News

In the past 30 years, the reliability of the finite element (FE) method for predicting the crash behaviour of vehicles has steadily improved. FE modeling improvements result in more and more detailed simulations with steadily increasing complexity of the data. Moreover, the growth of compute power allows larger numbers of simulations. Due to this data and complexity growth on the one hand and limited availability of engineering time on the other hand, simulation result data is often not explored in detail. Furthermore, the complexity and size of the simulation data often also prevents the direct application of standard machine learning (ML) methods on the simulation data.                                                                      In this talk I will present our recent developments in the analysis of data from numerical car crash simulations. I will cover simple tools that are in productive use at cara companies, advanced data exploration approaches as well concepts to collect and use further data from the virtual product development process.

The guest talk will take place on November 28, 2023 from 4:30 - 5:30 pm in S 106 (FAN C).