elib
DLR-Header
DLR-Logo -> http://www.dlr.de
DLR Portal Home | Imprint | Privacy Policy | Contact | Deutsch
Fontsize: [-] Text [+]

Prediction of fatigue failure in small-scale butt-welded joints with explainable machine learning

Braun, Moritz and Kellner, Leon and Schreiber, Sarah and Ehlers, Sören (2022) Prediction of fatigue failure in small-scale butt-welded joints with explainable machine learning. Procedia Structural Integrity, 38, pp. 182-191. Elsevier. doi: 10.1016/j.prostr.2022.03.019. ISSN 2452-3216.

[img] PDF - Published version
1MB

Official URL: https://www.sciencedirect.com/science/article/pii/S2452321622002323?via%3Dihub#!

Abstract

Butt-welded joints are common in many industries. The fatigue behavior of such joints depends on numerous factors, e.g. load level, local weld geometry, or parent material strength. To make things worse, these factors often interact, yet mutual influence can hardly be quantified by multivariate studies, i.e. varying one factor at a time out of many factors, due to the large number of required tests and the statistical nature of weld geometry. Consequently, fatigue assessment of such joints often deviates significantly between prediction and experimental result. Thus, alternative methods are desirable in order to take various influencing factors into account. To this end, machine learning techniques were used to predict failure locations and number of cycles to failure of fatigue tests performed on small-scale butt-welded joint specimens. In addition to accurate predictions, an understanding of importance and mutual influence of the factors is desired, e.g. a ranking of the most important factors; however, capturing the influence of several possibly interacting factors usually requires complex nonlinear machine learning models. We used gradient boosted trees. Since these are black box models, the SHapley Additive exPlanations (SHAP) framework was used to explain the predictions, i.e. identify influential features and their interactions. Lastly, the model explanations are linked back to domain knowledge.

Item URL in elib:https://elib.dlr.de/187553/
Document Type:Article
Title:Prediction of fatigue failure in small-scale butt-welded joints with explainable machine learning
Authors:
AuthorsInstitution or Email of AuthorsAuthor's ORCID iDORCID Put Code
Braun, MoritzHamburg University of Technology, Institute of Ship Structural Design and Analysishttps://orcid.org/0000-0001-9266-1698UNSPECIFIED
Kellner, LeonHamburg University of Technology Hamburg, Germanyhttps://orcid.org/0000-0001-9722-7508UNSPECIFIED
Schreiber, SarahHamburg University of Technology Hamburg, GermanyUNSPECIFIEDUNSPECIFIED
Ehlers, SörenUNSPECIFIEDhttps://orcid.org/0000-0001-5698-9354UNSPECIFIED
Date:2022
Journal or Publication Title:Procedia Structural Integrity
Refereed publication:Yes
Open Access:Yes
Gold Open Access:No
In SCOPUS:Yes
In ISI Web of Science:No
Volume:38
DOI:10.1016/j.prostr.2022.03.019
Page Range:pp. 182-191
Publisher:Elsevier
Series Name:Elsevier
ISSN:2452-3216
Status:Published
Keywords:Fatigue life predictionWelded joints Fatigue strength Machine learning models explainable AIgradient boosted trees SHAP
HGF - Research field:Energy
HGF - Program:other
HGF - Program Themes:E - no assignment
DLR - Research area:Energy
DLR - Program:E - no assignment
DLR - Research theme (Project):E - no assignment
Location: Geesthacht
Institutes and Institutions:Institute of Maritime Energy Systems
Deposited By: Piazza, Hilke Charlotte
Deposited On:17 Oct 2022 07:27
Last Modified:02 Dec 2022 09:26

Repository Staff Only: item control page

Browse
Search
Help & Contact
Information
electronic library is running on EPrints 3.3.12
Website and database design: Copyright © German Aerospace Center (DLR). All rights reserved.