it was turned on, and people were actually quite pleased. It was within a 10X of the requirement. The sensing and control system responsible for alignment, maintaining a residence of the interferometer have been essentially completed at Hanford. Electronics associated with 15-mode cleaner were obviously delivered to allow the management that have been done. There's a power transient protection on the mode cleaner photodiode involved in the electronic shutter that's been integrated. Pico motors in the vacuum that will be used for steering the beam out of the mode cleaner into the input optics, into the telescope that feeds the recycling there. Those have been set up. All the racks associated with controlling the 2-kilometer interferometer have been installed The alignment sensitive control subsystems have been tested and signals have been transferred through the ADC system to the acquisition and diagnostics for initial testing and analysis by the research scientists. We ended up picking a Pentium III Intel processor, the real-time digital control. There was a slight change in what our original VME design involved, and so we had to go through and demonstrate performance timing that we can keep up with the real-time control system using this particular processor with the signal processing that's required. The design is complete and the boards are now in the process of a two-kilometer parameter. We have a diagnostic system that's been worked on by Daniel Zieg at Hanford and John Spites in the Carl Tech, which involves data monarchy, There's a big software package that runs in real time that looks at all the critical signals and generates flags whenever things are out of bounds, for example.

27:30 It's been demonstrated to provide adequate type of bandwidth transmission across the local network. Software is being implemented in the stilter to do various types of bus signal conditioning. The frame-based data is available through the data server now to multiple users who control them in real time. All the drivers that are needed to do transfer functions have been completed and we're looking at interfacing it to a real-time data analysis package called ROOT for the online system. Let me, since my time has expired, I just want to say one last thing and I apologize for this. We recently performed successfully an incident run with the TAMA project. It's basically the swan saw to the 40 meter. It's the last run that the 40 meter will do as a green argon ion laser system. It was run over the middle weekend in September. We achieved in the end something on the order of 1 to 10 or minus 17 meters per root hertz for the 40 meter as a fully recycled . These numbers, I believe, are comparable to what Tom was reporting during the same period. I don't have with me the coincidence box sections, but from our side, we were able to maintain lock for 17 hours with a 90% duty cycle, so to speak. 15 hours had 80 to 90% duty cycle. And then about 10 hours were down around 70% lock sections. Let me finish with that. We thank Mr. Peter, Alba. And actually, I had to ask Alba to say a few words future LIGO, LIGO 2, I'll say LIGO 3, and so you'll have a few more minutes for that. Okay. The last slide I was going to show, which I ran out of time, is on the schedule. I just wanted to reiterate that this coming year is the year when we will be finishing the interferometers and then we will go into a shakedown phase, and the astrophysics run that we planned would be 2002 to 2004. So leading on to Oliver's question is what will happen or what does LIGO want to do beyond, let's say, 2004, the middle of the next decade? And what we recently produced in response to an NSF request, an Natural Science Foundation,

30:00 is to put together a conceptual design of what the LIGO 2 system might look like. And as you can see from the date, as soon as those of us who are going to the National Science Foundation next week, as soon as we leave this meeting, we go to Washington to present a conceptual review to a special emphasis panel, a blue ribbon panel in Washington that is going to consider making recommendations to the NSF as to how one might go beyond what currently is called LIGO 1 to LIGO 2. The basis of that involves a performance enhancement that takes the LIGO 1 curve and drops it more of 10. So the proposal is to go to a centimeter parameter that has sensitivity that has a sweet spot of something of 2.3 centimeters minus 4, 24 meters per root hertz. And the attendance factor of 10 increase in sensitivity translates off directly into a rate increase of 1,000 or so if you believe a uniform model of sources throughout the local universe. The model is essentially rating dominated in mid-range and involves resonance site and high-quality mirrors made of sapphire and all-glass suspension systems. It's a radical change for LIGO, and it adopts a lot of the technology that has come out of the GEO project. And in fact, the proposal to the NSF is that the GEO group in Europe would be partners in crime, so to speak, with LIGO in this upgraded LIGO 2, and there would be actual in-kind team contributions from the geo-collaboration in collaboration with LIGO and the LIGO scientific collaboration to then make LIGO 2 a reality sometime in the middle of the next decade. Yeah, thank you for these additional quotes. Any questions? It definitely was a very clear and concise talk that you gave. So it seemingly did not leave any questions of them. I thank you again, Howard. Perhaps this does have a slightly better resolution.

32:30 So while this is going on, let me introduce the next speaker. And we will give it all of you over the project. We all have five minutes late, so you don't have to try to make that answer. VIRGINIA VIRGINIA ZIPPED CONTINUES Oh, that's right, so. I think we saw the study makers likely advanced since last summer, but we still think that these types are relevant ones since all of it is my goal as to the history because it's the first in which topics as water construction and buildings for land acquisition by the government to be assigned to the experiment, but not mentioned at least explicitly these problems have unsolved there over. So let's talk a little bit about infrastructure. The infrastructure of the central button of the center are complete.

35:00 There is a system which is the central button chamber which is the system of button chambers installed in the center which we see, which is complete. The intermediate are now under test and some of them have already been tested. I will show what they are. And the tube construction on both sides, that one of the vacuum tube and the other side of the civil engineering around the arms are going on. and the detector of the center of the front, which is the first build of the van, which is actually a simple match of something. Power recycling is going on, and we are getting, in the next year, in the commissioning phase. We are presently in the assembly integration phase, which still will last until the next summer and manufacturing of large mirrors in progress. This is a view of the side, unfortunately here we don't have perspective so we don't have any computers to take care, sorry, but this is now started For instance, from this picture, the next one, almost no difference can be detected by eye, but actually the construction is going on. This is the most cleaner, this is the most cleaner, and this is the west arm, this is the north arm, and I will show you They are almost more than one kilometer, the construction, and it's going on in time. This was the bait-out, this was the first bait-out, which took place in Kashina in June, but they are all over now. We had an automatic system, we were calling the bait-out, and this was the first tower.

37:30 This is the game speaker tower and that behind it you see the system which is the detection bench tower and on the bottom there is one meter of flame to get inside the tower to a clean room which is below it. And this thing is rather the tower than the place, you see. This was when the towers, when the lowest part of the tower were put into place. And here, under this door here, there are between rooms to get inside. And after that, we built such a structure to work around. That was at the level of the intermediate vacuum chamber, which is a system that allows to keep the difference of vacuum between the region in which there is the optics of it and the region which would be mounted up from here, which is derivated in the super-automated installation. So this is the module, the transportation system which will be charged of this company, Italian company, by the two-piece construction. And the construction started already, and we last for one and a half year. And the batteries are already under construction and clean. And the light values are already to be planned, and the other is under construction. So this is the mode cleaner, which has been tested without so much surprises and successfully. And here there is the end tower of the Montlina, which has been installed.

40:00 Let me show here this place, which is the central building. The central upper chamber is the system composed by seven of such chambers It was connected by pipes, links, and what everything now has been seen. And this space is secretive because this is the detection tower chamber when that was in the space. And here, now we have a bench, an external bench for and this is a thin room which is below this chamber. For today, you know very well probably, everybody knows this thing, which is the system for super activator, and we have this room here, which is that one, which is kept in high-value conditions, And these are the legs. So these are the legs. And these are the points, which is suspended through this cable, to the last phase of the super-active cable. Because the masses of this multiple cable are not just masses, but they are vertical filters. And this is made of blades, anti-string systems, and dampers, tuners, and flexors of blades. So the system is very complex. And this is for the towers of the pre-belled lift table.

42:30 but we have also a shorter tower, which is for the other towers, and in this tower you see still that you have a stationary system, these are the legs, and it's here, and we We have still the same module, which is called filter zero for historical realism, and this is called filter seven, still for historical realism. We have it the same here and here. And there is a central vacuum chamber, which is a structure, a movement of the vacuum chamber, which is situated here, around here. And it allows, by means of conducting spying, to keep the difference of the state between this array and the upright array. It's very important to have the possibility to adjust the roof of separation, separating the roof between this region and this, because that means the alignment of the table, suspended to the table of the heritage. This is the prototype of the suspension in San Piero Grado, and this is here, this is These cables, they cannot be seen here, but they are positioning the center of the vertical This is a controlled experiment using this system, which is usually suspended to an inverted pendulum. The embedded engine was the engine here, and we also got a bit of time here. It's going to be right here.

45:00 Let me show here, that is a transfer function. That was measured in the prototype here, and measured stage by stage, when the transfer function was proposed to be the something which is probably the expected one that was simulated using siesta, the problem of simulation which is very hard in this case because we can compute the transfer function from any stage to any other stage and this for each degree of freedom. So, for instance, if one acts at the level of the marioneta, he will move the mirror this way, but we will see this kind of movement at the level of the theta zero, which is for the torsional motion. So that means, of course, that this kind of transfer function has to be known by the people who will construct and realize the composition of the focus of the argument. will be measured, and so I have some more slides here, this is called the last page, which could not be seen so clearly in the previous one, and this is the variable in the system with some magnets, and when I thought about the torsional and zeta-wide degree of pivot This is a mirror and we have a curious shape for the reaction mass which was intended that way in order to protect the mirror from the dust but it's not such a dark place or shouldn't be but in any way. This is the marionette in which it is also possible to store further blades. The wires go through the marionettes. And the coils are up here, here.

47:30 and I have also another auxiliary slide here this is a magnetic system and you see the center of this axis by this centering system. Okay, then we have ability to measure the position, the vertical position of cross-band in the section of the body of the filter. And that is, this is when the system, we have this, For instance, this is a tool to open this plane and to access the chamber. This is seen from the bottom of the chamber. And here is when the mechanical filter is put in the tower. So from the top side, we put inside the chamber the mechanical filter. And from the bottom, we put inside the trailer. This is the schedule for the suspension installation. And we are progressing here. We changed a little bit the order in our schedule. and the most linear suspension will be the first to be installed. And all the short suspensions will be installed by December 2000 and also the locked one. And the preparation of all the components is over. There is a lot of work to assemble such new long structures. And the suspension line definition here, which is the first stage of the alignment of the payload, of course, because the towers have been already aligned as structures, but inside probably the payload is going on.

50:00 is going on and here there is the prototype of the last type of suspension here and the benches to actuate the marionette to actuate on the system composed by reaction mass and mirror and the system of the bench are different you see here this is the last one I showed before and this is another one which is called present this is the prototype in which the integration of such prototype which was integrated in a system like this this is there are rnds going in parallel near the construction for some of these extensions in collaboration with that to the all components are very quickly have some pictures of the laser system that was in LAL it was successfully tested and we will have a lab of caching which is being installed in this space and here it is the injection bench prototype and now this the bench is already in place and caching because the suspension of injection batches in place, but it has to be, it needs some further arrangement, and this is the chapter of the center of the mirrored, which already have been prepared, and for the last For the last mirror, sorry, this mirror has been tested such as to me. And the environment, of course, are not so strict as those for the large mirrors.

52:30 You shall see these. The substrate has been already produced. One of them will be cut, it will be thinner, not thinner than the end mirrors, in order to have all the mirrors in the same shape, for the size of the system of the mirrors. So, what is going on? Because in New York there is a factory which was essentially Thank you very much. This is a man dying, I'm sure, here is a technician working at our table, and the local control of such a system was already tested. In the left, it goes back to zero. Okay, so let me skip to the sensitivity curves here, but I wanted just to mention that these sensitivity curves here are different from the center and the thermometer, which is post-commissioning will go on until the long arm will be ready. So we will use such an expression detector until the day. And you see, we have different purposes, because the mirrors will be small ones, and will be put inside the mirror holder with the shape of the real mirror here, in order not to change the suspension. This is why the mirrors are not very good.

55:00 the commission in 2001, when the arms were erased in 2001, December, and essentially the main constraints are the large mirror and the construction of the arms. And data taken, I think, here, end of 2002, it's, I have to say, 2003, but I have to say that because I'm planning that it's written. And finally, I want to finish with the gravitational wave data. This is a workshop we wrote on the end of December, from the North. That's all. Any questions? Can you tell us a little more detail about your board cleaner? And I'm interested in how you initially aligned with the board cleaner. The board cleaner, I have some, probably some more. The board cleaner was tested essentially in Orsay. And the control on the motor cleaner was operated just recently, I think in Ursa, in 1999. They have, I have seen it, and they have first position control using cameras, local controls, and this control seemed to work fine and I probably have some, I thought I had some picture of the control behavior. But essentially the system has been aligned, the suspension has been aligned. And once it is aligned, the direction of the injected beam to the interferometer can be adjusted using a telescope.

57:30 So the plane onto which the more linear triangular system is laying is defined and locally controlled through local control of the projection range. Then there is a further adjustment of the ejected beam direction, which is delved by the global control of the vehicle for the operation. Any more questions? Yes, Jeff? Can I ask a little bit about the beam splitter in the ejected? Is this the special, you know, supercell SV beam splitter? Actually, not 5 years, but it was expected to have standard facts and errors, but measurement done in the old showed high barricades, so it's very strange, it's very strange, that's why. You know, is it the one that's especially low loss, low water contact? Yes, yes, yes. Any further questions? This seems not to be the case. Thank you again very much. And next on our schedule, we have Beno Wilker, who is going to give an outline of the year 600. Oh, yes.

1:00:00 Thank you very much Albrecht. I'll give you a short introduction to the geostexembrical gravitational wave detector and as there are a couple of detailed talks which deal with special subsystems, I only flesh on those subsystems in this talk. The geostexembrical architecture is located in the center of Europe. We are in the north of Hannover in Germany. As you can see here, two 600m long arms. Here is the Lein River and north up here is Hannover. You can see the central building here. The building is one here and one over there. The geosexammering system team is based basically on five different groups. We have the group in Hannover, which is responsible for the curling, the vacuum system and the laser system. We have the group in Glasgow, and they do have a 10-meter prototype. They are responsible for the suspension, seismic acidation and ballistic suspension, in collaboration with Stanford University, she is working in Stanford now, part of her time, most of her time. They are also responsible for the computer control of the geosexample detector. In Geichen, we have a 30 meter prototype where we demonstrated the first fully suspended view recycled meter parameter last year, and Geichen is responsible for the optics. and the groups in Kodstam and Kadel are responsible for data acquisition and data analysis. Let me show you the design sensitivity curve of Geo. We are limited basically by three different noise sources, which is the seismic noise here, the thermal noise of the pendulum and the internal thermal noise here, and the shut noise here. And this is the first speciality of the geodetector that due to the georecycling we use we are able to shape the parent of the shot noise that comes from the readout noise. So we can run the detector in the broadband mode which you can see in the upper graph here.

1:02:30 and we are limited, in this mode we are, for example, limited only by thermal noise up to 300 thirds but we can also shape the shot noise by making the detector a narrow band in a way that we are already limited by shot noise in the low frequency range but are no longer limited by the shot noise here, for example at 600 thirds but are only limited by thermal noise here. We have the layout of the geosystem detector. It's also a little different from the other large scale in front of it. We start with an injection log laser system over here. Then we have two sequential mode cleaners. The light is after that injected into the power cycle in front of it. We do not have any cavities in the arms, but a two-fold delay line so that the light travels from the beam splitter to the one end mirror and back to the central rotation to the real end mirror goes back to the beam splitter and travels um 2,400 meters and it's way back to the beam splitter you can also see the single reciting mirror over here We first go back to the laser system. The laser system is developed as a laser set in Hanover, and we do have injection of the laser system. We have the laser laser here, which is a well-known monolithic We have a pump source, 8 times 10 nanometer injected in a monolithic UDMI crystal which has a size of about 2 cm. And the light travels around. We have an internal optical diode to ensure one-direction operation. And this light now is injected to the slave laser, as you can see here. Now we have two helium-yegg rods, which are anchored by fiber-coupled diode lasers. These lasers are able to deliver 30 watts, but we run them only at 17 watts to increase their lifetimes. The output of the laser is about 12 watts.

1:05:00 Here you can see a picture of the spacer, and here we have the two laser dart modules and the fibers that couples the light into the slave cavity, and we choose a quasi-monolithic slave design. So we have the slave cavity, and it's basically three mirrors that are attached to this spacer here and one adjustable mirror in this copper tube here which is sitting on a PZP, piezoelectric transducer, to change the length of the cavity and clocking master lasers. This laser now has to go into the vacuum system to avoid residual gas fluctuations in the output. And the G600 backing system is now almost finished. Over the last month we assembled the central cluster, which is here. Those tanks are 2 meter tall and 1 meter in diameter. And this also gives you an impression of the central building of GEO. It's somewhat smaller than the one of Lyward Urgo. So we have a central building which is 13 by 8 meters. and the clean room area in here. These are the gate valves and the beam tube, 60 cm diameter, 600 m long, of course, and we only use 0.8 mm wall thickness, which we can do because we have a corrugated structure of the beam tube. The beam tube is only pumped at the end, and we achieved the overall pressure in both tubes now, which is 1 times 10 minus 8 millibar, and in the central cluster we have 5 times 10 minus 8 millibar at the moment, which is due to the fact that we didn't make the central cluster yet. The optical layout that I showed you before is only a schematic layout, this is a real optical layout. It's not meant that you are able to read all the details, but this is the output of a program written by Roland Schilling, which is called OptoCAN, which is able to calculate all the Gaussian beams in the interferometer. We have the laser bench down here.

1:07:30 Then we have nine vacuum tanks in the central building here. So we inject light to the mode cleaners. Both mode cleaners have only two tanks. So for example, in this tank, we have eight suspended optics. All the optics are suspended as double pendulums. I'll show you a picture of that in a second. After that the light is injected into the power cycle cavity and here you can see a gap of 600 meter and here is the end mirrors, the light is reflected back to the inward mirror and back here is the beam to the other bench. As I mentioned, all Voxina suspensions are double pendulums. Here you can see a schematic picture of the double pendulums. we have a top plate which is sitting on the two layer passive stack. From there on we have some possibility to rotate the structure which is the suspension point of the pendulum and also the coil cordless for the local damping. We have the two wire suspension of intermediate mass here in blue and from there we have a two sling suspended mirror down here and we damp four degrees of of this whole pendulum by only feeding back with co-located feedback to the intermediate mass here. So we do not damp the roll mode and the vertical mode of this pendulum. Here you can see a picture of how it really looks like. It's hard to see, but up here we have the top plate. Here is the so-called revolving cable, which we use for rotation free alignment of the mirror. From there we have some coil holders. All the coil, the shadow sensor coil units are encapsulated because we try to keep our vacuum system completely free of hydrocarbons. Here you can see the intermediate mass. One of the intermediate masses and the other one is sitting here. And here we have some mirrors and sketches. One of the Motina mirrors has a reaction pendulum sitting actually in front of it, which you can see here. This is one mirror. Behind that mirror we have a reaction mass suspended in the same way. So we have a second double pendulum sitting behind the first one with so-called

1:10:00 reaction mass and here you can see the mirror and magnets glued to the mirror and this is the reaction mass. Here the encapsulated coils again and we use those to apply fast feedback to lock the boat cleaner. In phases here we lock the first boat cleaner in air and here Here you can see an error point factor density of the laser noise. The red one is the laser with respect to the rigid reference cavity and the blue curve is the laser with respect to the motor cleaner and you can see that the laser or the noise is dominated by acoustics in the tank because the lock is still in the air but above, around 3 kHz, noise dominated by the laser noise. Last week, oh, I forgot, after we locked the first multi-net air, we tried out the first auto-alignment system and we were able to close all the auto-alignment tubes for the first multi-net. After that we opened the tanks again and installed the second multi-net optics and last week we were able to lock the second multi-net. Let me now flash the main suspension. We'll have a talk about the main suspension system on Friday, and we also have a talk on the laser system by Sascha Brotzek on Friday too. The next time we will try to optimize the mode cleaner and at the same time install the main suspension system and on that way we made a very important step also last month which was the last main interface test of the main suspension. Here's the installation team, and this is the interface test where we first tried whether all pieces fit together when they come out of the shop. So this is the movement from the physics prototype to the real system. We have a two-layer stack system, an active-passive stack system, a triple pendulum with two vertical stages. All the pendulum modes are down at the upper mass. And here you can see a computer screen, a printout of the local damping control system after step excitation.

1:12:30 Here are the pictures of this system. You can see the stacks, which are not encapsulated here to show you all the details here, the rubber pieces for the passer player. You can see the cantilevers for the vertical stage, and this is the upper mass, the intermediate mass, and this is the lower mass. Eventually, Geo will use a silica monolithic lower mass expansion, where Sheila will get the torque on the surface. Again, one of the mirrors in each arm has to have fast feedback, and for that purpose, we have in one of the suspensions of the main mirrors, again, a double pendulum sitting with a reaction mass, so we have two triple pendulums in this tank, and we apply electrostatic feedback to the main test mass here, because we want to avoid the new magnets onto that lamp. The main optics for GEO, the main mirrors are made of Suprazil 1, 18cm diameter, 10cm thick. The recycling mirrors which have to transmit light, that's the low power level, are made from Suprazil 2, 250 mm diameter and 75 mm thick and the beam splitter is made of 311 SV with a diameter of 260 mm and 80 cm thick and lately we were quite happy because of the absorption of those beam splitters which were delivered is less than 0.5 ppm per cent. All main optics are polished or will be polished by General Optics and I show you a picture of one of the measurements done by General Optics and probably Jordan Campbell will show you more of them when he talked about the characterization of the light optics.

1:15:00 But you can see, especially on this measurement here, that the RMS of 0.4 angstroms and the peak-to-belly of 3.1 angstroms is quite impressive. All of the main optics are delivered. Eight out of 17 are polished and one is already coated. During the commissioning of the detector, GEO will use test optics, so whenever, or during the first steps of installation when we try out the lock systems, and when we are not sure how good the vacuum system is with all the suspensions, we will use only test optics, and those are also already delivered to college. The control philosophy of Geo is a little bit different than of the other long baseline detector due to the fact that we do not have cavities in the arm. We have also pre-stabilized data. We love the Mokin and the power recycling cavity with the normal RF modulation technique, but then we use the front-run schnupp modulation for the mechanism, which we send at the output out of one arm to get a narrow signal for the same recycling mirror. We will have all the lines of all the important optics of the microphone and the core cavities. All electronics, all feedback systems are basically analog. Only the active part of the cyclical oscillation system is a digital servo, but all the others are analog feedback servos which are guided by a lab view control program. alignment, gain adjustment, and offset control of these, of the analog electronics. This summer, we installed the first step of the GeoData acquisition system, which is basically the same as the LIGO data acquisition system. The only difference is basically that we do not have some reflective memory technology as people in LIGO have. So it's based on the EX12 works tornado operation system. We have some workstations, two redundant workstations

1:17:30 as the data position goes. We are able to record 64 channels at 8 kHz and also 64 channels and each station was a lower bandwidth, but we only used 25 channels that was 8 kHz to really send to Hannover and store there on tape. The data will be recorded on hard disk at the site and sent by a radio link to Hannover where we store them on tape, And from there we distribute the data by internet overnight to Potsdam and the tapes will be sent to Cardiff for the data analysis. We expect, we'll probably have a computer system of around 30 megaclops distributed between Potsdam, Cardiff and Hannibal. So we will have a small computer cluster in Hannover for the time-critical analysis and two bigger Beowulf systems in Cardiff and for the continuous wave search. But there's also a complete session on data analysis on Thursday. All right, for the next steps, we finish the first installation step, which is master laser. Oh, I forgot to mention that at the moment we only use a 1-watt master laser system at the side. So we do not have the 12-watt system at the side. The mode cleaners are also available. So we finish the first step with the master laser and the mode cleaners. over the next month we optimize the performance, vacuum then of course, and last week we started, about two weeks ago, we started with the installation of the first main suspension. The next step for GEO would be to set up the first 1200 meter cavity to learn a lot about the alignment of this cavity and of the seismic and these end buildings. We set up this 1,200 meter cavity, which will probably finish in April next year. In August, we hope to have the first hydrogen front in place.

1:20:00 And after that, in November, we hope to have the 12-watt laser at the site and the first due recycling front in there. And after that we will try to understand the system, get the noise curve as much down as we can with the test optics, then switch from test optics to main optics, have a little data position test run, and hope to be ready in the middle of the cell phone for data tape. Here's the default cell phone. Thank you. Thank you. I'm not sure about that. Are those homemade? No, those are the same as the ICS, but I think the same as you. We have 18. We have 18. If you make the same thing otherwise, you get a lot of your kids. Okay, so... Okay. We've got a question. So it seems to be so clear that our questions are you still have the microphone on here. To say a few words on the future European detector, which for some time have the name EGLO, not everybody likes that, but the next one calling it EURO is actually not really saying very much. Now, what about future large detectors?

1:22:30 We have just a few words from Albert on the LIGO II, which certainly will at some time be followed by a LIGO III. And in that case, the addition of the project is really quite far progressed. I think I'll switch back to technology and use this here. So here clearly the existing science will be used and some of it, as Albert has mentioned, will be triple suspension to achieve really high sensitivity. Massive mirrors are going to be used and certainly some advanced scheme like signal recycling of RSE. As for the plants in Japan, Professor Kuroda will tell us more about this very soon. This afternoon. what is intended is to have an installation at Kamioka, three kilometer arms, definitely it will make use of the cryogenic technology, perhaps even cryogenic suspensions, I don't know, and it will be deep underground. Now all of this is already rather well defined, but as for the duro, we are actually not as far yet. In the Spanish language, they put this inverted question mark at the point when they are just beginning to ask the question. And this is roughly the state that we are in. We are just thinking of what questions should we ask. Well, I was asked to give it one minute on. I will probably over drop by 200% or so. I will ask your forgiveness. So what is it going to have?

1:25:00 Three kilometer arms is a rather sure guess. but it's not a necessity. It will perhaps be deep underground. Certainly we will try to go for kilowatt lasers or so. Nevertheless, thinking of power recycling and certainly some advanced techniques like signal recycling or RSE. Whether cryogenics will be used in suspension as well as we're pulling the mirrors is not yet certain. The mirror materials might be silicon at low temperatures, might be sapphire, but perhaps also something like YAG where transmission of the beam is required. And as a further option, diffractive optics might perhaps be used. I think there was just one more. Well, at any rate, the concept of it is to build an parameter that is limited only by quantum noise, by photon noise and radiation pressure noise. What is the plan is supported led by the four funding authorities of Europe, that is INN, CNRS, Piedmont from Great Britain, and the Marx Marx Society from Germany. So one way, well, I'll have that.

1:27:30 We will have to go right now. The present state is that a panel has been installed to look into such questions as what material to use, how to reduce the seismic noise to the proper things. Also, what might be the most interesting sources. So to each of these, some problems, certain names have been attached who are to report on the feasibility and the main ideas by the end of this year. And this will then be followed by a two-year phase A study into various aspects of the thing and with results coming at the end of the year 2001. And the expected time for this euro might be in the year 2008. that can be safe out of this problem. Would you pull the break back up on an answer? Compared to the, what's the priority? There are some with a long ways on the difference. What are the scientific arguments that are being made? Well, the argument in one perspective is, of course, not to hold back behind what others are doing in the U.S. or perhaps in Australia, the other argument is, of course, that inside Europe, it would be a wonderful thing to have something that perhaps in its stages can be used together with the protocol to form a pair of of interferometers on the S1 sides to use

1:30:00 yes yes yes It's just the same. It's just what I was saying, it is not getting started because I want to not use the housing of the door. I think there's a real option. It's like cross-correlation. We need to detect as well. VIRGINIA Thank you. Thank you. I'd love to be standing up here and announce that the AFC, along with international funding agencies, have formed a collaboration to consider the construction of the southern people's feet affected, and what it should be, and when it should occur, at what time scale. Unfortunately, we're further behind the neuro now. Nevertheless, I'm going to summarize really the research development that's occurring in Australia at the moment.

1:32:30 of the KIGA, and KIGA being the Australian Consortium for Interfermetral Reputation of Astronomy, which consists of the Universities of Western Australia and Adelaide, along with the Australian National University. The KIGA has researched and developed under four major subsystems, laser and optics, configurations, isolation station, thermal noise, as well as data analysis. And we have constructed recently two new facilities One of them is the one shown on this slide here, which is in Jinjin, north of Perth. It's an hour by the way. And what we have shown here is the central laboratory station. We've got a 20m by 20m, so a 10m by building. And 80m down the line are two end stations which will allow us to construct some forms of interferometer in that building. The hope was that at some point is that sometime in the future, this area would become the site of the whole baseline project. It has up to four kilometres. So we're starting putting the infrastructure on site and this is where the UWA group are moving their research and development. Inside this building, to show you what's in the status of where things are going, this shows three tanks that have been constructed and put into the building, and a pipe here which we acknowledge from the Virgo group, which we're thinking about using a motor pipe. This gives you some idea of the extent of the building, and you could recognise this fellow here. and John will be giving a talk later in the week on the cultural advocacy isolation site. On the other side of the country, at the Australian National University, this one obviously is not on a university side, it's an hour away from UWA. On campus today, this new building is just being constructed. It's on the end of the physics department. This is the physics department on campus. This building here is 10 hours.

1:35:00 It's about a 10, I guess, 10 by 10, I don't know, six meters high building. This shows just playing inside. It's the IPX district down Shattuck. This facility is going to be used for R&D. So there's every retention of that attempt to turn it into an interprometal canvas. So sometime in the future we have a Saranli facility on the east coast and interprometal west coast. So what's going to happen with those facilities is a little bit up in the air, But I'll summarise with you now the type of research that's going on with the consortium under the principal subsystems. Firstly, the subsystems are interparameter configurations. We're looking at the control of resident sideline extraction device. This shows the principles of RSE. I'm sure many people know how RSE works. with the forks and control of resistance I've done extraction from the Japanese group, later in which. In principle, this mirror here forms a cavity, an input cavity for the signal into the arm cavities so you can effectively control the rate of activity or the storage time of the signals in the parameter independent of the carrier. So you can, if you wish, to have a very high power of the carrier and extract signals off of it. or extraction with resin sideband extraction or could also be referred to as signal recycling depending on the tuning of this mirror. There's a control, one of control. One has to control five degrees of freedom. The arm cavity common mode and differential modes, power cavity, power mirror, micrason, and the signal cavity. It's a complex control problem, one which is taxing the minds of at least four groups now. at the Brooklyn University of Florida, in Caltech, in ANU, and in Japan. So we've come up with one solution, what we've put out one possible solution is control issue. What the philosophy behind this is that we want to be able to independently tune the signal recycling mirror, so we don't want to fix its position and have it set there. So that means we have to look at a multiple modulation frequency scheme, and this shows that we're going to have

1:37:30 at least four modulations, as it's wrong. There's the power recycling is controlled by modulating the carrier at 75 megahertz. 75 megahertz with a one meter arm difference between the microns and here. 75 megahertz is completely reflective, so 75 megahertz modulation does not sense that mirror. You can use 75 megahertz and you can use a port here and a port there to get control over the arm cavity common bone and the microsoft common bone. We use a 15 megahertz subcarrier, sorry, we use a subcarrier which we then apply 15 megahertz sidelines to, to control the signal recycling, this coupled cavity between the power mirror and the signal mirror. So you have one, the subcarrier which resonates from this couple of the system, the modulation which reflect off, and you can use this hand and you can control it. And we have along with that 112.5 megahertz amplitude modulation sidebands because we use the things of the AM sidebands with the PM, the 75 megahertz PM to get control by . So that's the scheme that we're proposing. Where are we at with that experiment? Put this on the table, this is how you modulate laser to produce all of the side waves that you need for injection into the interbarometer. So we're currently in progress for the bench top experiment to look at the feasibility of that control scheme. and we'll get together with the rest with the other members of the LSC in March next year to come up with the control scheme to be demonstrated or tested on. Optics and lasers, well, the laser development is occurring at the University of Adelaide, the high power laser development, where they use a scheme of taking a master, one of the milliwatts, to injection of the medium-power slave, which is then used to injection of the high-power slave. The medium-power and the high-power laser are developed by the University of Adelaide. It's a multiple-stage. It uses a stable-unstable resonator design for high-power laser stage in comparison or contrast

1:40:00 to the application stage proposed by the The 5 watt laser, this is the status of it. It achieves the requirements for LIGO-1 intensity and frequency noise in the gravitational band. There's the box showing the intensity noise. So, it's measured to have the intensity noise requirements for injection locking, for injection locking. Unstable stabilizer. The status of the high power laser experiment. The Adelaide have demonstrated efficient lazing for 30 watts multi-load, 20 watts TM00, 100 watt, and come. Slope efficiency of 50%. They've been able to demonstrate control of thermal lensing, which was a real question about the design. This has been demonstrated in those who are at the Maldi meeting Peter Beecher and other slides to demonstrate that they contain control of thermal lensing, therefore put good mode quality on output. Every 100-watt laser, the diodes for the 100-watt laser in early November, and we get temperature control systems, and the diodes have been constructed, so around mid-November on to early December, we have to be putting in place the first prototype of the 100-watt hyper-laser. So we put progress there in hyper-laser development, and that is a collaboration within the with Stanford University and with Caltech through the University of Magdalene funded by the Australian researchers. In terms of lasers and optics, there was also a sapphire test mass project at the University of Western Australia. This is a collaborative project between UWA and Caltech. LIGO providing the sapphire measurements. The largest Q measured, 105 by 28, using wire suspensions. and measured Raleigh scattering of around 18 ppm per centimetre, which they compare with measurements of homogeneity birefringence measured to be less than 0.1 degrees per centimetre,

1:42:30 shown as low as three ppm per centimetre. Most samples have measured to come at 45 ppm per centimetre. They found that the absorption loss is affected by x-ray samples. So I'm a question last year about some strange measurements of absorption of the same pieces of AR groups. We have available enjoyment to play this within some slides and showing the measurements of the scattering and absorption of people who want to see them. From the Sapphire Test Mass Program, we then look at the Isolation and Suspension Program, again at the University of Western Australia, and we'll refer to John's talk later in the week. I'll give you a full description of this isolation system. Basically, the key features are the ultra-low frequency passive isolation stages, which the prediction is that with this two-stage of vertical and horizontal ultra-low frequency isolation, you'll be achieving around one nanometer RMS for test mass. This has great benefit, of course, for the control of the long-due control of the mass. It employs another innovative system called self-damping. And John will talk about how this can control, can damp out what are in an all-metal system is a high-Q type system. This could create problems by using self-damping and damp out some of these high-Q modes with an all-metal system. The predicted performance of that system is shown here, and as I said, John will talk about this over the week. The red curve here, as you know, is the... sorry, the black curve is the typical season before 10 to minus 6 over here squared. The red curve is the result of the five-stage filter, 2.5 metres high, showing the roll-off above 5 hertz. I hope you see these high stages, then successive application of stages of ultra-frequency isolation and control of tilt-the-tilt effect, if you get tilting your cross-company horizontal vertical control of the tilt,

1:45:00 predicts a final performance in the system showing in the green here, and that gives us a great hope. This gives us a, you could almost get to the stage where you may not have to apply, or the forces you apply to attest as any of the bottom of the isolation state would be so small as to not introduce any severe degradation cure. But John is going to give you a full description of that isolation state, at the ANU we're setting up a thermalized measurement system in collaboration now, in collaboration with the University of West Australia, with the Adelaide University, as well as the Virgo project. So this is an ARC funded project involving investigators from the Australian Government. Our plan, eventually, in our laboratory, looks like this one. This shows that we have a reference cavity, frequency stabilization, laser. We then have a test cavity, a very short test cavity compared to reference cavity. There are two which we then examine in the boundary before locking system, the noise readout test. There are two main features we should point out which are different from this system and other formalization systems. Firstly, the intensity noise stabilization which occurs in two stages. There's a stage where we take about 20% of the laser in the high gain servo to the first stage of intensity suppression. of noise suppression is to take all of the light which comes out of the test cavity onto a photo detector and put that into a feedback loop. And this means that we can use all of the radiation in a feedback loop and therefore squash the noise down, the intensity of noise down to shock the noise. The second unique item on this one is this split detector here, which uses a system we refer to as tilt-locking

1:47:30 of the laser on the cavity. In tilt-locking, you deliberately introduce a small tilt of the laser on the cavity, which means that when this light hits the cavity, you can decompose it using Gaussian mode into a TEM10, which is not resonant in the cavity and reflects straight off, and the TEM00, which is resonant. You treat the TEM10 as the phase reference for the TEM00. It's essentially then equivalent to what looks like a usual reflection locking that you use in spatial mode, but in those equivalents. So, the system looks like this one here with a split photo detector. You need that because when you have the, when the system is aligned, each side of the detector gives you zero. When the system is, when the 0-0 is not on resonance, it requires a slight phase tilt, phase shift, and then the two halves of the plane you take the show and the ACC. That's called tilt locking, demonstrated tilt locking, and this is the performance in comparison of the boundary pull system, and it works very well. One of the major benefits of tilt lockings, you have no phase modulators, it's all done. Okay, the thermal noise experiment then, the first stage of it, which we hope to complete by the end of next year, this is a tank with an isolation stack designed by the University of Western Australia installed at the bottom of that isolation stack, optical bench, controlling, containing the reference cavity and the test cavity. So this is the first stage, it does give us the full film on the street out. It does test out the motions at the bottom of the optical bench. The Adelaide University was responsible in this for the laser, stabilizing the laser. UWA is delivering the table and tanks on the optical bench, and it's being installed and operated for the new. The final program, I'll introduce briefly, is the data characterisation program at ANU, which is the latest program to be initiated by Akega. It's led by Dr. Susan Scott, and she's going to be giving a talk about the progress in this field later in the week, these are the main participants' preliminaries.

1:50:00 We basically are becoming familiar with the various operating systems that are being used in the rest of the world. We install programs such as GRASP. We've obtained access to our new mass data storage system. We've got access to 40 meter data and some of Glasgow data. Again, we thank those groups for allowing us access to that information. and we've obtained approval for using the Power Challenge computer system manual. And we've written a proposal for the national bandwidth for testing. So this is really two types of aspects to the data analysis program. One is noise characterization. The other is setting up, which is for the future, how data could be moved around Australia from the west to the east, stored and analyzed. So one is really looking right into the future, the other is looking right now. So what's been completed there, network connections for bandwidth testing have been completed. The tests for the AR, for the AARNet2, were completed in May. We've implemented data collection and data analysis algorithms, using analysis of the GRASP and FRAME codes for confoundability of the requirements of the group and of the systems that we use at ANU. Provided interfaces between MATLAB, FRAME and GRASP. and this is the work which is from the report on, is on statistical characterization of the 5.4 meter of the 40 meter data set. Measures in terms of the Gaussianity using likelihood ratio tests and part two statistics for that data. We've started looking at line removal algorithms, given the algorithm of development and we'll be comparing with taking other algorithms It doesn't have to be looking at the impact on underlying noise of these non-renewables. And that's going to be in progress between now and in about June next year. Before closing, I'd just refer to my talk, my own talk at the end of the week, which is a bit of an elective sensor talk on a good one use squeezing with a Fabri-Pro micrason rather than as a kind of abstraction.

1:52:30 I have two questions here related to the project. The first question is what is the limit factor of getting a higher power in TN00? Yeah. That light of diodes. The second question is this control of thermal lensing has been demonstrated. Can you elaborate that? Danny, can you repeat the question? The question was how is the thermal lensing controlled? Basically with heat elements. So you control the heat flow out of the central region. And what they have to demonstrate is that by cooling and heating, and so they actually get you control to get a reasonably flat note. The geometry of the hitting element. It's in the direct, it's an unstable, stable resonator design, so in one direction you have control via the resonator geometry, so in the other direction you have the hitting. I think it would be true to say they've been deferred to later, I'm not sure if anyone else wants to comment on that. The problem of longitudinal control is a difficult one, and once we come up, that'll be one, I think one of the questions which controls need to go for will be it's compatible with the angular. Sorry, these are fixed mass table types of tests. And which system the AMU or the Caltech or the Florida or the Japanese system has chosen will, angular control has to work with those tests. The evolution of tilt with time. This is a DC measurement, yes. How much tilt is about It's about the same amount, say 1%, which is no different to the amount of light you need to pump into the high roll mode, to the phase modulation mode, so you're not actually losing any more light from the beam, the tilts are very, very small, but in fact, if I put up, if I find this light, I'll try and tilt it now, I'll show you later, because the size of the area signal is very large.

1:55:00 And so the question of evolution of time is one that needs to be addressed. Other questions such as robustness to intensity noise is similar to . And so we're actually doing tests on those . Do you monitor the world structure? How do you know that ? If you have to not only to monitor the TEN10, but you've got to get the TN10 in the correct quadrature to correspond to angle rather than shift. So if you get rather than displacement, if you get a displacement you also get a signal which degrades the deployment. So you have to, but you can see that in the shape of the signal and you can zero that out. So tilt locking would actually work with a 2.0 mode as well. That's the major excitement with this on the 1.0. Yes, to the angle, to the direction of just... Okay, I'm happy to be here and see the progress of Kama. I want to talk briefly about bars. Let me make a very short reminder of what we are. That's the schematic. In principle, it's very simple, of course, because we have only two sources of noise, thermal noise,

1:57:30 by itself, and a final amplifier. The system is made of actually two, it's a two-mode mechanical resonator. the big mass which is catching the waves in a fundamental longitudinal mode of resonance and we have an electromechanical perspective. Oh, yeah. This one? Thank you. So, as I say, very simple system in long isolation, of course. so that basically in our system, the enforcement frequency is completed by external disadvantages of this origin. The current technology is of tons of mass, mechanical pure 10 to 7, like two meters, and temperatures for the so-called ultra-diogenic 1,6 at all .1 K. And I'd like to mention that the only factor of sensitivity is basically the same number, which is the same number of changes that the only factor of thermal noise controls because there is some contact that melts in this space. That's the bar, the second one, three dots, aluminum, in this case. The bar is the antenna pattern which is summarized with this graph. You can see that, for instance, if you want to look at the center of the galaxy, just as an example, So then, by the advantage of earth rotation, we have times during the day of very large cross-section and times of very small cross-section. So it's effective to take into consideration and actually look at some specific source. Of course, any disadvantage can be put to be advantage of time, and of course, this peculiar field may be of interest in a global network if the bars may fit before the detector's progress. And probably to solve the worst problem,

2:00:00 deconstructing an impulsive signaling, we need at least one bar in there, for the special signatures of the detector. I believe you in one graph, a little history, both in past and in future, I mean, this is just to have a glance, an incredible repeat. group is made over the years, both for one time, but from the very first So now, you can see here the spectral sensitivity expressed in noise per bed, going now to test low temperature bar comparable to a small interthrometer, and now we have ultra-ergenic and energetic bars in operation of the band, compared with the resistance of the inter-band in the sensitivity of the period of the strength of the region, we are in the region about 0.1 degrees, and the magic number is about close to 10 to minus 20, but that's expected sensitivity. Also, this tries to show very, very approximately how can be a prospect for enhancing sensitivity. A high-tech vehicle may go down like this, basically, in this vision perspective. And again, large mass parts of high-tech cinema linear, or spherical detectors, So basically, the two medals, they stay complemented in the future somehow. That's a summary of the parts of the ocean in their characteristics. As you can see, they are all similar, somehow, in performance, and some of them are some of the factor in one of all of the characteristics for instance, I believe Alejo is the best to decide with data for time, most of the time 97% more.

2:02:30 I hope this distinction becomes very high Q in the system. and the two pterogenic ones are pretty low spectra-dense. All of them are basically of the same. In fact, the sensitivity of short processes is the same. Okay, so let's just give a short outline of what is passed. Let me go to the central subject of my book. What are we doing? What are passed? Basically, what is the activity is the research for enforcing threats, of course, by filtering from here. This can probably be solved actually in two steps logically and obviously. First one, they try to clean out the single detector. By this I mean that trying to reduce the noise we observe in a quasi-stationary ocean part contribution, possibly modulate and possibly checking what we predict from the characteristics of the system, plus what is always found in unmodeled background of hopefully rare events. That's the best kind of operation we've ever had in Spain. And I'm going to tell you something about how we do this now. The second step, of course, is to keep the background with coincidences between many decades. I will talk about what we are doing with this perspective. So very shortly, I'm thinking what Arriga is doing. This is where we have the output of the system, as we see in our three-transform of the spectrum.

2:05:00 And of course, besides the two resonances And in the resonances of the two mechanical models, there are overfills, overfills, which spoil any kind of attempt to make a simple fit to a two-race-weighted model. So, my colleagues have been in that reduction of this solution to a point that by subsampling detector signal output the mirror filter after the mirror filter. So we can bypass this part, excluding all of our resonances. And that works very well, as you can see, the part is large, it's about 100 Hertz. And what happens is that we have only now only the two modes we need to examine and to use as a model for the noise analysis. And in fact, one can see that if we widen the data, we have very good widening. So basically, it says that the risk model will widen the data and reduce the simple Gaussian behavior of the system. Once you have done that, then of course the system will not be stable on the time, so no stationarity to which we try to adapt. So if you fit with the parameters of the model, the set of parameters, then you try to adaptively follow the parameters to keep the machine consistent. The crucial ones, the parameters entering the model, are of course, the resonant frequencies of the modes, refiltering bandwidth, basically the mechanical system which is very stable by resonant frequencies, but low temperatures are very stable. While the post-filtring bandwidth, which depends on the ratio between road band noise and error band noise, and that induces immediately an alteration to the wider spectrum whenever the model parameters are not like this.

2:07:30 So we actually need feedback on this to adapt. Feedback on the widely, if the widely is not enough, if we see an alteration widely, and we can adjust and adapt the modern parameters to fit the noise in the back of the water. This is done automatically, close, so effort. And now the last tool is the cushion, how much will come to be the statistics of the output. And you can see here, this is for instance, in real time is the amplitude coming out of the inner filter. You can see some of the glands, you can see a regular section, very regular salt and disturbance, lasting a few minutes, and then again the glands, to say, to kill. Of course this is political, so we try to make a quantitative examination of this, how to separate these obvious, these obvious measures, in such a way where the Jackson criteria to exclude this from data, if only goes back which language is good. So you can see that in this case, it is successful with a rather elaborate way of testing momentarily about short stretches or about buffers which contains short times like a few minutes. and then again looking at the Python filter communication check you can see that you have this is a distribution in amplitude for the middle filter against against the signal

2:10:00 to noise ratio and you can see Here a Gaussian behavior with small events outside. In fact, the white filter is very stable and bright. While in here, this section of data, you can see that the Gaussian behavior is quite and there is a lot of extra signals out of statistics. In addition, the widening is not. So basically, there is three of distribution in the two regions and choosing in the data which one is appropriate which signifies the analysis to say basically the quotient part adaptive and some events which are the background on the ceiling. So this is the first part of the rise. The last part is to try to take care of the receivable events which do not belong to the quotient statistics by examining if they come from a bar from somewhere else in the system. If an energy signal is injected in some part of this transfer chain, it will appear to be a different way by transmission and reflection from how different is started here or started somewhere else. So we apply for instance a bar signal to show like this after we have filmed. so called transducer signal, something coming in like this and something going into a screen insertion. So basically, one tries to distinguish by SQMS if the signal benefits in the positive is that the bar also has a system and then select according to the spoolie. And there's some success in that. I think that it can reduce considerably the noise so that whatever is left out of the push

2:12:30 and the country push is just a few events out of the push. As far as we expected, the number of, of the signals, I suppose, some pressure is comparable to what is expected from . So that's how we try to clean up the system for protection. And then, let me tell you briefly about the use. We have been signing an agreement for four experiments, five paths, that we're running continuously with different recycles for two years now. We have data for two years. We've been exchanging all this data. And we're starting, actually we already have a linear course in fourfold and threefold coincidences. so that by now we are about quarter days over two years in which we have been having at least three detectors on in the same time, So, it is about 25% of the site. What can be found in the case of a signal or in the case of a signal? This is a summary of the present capability of the observatory we got by Bars and understand that in the sources of stock market supernova we've done, so two sources that have never been able to be in the galaxy. So we have a coverage of the galactic use mass of about 75%. The minimum detectable signal, if we convert it in solar masses and galactic centers, now we'll find them to minus 3 solar masses and it can easily go to minus 5 with the upgrades which are in front of it for all detectables. Background is such red, we will have a probability of four-fold courses per year of 3.3-1.5, so that we know an idea about the level of confidence of the prospecting positive detection.

2:15:00 Also, the system network will solve the line of time of the course in the second of large and thus it may also put internal videos on propagation of our network of the signal. Of course, well, there is a technical range goes down because of polarization, but one nice thing that can't go so bad. And the source position in the sky is missing out of the world, we need to get some scale of each one here, each one there. There is now underground orinodectus in operation, so one can connect these two searches to see the possible consequences. The last thing is that for nearby bars of over 400 kilometers like and explore omega and others. The stochastic background limit which can be put is close to one. So if we don't have 60, I think we are now giving about 10 the problems with human. We can have a limit of omega . So that's . I think I'm in several times, so let me end up with objective updates that you mentioned, I'd say short term, 1, 2, 3 years, basically is to get fully thermal on a big area plot, the ground is actually infusing, some heating of the system above the aerodynamic temperature. then improve the amplifier, go closer to the quantum limit for the parallel amplifier, which will bring the minimum of both technical and signal to the ratio 1 around 3.0.

2:17:30 There are many options to learn, we have of course, one is to improve the speed and the other one is to try an optical defuser. And she does, it's progressing quite fast in the group that we saw in the next year. And by the same token, if we can get this, somehow go to this region, then we also get an open-locked bank. One else, which is a pleasant bank, it's a post-action bank, we may need to do about a few tens of this. That's all. Thank you very much. Are there any questions? I'd like to ask a technical question about how you identify the non-Gaussian segments. You said if you had 20-minute stretches with more than 40% bad buffers. Could you explain the algorithm by which you determine if a buffer is bad? See, this gives us steps to take what we do to separate and see. We examine the wetland buffers to estimate the spectral density noise and the variance of wetland data. And we select for the statistical properties of both. So data must be increased, correlation must be bad, and we discard by Schroeder principle, by Schroeder selection, which may last one or two steps, more than that, we select events, otherwise we have one or two or three, which may make sense. Doing that, after the selection, if the buffer we still have 9% of sample, then we accept long. Then we look at the filter buffers to estimate the variance of noise, final noise.

2:20:00 And again, we make an estimate of the statistical characteristics to Soled. After this is done, if we surpassed, so if so many as 80% of buffers to one hour will push in this sense, may we accept it and may implement the outcome. Thank you. Does anyone have any questions? Okay, thank you very much. Let's close our session.