Distributed Brillouin sensing using polarization-maintaining fibers with high measurement accuracy

This thesis is a collection of studies on a distributed sensor based on Brillouin scattering using polarization maintaining (PM) fibers as well as single mode fibers. With a specially designed optical fiber circuit consisting of two polarizing beam splitters and polarization controllers, PM fibers were successfully used as sensing media for the Brillouin distributed sensor developed at the University of Ottawa. The use of a single axis of a PM fiber minimizes polarization effects and therefore improves the signal to noise ratio of the Brillouin spectrum and hence, the measurement resolution. The strain and temperature dependence of Brillouin frequency, intensity and bandwidth of three types of commercial PM fibers: PANDA, Bow-tie and Tiger were investigated. The main findings from these investigations are as follows: (1) in a PM fiber, the temperature dependence of the Brillouin frequency in the slow axis is the same as that in the fast axes, and so is the strain dependence, (2) the Brillouin spectrum bandwidth is strain and temperature dependent, and (3) PANDA PM fiber shows the greatest promise for use in sensing applications and Tiger fiber is not suitable for Brillouin sensing. The feasibility of simultaneously measuring strain and temperature using PM fibers was examined. The 2nd order partial derivative of the Stokes intensity with respect to frequency and position shows a maximum or minimum at the boundary between two different strained sections. This idea was used to locate the boundary between different strain regions. This provides a spatial location accuracy between 5 cm to 10 cm, which is better than the 20 cm pulse length used in the experiment. The influences of transient phonon (defined as when pulse width is shorter than the phonon relaxation time) relaxation on the Brillouin loss spectrum have been observed and investigated. The transient phonon field lasts longer than the nature phonon lifetime (∼10 ns). This transient phonon relaxation time is pump power and probe ER dependent, and it affects the strain, temperature, and location measurement accuracy