TY - GEN

T1 - Optimal Fiber-Optic Sensor Placement Framework for Structural Health Monitoring of an Aircraft’s Wing Spar

AU - Razzini, Adrielly Hokama

AU - Todd, Michael D.

AU - Kressel, Iddo

AU - Ofir, Yoav

AU - Tur, Moshe

N1 - Publisher Copyright:
© 2023, The Society for Experimental Mechanics, Inc.

PY - 2023

Y1 - 2023

N2 - In this work, an optimal fiber-optic sensor placement framework is proposed for structural health monitoring (SHM) of an aircraft’s wing spar. The spar is entirely made of composite materials and the damage of interest is debonding between laminates. A high-fidelity finite element (FE) model is used as a synthetic data generator for this problem. The inputs are debonding damage size and location, and the outputs are uniaxial strain measurements. To overcome the high computational costs of each FE evaluation, “run time” surrogate models are created using different machine learning methods. Then, Bayesian inference is used to estimate the damage parameters given strain measured at candidate sensor locations. These estimations are used to assess damage criticality and transformed into decisions. The Bayesian optimization uses a utility function that selects the most valuable candidates by maximizing the expected information gain between prior and posterior probability distributions of the estimated damage parameters. The total cost of the design is calculated for each candidate sensor array, consisting of the costs associated with making a certain decision plus the costs of acquiring and installing the SHM hardware (sensors, data acquisition system, etc.). The candidate with the lowest cost is selected. The resulting optimal sensor configuration is presented.

AB - In this work, an optimal fiber-optic sensor placement framework is proposed for structural health monitoring (SHM) of an aircraft’s wing spar. The spar is entirely made of composite materials and the damage of interest is debonding between laminates. A high-fidelity finite element (FE) model is used as a synthetic data generator for this problem. The inputs are debonding damage size and location, and the outputs are uniaxial strain measurements. To overcome the high computational costs of each FE evaluation, “run time” surrogate models are created using different machine learning methods. Then, Bayesian inference is used to estimate the damage parameters given strain measured at candidate sensor locations. These estimations are used to assess damage criticality and transformed into decisions. The Bayesian optimization uses a utility function that selects the most valuable candidates by maximizing the expected information gain between prior and posterior probability distributions of the estimated damage parameters. The total cost of the design is calculated for each candidate sensor array, consisting of the costs associated with making a certain decision plus the costs of acquiring and installing the SHM hardware (sensors, data acquisition system, etc.). The candidate with the lowest cost is selected. The resulting optimal sensor configuration is presented.

KW - Bayesian inference

KW - Finite element

KW - Optimal sensor placement

KW - Structural health monitoring

KW - Surrogate model

UR - http://www.scopus.com/inward/record.url?scp=85180531999&partnerID=8YFLogxK

U2 - 10.1007/978-3-031-34946-1_16

DO - 10.1007/978-3-031-34946-1_16

M3 - ???researchoutput.researchoutputtypes.contributiontobookanthology.conference???

AN - SCOPUS:85180531999

SN - 9783031349454

T3 - Conference Proceedings of the Society for Experimental Mechanics Series

SP - 129

EP - 132

BT - Data Science in Engineering, Volume 10 - Proceedings of the 41st IMAC, A Conference and Exposition on Structural Dynamics 2023

A2 - Madarshahian, Ramin

A2 - Hemez, François

PB - Springer

T2 - 41st IMAC, A Conference and Exposition on Structural Dynamics, 2023

Y2 - 13 February 2023 through 16 February 2023

ER -