Predictive Models for the Characterization of Internal Defects in Additive Materials from Active Thermography Sequences Supported by Machine Learning Methods

The present article addresses a generation of predictive models that assesses the thickness and length of internal defects in additive manufacturing materials. These modes use data from the application of active transient thermography numerical simulation. In this manner, the raised procedure is an...

Full description

Bibliographic Details
Main Authors: Manuel Rodríguez-Martín, José G. Fueyo, Diego Gonzalez-Aguilera, Francisco J. Madruga, Roberto García-Martín, Ángel Luis Muñóz, Javier Pisonero
Format: Article
Language:English
Published: MDPI AG 2020-07-01
Series:Sensors
Subjects:
Online Access:https://www.mdpi.com/1424-8220/20/14/3982
_version_ 1797562223607939072
author Manuel Rodríguez-Martín
José G. Fueyo
Diego Gonzalez-Aguilera
Francisco J. Madruga
Roberto García-Martín
Ángel Luis Muñóz
Javier Pisonero
author_facet Manuel Rodríguez-Martín
José G. Fueyo
Diego Gonzalez-Aguilera
Francisco J. Madruga
Roberto García-Martín
Ángel Luis Muñóz
Javier Pisonero
author_sort Manuel Rodríguez-Martín
collection DOAJ
description The present article addresses a generation of predictive models that assesses the thickness and length of internal defects in additive manufacturing materials. These modes use data from the application of active transient thermography numerical simulation. In this manner, the raised procedure is an ad-hoc hybrid method that integrates finite element simulation and machine learning models using different predictive feature sets and characteristics (i.e., regression, Gaussian regression, support vector machines, multilayer perceptron, and random forest). The performance results for each model were statistically analyzed, evaluated, and compared in terms of predictive performance, processing time, and outlier sensibility to facilitate the choice of a predictive method to obtain the thickness and length of an internal defect from thermographic monitoring. The best model to predictdefect thickness with six thermal features was interaction linear regression. To make predictive models for defect length and thickness, the best model was Gaussian process regression. However, models such as support vector machines also had significative advantages in terms of processing time and adequate performance for certain feature sets. In this way, the results showed that the predictive capability of some types of algorithms could allow for the detection and measurement of internal defects in materials produced by additive manufacturing using active thermography as a non-destructive test.
first_indexed 2024-03-10T18:25:11Z
format Article
id doaj.art-e960bbb40ab74899b88b677d8e4f62c9
institution Directory Open Access Journal
issn 1424-8220
language English
last_indexed 2024-03-10T18:25:11Z
publishDate 2020-07-01
publisher MDPI AG
record_format Article
series Sensors
spelling doaj.art-e960bbb40ab74899b88b677d8e4f62c92023-11-20T07:06:42ZengMDPI AGSensors1424-82202020-07-012014398210.3390/s20143982Predictive Models for the Characterization of Internal Defects in Additive Materials from Active Thermography Sequences Supported by Machine Learning MethodsManuel Rodríguez-Martín0José G. Fueyo1Diego Gonzalez-Aguilera2Francisco J. Madruga3Roberto García-Martín4Ángel Luis Muñóz5Javier Pisonero6Department of Mechanical Engineering, Universidad de Salamanca, 37008 Salamanca, SpainDepartment of Mechanical Engineering, Universidad de Salamanca, 37008 Salamanca, SpainDepartment of Cartographic and Land Engineering, Universidad de Salamanca, 05003 Ávila, SpainPhotonics Engineering Group, CIBER-BBN and IDIVAL, Universidad de Cantabria, 39005 Santander, Cantabria, SpainDepartment of Mechanical Engineering, Universidad de Salamanca, 37008 Salamanca, SpainDepartment of Cartographic and Land Engineering, Universidad de Salamanca, 05003 Ávila, SpainDepartment of Cartographic and Land Engineering, Universidad de Salamanca, 05003 Ávila, SpainThe present article addresses a generation of predictive models that assesses the thickness and length of internal defects in additive manufacturing materials. These modes use data from the application of active transient thermography numerical simulation. In this manner, the raised procedure is an ad-hoc hybrid method that integrates finite element simulation and machine learning models using different predictive feature sets and characteristics (i.e., regression, Gaussian regression, support vector machines, multilayer perceptron, and random forest). The performance results for each model were statistically analyzed, evaluated, and compared in terms of predictive performance, processing time, and outlier sensibility to facilitate the choice of a predictive method to obtain the thickness and length of an internal defect from thermographic monitoring. The best model to predictdefect thickness with six thermal features was interaction linear regression. To make predictive models for defect length and thickness, the best model was Gaussian process regression. However, models such as support vector machines also had significative advantages in terms of processing time and adequate performance for certain feature sets. In this way, the results showed that the predictive capability of some types of algorithms could allow for the detection and measurement of internal defects in materials produced by additive manufacturing using active thermography as a non-destructive test.https://www.mdpi.com/1424-8220/20/14/3982active thermography (AT)finite element method (FEM)non-destructive testing (NDT)quality assessment (QA)machine learning (ML)additive materials (AM)
spellingShingle Manuel Rodríguez-Martín
José G. Fueyo
Diego Gonzalez-Aguilera
Francisco J. Madruga
Roberto García-Martín
Ángel Luis Muñóz
Javier Pisonero
Predictive Models for the Characterization of Internal Defects in Additive Materials from Active Thermography Sequences Supported by Machine Learning Methods
Sensors
active thermography (AT)
finite element method (FEM)
non-destructive testing (NDT)
quality assessment (QA)
machine learning (ML)
additive materials (AM)
title Predictive Models for the Characterization of Internal Defects in Additive Materials from Active Thermography Sequences Supported by Machine Learning Methods
title_full Predictive Models for the Characterization of Internal Defects in Additive Materials from Active Thermography Sequences Supported by Machine Learning Methods
title_fullStr Predictive Models for the Characterization of Internal Defects in Additive Materials from Active Thermography Sequences Supported by Machine Learning Methods
title_full_unstemmed Predictive Models for the Characterization of Internal Defects in Additive Materials from Active Thermography Sequences Supported by Machine Learning Methods
title_short Predictive Models for the Characterization of Internal Defects in Additive Materials from Active Thermography Sequences Supported by Machine Learning Methods
title_sort predictive models for the characterization of internal defects in additive materials from active thermography sequences supported by machine learning methods
topic active thermography (AT)
finite element method (FEM)
non-destructive testing (NDT)
quality assessment (QA)
machine learning (ML)
additive materials (AM)
url https://www.mdpi.com/1424-8220/20/14/3982
work_keys_str_mv AT manuelrodriguezmartin predictivemodelsforthecharacterizationofinternaldefectsinadditivematerialsfromactivethermographysequencessupportedbymachinelearningmethods
AT josegfueyo predictivemodelsforthecharacterizationofinternaldefectsinadditivematerialsfromactivethermographysequencessupportedbymachinelearningmethods
AT diegogonzalezaguilera predictivemodelsforthecharacterizationofinternaldefectsinadditivematerialsfromactivethermographysequencessupportedbymachinelearningmethods
AT franciscojmadruga predictivemodelsforthecharacterizationofinternaldefectsinadditivematerialsfromactivethermographysequencessupportedbymachinelearningmethods
AT robertogarciamartin predictivemodelsforthecharacterizationofinternaldefectsinadditivematerialsfromactivethermographysequencessupportedbymachinelearningmethods
AT angelluismunoz predictivemodelsforthecharacterizationofinternaldefectsinadditivematerialsfromactivethermographysequencessupportedbymachinelearningmethods
AT javierpisonero predictivemodelsforthecharacterizationofinternaldefectsinadditivematerialsfromactivethermographysequencessupportedbymachinelearningmethods