Vis enkel innførsel

dc.contributor.advisorHubert Choux, Martin Marie
dc.contributor.advisorBilal, Muhammad Talha
dc.contributor.authorKorneliussen, Jostein T.
dc.contributor.authorZubaidi, Orwah
dc.date.accessioned2023-06-06T16:23:31Z
dc.date.available2023-06-06T16:23:31Z
dc.date.issued2022
dc.identifierno.uia:inspera:109927222:70022748
dc.identifier.urihttps://hdl.handle.net/11250/3070216
dc.description.abstractThis thesis examines a study of The Litium-Ion Battery (LIB) from a electric vehicle, and it’s recycling processes. A Battery Module (BM) from the LIB is shredded when considered an End-Of-Life product, and motivates for automated dismantling concepts to separate the components to save raw materials. From State-of-the-art (SoA) research projects and background theory, automatic module dis- mantling concepts have been evaluated for a Volkswagen E-Golf 2019 battery module. The presence of irreversible fasteners make the use off destructive dismantling techniques neces- sary. This study evaluates two different concepts to disconnect laser welds holding together the compressive plates made of steel. A hydraulic actuated concept is first investigated to separate the welded compressive plates within the casing. A FEM analysis with different configurations is performed to evaluate the most effective hydraulic solution when analysing the Von Mises stress. This solution is further compared with another automatic dismantling concept, namely milling. For the purpose of an automated milling concept, manipulators from ABB are assessed and the feasibility is verified based on results from manual milling operation. The proposed dismantling operation is made possible by developing a system architecture combining robotic control and computer vision. Open source software based on Robot Op- erating System (ROS) and MoveIt connect and control an ABB IRB4400 industrial robot whereas the computer vision setup involves a cutting edge 3D camera, Zivid, and object detection algorithm YOLOv5 best suited for this task. Adjustable acquisition settings in services from Zivid’s ROS driver are tested to accomplish the optimal capture configuration. Two datasets generated with RoboFlow were exported in the YOLOv5 PyTorch format. Custom object detection models with annotated components from the BM was trained and tested with image captures. All in all, this study demonstrates that the automatic dismantling of battery modules can be achieved even though they include irreversible fasteners. The proposed methods are verified on a specific battery module (Egolf 2019) but are flexible enough to be easily extended to a large variety of EV battery modules.
dc.description.abstract
dc.language
dc.publisherUniversity of Agder
dc.titleElectric Vehicle Battery Module Dismantling "Analysis and Evaluation of Robotic Dismantling Techniques for Irre- versible Fasteners, including Object Detection of Components."
dc.typeMaster thesis


Tilhørende fil(er)

Thumbnail

Denne innførselen finnes i følgende samling(er)

Vis enkel innførsel