Altair physicsAI learns from “historical simulation data” to provide rapid predictions of physical behavior. This AI-enhanced simulation and design technology can now be run through HyperWorks. You can use any existing simulation data (including results from legacy designs, similar parts, or other non-Altair simulation tools) to train the model.
Using physicsAI for Fast Prediction of CAE Analysis Results
...... PhysicsAI is not constrained by parametric studies and can act directly on meshes and even CAD models. Powered by state-of-the-art geometric deep learning (GDL), physicsAI can learn the relationship between geometry and any physical response.
After training is complete, using a physicsAI model for prediction can be 10 to 100 times more efficient than traditional solver runs, enabling design and development teams to evaluate many more concepts (within the same schedule) and make better design decisions.
A key advantage of using physicsAI is that it builds fast prediction models from your CAE history data and can accept any physical data as input, even when the CAE model mesh parameters are inconsistent, or when the training data contains no defined design variables. Once training is finished, the model immediately delivers its efficiency benefits.
Altair physicsAI has been seamlessly integrated into the Altair® HyperWorks® product family, becoming a powerful tool that provides Altair users with AI-enhanced simulation and generative design capabilities.
We use cookies to collect and analyze information on site performance and usage. By Clicking "Continue" or by clicking into any content on this site, you agree to allow cookies to be placed. To find out more, please visit our privacy policy。
