School of Mechanical Engineering at Dublin City University (DCU), Dublin, IrelandAdditive manufacturing; Laser processing; Materials engineering; Near net shape forming; Data capture, Analysis and control; AI for process and part design
Dermot Brabazon holds a Full Professorship of Materials Science and Engineering in the School of Mechanical Engineering at Dublin City University (DCU). He received his BEng (Mechanical Engineering) and PhD (Materials Science) from University College Dublin. From 1995 to 2000 he worked with Materials Ireland, a state materials science research centre. From 2000 he was appointed as a lecture at Dublin City University, promoted to Senior Lecturer in 2006, Deputy Head of School in 2007, Associate Dean for Research in 2009, Professor in 2014, and Full Professor in 2020. In recognition of his academic achievements and contributions to development of engineering technologies, he was conferred with the President’s Award for Research in 2009, Fellow of the Institute of Mechanical Engineering in 2015, and received Invent Commercialization awards in 2015, 2017 and 2018. Since 2012, he has been Director of the Advanced Processing Technology Research Centre at DCU and since 2017 he co-founded and was appointed Deputy-Director of I-Form, the national centre for Advanced Manufacturing Research.
He has published over 250 internationally peer reviewed papers and supervised more than 35 researchers to project completion. His research is focused in the areas of materials and processing technologies, including Additive Manufacturing, Near Net Shape Forming, Laser Processing and Separation technologies. These overlapping activities are focused toward the development of advanced materials science and engineering knowledge to enable improved product and production, capability and quality, for the benefit of companies and the broader society.
Artificial Intelligence is being used more and more for the analysis of the datasets that are produced from the Additive Manufacturing Process. These data sets include IR, acoustic, eddy current, oxygen, composition and part density data. Often fundamental models can be produced which predict the melt pool width and height, the part density, pore formation, thermal fields, resulting microstructure, and part mechanical properties. While these analytical and FEA models can give a good understanding of the process. However, in many situations they are too slow for real time quality or close loop process control. Surrogate models (also called metamodels) are commonly being used instead for these predictions where faster analysis is required. Where the analytical and FEA models are validated within defined bounds, they can be used to generate a lot of process input and output simulation data. This data can be used in conjunction with the experimental data to train the AI models and thereby make them available for more real time process quality prediction or control.
Original research articles and reviews that fall within this topic area are welcomed. In this special issue and in the context of additive manufacturing, articles will be accepted which cover one of the following areas:
- the generation of experimental data that can be used for AI model development;
- the generation of model data that can be used for AI model development;
- the use of experimental, modelling or combination of data sets for the generation of AI models;
- the analysis of AI models for the prediction of process quality and/or control.
Prediction of wall geometry for cold-metal-transfer-based wire-arc additive manufacturing
Layer porosity in powder-bed fusion prediction using regression machine learning models and time-series features