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

Application of AI-based methods for the evaluation of a joining process for multi-material joints

Schulze, Julian and Greß, Alexander (2020) Application of AI-based methods for the evaluation of a joining process for multi-material joints. Master's, Hochschule Albstadt-Sigmaringen.

This is the latest version of this item.

Full text not available from this repository.


For the automotive industry, lightweight concepts become more and more important due to economic and ecologic constraints, therefore often multi-material structures are used to counteract this problem. An innovative joining solution for such structures is the flow drill screwing process. With increasing digitalization, promising technologies like artificial intelligence are implemented in manufacturing to increase productivity and gain competitive advantage. In this work, the application of different machine learning models for flow drill screwing joints is investigated, with focus on giving an overview about different methods and transparency of these. The strategy is first to generate a dataset. For this purpose, the used materials steel and aluminium are characterized by a literature review and with mechanical tests. Afterwards, single-lap-shear joints are generated using the flow drill screwing process. The joints are further investigated regarding their failure behaviour under shear tension. Selected mechanical and geometrical values of the materials are defined as input variables and the maximum force of the resulting joint is defined as the output/target variable. Finally, different machine learning models and data preparation techniques are proposed to find the most promising model. The results give an overview of the performance of different models based on the used data preparation strategies. The artificial neural network shows the best performance for this dataset. This model is then fine-tuned and evaluated regarding the effect of the training set size on the performance. Further methods concerning explainability and structure of the model are shown.

Item URL in elib:https://elib.dlr.de/142060/
Document Type:Thesis (Master's)
Title:Application of AI-based methods for the evaluation of a joining process for multi-material joints
AuthorsInstitution or Email of AuthorsAuthor's ORCID iD
Greß, AlexanderAlexander.Gress (at) dlr.deUNSPECIFIED
Date:15 December 2020
Refereed publication:Yes
Open Access:No
Gold Open Access:No
In ISI Web of Science:No
Keywords:Flow drill screwing, multi-material design, artificial intelligence, machine learning, regression, crisp-dm
Institution:Hochschule Albstadt-Sigmaringen
Department:Fakultät Engineering
HGF - Research field:Aeronautics, Space and Transport
HGF - Program:Transport
HGF - Program Themes:Road Transport
DLR - Research area:Transport
DLR - Program:V ST Straßenverkehr
DLR - Research theme (Project):V - NGC Fahrzeugstruktur II
Location: Stuttgart
Institutes and Institutions:Institute of Vehicle Concepts > Material and Process Applications for Road and Rail Vehicles
Deposited By: Greß, Alexander
Deposited On:29 Apr 2021 16:55
Last Modified:06 May 2021 12:35

Available Versions of this Item

Repository Staff Only: item control page

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