Current status of the CLIO project
K Yamamoto, T Uchiyama, S Miyoki, M Ohashi, K Kuroda, H Ishitsuka, T Akutsu, S Telada, T Tomaru, T Suzuki, N Sato, Y Saito, Y Higashi, T Haruyama, A Yamamoto, T Shintomi, D Tatsumi, M Ando, H Tagoshi, N Kanda, N Awaya, S Yamagishi, H Takahashi, A Araya, A Takamori, S Takemoto, T Higashi, H Hayakawa, W Morii, J Akamatsu
aa r X i v : . [ g r- q c ] M a y Current status of the CLIO pro ject
K Yamamoto , T Uchiyama , S Miyoki , M Ohashi , K Kuroda ,H Ishitsuka , T Akutsu , S Telada , T Tomaru , T Suzuki , N Sato ,Y Saito , Y Higashi , T Haruyama , A Yamamoto , T Shintomi ,D Tatsumi , M Ando , H Tagoshi , N Kanda , N Awaya ,S Yamagishi , H Takahashi , A Araya , A Takamori , S Takemoto ,T Higashi , H Hayakawa , W Morii and J Akamatsu Institute for Cosmic Ray Research, The University of Tokyo, Kashiwa, Chiba 277-8582,Japan National Institute for Advanced Industrial Science and Technology, Tsukuba, Ibaraki305-8563, Japan High Energy Accelerator Research Organization, Tsukuba, Ibaraki 305-0801, Japan Advanced Research Institute for the Sciences and Humanities, Nihon University,Chiyoda-ku, Tokyo 102-0073, Japan National Astronomical Observatory of Japan, Mitaka, Tokyo 181-8588, Japan Department of Physics, The University of Tokyo, Bunkyo-ku, Tokyo 113-0033, Japan Department of Earth and Space Science, Graduate School of Science, Osaka University,Toyonaka, Osaka 560-0043, Japan Department of Physics, Graduate School of Science, Osaka City University, Sumiyoshi-ku,Osaka, Osaka 558-8585, Japan Department of Management and Information Systems Science, Nagaoka University ofTechnology, Nagaoka, Niigata 940-2188, Japan Earthquake Research Institute, The University of Tokyo, Bunkyo-ku, Tokyo 113-0032,Japan Department of Geophysics, Kyoto University, Sakyo-ku, Kyoto, Kyoto 606-8502, Japan Disaster Prevention Research Institute, Kyoto University, Uji, Kyoto 611-0011, JapanE-mail: [email protected]
Abstract.
CLIO (Cryogenic Laser Interferometer Observatory) is a Japanese gravitationalwave detector project. One of the main purposes of CLIO is to demonstrate thermal-noisesuppression by cooling mirrors for a future Japanese project, LCGT (Large-scale CryogenicGravitational Telescope). The CLIO site is in Kamioka mine, as is LCGT. The progress ofCLIO between 2005 and 2007 (room- and cryogenic-temperature experiments) is introduced inthis article. In a room-temperature experiment, we made efforts to improve the sensitivity. Thecurrent best sensitivity at 300 K is about 6 × − / √ Hz around 400 Hz. Below 20 Hz, the strain(not displacement) sensitivity is comparable to that of LIGO, although the baselines of CLIOare 40-times shorter (CLIO: 100m, LIGO: 4km). This is because seismic noise is extremelysmall in Kamioka mine. We operated the interferometer at room temperature for gravitationalwave observations. We obtained 86 hours of data. In the cryogenic experiment, it was confirmedthat the mirrors were sufficiently cooled (14 K). However, we found that the radiation shieldducts transferred 300K radiation into the cryostat more effectively than we had expected. Weobserved that noise caused by pure aluminum wires to suspend a mirror was suppressed bycooling the mirror. . Introduction
Observations using several interferometric gravitational wave detectors (LIGO [1], Virgo [2],GEO [3], TAMA [4]) on the ground are presently in progress. It is possible for the current LIGOinterferometers to detect a chirp wave of a neutron-star binary coalescence at about 15 Mpcfar from Earth [5]. Since this observational distance must be extended to 200 Mpc in order todetect a few chirp events every year, some future projects have been proposed: Advanced LIGO(U.S.A.) [6] and LCGT (Japan) [7]. The LCGT (Large-scale Cryogenic Gravitational Telescope)project has some features that the current and other future projects do not have. One of these iscryogenic mirrors (below 20 K) and suspensions in order to reduce thermal noise. Another featureis that the site is 1000 m underground (Kamioka mine) because of small seismic motion. As aprototype of LCGT, the CLIO (Cryogenic Laser Interferometer Observatory) project [8, 9, 10]is now in progress. The goal of CLIO is to construct an interferometer in Kamioka mine and todemonstrate thermal-noise suppression by cooling the mirrors. The CLIO and LCGT projectsinclude the construction and operation of apparatus for geophysics [8, 9, 10, 11, 12].Details of the CLIO interferometer are given in Refs. [8, 9, 10]. An outline is introduced here.The site, Kamioka mine (Hida city, Gifu prefecture, Japan) is 220 km west of Tokyo (TAMA site).In this mine, there is the world-famous water Cherenkov neutrino detector, Super-Kamiokande[13]. The vertical distance from the level of Super-Kamiokande and CLIO to the top of thismountain is about 1000 m. The horizontal distance from a mine entrance to the CLIO siteis about 2000 m. Seismic vibration in Kamioka mine is about 100-times less than that in thesuburbs of Tokyo [14, 15]. At this silent site, stable operation with lower seismic noise is possible[16]. The CLIO interferometer has two 100 m length Fabry-Perot cavities (LCGT: 3 km) whichconsist of four sapphire mirrors. These mirrors and their suspensions are cooled by silent pulse-tube cryocoolers developed for CLIO [15, 17, 18, 19]. The temperature of these mirrors mustbe below 20 K. Light from a laser source passes through a 10 m mode cleaner, and is dividedby a beam splitter for the two cavities. A power-recycling mirror does not exist. The power onthe beam splitter is 0.5 W. Reflected beams by the two cavities are not recombined. The inlinecavity, into which the light transmitted by the beam splitter goes, is used for laser frequencystabilization. The length of the other cavity, the perpendicular one, is controlled so as to keepthe light in this cavity resonant. Feedback of this control system includes gravitational wavesignals [20]. Figure 1 shows the limit sensitivity of the CLIO interferometer. This limit consistsof the fundamental noise of the interferometers: shot noise, seismic noise, thermal noise of thesuspensions and mirrors. The thick black dashed and the solid lines are the limit sensitivityat room and cryogenic temperatures, respectively. At room temperature, the limit sensitivitybelow 30 Hz and above 400 Hz is dominated by the seismic and shot noise, respectively. Between30 Hz and 400 Hz, the sensitivity is limited by the thermal noise. The limit sensitivity in thisfrequency region becomes about 10-times better when the mirrors and suspensions are cooled.The CLIO project began in 2002. In autumn of this year, tunnels for CLIO were completed.Just before the previous Amaldi conference (June 2005), all vacuum chambers, pipes, pumps,cryostat, and cryocoolers were installed and assembled [10, 21]. After the previous Amaldiconference, optical and suspension systems were installed. Moreover, we made efforts for room-and cryogenic-temperature experiments. In the room-temperature experiment, we improved thesensitivity and operated the interferometer for gravitational wave observation. In the cryogenicexperiment, we checked whether the cryostat systems cool the mirrors sufficiently, and foundthat cooling a mirror reduced the noise. In this article, the progress of the CLIO project afterthe previous Amaldi conference is introduced. -20 -19 -18 -17 -16 -15 -14 -13 -12 D i s p l a c e m en t [ m / H z / ] Figure 1.
Limit sensitivity of the CLIO in-terferometer. This limit consists of the fun-damental noise of the interferometers: shotnoise, seismic noise, thermal noise of the sus-pensions and mirrors. The thick black dashedand the solid lines are the limit sensitivityat room and cryogenic temperatures, respec-tively. The limit sensitivity between 30 Hzand 400 Hz becomes about 10-times betterwhen the mirrors and suspensions are cooled.
2. Room-temperature experiment
After June 2005, we installed optics and suspensions. A one-arm cavity experiment began onSeptember 2005. On 18 February 2006, the CLIO interferometer was fully locked. After July2006, we made many efforts to improve the sensitivity. On 21 and 22 November, 2006, thedata-acquisition systems were tested. On 13 December 2006, we obtained the current bestsensitivity of CLIO. On February 2007, we operated the CLIO interferometer for one week toconduct gravitational wave observations. Here, the current best sensitivity and observations arediscussed.
Figure 2 shows the current best displacement sensitivity of the CLIO interferometer. The thicksolid (red in online) and dashed (black in online) lines are the current best and limit sensitivity,respectively. The strain best sensitivity around 400 Hz is about 6 × − / √ Hz. Above 300Hz, photo-current noise limits the sensitivity. This noise is 3-times larger than the ideal shotnoise. Studies after this conference revealed that this was shot noise. The reason why the actualshot noise is different from the ideal one is now being investigated. Between 20 Hz and 300Hz, the sensitivity is inverse proportional to the square of the frequency. This noise is unknownand several-times larger than the goal sensitivity. Below 20 Hz, the sensitivity is limited by theseismic noise. There is no alignment control noise, because alignment control systems had notbeen installed at that time. Figure 3 shows the best strain sensitivity of the CLIO interferometerin a low-frequency region. Below 20 Hz, the strain (not displacement) sensitivity of CLIO iscomparable to that of LIGO (thin solid line, blue in online) [5]; nevertheless, the arm length ofCLIO is 40-times shorter (LIGO: 4 km). The Virgo strain sensitivity (thin dashed line, greenin online) [22] is better than that of CLIO. However, the displacement sensitivity of CLIO isbetter between 20 Hz and 30 Hz (Virgo arm length: 3 km), although CLIO does not adopt low-frequency vibration-isolation systems, as do Super Attenuators of Virgo. Such good sensitivityof the CLIO interferometer in the lower frequency region is because of extremely small seismicvibrations in Kamioka mine.
We operated the CLIO interferometer at room temperature for observations. The data obtainedat 300 K will be compared with those at the cryogenic temperature. The observation term wasbetween 12 and 18 of February, 2007. The sensitivity in the observation was almost the sameas the current best one. The observable inspiral range of neutron-star binaries was 49.5 kpc (in -20 -19 -18 -17 -16 -15 -14 -13 -12 D i s p l a c e m en t [ m / H z / ] Frequency [Hz]Displacement sensitivity of CLIO Current best sensitivity (13 December 2006) Limit sensitivity (300K)Photo-current noise(Ideal shot noise * 3) Unknown f -2 noise Figure 2.
Current best displacementsensitivity of the CLIO interferometer atroom temperature. The thick solid (red inonline) and dashed (black in online) linesare the current best and limit sensitivity,respectively. Above 300 Hz, photo-currentnoise limits the sensitivity. This noise is3-times larger than the ideal shot noise.Between 20 Hz and 300 Hz, unknown noiselimits the sensitivity. Below 20 Hz, thesensitivity is limited by the seismic noise. -21 -20 -19 -18 -17 -16 -15 -14 -13 -12 -11 S t r a i n [/ H z / ] Strain sensitivity CLIO(current best, 13 December 2006) CLIO(300K limit) LIGO(Hanford 4 km, 13 March 2006) Virgo(15 March 2007)
Figure 3.
Current best strain sensitivity ofthe CLIO interferometer in a lower frequencyregion at room temperature. The thick solid(red in online) and dashed (black in online)lines are the current best and limit sensitivityof CLIO, respectively. The thin solid (blue inonline) and dashed (green in online) lines arethe sensitivity of LIGO [5] and Virgo [22].
Figure 4.
CLIO interferometer status duringthe observation on February 2007. Thehorizontal axis shows time from 5 am on oneday to 6 am on the next day. The blue partsimply that the interferometer was locked. Onthe right-hand side of this table, the dutycycle on each day is shown. The total dutycycle was 51%. The longest lock was about 9hours (in the afternoon on 13 February).a case of a optimum direction. The threshold of the signal-to-noise ratio was 10). The storagedata length was 86 hours. Data analysis is now in progress. Figure 4 shows the status of theinterferometer during the observation. The horizontal axis shows time. The blue parts implythat the interferometer was locked. The longest lock was about 9 hours (in the afternoon on 13February). The duty cycle on each day is shown in fig. 4. The total duty cycle was 51%. Thisduty cycle was not a disappointing result, because auto lock and alignment control systems werenot installed.Any improvement for stable operation is necessary. For example, a lock of the interferometerwas broken before 2 am every night, and never recovered. From 10 pm to 8 am, nobody wasnear the interferometer (at least two operators were near the interferometer from 8 am to 10pm everyday). An auto lock systems must be installed for operation without operators. On5 February, it was difficult to acquire a lock because a terrible storm occurred. Strong windincreased the seismic noise between 0.1 Hz and 1 Hz. The duty cycle on this day was only20%. For lock acquisition and stable locks on such a terrible storm day, systems to tune thecoil-magnet actuator efficiency and to control the upper mass (not the mirror) are necessary.Alignment control systems for slow drift cancellation to keep the best sensitivity must also beconsidered.
3. Cryogenic-temperature experiment
In the room-temperature experiment, the mirrors were suspended by 400 mm length Bolfur[23] wires. This material has a high tensile strength. Bolfur wire is not useful in cryogenicexperiments because of low thermal conductivity. In the final step of the cryogenic experiment,sapphire fibers will be adopted as LCGT suspensions [7]. A sapphire fiber has high thermalconductivity [24] and small mechanical loss [25]. However, it is expensive and fragile. Thus, inthe first step of the cryogenic experiment, pure aluminum wires that are inexpensive and notfragile are used [21]. A pure aluminum fiber has a high thermal conductivity. A fault is largemechanical loss. We are now in the first step of the cryogenic experiment. Two interestingtopics found in this first-step experiment are introduced: a mirror cooling test and a sensitivityimprovement.
We operated the cryocoolers in order to confirm that the mirrors could be cooled sufficiently[21]. The results are summarized in Table 1 (the perpendicular front mirror was cooled afterthis Amaldi conference). All of the mirror temperatures became below 14 K within about oneweek (168 hours). Since the mirror temperature must be less than 20 K, the cryogenic systemworked well (other parts of the cryostat were also sufficiently cooled).
Table 1.
Result of the mirror cooling test.Cooling time Mirror temperature Heat into mirrorInline front mirror 174 hours 13.4 K 36 mWInline end mirror 176 hours 13.5 K 40 mWPerpendicular front mirror 193 hours 13.8 K 29 mWPerpendicular end mirror 164 hours 12.5 K 62 mWAlthough the mirrors were cooled sufficiently, we found a problem of heat into the mirrors.The heat evaluated from this experiment was about 1000-times larger than the expected value(the heat in Table 1 was obtained after extra baffles and radiation shields were installed in thecryostat). When the expected value was calculated, we considered the model shown in fig. 5.A mirror and a part of the optical path are surrounded by a radiation shield to prevent 300Kradiation from attacking the mirror. We considered only 300K radiation entering the mirrordirectly, as shown in fig. 5. However, our recent investigation [26] revealed that the shield doesnot absorb, but reflects radiation (shields are not black bodies). 300K radiation reflected by theduct goes into the mirror (fig. 6). The contribution of this reflected radiation is large, althoughwe have neglected it. We must consider this problem in the design of the cryostat for LCGT.Details of this problem and a solution for LCGT are introduced in Ref. [26]. irror14 K Outer radiation shield 40 K 300 KOptical axisof cavity
Figure 5.
Model used in a previousestimation of heat into a mirror. Themirror and a part of the optical path aresurrounded by a radiation shield to prevent300K radiation from attacking the mirror. Weconsidered only 300K radiation entering themirror directly.
Mirror14 K Outer radiation shield 40 K 300 KOptical axisof cavity
Figure 6.
Revised model for estimatingthe heat into the mirror. The shield doesnot absorb, but reflects radiation, becauseshields are not black bodies. 300K radiationreflected by the duct goes into the mirror.This radiation power is 1000-times larger thanthat of the direct radiation shown in fig. 5,although this contribution was neglected in aprevious estimation [26].
In order to investigate the sensitivity at cryogenic temperature, we replaced the Bolfur wires bypure aluminum wires (1 mm in diameter) in a suspension. Only this suspension was cooled. Wemeasured the sensitivity of the interferometer. The results are summarized in figs. 7 and 8. Thethick black line in figs. 7 and 8 is the current best sensitivity with the Bolfur wires at 300 K.The thin grey line in fig. 7 is the room-temperature sensitivity with the pure aluminum wires.We found that the noise between 50 Hz and 800 Hz increased. This noise was caused by thepure aluminum wires. The thin grey line in fig. 8 is the sensitivity when the mirror temperaturewas 14 K. The noise caused by the aluminum wires between 50 Hz and 800 Hz was suppressedby cooling this mirror. It must be noted that the floor level at 14 K was comparable to the bestsensitivity at 300 K, although the cooled mirror and suspension had heat links [21] and lines forpower supplies and signal probes of thermometers.We found that the noise produced by the pure aluminum wires decreased when the mirrorand suspension were cooled. In order to answer whether we observed the suspension thermalnoise and the reduction by cooling, an investigation is necessary. This is because we do notunderstand why many peaks appear in the spectrum density in figs. 7 and 8, i.e. what resonantmodes they represent.After this conference, all of the sapphire mirrors were suspended by 0.5 mm diameter purealuminum wires and cooled. The (preliminary) sensitivity with the four cooled mirrors is shownin fig. 9 (thin grey line). This is the first obtained sensitivity of a fully cooled interferometricgravitational wave detector. Although there were many peaks, the floor level between 40 Hzand 100 Hz was comparable to the current best sensitivity at room temperature (thick blackline). The current best sensitivity at 300 K is not limited by the thermal noise in this frequencyregion, probably.
4. Future work
There are two kinds of future work: sensitivity improvement and long-term operation atcryogenic temperature. Regarding the former, we will try to reduce the shot noise (above300Hz) and to investigate the unknown noise (between 20 Hz and 300 Hz). In the latter, wemust install some apparatus for long observations: auto lock, alignment control, drift cancel,digital control systems, etc. -19 -18 -17 -16 -15 -14 -13 -12 D i s p l a c e m en t [ m / H z / ] Frequency [Hz]CLIO interferometer Current best sensitivity (300 K mirror, Bolfur wire) One mirror with pure Al wires (300 K)
Figure 7.
Sensitivity at room temperature.The thick black and thin grey lines arethe current best sensitivity with the Bolfurwires at 300 K and the room-temperaturesensitivity when a mirror was suspended bypure aluminum wires (1 mm in diameter). -19 -18 -17 -16 -15 -14 -13 -12 D i s p l a c e m en t [ m / H z / ] Frequency [Hz]CLIO interferometer Current best sensitivity (300 K mirror, Bolfur wire) One mirror at 14 K with pure Al wires
Figure 8.
Sensitivity at room and cryogenictemperatures. The thick black and thin greylines are the current best sensitivity with theBolfur wires at 300 K and the sensitivity whena mirror was suspended by aluminum wires (1mm in diameter) and cooled (14 K). -19 -18 -17 -16 -15 -14 -13 -12 D i s p l a c e m en t [ m / H z / ] Frequency [Hz]
Preliminary
CLIO interferometer Current best sensitivity (300 K mirror, Bolfur wire) All mirrors at 14 K with pure Al wires
Figure 9. (Preliminary) sensitivity with allfour cooled mirrors. The thick black and thingrey lines are the current best sensitivity withthe Bolfur wires at 300 K and the sensitivitywhen all of the mirrors were cooled withpure aluminum wires (0.5 mm in diameter),respectively.
5. Summary
The CLIO (Cryogenic Laser Interferometer Observatory) project [8, 9, 10] is now in progress.One of main purposes of CLIO is to demonstrate thermal-noise suppression by cooling mirrorsfor the future Japanese project, LCGT [7]. The CLIO and LCGT sites are in Kamioka mine,where seismic noise is extremely small. After the previous Amaldi conference (recent two years),we made efforts in room- and cryogenic-temperature experiments.In the room-temperature experiment, we improved the sensitivity and operated theinterferometer for gravitational wave observations. At the end of 2006, we obtained the currentbest sensitivity. The strain sensitivity around 400 Hz is about 6 × − / √ Hz. Above 20Hz, the current best sensitivity is a few or several-times larger than the limit sensitivity atroom temperature. Since the seismic vibrations are small in Kamioka mine, the strain (notdisplacement) sensitivity below 20 Hz is comparable to that of LIGO, although the baselinelength of CLIO is 40-times shorter. In February 2007, we operated the CLIO interferometer atroom temperature for observations. We obtained 86 hours of data. These data are now beinganalyzed.In the cryogenic experiment, we confirmed that the mirror temperatures became about 14 K,hich was lower than the goal temperature of 20 K. However, 300K radiation invading throughthe shield ducts is larger than that in the original design. This problem and a solution for LCGTare introduced in Ref. [26]. We found that noise caused by aluminum wires used to suspenda mirror could be suppressed by cooling this mirror. An investigation is necessary to checkwhether we observed the suspension thermal noise and the suppression by cooling.The main topics of future work are sensitivity improvement and long-term observation atcryogenic temperature.
Acknowledgments
This project was supported in part by Grant-in-Aid for Scientific Research on Priority Areas ofthe Ministry of Education, Culture, Sports, Science and Technology (13048101) and is supportedin part by Grant-in-Aid for Scientific Research (A) of Japan Society for the Promotion of Science(18204021).
References [1] Abramovici A et al
Science et al
Nucl. Instrum. Methods Phys. Res.
Sect A et al
Class. Quantum Grav. et al Phys. Rev. Lett. Proc. of the 7th Edoardo Amaldi Conference On Gravitational Waves (Bristol: Institute of PhysicsPublishing)[6] Wilkinson C A 2008
Proc. the 7th Edoardo Amaldi Conference On Gravitational Waves (Bristol: Institute ofPhysics Publishing)[7] Kuroda K 2008
Proc. of the 7th Edoardo Amaldi Conference On Gravitational Waves (Bristol: Institute ofPhysics Publishing)[8] Ohashi M et al
Class. Quantum Grav . S599[9] Miyoki S et al
Class. Quantum Grav . S1173[10] Miyoki S et al
Class. Quantum Grav . S231[11] Takemoto S et al
J. Geodynamics et al J. Geodynamics et al Phys. Rev. Lett . Rev. Sci. Instrum . et al J. Phys.: Conf. Ser. Phys. Rev. D Class. Quantum Grav . S1005[18] Tomaru T et al
Cryocoolers Cryocoolers et al Phys. Lett. A et al J. Phys.: Conf. Ser. Proc. of the 7th Edoardo Amaldi Conference On Gravitational Waves