In its annual general meeting at the beginning of December, the IC SimTech e. V., the industrial consortium of the Stuttgart Center for Simulation Science, awarded outstanding final theses of the year 2023. The prize, endowed with 500 EUR each, goes to the following recipients this year:
Bachelor's Thesis: Annalena Riffelt with "Simulating the Adsorption of Formaldehyde on Interstellar Olivine" (Institute of Theoretical Chemistry, Prof. Johannes Kästner)
The goal of the bachelor's thesis was to investigate whether the transport of formaldehyde through interstellar space is possible. If such transport is possible, formaldehyde, upon reaching our solar system, can react with larger molecules, such as amino acids, bringing us closer to the ultimate answer to how life on Earth could have originated. However, for formaldehyde to be transported through interstellar space, it must adsorb to larger solid bodies to withstand the harsh conditions of outer space during its journey. Therefore, simulations were used to determine the binding energy of formaldehyde and olivine, a possible component of such a solid body, as it serves as a measure of the strength of adsorption.
Master's Thesis: Benjamin Castellaz with "Theoretical Foundations for and Advancements in Eulerian Parameter Inference" (Institute of Stochastics and Applications, Prof. Nicole Radde)
The master's thesis focused on a novel method for solving so-called stochastic inverse problems. The most important application is the calibration of parameters in simulation models using observations or experiments. The process begins with a simulation model and some measurement data, ideally providing information on how the parameters must have been distributed to explain the observations. Originally, I was supposed to examine some mathematical properties of the method, but I accidentally discovered a way to directly evaluate the marginal distributions of the parameters in certain cases. This significantly reduces the computation time needed by the method in many cases. Additionally, together with Vincent Wagner and Prof. Radde, I attempted to determine how to prepare the measurement data to obtain even more accurate results with this method. As is customary in most SimTech final theses, I not only worked on the theory but also incorporated the results into the open-source software developed by the research group (feel free to try out the package: https://pypi.org/project/eulerpi/).
Doctoral Thesis: Fabian Kempter with "Validation Approaches of Active Human Models Considering Human Variability in Experiment and Simulation" (Institute of Technical and Numerical Mechanics, Prof. Jörg Fehr)
Modern safety systems protect vehicle occupants and other road users in potential accident scenarios. Computer simulations of standardized crash scenarios play a crucial role in their development to optimize virtual protective effects in early stages. The simulation models of vehicle occupants, typically used in these highly dynamic simulations, do not directly represent the human body but rather are models of anthropometric test dummies used as mechatronic test objects in experimental crash tests. These models, also referred to as dummies, significantly abstract human anatomy and are only suitable for one or a few specific standard crash scenarios. However, more detailed simulation models of humans are required concerning the increasing intervention possibilities of vehicles before a potential accident, such as emergency braking systems, on the one hand, and non-driving activities of occupants during the journey on the other hand. Particularly, the consideration of the active properties of human musculature becomes crucial as they influence pre-accident behavior and injury risks. Addressing active human behavior introduces new challenges in the areas of modeling, simulation, and validation. This work presents three aspects of considering active human behavior and its variability in simulation. In addition to experimental investigations through a designed driving simulator test bench, simulative studies on replicating active human muscle interventions are presented.
Congratulations to all the award winners!