A theory-guided probabilistic machine learning framework for accelerated prediction of process-induced deformations in advanced composites - read online free

Aug 5, 2026

Upload by Noah

This paper presents a theory-guided probabilistic machine learning framework aimed at efficient analysis of composite manufacturing processes and associated phenomena. The approach integrates sparse probabilistic characterizations, multi-fidelity simulation schemes, and a limited set of experiments to train surrogate ML models. Using Spatially Weighted Gaussian Process Regression (SWGPR) guided by probabilistic reasoning, predictive models are built from multi-fidelity data to enable rapid and accurate manufacturing assessments. Results show effective prediction of process-induced deformations (PIDs) for L-shaped composite parts with minimal experimental effort, targeting cost-efficient and broadly applicable PID mitigation and related problems.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages10

Reading Time25 min

AccessFree to download

Abstract

1. Introduction

POLITECNICO DI TORINO Repository ISTITUZIONALE A theory-guided probabilistic machine learning framework for accelerated prediction of process-induced deformations in advanced composites Original A theory-guided probabilistic machine learning framework for accelerated prediction of process-induced deformations in advanced composites / Schoenholz, C. ; Zappino, E. ; Petrolo, M. ; Zobeiry, N.. -ELETTRONICO. -8:(2024) . (Intervento presentato al convegno 21st European Conference on Composite Materials (ECCM21) tenutosi a Nantes (FRA) nel 2-5 July 2024) [10 .60691/yj56-np80] . Availability: This version is available at: 11583/2990058 since: 2024-07-01T12:03:35Z Publisher: The European Society for Composite Materials (ESCM) and the Ecole Centrale de Nantes Published DOI:10.60691/yj56-np80 Terms of use: This article is made available under terms and conditions as specified in the corresponding bibliographic description in the repository Publisher copyright (Article begins on next page) 28 March 2025 Proceedings of the 21st European Conference on Composite Materials Volume 8-Special Sessions A THEORY-GUIDED PROBABILISTIC MACHINE LEARNING FRAMEWORK FOR ACCELERATED PREDICTION OF PROCESSINDUCED DEFORMATIONS IN ADVANCED COMPOSITES Caleb Schoenholz 1,a, Enrico Zappino 2,b, Marco Petrolo 2,c, Navid Zobeiry 1,d,* 1 Materials Science & Engineering Department, University of Washington, 302 Roberts Hall, Box 352120, Seattle, WA, 98195-2120, U. S.A. [Email:](Email: a css1119@uw.edu)[ a](Email: a css1119@uw.edu)[ css1119@uw.edu](Email: a css1119@uw.edu), [d](d navidz@uw.edu)[ navidz@uw.edu](d navidz@uw.edu), Web Page: [https://composites.uw.edu/](https://composites.uw.edu/) 2 MUL2 Lab, Department of Mechanical and Aerospace Engineering, Politecnico di Torino, Corso Duca degli Abruzzi 24, 10129 Torino, Italy Email: b enrico.zappino@polito.it, c marco.petrolo@polito.it, Web Page: [http://www.mul2.polito.it/](http://www.mul2.polito.it/) * Corresponding Author Keywords: Aerospace composites manufacturing, process-induced deformations (PIDs), theory-guided probabilistic machine learning, Carrera Unified Formulation (CUF) 1412 1420 CONTENTS Abstract This paper introduces an innovative framework for efficient analysis of composites manufacturing processes and phenomena. The method combines sparse probabilistic characterizations, multi-fidelity simulation schemes, and limited experiments to train surrogate machine learning (ML) models. Guided by a probabilistic technique, Spatially Weighted Gaussian Process Regression (SWGPR), predictive models are constructed from multi-fidelity data to perform rapid and accurate manufacturing assessments. This study demonstrates the effectiveness of the framework in accurately predicting process-induced deformations (PIDs) for L-shaped composite parts using minimal experimental efforts. The method introduced in this work aims to offer a cost-efficient and broadly applicable framework for potentially mitigating PIDs and solving other composites manufacturing problems. 1. Introduction While carbon fiber-reinforced polymer (CFRP) composites have seen widespread use throughout the aerospace industry, manufacturers continue to face several challenges. One such challenge is the prediction and control of process-induced deformations (PIDs) in composite parts [1–3] . During autoclave processing, residual stresses form due to complex material and manufacturing phenomena [3] . Upon demolding, some of these stresses may be released through deformations such as angle alterationsat transition points (e.g. , spring-in) or warping of initially flat sections, as schematically illustrated in Figure 1 [4] . Consequently, these PIDs may induce joining gaps during assembly, prolong production timelines, and compromise the final structure’s mechanical efficiency [5] . Despite having a general understanding of PIDs, manufacturers often struggle to accurately predict deforma

FAQs

What problem does the paper address in composite manufacturing?

It targets the prediction and control of process-induced deformations (PIDs) in composite parts, which can cause issues such as joining gaps and reduced mechanical efficiency after demolding.

How does the framework combine simulations and experiments?

It uses multi-fidelity simulation schemes together with a limited experimental dataset to train surrogate machine learning models, reducing the fidelity–time/cost trade-off.

Which machine learning method is central to the proposed approach?

Spatially Weighted Gaussian Process Regression (SWGPR) is used to build predictive models from multi-fidelity data, guided by probabilistic reasoning.

Aug 5, 2026

Upload by Noah