Demonstration of Large Curvature Radius Shape Sensing Using Optical Frequency Domain Reflectometry in Multi-Core Fibers

In distributed shape sensing, the shape reconstruction error is more and more sensitive to the strain measuring error along with curvature radius of reconstructed shape increasing, which causes a notable challenge for large curvature radius shape reconstructing. In this paper, we demonstrate a large curvature radius shape sensing using optical frequency domain reflectometry (OFDR) in multi-core fibers. We construct a theoretical model of strain measuring error and curvature radius reconstructing errors under different curvature radius. In the experiment, by this reconstructing error model, we optimally select the measurable strain resolution and the sensing spatial resolution to realize the shape reconstruction with a large curvature radius and reconstruct two-dimensional (2D) circle shapes of curvature radii from 5 cm to 100 cm. To verify the accuracy of three-dimensional (3D) shape reconstruction, we present a 3D shape sensing validation method based on 3D printing technology. We fabricate a 3D phantom containing a groove with a variable curvature radius of 5 cm to 100 cm. The presented distributed shape sensing system realize to reconstruct this complicated 3D shape. The root-mean-square error of curvature radius between the reconstructed and designed 3D space curves is 7.2 mm and the mean Euclidean distance is 3.4 mm