Concept for Predictive Quality in Carbon Fibre Manufacturing

Remarkable mechanical properties make carbon fibres attractive for many industrial applications. However, up to today, carbon fibres come with a significant environmental backpack, undermining their advantages in light of a strong demand for absolute sustainability of new industrial products. Conseq...

Full description

Saved in:
Bibliographic Details
Main Authors: Sebastian Gellrich, Thomas Groetsch, Maxime Maghe, Claudia Creighton, Russell Varley, Anna-Sophia Wilde, Christoph Herrmann
Format: Article
Language:English
Published: MDPI AG 2024-11-01
Series:Journal of Manufacturing and Materials Processing
Subjects:
Online Access:https://www.mdpi.com/2504-4494/8/6/272
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1846104152998936576
author Sebastian Gellrich
Thomas Groetsch
Maxime Maghe
Claudia Creighton
Russell Varley
Anna-Sophia Wilde
Christoph Herrmann
author_facet Sebastian Gellrich
Thomas Groetsch
Maxime Maghe
Claudia Creighton
Russell Varley
Anna-Sophia Wilde
Christoph Herrmann
author_sort Sebastian Gellrich
collection DOAJ
description Remarkable mechanical properties make carbon fibres attractive for many industrial applications. However, up to today, carbon fibres come with a significant environmental backpack, undermining their advantages in light of a strong demand for absolute sustainability of new industrial products. Consequently, there is considerable demand for high-quality carbon fibre manufacturing, low waste production, or alternative precursor systems allowing minimization of environmental impacts. Therefore, this paper investigates the capabilities of data analytics with a special emphasis on predictive quality in order to advance the quality management of carbon fibre manufacturing. Although existing research supports the applicability of machine learning in carbon fibre production, there is a notable scarcity of case studies and a lack of a structured repetitive data analytics concept. To address this gap, the study proposes a holistic framework for predictive quality in carbon fibre manufacturing that outlines specific data analytics requirements based on the process properties of carbon fibre production. Additionally, it introduces a systematic method for processing trend data. Finally, a case study of polyacrylonitrile (PAN)-based carbon fibre manufacturing exemplifies the concept, giving indications on feature importance and sensitivity related to the expected fibre properties. Future research can build on the comprehensive overview of predictive quality potentials and its implementation concept by extending the underlying data set and investigating the transfer to alternative precursors.
format Article
id doaj-art-85ca7eb4c8ea4acf9db4ff0fc9c9ba04
institution Kabale University
issn 2504-4494
language English
publishDate 2024-11-01
publisher MDPI AG
record_format Article
series Journal of Manufacturing and Materials Processing
spelling doaj-art-85ca7eb4c8ea4acf9db4ff0fc9c9ba042024-12-27T14:32:51ZengMDPI AGJournal of Manufacturing and Materials Processing2504-44942024-11-018627210.3390/jmmp8060272Concept for Predictive Quality in Carbon Fibre ManufacturingSebastian Gellrich0Thomas Groetsch1Maxime Maghe2Claudia Creighton3Russell Varley4Anna-Sophia Wilde5Christoph Herrmann6Institute of Machine Tools and Production Technology (IWF), Technische Universität Braunschweig, Langer Kamp 19b, 38106 Braunschweig, GermanyCarbon Nexus at the Institute for Frontier Materials, Deakin University, Geelong, VIC 3216, AustraliaCarbon Nexus at the Institute for Frontier Materials, Deakin University, Geelong, VIC 3216, AustraliaCarbon Nexus at the Institute for Frontier Materials, Deakin University, Geelong, VIC 3216, AustraliaCarbon Nexus at the Institute for Frontier Materials, Deakin University, Geelong, VIC 3216, AustraliaInstitute of Machine Tools and Production Technology (IWF), Technische Universität Braunschweig, Langer Kamp 19b, 38106 Braunschweig, GermanyInstitute of Machine Tools and Production Technology (IWF), Technische Universität Braunschweig, Langer Kamp 19b, 38106 Braunschweig, GermanyRemarkable mechanical properties make carbon fibres attractive for many industrial applications. However, up to today, carbon fibres come with a significant environmental backpack, undermining their advantages in light of a strong demand for absolute sustainability of new industrial products. Consequently, there is considerable demand for high-quality carbon fibre manufacturing, low waste production, or alternative precursor systems allowing minimization of environmental impacts. Therefore, this paper investigates the capabilities of data analytics with a special emphasis on predictive quality in order to advance the quality management of carbon fibre manufacturing. Although existing research supports the applicability of machine learning in carbon fibre production, there is a notable scarcity of case studies and a lack of a structured repetitive data analytics concept. To address this gap, the study proposes a holistic framework for predictive quality in carbon fibre manufacturing that outlines specific data analytics requirements based on the process properties of carbon fibre production. Additionally, it introduces a systematic method for processing trend data. Finally, a case study of polyacrylonitrile (PAN)-based carbon fibre manufacturing exemplifies the concept, giving indications on feature importance and sensitivity related to the expected fibre properties. Future research can build on the comprehensive overview of predictive quality potentials and its implementation concept by extending the underlying data set and investigating the transfer to alternative precursors.https://www.mdpi.com/2504-4494/8/6/272data analyticsmachine learningcarbon fibre manufacturingPANquality predictiontrend data pre-processing
spellingShingle Sebastian Gellrich
Thomas Groetsch
Maxime Maghe
Claudia Creighton
Russell Varley
Anna-Sophia Wilde
Christoph Herrmann
Concept for Predictive Quality in Carbon Fibre Manufacturing
Journal of Manufacturing and Materials Processing
data analytics
machine learning
carbon fibre manufacturing
PAN
quality prediction
trend data pre-processing
title Concept for Predictive Quality in Carbon Fibre Manufacturing
title_full Concept for Predictive Quality in Carbon Fibre Manufacturing
title_fullStr Concept for Predictive Quality in Carbon Fibre Manufacturing
title_full_unstemmed Concept for Predictive Quality in Carbon Fibre Manufacturing
title_short Concept for Predictive Quality in Carbon Fibre Manufacturing
title_sort concept for predictive quality in carbon fibre manufacturing
topic data analytics
machine learning
carbon fibre manufacturing
PAN
quality prediction
trend data pre-processing
url https://www.mdpi.com/2504-4494/8/6/272
work_keys_str_mv AT sebastiangellrich conceptforpredictivequalityincarbonfibremanufacturing
AT thomasgroetsch conceptforpredictivequalityincarbonfibremanufacturing
AT maximemaghe conceptforpredictivequalityincarbonfibremanufacturing
AT claudiacreighton conceptforpredictivequalityincarbonfibremanufacturing
AT russellvarley conceptforpredictivequalityincarbonfibremanufacturing
AT annasophiawilde conceptforpredictivequalityincarbonfibremanufacturing
AT christophherrmann conceptforpredictivequalityincarbonfibremanufacturing