This Website is not fully compatible with Internet Explorer.
For a more complete and secure browsing experience please consider using Microsoft Edge, Firefox, or Chrome

CAE & AI: An Overview and Field Report on Various Applications


Dr. Andreas Kuhn from ANDATA delivered a presentation titled "CAE & AI: An Overview and Field Report on Various Applications." The focus was on integrating Artificial Intelligence (AI) with Computer-Aided Engineering (CAE), exploring various applications and discussing the implications of AI in engineering. ANDATA, established in 2004 and based in Hallein, Austria, specializes in complex modeling and simulation operations. The company is a pioneer in utilizing AI and machine learning in the realm of CAE. Dr. Kuhn addressed the fundamental question of what constitutes "Intelligence" and "Artificial Intelligence," citing definitions from established AI literature, including the Turing Test. The presentation then transitioned to practical applications of AI in CAE. One of the key highlights was a problem statement on weld spot failure, where AI methods were employed to forecast models for characteristic features like maximum force. These forecasts were based on various factors such as material, load, and geometry, and were essential for inserting features into Finite Element material models. A significant part of the presentation was dedicated to the forecast of maximum weld spot force using Neural Networks. This involved extensive data collection from material combinations and single tests under various load conditions. The Neural Network model's training, validation, and test data were showcased. Dr. Kuhn emphasized the quick adaptability and learning capabilities of these models, particularly when dealing with high-strength steels. He highlighted the ease of adaptation by extending training data and the efficiency of the process. The presentation also covered the identification of best-fit material/element parameters to match validation test data using methods like evolutionary approaches for multicriteria optimization and Monte Carlo simulation for robustness/sensitivity analysis. Another significant aspect discussed was the example-based material model formulation using Artificial Neural Networks. This approach allows for a flexible and precise formulation of material/element/component behavior, which is essential for operational design domain assurance. The talk concluded with a discussion on the generalization for arbitrary anomalies detection in post-processing and the integration of data plausibilization in Simulation Data Management Systems (SDMS). Dr. Kuhn presented various examples and use-cases where AI and machine learning significantly contributed to enhancing CAE processes.

Document Details

AuthorsKuhn. A
Date 25th October 2023


Back to Previous Page