PromoAdd3D Additive Manufacturing


Process monitoring and quality assurance for additive manufacturing processes.

Additive manufacturing processes enable the production of three-dimensional objects
Project Management: Prof. Dr. Michael Möckel
University: Aschaffenburg University of Applied Sciences, Germany
Research Area: Functional Materials
Project Duration: 01.10.2018 - 31.12.2021

IntelligentSystems   RoboticsAutomation   ArtifialIntelligenceDataScience   IntelligentSensorsSignals   Materials   MaterialTestingSensorTechnology   InnovativeMaterialProcessing   Prof.Dr.Moeckel   Prof.Dr.Hellmann


Background

Additive manufacturing processes enable the direct, tool-free and therefore flexible production of three-dimensional objects based on digital information (Industry 4.0). Products can have almost any complex geometry and internal structures and can be produced efficiently even for small lot sizes (individualized production).


Objectives

The aim of the project is to further improve the selective laser melting (SLM) metal manufacturing process by developing advanced process monitoring techniques. Those are necessary to achieve a high level of automatic process control. For this purpose, new approaches in anomaly detection based both on process simulations and machine learning techniques are investigated. Early indicators for failure will be extracted from multimodal sensor data of the SLM process. The aim is to design a hybrid model which combines existing knowledge on the process with machine learning and can be used for surveillance, quality control and an accelerated approach towards model based dynamic process control.


Methods

The hybrid "physics based and data driven" model exploits additional high-dimensional correlations between sensor data for detecting early signatures for an upcoming process irregularities. Different topologies for hybrid models following established paradigms of the field will be developed and assessed for their applicability in process monitoring, for in situ quality control and for dynamic process control.

The project implementation takes place in the Lab for Medical Informatics and Simulation (German information available).


Project Members

Prof. Dr. Michael Möckel, Aschaffenburg UAS
Prof. Dr. Ralf Hellmann, Aschaffenburg UAS
Jorrit Voigt, PhD student (cooperative PhD with FAU Erlangen) and member of our iDok Doctoral Program, located at ZeWiS / ICO Obernburg



Cooperation

Are you doing research on integrating AI methods in physical or technical simulations and models as professor, postgraduate or student? If you are interested in research cooperations we will be happy to identify our cooperation potential. Please feel free to contact the research office for general enquiries or for information on currently available financed positions or PhD programmes. Prof. Dr. Möckel is happy to respond to research related questions (preferably by email to Michael.Moeckel@th-ab.de).


more on research activities on Intelligent Systems

more on research activities on Materials