Bayesian Statistic Based Multivariate Gaussian Process Approach for Offline/Online Fatigue Crack Growth Prediction

S. Mohanty, Aditi Chattopadhyay, Pedro Peralta, S. Das

Research output: Contribution to journalArticle

22 Citations (Scopus)

Abstract

Offline and online fatigue crack growth prediction of Aluminum 2024 compact-tension (CT) specimens under variable loading has been modeled, using multivariate Gaussian Process (GP) technique. The GP model is a Bayesian statistic stochastic model that projects the input space to an output space by probabilistically inferring the underlying nonlinear function. For the offline prediction, the input space of the model is trained with parameters that affect fatigue crack growth, such as the number of fatigue cycles, minimum load, maximum load, and load ratio. For the online prediction, the model input space is trained using piezoelectric sensor signal features rather than training the input space with loading parameters, which are difficult to measure in a real time scenario. Principal Component Analysis (PCA) is used to extract the principal features from sensor signals. In both the offline and online case, the output space is trained with known associated crack lengths or crack growth rates. Once the GP model is trained, a new output space for which the corresponding crack length or crack growth rate is not known, is predicted using the trained GP model. The models are validated through several numerical examples.

Original languageEnglish (US)
Pages (from-to)833-843
Number of pages11
JournalExperimental Mechanics
Volume51
Issue number6
DOIs
StatePublished - Jul 2011

Fingerprint

Fatigue crack propagation
Statistics
Crack propagation
Cracks
Sensors
Stochastic models
Principal component analysis
Fatigue of materials
Aluminum

Keywords

  • Fatigue life prediction
  • Gaussian process
  • Probabilistic approach
  • Statistical method
  • Structural health monitoring
  • Variable loading

ASJC Scopus subject areas

  • Mechanical Engineering
  • Mechanics of Materials
  • Aerospace Engineering

Cite this

Bayesian Statistic Based Multivariate Gaussian Process Approach for Offline/Online Fatigue Crack Growth Prediction. / Mohanty, S.; Chattopadhyay, Aditi; Peralta, Pedro; Das, S.

In: Experimental Mechanics, Vol. 51, No. 6, 07.2011, p. 833-843.

Research output: Contribution to journalArticle

@article{8c70bbdb245946b3b656fb9cb3cdeb5d,
title = "Bayesian Statistic Based Multivariate Gaussian Process Approach for Offline/Online Fatigue Crack Growth Prediction",
abstract = "Offline and online fatigue crack growth prediction of Aluminum 2024 compact-tension (CT) specimens under variable loading has been modeled, using multivariate Gaussian Process (GP) technique. The GP model is a Bayesian statistic stochastic model that projects the input space to an output space by probabilistically inferring the underlying nonlinear function. For the offline prediction, the input space of the model is trained with parameters that affect fatigue crack growth, such as the number of fatigue cycles, minimum load, maximum load, and load ratio. For the online prediction, the model input space is trained using piezoelectric sensor signal features rather than training the input space with loading parameters, which are difficult to measure in a real time scenario. Principal Component Analysis (PCA) is used to extract the principal features from sensor signals. In both the offline and online case, the output space is trained with known associated crack lengths or crack growth rates. Once the GP model is trained, a new output space for which the corresponding crack length or crack growth rate is not known, is predicted using the trained GP model. The models are validated through several numerical examples.",
keywords = "Fatigue life prediction, Gaussian process, Probabilistic approach, Statistical method, Structural health monitoring, Variable loading",
author = "S. Mohanty and Aditi Chattopadhyay and Pedro Peralta and S. Das",
year = "2011",
month = "7",
doi = "10.1007/s11340-010-9394-7",
language = "English (US)",
volume = "51",
pages = "833--843",
journal = "Experimental Mechanics",
issn = "0014-4851",
publisher = "Springer New York",
number = "6",

}

TY - JOUR

T1 - Bayesian Statistic Based Multivariate Gaussian Process Approach for Offline/Online Fatigue Crack Growth Prediction

AU - Mohanty, S.

AU - Chattopadhyay, Aditi

AU - Peralta, Pedro

AU - Das, S.

PY - 2011/7

Y1 - 2011/7

N2 - Offline and online fatigue crack growth prediction of Aluminum 2024 compact-tension (CT) specimens under variable loading has been modeled, using multivariate Gaussian Process (GP) technique. The GP model is a Bayesian statistic stochastic model that projects the input space to an output space by probabilistically inferring the underlying nonlinear function. For the offline prediction, the input space of the model is trained with parameters that affect fatigue crack growth, such as the number of fatigue cycles, minimum load, maximum load, and load ratio. For the online prediction, the model input space is trained using piezoelectric sensor signal features rather than training the input space with loading parameters, which are difficult to measure in a real time scenario. Principal Component Analysis (PCA) is used to extract the principal features from sensor signals. In both the offline and online case, the output space is trained with known associated crack lengths or crack growth rates. Once the GP model is trained, a new output space for which the corresponding crack length or crack growth rate is not known, is predicted using the trained GP model. The models are validated through several numerical examples.

AB - Offline and online fatigue crack growth prediction of Aluminum 2024 compact-tension (CT) specimens under variable loading has been modeled, using multivariate Gaussian Process (GP) technique. The GP model is a Bayesian statistic stochastic model that projects the input space to an output space by probabilistically inferring the underlying nonlinear function. For the offline prediction, the input space of the model is trained with parameters that affect fatigue crack growth, such as the number of fatigue cycles, minimum load, maximum load, and load ratio. For the online prediction, the model input space is trained using piezoelectric sensor signal features rather than training the input space with loading parameters, which are difficult to measure in a real time scenario. Principal Component Analysis (PCA) is used to extract the principal features from sensor signals. In both the offline and online case, the output space is trained with known associated crack lengths or crack growth rates. Once the GP model is trained, a new output space for which the corresponding crack length or crack growth rate is not known, is predicted using the trained GP model. The models are validated through several numerical examples.

KW - Fatigue life prediction

KW - Gaussian process

KW - Probabilistic approach

KW - Statistical method

KW - Structural health monitoring

KW - Variable loading

UR - http://www.scopus.com/inward/record.url?scp=79957919495&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=79957919495&partnerID=8YFLogxK

U2 - 10.1007/s11340-010-9394-7

DO - 10.1007/s11340-010-9394-7

M3 - Article

VL - 51

SP - 833

EP - 843

JO - Experimental Mechanics

JF - Experimental Mechanics

SN - 0014-4851

IS - 6

ER -