Test mass suspension Q measurement and excess loss diagnosis for gravitational wave interferometers

Date of Award


Degree Type


Degree Name

Doctor of Philosophy (PhD)




Peter R. Saulson


suspension Q, thermoelastic dampening, Astronomy, Astrophysics, Mechanical engineering

Subject Categories

Astrophysics and Astronomy


For gravitational wave interferometers to work as planned, they must exhibit low levels of thermal noise. Thermal noise performance demands the test mass suspension in the detectors has unprecedentedly high quality factors Q for all of its vibration modes in the bandwidth of ${\sim} 1$ Hz to a few kHz. A key feature in the design of such suspensions is to build a pendulum with fine wire loaded at high stress to benefit from the "dissipation dilution" effect. This amplifies the suspension Q by a large factor determined by the ratio between the wire length and the small bending length under high tension.

A major concern about this feature is the possibility that large tension may degrade suspension Q by degrading wire's internal friction. These need to be tested before implementation of the pendulum in the detectors.

In addition to the internal friction of wire, suspension Q is also limited by many excess loss mechanisms resulting from cross coupling between the measured suspension mode and all other degrees of freedom in the system. They prevent the true Q value, limited only by wire's internal friction, from being revealed in measurements.

The noval feature of this work are following: we will present a precise test of the dissipation dilution effect against most of the foreseeable excess loss mechanisms under various tensions in different wire-test mass clamping schemes. We found that the dissipation dilution effect works very well at stress levels up to the breaking stress when excess loss is as low as ${\sim} 10\sp{-6}.$ Discrepancies up to a factor of 2 to 4 still exist for predicted Q's higher than $10\sp6.$

We will develop a clear diagnosis for microslip sliding friction damping at the clamping contact. We will also discover that the classical model of thermoelasticity breaks down in wire under large tension: a new theory we developed can provide a better explanation for the measured data.

Seven excess loss mechanisms will be examined in some detail. While a few of them may contribute at or close to the ${\sim} 10\sp{-6}$ level, the sliding friction effect in the linear form appears to be the dominant one and deserves more attention.


Surface provides description only. Full text is available to ProQuest subscribers. Ask your Librarian for assistance.