| About this Abstract |
| Meeting |
MS&T24: Materials Science & Technology
|
| Symposium
|
Materials Processing and Fundamental Understanding Based on Machine Learning and Data Informatics
|
| Presentation Title |
High-throughput, Ultra-fast Laser Sintering of Ceramics and Machine-learning-Based Prediction on Processing-Microstructure-Property Relationships |
| Author(s) |
Xiao Geng, Jianan Tang, Ningxuan Wen, Siddhartha Sarkar, Rajendra K Bordia, Jianhua Tong, Dongsheng Li, Hai Xiao, Fei Peng |
| On-Site Speaker (Planned) |
Fei Peng |
| Abstract Scope |
We report (1) high-throughput, ultra-fast laser sintering of alumina as a tool to efficiently establish large quantities of data on the processing-microstructure-property relations and (2) machine learning approaches to predict such relations. Using ultra-fast laser sintering, we fabricated an alumina sample array that contains ~90 individual sample units, in one laser scan. A microstructure-sensitive property, hardness, of the units in a large sample array was measured using micro-indentation. We developed machine learning (ML) algorithms to predict (1) the microstructure under arbitrary laser power, (2) the ceramic's microstructure based on laser spot brightness during laser sintering, as an online monitoring tool, (3) the microstructure for a target hardness. We also demonstrated a reinforcement learning (RL)-based approach to efficiently design a material's structure for target thermal properties. The RL approach demonstrated the efficiency advantage in achieving multiple goals simultaneously without exhausting all the design parameters (over 1 million possible parameter combinations) |