The Web Newsletter Front Page Hanford Livingston MIT

LIGO Caltech News

Thermal Noise Interferometer Coming Together

Thermal Noise Interferometer Coming Together

- Contributed by Ken Libbrecht, Eric Black, Shanti Rao, and Antal Gyori

The thermal noise arising from mechanical dissipation in the test mass (a kind of internal friction), as well as from dissipation in the test mass suspension, is a fundamental noise source limiting the sensitivity of LIGO detectors. Unfortunately we have only a rather poor understanding of the fundamental physics responsible for these noise sources, for example no microscopic model realistically describes measured material loss functions. The thermal noise arising from optical coatings and non-Gaussian noise--from material creak and creep--are two significant unknowns which must be explored in order to further improve the LIGO detectors.

The goal of the Thermal Noise Interferometer (TNI) is to measure directly the displacement noise in a suspended interferometer; to reduce the noise to the lowest possible level; and to fully characterize the thermal and non-thermal noise spectrum. The TNI will be optimized for displacement noise measurements using a short test cavity length, high finesse coatings for maximum power build-up, mounting on a common seismic isolation stack, etcetera. This not only diminishes seismic and shot-noise levels, it also reduces a number of design requirements relative to a long-arm interferometer, for example the test mass pointing.

Figure 1. Frequency Noise in the TNI

Progress to date on the TNI has been excellent, due in large part to the concentrated expertise in interferometer building that exists within the LIGO group. The TNI project began in mid-1997 by assembling a brand new lab (basically an optical table with clean-room enclosure) in the high-bay Synchrotron Hall at Caltech. By the end of 1997 we had purchased an NPRO laser, built a reference cavity, and locked our Pre-Stabilized Laser (PSL) following the LIGO PSL design. The TNI PSL looks essentially like the LIGO version, and the in-loop error signal gives a frequency noise of approximately 20 mHz/sqrt-Hz from ~200 Hz to ~2 kHz. Figure 1 at right shows the measurement of the in-loop error signal. On this plot, -120 dBVrms/rt-Hz corresponds to 10 mHz/rt-Hz in frequency space, and the scale is 5 dB/div. Our frequency noise stays below 30 mHz/rt-Hz from about 200 Hz up to 5 kHz. We have not yet measured the noise of our PSL outside the loop, but plan to do so using a separate cavity this summer.

Our recent efforts have focused mainly on details of the TNI design and the purchase of optics for the test mass cavities and for a suspended mode cleaner. For the latter we are fortunate to have available many parts first used for a prototype 12-meter suspended mode cleaner which was built in the LIGO optical lab. We need to replace the optics, since the prototype mode cleaner was designed for an argon-ion laser, and we intend to reduce the cavity length a fair amount in the process. But the servo controls for the suspended ring cavity can all be taken directly from the prototype, saving a great deal of time and effort.

Our initial set of test masses will be made from 4-inch-diameter fused silica, which will each be suspended by a single loop of steel piano wire, much like the LIGO-I suspensions. We intend to upgrade the TNI to an all-fused-silica suspension in the near future but the wire suspension should allow us to build a working interferometer more quickly, while still getting down to the 10-18 m/sqrt-Hz noise level. Our particular aim with the initial TNI is to examine and characterize the non-thermal displacement noise spectrum of a LIGO-like interferometer.