Optimal selection of engineering materials (Metals, Ceramics, Polymers, etc..) is important in manufacturing process. My basic idea is to let machine learn the materials so that we can computationally predict composite/hybrid materials property (e.g. Maximum tensile stress, Elastic modulus, Thermal expansion, Thermal/Electrical conductivity, etc.).
Approach
We could apply Machine Learning techniques in order to train artificial neural network. The overall process may look like the following.
Categorization of the materials Selection of the group of materials Raw data collection Feature engineering Training, Validation, and Testing with Neural Network Reverse engineering with optimization techniques
Expected Results
1. Providing the optimal selection of engineering materials needed for manufacturing process
2. New hybrid/composite materials design and discovery