Data-driven and physics-based modelling of process behaviour and deposit geometry for friction surfacing

Publikation: Beiträge in ZeitschriftenZeitschriftenaufsätzeForschung

Standard

Data-driven and physics-based modelling of process behaviour and deposit geometry for friction surfacing. / Bock, Frederic E.; Kallien, Zina; Huber, Norbert et al.
in: Computer Methods in Applied Mechanics and Engineering, Jahrgang 418, Nr. Part A, 116453, 01.01.2024.

Publikation: Beiträge in ZeitschriftenZeitschriftenaufsätzeForschung

Harvard

APA

Vancouver

Bibtex

@article{16a6034cb2d24a4cb82f294d8e3ea7b2,
title = "Data-driven and physics-based modelling of process behaviour and deposit geometry for friction surfacing",
abstract = "In the last decades, there has been an increase in the number of successful machine learning models that have served as a key to identifying and using linkages within the process-structure–property-performance chain for vastly different problems in the domains of materials mechanics. The consideration of physical laws in data-driven modelling has recently been shown to enable enhanced prediction performance and generalization while requiring less data than either physics-based or data-driven modelling approaches independently. In this contribution, we introduce a simulation-assisted machine learning framework applied to the solid-state layer deposition technique friction surfacing, suitable for solid-state additive manufacturing as well as repair or coating applications. The objective of the present study is to use machine learning algorithms to predict and analyse the influence of process parameters and environmental variables, i.e. substrate and backing material properties, on process behaviour and deposit geometry. The effects of maximum process temperatures supplied by a numerical heat transfer model on the predictions of the targets are given special attention. Numerous different machine learning algorithms are implemented, optimized and evaluated to take advantage of their varied capabilities and to choose the optimal one for each target and the provided data. Furthermore, the input feature dependence for each prediction target is evaluated using game-theory related Shapley Additive Explanation values. The experimental data set consists of two separate experimental design spaces, one for varying process parameters and the other for varying substrate and backing material properties, which allowed to keep the experimental effort to a minimum. The aim was to also represent the cross parameter space between the two independent spaces in the predictive model, which was accomplished and resulted in an approximately 44 % reduction in the number of experiments when compared to carrying out an experimental design that included both spaces.",
keywords = "Machine learning, Feature selection, Numerical modelling, Heat transfer, Design of experiments, Explainable AI, Engineering",
author = "Bock, {Frederic E.} and Zina Kallien and Norbert Huber and Benjamin Klusemann",
note = "This project has received funding from the European Research Council (ERC) under the European Union's Horizon 2020 research and innovation program (grant agreement No 101001567). Publisher Copyright: {\textcopyright} 2023 The Author(s)",
year = "2024",
month = jan,
day = "1",
doi = "10.1016/j.cma.2023.116453",
language = "English",
volume = "418",
journal = "Computer Methods in Applied Mechanics and Engineering",
issn = "0045-7825",
publisher = "Elsevier B.V.",
number = "Part A",

}

RIS

TY - JOUR

T1 - Data-driven and physics-based modelling of process behaviour and deposit geometry for friction surfacing

AU - Bock, Frederic E.

AU - Kallien, Zina

AU - Huber, Norbert

AU - Klusemann, Benjamin

N1 - This project has received funding from the European Research Council (ERC) under the European Union's Horizon 2020 research and innovation program (grant agreement No 101001567). Publisher Copyright: © 2023 The Author(s)

PY - 2024/1/1

Y1 - 2024/1/1

N2 - In the last decades, there has been an increase in the number of successful machine learning models that have served as a key to identifying and using linkages within the process-structure–property-performance chain for vastly different problems in the domains of materials mechanics. The consideration of physical laws in data-driven modelling has recently been shown to enable enhanced prediction performance and generalization while requiring less data than either physics-based or data-driven modelling approaches independently. In this contribution, we introduce a simulation-assisted machine learning framework applied to the solid-state layer deposition technique friction surfacing, suitable for solid-state additive manufacturing as well as repair or coating applications. The objective of the present study is to use machine learning algorithms to predict and analyse the influence of process parameters and environmental variables, i.e. substrate and backing material properties, on process behaviour and deposit geometry. The effects of maximum process temperatures supplied by a numerical heat transfer model on the predictions of the targets are given special attention. Numerous different machine learning algorithms are implemented, optimized and evaluated to take advantage of their varied capabilities and to choose the optimal one for each target and the provided data. Furthermore, the input feature dependence for each prediction target is evaluated using game-theory related Shapley Additive Explanation values. The experimental data set consists of two separate experimental design spaces, one for varying process parameters and the other for varying substrate and backing material properties, which allowed to keep the experimental effort to a minimum. The aim was to also represent the cross parameter space between the two independent spaces in the predictive model, which was accomplished and resulted in an approximately 44 % reduction in the number of experiments when compared to carrying out an experimental design that included both spaces.

AB - In the last decades, there has been an increase in the number of successful machine learning models that have served as a key to identifying and using linkages within the process-structure–property-performance chain for vastly different problems in the domains of materials mechanics. The consideration of physical laws in data-driven modelling has recently been shown to enable enhanced prediction performance and generalization while requiring less data than either physics-based or data-driven modelling approaches independently. In this contribution, we introduce a simulation-assisted machine learning framework applied to the solid-state layer deposition technique friction surfacing, suitable for solid-state additive manufacturing as well as repair or coating applications. The objective of the present study is to use machine learning algorithms to predict and analyse the influence of process parameters and environmental variables, i.e. substrate and backing material properties, on process behaviour and deposit geometry. The effects of maximum process temperatures supplied by a numerical heat transfer model on the predictions of the targets are given special attention. Numerous different machine learning algorithms are implemented, optimized and evaluated to take advantage of their varied capabilities and to choose the optimal one for each target and the provided data. Furthermore, the input feature dependence for each prediction target is evaluated using game-theory related Shapley Additive Explanation values. The experimental data set consists of two separate experimental design spaces, one for varying process parameters and the other for varying substrate and backing material properties, which allowed to keep the experimental effort to a minimum. The aim was to also represent the cross parameter space between the two independent spaces in the predictive model, which was accomplished and resulted in an approximately 44 % reduction in the number of experiments when compared to carrying out an experimental design that included both spaces.

KW - Machine learning

KW - Feature selection

KW - Numerical modelling

KW - Heat transfer

KW - Design of experiments

KW - Explainable AI

KW - Engineering

UR - http://www.scopus.com/inward/record.url?scp=85172876453&partnerID=8YFLogxK

U2 - 10.1016/j.cma.2023.116453

DO - 10.1016/j.cma.2023.116453

M3 - Journal articles

VL - 418

JO - Computer Methods in Applied Mechanics and Engineering

JF - Computer Methods in Applied Mechanics and Engineering

SN - 0045-7825

IS - Part A

M1 - 116453

ER -

DOI

Zuletzt angesehen

Aktivitäten

  1. Project Workshop on "Worker Flows, Match Quality, and Productivity" - 2019
  2. 13th IFAC and IEEE Conference on Programmable Devices and Embedded Systems - PDeS 2015
  3. Applied Econometrics with Stata for PhD Students
  4. Princeton University
  5. Thinking of Time - A Resource which Should be Allocated Equally
  6. Changing learning environments at university? Comparing the learning strategies of non-traditional European students engaged in lifelong learning.
  7. Predicting negotiation success with a multitude of negotiators’ inter-individual differences—a latent personality model of the successful negotiator
  8. Visualizing and analyzing big data sets: Results from the Student Bodies-Eating Disorders study
  9. "Crowds and Party" Reading Workshop with Jody Dean - 2019
  10. European University Institute
  11. Where is language use in the description of the Englishes? - ESSE 2006
  12. Comfort and Adaptive Cruise Control in Highly Automated Vehicles
  13. Panel Cointegration Testing with Time Trend and Analysis of Money Demand in OECD Countries
  14. Navigating Educational Ambidexterity: Exploring Leader-Member Exchange and Open Innovation
  15. Digital Capitalism meets Leberkaspeppi: Temporal Orientations in Business Models as a Source of Platform Power in Mature Industries
  16. THE EUROPEAN CONFERENCE ON MACHINE LEARNING & PRINCIPLES AND PRACTICE OF KNOWLEDGE DISCOVERY IN DATABASES 2017
  17. Ghosting the City – Zooming in on Otherwise Publics in Virtual Worlds
  18. 6th Workshop on Unintended Consequences
  19. Internet-based guided self-help to reduce depressive symptoms in teachers: Results from a randomized controlled trial
  20. Modelling biodegradability based on OECD 301D data for the design of mineralising ionic liquids
  21. Mathematical and Computational Applications (Fachzeitschrift)
  22. Response to Philip Hogh: Suffering, Pain and the Idea of Progress

Publikationen

  1. Cue predictability changes scaling in eye-movement fluctuations
  2. Service Level Driven Stock Allocation
  3. Digging into the roots
  4. Design for Product Care—Development of Design Strategies and a Toolkit for Sustainable Consumer Behaviour
  5. The effect of structural complexity on large mammal occurrence in revegetation
  6. Towards Advanced Learning in Dispatching Rule-Based Scheuling
  7. Integrating inductive and deductive analysis to identify and characterize archetypical social-ecological systems and their changes
  8. "Introduction," communication +1
  9. What Makes for a Good Theory? How to Evaluate a Theory Using the Strength Model of Self-Control as an Example
  10. Embedding Evidence on Conservation Interventions Within a Context of Multilevel Governance
  11. Drafts in Action
  12. Development and comparison of processing maps of Mg-3Sn-1Ca alloy from data obtained in tension versus compression
  13. Modeling of temperature- and strain-driven intermetallic compound evolution in an Al-Mg system via a multiphase-field approach with application to refill friction stir spot welding
  14. Discriminative clustering for market segmentation
  15. From Open Access to Open Science
  16. How to move the transition to sustainable food consumption towards a societal tipping point
  17. Integrating business models and enterprise architecture
  18. Concepts, Formats, and Methods of Participation
  19. Effective digital practice in the competence-oriented English as a foreign language classroom in Germany
  20. Trajectory tracking using MPC and a velocity observer for flat actuator systems in automotive applications