Math SOMMa Junior Meeting

The automotive design process is complex and time-consuming, requiring collaboration between engineers from different areas of expertise to predict and optimize every aspect of the product. Meeting numerous targets and regulations while keeping up with the fast-paced global market adds further challenges. To remain competitive, automotive companies must improve development efficiency by reducing time-to-market and production costs without compromising quality. Simulation-based studies are crucial in the early design phase, but the extensive computational requirements of detailed vehicle models make them impractical for exploring the entire design space. Therefore, the industry persists in employing trial-and-error approaches, limiting design exploration to a few thousand variants out of billions of possibilities, thus potentially missing better solutions. A possible solution is represented by ROM technique that can be employed to allow multidimensional design exploration.

This work introduces an innovative methodology to assess the noise and vibration (NVH) performance of car-body structures. NVH simulation engineering plays a critical role in ensuring that the product meets noise and vibration criteria, thereby improving comfort, quality, and customer satisfaction. This target is highly influenced by the global static and dynamic stiffness of the vehicle body structure, making it extremely sensitive to changes in design parameters. To address this challenge, the proposed approach extends the well-known Proper Generalized Decomposition (PGD) technique to solve parametrically the dynamic and static analysis of a structure characterized by material and/or geometric design variables. The main idea is to construct a parametric FE model of the structure and then employ a PGD-based parametric solver. By utilizing the proposed method, engineers can develop predictive models capable of rapidly evaluating multiple design configurations, facilitating the identification of optimal choices that meet NVH requirements. The proposed method was developed in the context of the doctoral thesis “Static and dynamic global stiffness analysis for automotive pre-design” (more info at: https://www.lacan.upc.edu/ProTechTion/ ). The thesis was supervised by the Universitat Politècnica de Catalunya (UPC) and Swansea University (SU), in collaboration with the automotive company SEAT S.A. as industrial partner. The project is currently in a “Proof of Concept” stage at SEAT, such that it can be included in the current SEAT workflow.

I am an assistant professor (professor lector) at Universitat de Barcelona and a member of the Centre de Recerca Matemàtica (CRM-Barcelona).
I am a member of a research group in holomorphic dynamics at UB called HoloDyn.
I do research in geometry and dynamics. Some keywords are: holomorphic dynamics, hyperbolic dynamics and geodesic flows, rigidity of dynamical systems, renormalization theory, Riemannian and convex geometry, isoperimetric inequalities.

Web: https://sites.google.com/view/josecondealonso 

Web:  https://mehidi.pages.math.cnrs.fr/siteweb/ 

The impact of “Artificial Intelligence” in science and technology in recent years has been enormous, giving rise to a bubble of overhyped claims and promises that heavily influence the research landscape in fields as diverse as chemistry, medicine or energy, to name just a few. However, it is increasingly evident that there is a dissonance between what AI does and what AI gurus tell us it can do.

In this talk, we will show why AI, and more specifically Machine Learning, is a perfect tool for self-deception, even if we are extremely careful with our experimental methodology. We will discuss the main issues that can lead to overestimated performance results, and illustrate how they appear subtly in more real and “top-level” scientific studies than one would expect.

Web: https://orcid.org/0000-0002-2175-7382