Improving mechanical property consistency in fused layer modeling through deep learning-based material flow control

Improving mechanical property consistency in fused layer modeling through deep learning-based material flow control | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Improving mechanical property consistency in fused layer modeling through deep learning-based material flow control Lukas Bauer, Tobias Laumer This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8721830/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Fused layer modelling (FLM) is an additive manufacturing technique of growing relevance. Therefore, quality requirements regarding component properties, reproducibility and waste reduction are also rising. To achieve this, process control is required to react to errors by adjusting process parameters. The material flow rate is a relevant process parameter with significant impact on component quality. Deviations from the optimal material flow rate lead to under- or overextrusion, affecting the resulting component properties. This work investigates material flow control through optical measurements using neural networks based on RGB images. Underextrusion and component density are measured via semantic segmentation. Overextrusion is analyzed using a material flow rate regression approach. The approaches are validated via suitable methods such as X-ray computed tomography scans, microscopy of microsections, and weight measurement. The segmentation model achieves a mean intersection over union of 86.9% and a mean dice score of 91.4%. The regression network achieves a mean error of 2.13% and a standard deviation of 2.60%. Density measurement via segmentation deviates by 6–8% from density determined via XCT scans. Measurement of overextrusion yields a deviation of 1%. The influence of the control method on the tensile strength and elongation at yield is investigated for different initial material flow rates. An increase in tensile strength of 23% and 12% was achieved for initial material flow rates of 90% and 95%. For initial material flow rates greater than or equal to 100%, a decrease of 4.5% to 7.5% was observed. No significant influences on elongation at yield were observed. error detection process control fused layer modeling deep learning material flow rate component properties Full Text Additional Declarations No competing interests reported. Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-8721830","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":627101675,"identity":"702764d2-6069-4fa3-adee-7d7c94e35e0e","order_by":0,"name":"Lukas Bauer","email":"data:image/png;base64,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","orcid":"","institution":"Regensburg University of Applied Sciences","correspondingAuthor":true,"prefix":"","firstName":"Lukas","middleName":"","lastName":"Bauer","suffix":""},{"id":627101676,"identity":"c496f47a-f382-4121-b77c-35808418e2ae","order_by":1,"name":"Tobias Laumer","email":"","orcid":"","institution":"Regensburg University of Applied Sciences","correspondingAuthor":false,"prefix":"","firstName":"Tobias","middleName":"","lastName":"Laumer","suffix":""}],"badges":[],"createdAt":"2026-01-28 14:10:44","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8721830/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8721830/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":108180956,"identity":"ed1a1ede-d4b0-41bc-a194-42fb5b2c1e39","added_by":"auto","created_at":"2026-04-30 08:55:41","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1593699,"visible":true,"origin":"","legend":"","description":"","filename":"Manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8721830/v1_covered_da4ef46b-8901-4f3a-81c8-dec53667aafb.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Improving mechanical property consistency in fused layer modeling through deep learning-based material flow control","fulltext":[],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":true,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":true,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"error detection, process control, fused layer modeling, deep learning, material flow rate, component properties","lastPublishedDoi":"10.21203/rs.3.rs-8721830/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8721830/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eFused layer modelling (FLM) is an additive manufacturing technique of growing relevance. Therefore, quality requirements regarding component properties, reproducibility and waste reduction are also rising. To achieve this, process control is required to react to errors by adjusting process parameters. The material flow rate is a relevant process parameter with significant impact on component quality. Deviations from the optimal material flow rate lead to under- or overextrusion, affecting the resulting component properties. This work investigates material flow control through optical measurements using neural networks based on RGB images. Underextrusion and component density are measured via semantic segmentation. Overextrusion is analyzed using a material flow rate regression approach. The approaches are validated via suitable methods such as X-ray computed tomography scans, microscopy of microsections, and weight measurement. The segmentation model achieves a mean intersection over union of 86.9% and a mean dice score of 91.4%. The regression network achieves a mean error of 2.13% and a standard deviation of 2.60%. Density measurement via segmentation deviates by 6\u0026ndash;8% from density determined via XCT scans. Measurement of overextrusion yields a deviation of 1%. The influence of the control method on the tensile strength and elongation at yield is investigated for different initial material flow rates. An increase in tensile strength of 23% and 12% was achieved for initial material flow rates of 90% and 95%. For initial material flow rates greater than or equal to 100%, a decrease of 4.5% to 7.5% was observed. No significant influences on elongation at yield were observed.\u003c/p\u003e","manuscriptTitle":"Improving mechanical property consistency in fused layer modeling through deep learning-based material flow control","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-04-27 10:46:01","doi":"10.21203/rs.3.rs-8721830/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"62d00775-b9ca-488b-86e3-11b75929c98f","owner":[],"postedDate":"April 27th, 2026","published":true,"recentEditorialEvents":[{"type":"editorInvitedReview","content":"","date":"2026-05-12T13:47:34+00:00","index":17,"fulltext":""}],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2026-04-27T10:46:02+00:00","versionOfRecord":[],"versionCreatedAt":"2026-04-27 10:46:01","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8721830","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8721830","identity":"rs-8721830","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}