Predicting Bicycle Fork CAE Results Using AI Data Analytics Models｜physicsAI

 Software Used: physicsAI

[ Background ]

       The customer in this case is a bicycle design and manufacturing company. To accelerate the development process in the product design stage, they wanted to introduce AI big data prediction tools to assist with rapid prediction. Richin helped the customer implement Altair physicsAI, using the customer’s historical CAE simulation results as the basis to train an AI prediction model. In the early stage of product design, Altair  physicsAI is then used to predict the physical field distribution of new designs.

 

[ Results ]

• Richin used Altair physicsAI analysis software and the customer’s historical fork CAE simulation results to train the AI model.

• Leveraging our secondary development capabilities, Richin helped the customer quickly generate geometric models for various new designs and perform prediction of CAE analysis results.

• It was verified that after importing different new fork geometry models, physicsAI can rapidly predict and output structural stress contour plots within 10 seconds.

• By comparing with traditional CAE model analysis results, it was confirmed that physicsAI can successfully predict accurate structural stress contour plots.

• This enables faster and more accurate structural analysis and design optimization, thereby improving product performance and safety while accelerating time-to-market.

[ Technical Features ]

• Uses AI deep learning technology, taking historical physical field analysis results as training data to build the AI model.

• Imports new CAD geometry or FEM mesh models as input data and uses the trained AI model for prediction, quickly obtaining predicted physical field distributions of new designs without running CAE simulations.

[ YouTube – Technical Insights ]

Further reading: Product Introduction – Altair PhysicsAI