Cargando…

Elastomer-Embedded Multiplexed Optical Fiber Sensor System for Multiplane Shape Reconstruction

This paper presents the development and application of a multiplexed intensity variation-based sensor system for multiplane shape reconstruction. The sensor is based on a polymer optical fiber (POF) with sequential lateral sections coupled with a flexible light-emitting diode (LED) belt. The optical...

Descripción completa

Detalles Bibliográficos
Autores principales: Leal-Junior, Arnaldo, Macedo, Leandro, Avellar, Leticia, Frizera, Anselmo
Formato: Online Artículo Texto
Lenguaje:English
Publicado: MDPI 2023
Materias:
Acceso en línea:https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9863505/
https://www.ncbi.nlm.nih.gov/pubmed/36679789
http://dx.doi.org/10.3390/s23020994
Descripción
Sumario:This paper presents the development and application of a multiplexed intensity variation-based sensor system for multiplane shape reconstruction. The sensor is based on a polymer optical fiber (POF) with sequential lateral sections coupled with a flexible light-emitting diode (LED) belt. The optical source modulation enables the development of 30 independent sensors using one photodetector, where the sensor system is embedded in polydimethylsiloxane (PDMS) resin in two configurations. Configuration 1 is a continuous PDMS layer applied in the interface between the flexible LED belt and the POF, whereas Configuration 2 comprises a 20 mm length PDMS layer only on each lateral section and LED region. The finite element method (FEM) is employed for the strain distribution evaluation in different conditions, including the strain distribution on the sensor system subjected to momentums in roll, pitch and yaw conditions. The experimental results of pressure application at 30 regions for each configuration indicated a higher sensitivity of Configuration 1 (83.58 a.u./kPa) when compared with Configuration 2 (40.06 a.u./kPa). However, Configuration 2 presented the smallest cross-sensitivity between sequential sensors (0.94 a.u./kPa against 45.5 a.u./kPa of Configuration 1). Then, the possibility of real-time loading condition monitoring and shape reconstruction is evaluated using Configuration 1 subjected to momentums in roll, pitch and yaw, as well as mechanical waves applied on the sensor structure. The strain distribution on the sensor presented the same pattern as the one obtained in the simulations, and the real-time response of each sensor was obtained for each case. In addition, the possibility of real-time loading condition estimation is analyzed using the k-means algorithm (an unsupervised machine learning approach) for the clusterization of data regarding the loading condition. The comparison between the predicted results and the real ones shows a 90.55% success rate. Thus, the proposed sensor device is a feasible alternative for integrated sensing in movement analysis, structural health monitoring submitted to dynamic loading and robotics for the assessment of the robot structure.