Stress-field reconstruction from laser interferometric strain measurements

Data routinely collected from distributed fiber-optic cables embedded in civil structures can be repurposed as measurements for structural health monitoring (SHM). However, effectiveness depends critically on fiber-structure coupling and on the ability to infer meaningful structural parameters from the recorded signals. This proof-of-concept study proposes a framework to transform fiber-optic signals, acquired by a recently developed coherent Laser Interferometry (LI) technique, into distributed structural stress fields by combining a calibrated digital twin with surrogate modeling. A finite-element (FE) model is first calibrated using fiber-derived natural frequencies and fiber-averaged strain histories under different fiber-structure coupling conditions, providing a physics-based reference for subsequent strain-to-stress inference. Based on the calibrated FE response, surrogate models are identified via a local response function approach to map fiber-averaged strain to the stress distribution along the beam. Numerical damage scenarios show sensitivity to localized stiffness loss, with the surrogate model reconstructing damage-induced stress redistributions and phase shifts from the fiber signal alone. The proposed framework highlights the potential of repurposing existing communication fiber networks to recover distributed strain and stress fields over large-scale structures when combined with appropriate digital surrogate modeling.