Strain Monitoring with Embedded Fiberoptic Sensors and Modeling with Machine Learning (ML) Tools
Conference
Laurent, M, Arias, J, Rodriguez, P et al. (2025). Strain Monitoring with Embedded Fiberoptic Sensors and Modeling with Machine Learning (ML) Tools
. 1712-1721. 10.13182/NPICHMIT25-46959
Laurent, M, Arias, J, Rodriguez, P et al. (2025). Strain Monitoring with Embedded Fiberoptic Sensors and Modeling with Machine Learning (ML) Tools
. 1712-1721. 10.13182/NPICHMIT25-46959
This paper presents a novel approach to structural health monitoring (SHM) in heat pipe micronuclear reactors (HPMRs) through the integration of embedded fiber optic cables (FOC) during metal 3D printing of reactor monoliths. The study addresses critical limitations in current monitoring systems by developing and validating a scalable method for internal stress measurement. A scaled-down monolith prototype was designed and manufactured using 17-4 stainless steel, incorporating helical channels for FOC placement; The use of additive manufacturing allowed for the implementation of these channels, and covered more area along the monolith, which gave more data to the machine learning algorithms. Results indicate that subsurface, helical FOC placement yields superior stress prediction capabilities compared to surface, rectilinear mounting, with R2 values of 0.622 and 0.337, respectively. Additionally, the relationship between scaled and full-size models was validated using Random Forest regression, achieving an R2 value of 0.94, while Gradient Boosting models evaluated sensor placement effectiveness with an R2 value of 0.67. Testing demonstrated that embedded FOCs provided more accurate internal stress measurements compared to external mounting, with deviations remaining within 8% of simulated values versus 15% for external sensors. This research advances the state-of-the-art HPMR monitoring by establishing a reliable method for real-time internal stress measurement, which is essential for ensuring operational safety and maintenance optimization in next-generation nuclear systems.
