Efficient Simulation Chains using Artificial Intelligence
Dr. Hannes Grillenberger
Value for the audience
This presentation gives opinions and shows possibilities on the usage of methods of artificial intelligence in CAE tools for engineering design, especially in simulation program usage.
Methods of Artificial Intelligence are more and more introduced in our daily life of program, app and internet use. This presentation gives opinions and shows possibilities on the usage and workflows of some of these methods in CAE tools.
Simulation, computer aided design and engineering are becoming more and more important and are thus their usage needs to be re-invented in the age of AI. Simulations are crucial in the current design process as they enable a huge number of design variants to be tested and optimized before starting physical tests. This reduces both development time and costs. For this, good and powerful simulation tools returning a result within reasonable simulation time are necessary. These programs are in the simulation software development focus. For example, the Bearinx Simulation Suite or the FVA Workbench are tool packages, that focus on simulations of bearings in systems. These programs are constantly expanded to better predict bearing performance – like friction, NVH, dynamics, life rating, etc.
However, for highly integrated simulations typically an expert is needed to build and verify the models, simulation chains, result generation, data processing and interpretation. In many possibly simulation driven design cases, this is the bottleneck as this takes time, needs specific simulation and automation knowledge, and these experts are rare to find. In consequence, less simulations are done which might lead to higher overall development cost or in a sufficient but not optimal design.
The acceleration and facilitation of this process leads to faster results, more simulations of variants and possibly higher design quality, and a relief in the workload for simulation experts. Large scaling effects on the use of simulation as more people can perform simulations, and the expert himself can focus on the complex situations. Methods of AI can make an important contribution to accelerate and simplify this process.