Developments Towards Low Loss Suspensions for Laser Interferometric Gravitational Wave Detectors

Einstein, in his General Theory of Relativity, predicted that fluctuations in gravitational effects propagate as waves at the speed of light through space-time. Currently there has only been indirect evidence for the existence of these elusive gravitational waves. Scientists working on a number of large-scale research projects around the world are concentrating their efforts on detecting gravitational waves directly. Not only will their detection provide a test of some of the predictions of Einstein's theory but also much information about the astrophysical processes and sources that produce them. Gravitational waves are quadrupole in nature and produce a tidal strain in space. However their interaction with matter is very weak, making them difficult to detect. Gravitational waves emitted by violent astrophysical events are predicted to produce strains at the Earth of the order of ~10-21 to ~10-22 at frequencies accessible to ground- based detectors (i.e. a few tens of Hz to a few kHz). All detection schemes involve trying to sense these strains in space. One such detection scheme is based on the laser interferometer. Researchers at the University of Glasgow are working with collaborators from Germany on a project called GEO 600. GEO 600 is an interferometer with arms of 600 m in which light from a Nd:YAG laser is reflected between mirrors suspended as pendulums in a vacuum system. Scientists from the Max-Planck-Institut fur Quantenoptik at Hannover and Garching, the University of Cardiff and the Albert Einstein Institut at Potsdam also work on this project, together with collaborators from the Laser-Zentrum, Hannover. The design criteria of the GEO 600 detector defines that the thermal noise associated with internal modes of the fused silica test masses will limit the low frequency sensitivity of the detector. Taking a value of the intrinsic loss factor of &phis;internal (o) = 2 x 10-7 for fused silica, we can show that the thermal noise due to this loss results in a motion of one test mass of &phis;internal(o) = 7 x 10-20m/√Hz at 50 Hz. The loss factors associated with the modes of the suspension, including the presence of any additional external loss mechanisms, can increase this minimum level of thermal noise and consequently decrease the detector sensitivity. In order to prevent the achievable level of thermal noise being degraded by contributions from these other sources of thermal noise a design specification is set. Our aim is to achieve a level of thermal noise displacement of no greater than &phis;pend(o) = 2 X 10-20m/√Hz at 50 Hz at each test mass from the source of thermal noise associated with the 1 Hz pendulum mode. This corresponds to an acceptable loss factor of &phis;pendtotal(o)= 3.7 x 10-8 at 50 Hz for a GEO 600 sized test mass. The work contained in this thesis covers various methods of reducing the levels of loss in the test mass suspensions and thereby minimising the impact of thermal noise on the overall detector sensitivity. Suitable techniques have been developed towards the final design of a test mass suspension for GEO 600. To achieve a low level of loss factor associated with the pendulum mode requires a suspension wire or fibre material which is itself of low intrinsic loss. The projected material loss at 50 Hz for samples of carbon steel wire is ~40x higher than that set by the GEO 600 thermal noise specification. Fused quartz is known to be a material of low loss and so tests were performed on samples of ribbon fibres. Measurements on fused quartz ribbons, corrected for thermoelastic damping, yielded values for the intrinsic loss factor of fused quartz of &phis;matintrinsic(o) = (0.5 - 1) x 10-6 in the frequency range 6 Hz to 160 Hz. This is potentially a factor of 10 better than the maximum acceptable material loss factor for the suspension fibres as set by the GEO 600 design specification. It was found that the internal loss mechanism exhibited by 'standard grade' fused quartz appeared to be consistent with that of structural damping. These measurements were especially interesting because they were made over a frequency range that lies partially within the detection band of GEO 600. The loss factor associated with the pendulum mode of various masses suspended from two fused quartz fibres was measured. When the fibres were attached to the points of suspension by mechanical clamps, the level of loss factor observed was found to be equal to that determined by known losses of the system (i.e. the sum of the losses associated with the material of the suspension fibres and a contribution due to recoil damping - which results from the pendulum being suspended from a lossy support structure). All excess loss mechanisms were therefore eliminated. An all-welded suspension was also constructed. The loss factor was again found to be set by the known losses in the system. By subtracting the recoil loss from the measured loss, we could use the value of the remaining loss to predict the loss factor achievable by a full sized GEO 600 suspension. (Abstract shortened by ProQuest.)