Developing and Characterizing Novel Atom Interferometer Gyroscopes for Applications in Inertial Navigation

The Global Positioning System (GPS) reliably determines a unique position on the Earth’s surface if the receiver can communicate with at least four satellites in orbit around the earth simultaneously. However, the entire system breaks if the weak signal from the distant satellite fails to travel in a straight line from the satellite to the receiver. Due to difficulties operating GPS in every environment and circumstance, civilian and defense applications also rely on traditional position-finding methods using Inertial Navigation Systems (INS) that are uninterruptible. An INS replaces external references with internally-determined acceleration and rotation measurements to compute a platform’s change in position. Precise location determination depends on exquisite accuracy in the inertial sensors as sensor errors integrate over time. Currently, the rotation sensing component of high-performance INS use classical, light-based gyroscopes that rely upon Sagnac interferometry. While these sensors offer practical precision, they experience drift leading to large inaccuracies in calculations if denied GPS for long periods of time.

Sagnac atom interferometers are a promising technique for high-performance rotation sensing because atoms offer intrinsic stability and precision for measurements of inertial forces. The use of trapped atoms for the interferometer avoids the need for long free-fall distances that would be incompatible with a navigation apparatus. Rotation sensing with a dual Sagnac atom interferometer gyroscope was achieved in a previous experimental apparatus. The measurement cycle begins by producing a Bose- Einstein condensate (BEC) in a cylindrically harmonic trap. The BEC is then split and recombined by standing-wave Bragg laser pulses, with the magnetic trap guiding the atoms to enclose a circle. A feature of this method is the use of two counter-propagating interferometers to cancel common-mode noise that can mask the rotation signal.

This dissertation documents the construction and characterization of a new instrument with improved performance. In the new apparatus, a Sagnac area of 8 mm^2 was achieved using multiple orbits of the BEC wave packets, giving a calculated sixteen-fold improvement in sensitivity over the previous work. The interferometer operation is sufficiently stable to operate for a day or longer continuously. The Allan deviation was measured over a 26-hour period, and exhibited favorable 1/√τ scaling over averaging times τ up to 10^4 s. At 10^4 s, the resulting rotation sensitivity is 7x10^-6 rad/s, or about 0.1 revolutions per day.

While this new instrument is an improvement on the previous iteration, it is still an impractical device for rotation sensing in INS due to its size and complexity. The design and evaluation of a compact instrument that uses a volume of only 50 liters for all optics, vacuum chamber and magnetic coils is also reported. This system features an atom chip that promises to speed up the measurement cycle by a factor of ten by decreasing the BEC preparation time. A MOT with around 10^7 atoms was produced in the compact instrument and new techniques were developed for its operation. In initial experiments, this system bottlenecked at the MOT stage. Potential solutions to this problem will be presented along with a discussion of improving the short-term stability of the laboratory-sized apparatus.