Multidimensional Frame Optimization Using AI Data Analytics｜ExpertAI

Software used : Altair ExpertAI、HyperStudy

【 Background 】

       The client in this case wanted to improve the structural stiffness of the frame and introduce AI big data analysis to accelerate the development process. With traditional optimization analysis, it is not possible to optimize for a specific stress distribution, so the resulting optimized design cannot achieve the desired deformation pattern.
       Richin assisted the client in implementing Atair ExpertAI, allowing them to consider deformation distribution first and then perform optimization, ultimately achieving a more extreme level of optimization.

【 Results 】

• Used Atair ExpertAI to perform clustering analysis on the stress distribution, and then used HyperStudy for optimization.

• Atair ExpertAI optimization compared with traditional optimization results:

1. The frame stress result was reduced by about 10 MPa.

2. The frame mass was further reduced by 20% (from 8.2 kg down to 5.6 kg).

【 Technical Features 】

• Optimization is no longer limited to a single physical quantity.

• Altair ExpertAI uses AI unsupervised-learning clustering techniques to identify stress distributions, deformation trends, and other physical quantity distribution patterns, and then performs clustering.

• By using the deformation mode or the type of physical field distribution as a constraint in the optimization setup, multi-dimensional optimization can be achieved.

【 Youtube Tech Deep Dive 】

Further reading : PCB mounting location optimization using AI big data ｜ ExpertAI