Comparison of deep learning and physics-aware surrogate models for melt-pool temperature prediction in laser welding

Comparison of deep learning and physics-aware surrogate models for melt-pool temperature prediction in laser welding | 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 Comparison of deep learning and physics-aware surrogate models for melt-pool temperature prediction in laser welding Amena Darwish, Akash Meena, Stefan Ericson, Kent Salomonsson This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9356611/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 5 You are reading this latest preprint version Abstract Nonlinear multiphysics systems, such as the melt-pool temperature field in laser welding, are difficult to predict efficiently: high-fidelity CFD solvers are accurate but computationally expensive, and the dynamics are non-stationary and characterized by sharp thermal gradients. This work introduces and compares two surrogate models for full-field, time-resolved melt-pool temperature prediction from laser power, speed, and time. The first is a black-box deep learning surrogate that treats each spatial location as a pixel trajectory and uses a two-branch architecture with Temporal Convolutional Network (TCN) blocks followed by a BiLSTM to capture history-driven dynamics, in parallel with a Multi-Layer Perceptron (MLP) that only sees the last time step. A hot-zone-weighted mean-squared error loss emphasizes accuracy in the melt pool. The second is a physics-aware surrogate that applies sliding-window Dynamic Mode Decomposition (SW-DMD) to separate slow and fast thermal modes, projects the fields onto a shared POD basis, and uses two small MLPs to map (power, speed, time) to the complex modal coefficients before reconstructing the temperature field. The novelty of this work lies in the deep-learning design, where an extra MLP branch and a hot-zone-weighted loss increase melt-pool fidelity in the first model, and in the second model, which uses a physics-aware modal representation that enables the network to learn directly on the coefficients of the underlying dynamical modes. Both surrogates are trained on eight Flow-3D laser-weld simulations of AA1050, resampled to 50 snapshots per power–speed combination. On a held-out high-power weld, the SW-DMD–MLP surrogate halves the average RMSE in the 700–950 K solidification range compared with the TCN–BiLSTM+MLP model (≈48 K vs ≈98 K), achieves high overlap of the hot region, and is about four orders of magnitude faster to train. These results show that embedding physics-aware modal structure before learning leads to more accurate and resource-efficient surrogates than black-box data-driven models for this nonlinear weld-dynamics problem. Laser beam welding (LBW) Melt-pool temperature field Surrogate model Physics-aware surrogate model Sliding-window Dynamic Mode Decomposition (SW-DMD) Full Text Supplementary Files SWdmdreconstructionscalar6000545p5.gif fullweldpredvsorig.mp4 Cite Share Download PDF Status: Under Review Version 1 posted Reviewers agreed at journal 30 Apr, 2026 Reviewers invited by journal 14 Apr, 2026 Editor invited by journal 13 Apr, 2026 Editor assigned by journal 09 Apr, 2026 First submitted to journal 08 Apr, 2026 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-9356611","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":623167526,"identity":"0000a2f5-02fa-4fb1-ac43-0ab272515d45","order_by":0,"name":"Amena Darwish","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA50lEQVRIiWNgGAWjYPACZiBmbGBmqDjAQ6qWM6RpAZKMbQcIq+WfdvjYgx8M1vLmM5KbPxfOuyMj38D+8AE+LRK309INexjSDefcSGyTnrntGY/BAR5jA7zW3M4xk+BhOMw4QzqxjZl322EeAwYeNgl8OuRv53+T/MNw2B6opfkz75zDPECHPf+BT4vB7Rw2aaAtiUAtDdK8DYd5GA4wmOF1l+HtNDNpGYP05BnyD9ukeY4BHXaYxxivw+RuJz+TfFNhbTuD5/jjzzw1h+3l29sffsBrDcR5yBxmwupHwSgYBaNgFBAAAFj0Rb08hz96AAAAAElFTkSuQmCC","orcid":"https://orcid.org/0000-0002-3052-9277","institution":"University of Skövde: Hogskolan i Skovde","correspondingAuthor":true,"prefix":"","firstName":"Amena","middleName":"","lastName":"Darwish","suffix":""},{"id":623167527,"identity":"c305be0b-0a38-44e6-91b2-63ba08fcc700","order_by":1,"name":"Akash Meena","email":"","orcid":"","institution":"Danmarks Tekniske Universitet","correspondingAuthor":false,"prefix":"","firstName":"Akash","middleName":"","lastName":"Meena","suffix":""},{"id":623167528,"identity":"dc5a6860-9fdf-4b6f-b365-1cab43ea7282","order_by":2,"name":"Stefan Ericson","email":"","orcid":"","institution":"University of Skövde: Hogskolan i Skovde","correspondingAuthor":false,"prefix":"","firstName":"Stefan","middleName":"","lastName":"Ericson","suffix":""},{"id":623167529,"identity":"85b27b9f-de7c-4dd7-bf15-b835e00ab579","order_by":3,"name":"Kent Salomonsson","email":"","orcid":"","institution":"University of Skövde: Hogskolan i Skovde","correspondingAuthor":false,"prefix":"","firstName":"Kent","middleName":"","lastName":"Salomonsson","suffix":""}],"badges":[],"createdAt":"2026-04-08 11:55:19","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-9356611/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9356611/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":107710196,"identity":"323154ac-6bd1-4647-b2f2-002e53a71273","added_by":"auto","created_at":"2026-04-24 09:39:54","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":846395,"visible":true,"origin":"","legend":"","description":"","filename":"Manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9356611/v1_covered_5dfa6fa4-d687-4428-8185-bcc3b3288d8d.pdf"},{"id":107522796,"identity":"95367712-38c6-45ca-831f-de42a35b6eb7","added_by":"auto","created_at":"2026-04-22 09:16:18","extension":"gif","order_by":5,"title":"","display":"","copyAsset":false,"role":"supplement","size":26523956,"visible":true,"origin":"","legend":"","description":"","filename":"SWdmdreconstructionscalar6000545p5.gif","url":"https://assets-eu.researchsquare.com/files/rs-9356611/v1/1cbc71e4c1f6ad6517399ecd.gif"},{"id":107705808,"identity":"f1e2ccfc-9a82-4029-9a59-c87bc42f10e5","added_by":"auto","created_at":"2026-04-24 09:15:22","extension":"mp4","order_by":6,"title":"","display":"","copyAsset":false,"role":"supplement","size":1489016,"visible":true,"origin":"","legend":"","description":"","filename":"fullweldpredvsorig.mp4","url":"https://assets-eu.researchsquare.com/files/rs-9356611/v1/5c513dce2521e4515766d05c.mp4"}],"financialInterests":"","formattedTitle":"Comparison of deep learning and physics-aware surrogate models for melt-pool temperature prediction in laser welding","fulltext":[],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"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":"welding-in-the-world","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"witw","sideBox":"Learn more about [Welding in the World](https://www.springer.com/journal/40194)","snPcode":"40194","submissionUrl":"https://www.editorialmanager.com/witw/","title":"Welding in the World","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"Laser beam welding (LBW), Melt-pool temperature field, Surrogate model, Physics-aware surrogate model, Sliding-window Dynamic Mode Decomposition (SW-DMD)","lastPublishedDoi":"10.21203/rs.3.rs-9356611/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9356611/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eNonlinear multiphysics systems, such as the melt-pool temperature field in laser welding, are difficult to predict efficiently: high-fidelity CFD solvers are accurate but computationally expensive, and the dynamics are non-stationary and characterized by sharp thermal gradients. This work introduces and compares two surrogate models for full-field, time-resolved melt-pool temperature prediction from laser power, speed, and time. The first is a black-box deep learning surrogate that treats each spatial location as a pixel trajectory and uses a two-branch architecture with Temporal Convolutional Network (TCN) blocks followed by a BiLSTM to capture history-driven dynamics, in parallel with a Multi-Layer Perceptron (MLP) that only sees the last time step. A hot-zone-weighted mean-squared error loss emphasizes accuracy in the melt pool. The second is a physics-aware surrogate that applies sliding-window Dynamic Mode Decomposition (SW-DMD) to separate slow and fast thermal modes, projects the fields onto a shared POD basis, and uses two small MLPs to map (power, speed, time) to the complex modal coefficients before reconstructing the temperature field. The novelty of this work lies in the deep-learning design, where an extra MLP branch and a hot-zone-weighted loss increase melt-pool fidelity in the first model, and in the second model, which uses a physics-aware modal representation that enables the network to learn directly on the coefficients of the underlying dynamical modes. Both surrogates are trained on eight Flow-3D laser-weld simulations of AA1050, resampled to 50 snapshots per power–speed combination. On a held-out high-power weld, the SW-DMD–MLP surrogate halves the average RMSE in the 700–950 K solidification range compared with the TCN–BiLSTM+MLP model (≈48 K vs ≈98 K), achieves high overlap of the hot region, and is about four orders of magnitude faster to train. These results show that embedding physics-aware modal structure before learning leads to more accurate and resource-efficient surrogates than black-box data-driven models for this nonlinear weld-dynamics problem.\u003c/p\u003e","manuscriptTitle":"Comparison of deep learning and physics-aware surrogate models for melt-pool temperature prediction in laser welding","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-04-22 09:16:13","doi":"10.21203/rs.3.rs-9356611/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"reviewerAgreed","content":"","date":"2026-04-30T14:36:54+00:00","index":0,"fulltext":""},{"type":"reviewersInvited","content":"","date":"2026-04-14T19:01:54+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"Welding in the World","date":"2026-04-13T19:41:43+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-04-09T04:19:19+00:00","index":"","fulltext":""},{"type":"submitted","content":"Welding in the World","date":"2026-04-08T07:53:13+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"welding-in-the-world","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"witw","sideBox":"Learn more about [Welding in the World](https://www.springer.com/journal/40194)","snPcode":"40194","submissionUrl":"https://www.editorialmanager.com/witw/","title":"Welding in the World","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"00abc5c3-bf29-4f41-a205-9722a16e6eee","owner":[],"postedDate":"April 22nd, 2026","published":true,"recentEditorialEvents":[{"type":"reviewerAgreed","content":"","date":"2026-04-30T14:36:54+00:00","index":0,"fulltext":""}],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[],"tags":[],"updatedAt":"2026-04-22T09:16:14+00:00","versionOfRecord":[],"versionCreatedAt":"2026-04-22 09:16:13","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9356611","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9356611","identity":"rs-9356611","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}