Cluster Expertise at the ECCOMAS Congress in Oslo
At this year's ECCOMAS Congress 2022 in Oslo, Norway, Daniel Hilger of the Chair of Computational Analysis of Technical Systems (CATS) was given the opportunity to present his research to an expert audience under the title " A Data-Driven Reduced Order Modeling Approach Applied in Context of Numerical Analysis and Optimization of Plastic Profile Extrusion". The conference, organized by the European Community on Computational Methods in Applied Sciences took place in the Norwegian capital from June 5-10, 2022.
What impressions did you personally take away from the conference?
Finally, after almost two years of the Corona pandemic and the associated cancellation or postponement of conferences to the digital world, there was again the opportunity to exchange views on current research topics in person. With about 1700 conference contributions, ECCOMAS is one of the largest conferences in the field of computational mechanics on a European level. Especially in direct comparison to the last conference, which took place in Berlin, the extreme growth and interest in contributions from the field of ML can be clearly seen. For me, there were many new impressions to collect concerning my research topic. For me, this conference has again underlined the importance of in presence events in the scientific context. Furthermore, the informal discussions that took place with other conference topics away from the lectures were also part of this. Overall, I had the feeling that the participants really enjoyed dealing with their own research away from the digital world and thinking outside the box.
How can you briefly break down your presentation topic and what is its relevance for the scientific community and how is your research topic embedded in the complex of IoP?
In my case, I presented a data-driven reduced model, which was exemplarily applied to the use case of plastic profile extrusion. More precisely, it was about the prediction of the temperature curve in the calibration phase. Reduced models are essential wherever additional information about the process is needed, which can only be partially or incompletely collected by measurements. Reduced models offer the possibility to obtain additional information about the process in process time. This information or the model itself can then be directly integrated into structures such as a digital shadow.