• equivalence principle
• experimental tests of gravitation
• universality of free fall

The Weak Equivalence Principle is at the basis of General Relativity and for this reason it is very important that its validity be experimentally verified as accurately as possible. The best experimental results have been obtained with test masses of different composition suspended on slowly rotating torsion balances. The rotation of the balance is crucial to up-convert the signal (which is DC in the field of the Earth and at the diurnal frequency in the field of the Sun) to higher frequency where both electronics and thermal noise are lower.

Inside a spacecraft orbiting the Earth at low altitude test masses suspended similarly to a torsion balance are subject to a driving acceleration signal from the Earth about 3 orders of magnitude stronger and can therefore significantly improve ground based results. The proposed GG space experiment has the additional advantage of a rotation frequency much higher than that of ground balances, with consequent lower electronics and thermal noise. This is achieved with the test masses weakly coupled and in rapid rotation to form a differential accelerometer in “supercritical regime” (namely with a rotation frequency higher than their coupling frequency) sensitive in both durections of the plane perpendiclar to the rotation axis. This new type of sensor is tested in the laboratory with the rotating GGG accelerometer.

GGG has the same number of degrees of freedom and the same dynamical structure as the GG sensor, and it has large mass test bodies as in space of 10 kg each. Such large masses cannot be suspended and coupled with the same weak suspensions that can be used in absence of weight, hence the GGG accelerometer cannot be as sensitive as the GG accelerometer sensor in space. Moreover, GGG is subject to microseismic noise of the local terrain and to the noise from motor and bearings, both absent in space. These noise sources must be reduced for GGG to be sensitive to a very small differential acceleration at the low frequency at which the GG sensor in space would detect the signal of an Equivalence Principle violation in the field of the Earth, namely the orbital frequency of about 1.7x10^-4 Hz, up-converted to high frequency by rotation. In rotation is provided by the whole spacecraft without a motor (by angular momentum conservation after initial set-up) while in GGG it requires motor and bearings, whose noise must also be reduced.

During this thesis passive attenuation of tilt and horizontal acceleration noise due to local microseismicity has been implemented whose expected advantage over active control has been confirmed. At first a non rotating 2D suspension was designed, made in separate components with clamped flexures. The thesis work started by setting up an apparatus for measuring the level of tilt attenuation achievable with the flexures of the non rotating suspension, and with a procedure for measuring the level of tilt attenuation on the real system. A degradation of performance was observed due to clamping. A monolithic 2D suspension was designed, which in addition could be mounted on the rotating shaft below the ball bearings so as to attenuate also tilts of the shaft due to balls and rings irregularities. The elastic constants of the suspension in both directions have been measured on bench by setting up a specific apparatus. A 1-month experimental run with this suspension mounted on GGG has provided good results reported in a review paper on GG which has appeared in August 2012 in a CQG issue dedicated to the Weak Equivalence Principle.

Encouraged by these results and in order to further reduce low frequency tilt noise we have concentrated on improving the design of all GG monolithic CuBe suspensions. Despite the need to suspend large masses, an appropriate design allows the relevant stiffness to be reduced so as to improve the sensitivity to differential accelerations as well as the capability of attenuating tilts. All suspensions have been heat treated during manufacturing in order to improve their mechanical quality. We have measured the elastic constants of all new joints showing the improvement with respect to the old ones. Then, the GGG apparatus has been dismantled and reassembled with the new joints, and after a commissioning phase an experimental run has started and is ongoing. Finally, an additional set of capacitance plates has been designed and manufactured to read the relative displacements between the shaft and the arm which couples the test cylinders (both rotating); by an appropriate adjustment of the differential period such measurement would be (almost) unaffected by low frequency tilts of the shaft (they would be rejected) while being sensitive to differential accelerations between the test cylinders.