Repeatability precision error analysis of the distributed fiber optic strain monitoring

Abstract

The precision error of the Brillouin optical time domain reflectometry (BOTDR)-based distributed fiber optic strain measurement is normally evaluated based on strain change from the initial zero strain state. In many structural health monitoring applications, however, there is initial strain caused by the installation process of a fiber optic sensor cable to a structure. Engineers are interested in the incremental strain profile from the initial strain profile to assess the performance of the structure. The initial strain profile is often not constant throughout the cable length due to the manner that the fiber optic cables are installed (e.g., gluing, clamping, or embedding). This uneven strain distribution causes precision error in the strain incremental values, which in turn leads to difficulty in data interpretation. This paper discusses why large initial strain variation (or initial strain gradient) increases the precision error of the subsequent incremental strain reading and how to evaluate the magnitude of such precision error. A relationship between strain gradient and precision error is demonstrated. A sectional shift method is proposed to minimize the precision error. Results from laboratory tests and a field case study show that the method can reduce the precision error approximately 50% when the strain gradient is large.