Physics-Informed Neural Initialization for Robust Multi-Fidelity Coupled Simulation of Advanced Propulsion Systems

Physics-Informed Neural Initialization for Robust Multi-Fidelity Coupled Simulation of Advanced Propulsion Systems | 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 Physics-Informed Neural Initialization for Robust Multi-Fidelity Coupled Simulation of Advanced Propulsion Systems Xin Yang, Pengfu Xie, Xuezhi Dong, Wenchao Sun, Xiyang Liu, Hulan Yang, and 1 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6812843/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 02 Feb, 2026 Read the published version in Engineering with Computers → Version 1 posted 7 You are reading this latest preprint version Abstract Traditional zero-dimensional (0D) engine performance analysis relies on dimensionless characteristic maps, which cannot capture complex nonlinear phenomena such as shocks and flow separation in transonic regimes. Integrating high-fidelity components with complex features can invalidate previous constrains, turning the system from a single-solution to a multi-solution domain. Conventional solvers, such as the Newton-Raphson method and its variants struggle in these cases, often converging to non-physical solutions and causing pseudo-convergence. While machine learning methods can model nonlinearities, their generalization is limited when training data is sparse, increasing computational costs and risking design errors. To address these unresolved issues, this study presents a novel physics-informed neural initialization algorithm that tightly integrates physics-based nonlinear solvers with data-driven surrogate models. The proposed method employs a multilayer feedforward neural network (MLF) to generate initial values for the iterative solution process, which are further refined and constrained by physical model consistency. Multi-fidelity simulations are then performed with these calibrated values, and their results are used to iteratively refine the surrogate model. Validation on a twin-spool turbofan engine with a vectoring nozzle shows that the algorithm matches the predictive accuracy of traditional methods under normal conditions. At extreme points, such as large deflection angles and off-design operations, it effectively suppresses pseudo-convergence and reduces the number of 3D CFD model iterations by over 60%. Overall, this method enables accurate modeling of complex nozzle regulation, enhances solver robustness under high-fidelity coupling, and significantly reduces the dependence of machine learning on large-scale datasets. Aero engines multi-fidelity simulation machine learning neural network hybrid modeling methods Full Text Additional Declarations No competing interests reported. Cite Share Download PDF Status: Published Journal Publication published 02 Feb, 2026 Read the published version in Engineering with Computers → Version 1 posted Editorial decision: Revision requested 16 Oct, 2025 Reviews received at journal 12 Aug, 2025 Reviewers agreed at journal 13 Jul, 2025 Reviewers invited by journal 25 Jun, 2025 Editor assigned by journal 04 Jun, 2025 Submission checks completed at journal 04 Jun, 2025 First submitted to journal 03 Jun, 2025 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-6812843","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":476753476,"identity":"3f09c97e-60ba-4b71-a4c2-e063bb788c87","order_by":0,"name":"Xin Yang","email":"","orcid":"","institution":"Tsinghua University","correspondingAuthor":false,"prefix":"","firstName":"Xin","middleName":"","lastName":"Yang","suffix":""},{"id":476753477,"identity":"efddb059-2900-4941-b75a-d567f878571b","order_by":1,"name":"Pengfu Xie","email":"","orcid":"","institution":"Tsinghua University","correspondingAuthor":false,"prefix":"","firstName":"Pengfu","middleName":"","lastName":"Xie","suffix":""},{"id":476753478,"identity":"61e65385-dc7f-48c1-b14c-808eac28351a","order_by":2,"name":"Xuezhi Dong","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABEklEQVRIie3RPUvDQBjA8QsHcbm063XxMzyl0FKU5KukBJLFoVPmSOGmdo/0U3QSt8gzZAlxzdAhXVx0SHAwg6IPrUOHNOomeH8O7oX7wcExptP9xZKvGTinJewHZf6ImLPol4SJYXRYfEN66QqfLHXpTM7E666Zb+1b6ZWsCpH111ErGWS5P71RPp8urM1CwKN3F/tgxDkyuU1aCRRXY6gVmoBEGKAHhQvcUshAuqfJTH0IQLG7bvYkqPh7NxmVtUokESMiaNMJcKODDLJsTC/3ANAcrom4kD3P75d5IGTRTnrpcvQiQtuBByzr5g0dSINN2YQX5/24nVCmPN7Rfx4+S5y6T/HqeOd03NTpdLp/2ifid2XA2l5pUgAAAABJRU5ErkJggg==","orcid":"","institution":"Tsinghua University","correspondingAuthor":true,"prefix":"","firstName":"Xuezhi","middleName":"","lastName":"Dong","suffix":""},{"id":476753479,"identity":"0f1c0fc9-dd65-49ef-9fa2-827b2a0f6439","order_by":3,"name":"Wenchao Sun","email":"","orcid":"","institution":"Tsinghua University","correspondingAuthor":false,"prefix":"","firstName":"Wenchao","middleName":"","lastName":"Sun","suffix":""},{"id":476753480,"identity":"a6d22865-f70f-4b8c-a9c8-0669c3a1747a","order_by":4,"name":"Xiyang Liu","email":"","orcid":"","institution":"Tsinghua University","correspondingAuthor":false,"prefix":"","firstName":"Xiyang","middleName":"","lastName":"Liu","suffix":""},{"id":476753481,"identity":"0aae01e7-692f-4482-9030-91e5fe44ffc0","order_by":5,"name":"Hulan Yang","email":"","orcid":"","institution":"Nanjing Tianfu Software Co., Ltd, Jiangning Development Zone","correspondingAuthor":false,"prefix":"","firstName":"Hulan","middleName":"","lastName":"Yang","suffix":""},{"id":476753482,"identity":"cbfe7ed5-1384-45a9-ac41-d4725d1d82ec","order_by":6,"name":"Chunqing Tan","email":"","orcid":"","institution":"Tsinghua University","correspondingAuthor":false,"prefix":"","firstName":"Chunqing","middleName":"","lastName":"Tan","suffix":""}],"badges":[],"createdAt":"2025-06-03 15:23:27","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6812843/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6812843/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1007/s00366-026-02279-4","type":"published","date":"2026-02-02T15:57:18+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":102233999,"identity":"5011abe3-bdb3-4983-9398-fc5c4bcb4a82","added_by":"auto","created_at":"2026-02-09 16:02:46","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1582882,"visible":true,"origin":"","legend":"","description":"","filename":"25.06.03Physics.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6812843/v1_covered_4e9ec32e-875e-44e0-81bd-310e131c8ac1.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Physics-Informed Neural Initialization for Robust Multi-Fidelity Coupled Simulation of Advanced Propulsion Systems","fulltext":[],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":true,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":true,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"engineering-with-computers","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"ewco","sideBox":"Learn more about [Engineering with Computers](http://link.springer.com/journal/366)","snPcode":"366","submissionUrl":"https://submission.nature.com/new-submission/366/3","title":"Engineering with Computers","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"Aero engines, multi-fidelity simulation, machine learning, neural network, hybrid modeling methods","lastPublishedDoi":"10.21203/rs.3.rs-6812843/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6812843/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eTraditional zero-dimensional (0D) engine performance analysis relies on dimensionless characteristic maps, which cannot capture complex nonlinear phenomena such as shocks and flow separation in transonic regimes. Integrating high-fidelity components with complex features can invalidate previous constrains, turning the system from a single-solution to a multi-solution domain. Conventional solvers, such as the Newton-Raphson method and its variants struggle in these cases, often converging to non-physical solutions and causing pseudo-convergence. While machine learning methods can model nonlinearities, their generalization is limited when training data is sparse, increasing computational costs and risking design errors. To address these unresolved issues, this study presents a novel physics-informed neural initialization algorithm that tightly integrates physics-based nonlinear solvers with data-driven surrogate models. The proposed method employs a multilayer feedforward neural network (MLF) to generate initial values for the iterative solution process, which are further refined and constrained by physical model consistency. Multi-fidelity simulations are then performed with these calibrated values, and their results are used to iteratively refine the surrogate model. Validation on a twin-spool turbofan engine with a vectoring nozzle shows that the algorithm matches the predictive accuracy of traditional methods under normal conditions. At extreme points, such as large deflection angles and off-design operations, it effectively suppresses pseudo-convergence and reduces the number of 3D CFD model iterations by over 60%. Overall, this method enables accurate modeling of complex nozzle regulation, enhances solver robustness under high-fidelity coupling, and significantly reduces the dependence of machine learning on large-scale datasets.\u003c/p\u003e","manuscriptTitle":"Physics-Informed Neural Initialization for Robust Multi-Fidelity Coupled Simulation of Advanced Propulsion Systems","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-06-30 11:55:34","doi":"10.21203/rs.3.rs-6812843/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2025-10-16T09:38:33+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-08-12T17:50:13+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"190383270502881348103481715265513649875","date":"2025-07-14T03:03:44+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-06-25T21:40:18+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-06-04T05:57:50+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-06-04T05:23:36+00:00","index":"","fulltext":""},{"type":"submitted","content":"Engineering with Computers","date":"2025-06-03T15:18:29+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"engineering-with-computers","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"ewco","sideBox":"Learn more about [Engineering with Computers](http://link.springer.com/journal/366)","snPcode":"366","submissionUrl":"https://submission.nature.com/new-submission/366/3","title":"Engineering with Computers","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"20075dd2-7128-4a9c-b404-636642cecf68","owner":[],"postedDate":"June 30th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[],"tags":[],"updatedAt":"2026-02-09T16:00:52+00:00","versionOfRecord":{"articleIdentity":"rs-6812843","link":"https://doi.org/10.1007/s00366-026-02279-4","journal":{"identity":"engineering-with-computers","isVorOnly":false,"title":"Engineering with Computers"},"publishedOn":"2026-02-02 15:57:18","publishedOnDateReadable":"February 2nd, 2026"},"versionCreatedAt":"2025-06-30 11:55:34","video":"","vorDoi":"10.1007/s00366-026-02279-4","vorDoiUrl":"https://doi.org/10.1007/s00366-026-02279-4","workflowStages":[]},"version":"v1","identity":"rs-6812843","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6812843","identity":"rs-6812843","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}