Learning structural damage fields from time-domain wave propagation data with Bayesian neural networks
Abstract
This study investigates the application of Bayesian Neural Networks for regression tasks aimed at recovering structural damage parameters from time-domain wave propagation data. The raw wave propagation data are simulated in a one-dimensional elastic bar using the Spectral Finite Element Method, considering data collection from a limited number of sensors. The damage fields are modeled as rectangular profiles, described by three parameters: position, size (or width), and magnitude. The inverse problem of damage recovery is formulated within a Bayesian framework, where the wavefield data collected by the sensors serve as inputs, while the outputs consist of probabilistic distributions of the damage parameters, allowing the quantification of uncertainty in the damage parameters and providing a stochastic characterization of the damage state. The training data include deterministic values for system properties, while the validation and test datasets incorporate uncertainties in the elastic bar’s properties, thereby simulating modeling errors during the inference stage of the inverse model. The main contribution of this work is to demonstrate the effectiveness of Bayesian Neural Networks as surrogate inverse models for damage characterization, addressing challenges related to model training and generalization. The results highlight the potential of this approach for reliable damage assessment in Structural Health Monitoring applications under uncertainty.
Related
- Fascia anisotropic damage propagation with microstructure- level descriptors
- Effect of damage on Rayleigh wave propagation in second gradient lattice materials
- Automatic-Differentiation-Enabled Finite Element Model Updating: Learning Constitutive Neural Networks from Structural Deformation Data
- Analysis of nonlinear wave propagation within architected materials consisting of nonlinear Timoshenko beam structural elements
- Energy Transmission and Harmonic Propagation in Unidirectionally Coupled Plasma Oscillator Networks
BibTeX
@article{junqueira2026learning,
title = {{Learning structural damage fields from time-domain wave propagation data with Bayesian neural networks}},
author = {Bernardo Feijo Junqueira and Gabriel L Silva and Rafaël Teloli and Ricardo Leiderman and Daniel Alves Castello},
journal = {Journal of the Brazilian Society of Mechanical Sciences and Engineering},
year = {2026},
month = sep,
url = {https://hal.science/hal-05805109v1},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - Learning structural damage fields from time-domain wave propagation data with Bayesian neural networks
AU - Bernardo Feijo Junqueira
AU - Gabriel L Silva
AU - Rafaël Teloli
AU - Ricardo Leiderman
AU - Daniel Alves Castello
PY - 2026
DA - 2026/09/02
JO - Journal of the Brazilian Society of Mechanical Sciences and Engineering
UR - https://hal.science/hal-05805109v1
AB - This study investigates the application of Bayesian Neural Networks for regression tasks aimed at recovering structural damage parameters from time-domain wave propagation data. The raw wave propagation data are simulated in a one-dimensional elastic bar using the Spectral Finite Element Method, considering data collection from a limited number of sensors. The damage fields are modeled as rectangular profiles, described by three parameters: position, size (or width), and magnitude. The inverse problem of damage recovery is formulated within a Bayesian framework, where the wavefield data collected by the sensors serve as inputs, while the outputs consist of probabilistic distributions of the damage parameters, allowing the quantification of uncertainty in the damage parameters and providing a stochastic characterization of the damage state. The training data include deterministic values for system properties, while the validation and test datasets incorporate uncertainties in the elastic bar’s properties, thereby simulating modeling errors during the inference stage of the inverse model. The main contribution of this work is to demonstrate the effectiveness of Bayesian Neural Networks as surrogate inverse models for damage characterization, addressing challenges related to model training and generalization. The results highlight the potential of this approach for reliable damage assessment in Structural Health Monitoring applications under uncertainty.
SN - 1678-5878
SN - 1806-3691
ER -