About this Abstract |
| Meeting |
MS&T25: Materials Science & Technology
|
| Symposium
|
Advances in Ferrous Process Metallurgy
|
| Presentation Title |
Enhancing Property Prediction in Steel Alloys through Quantitative Microstructural Data |
| Author(s) |
Malavikha Rajivmoorthy, Patrick J. Cleaver |
| On-Site Speaker (Planned) |
Malavikha Rajivmoorthy |
| Abstract Scope |
Steel alloy development has traditionally relied on expert interpretation of microstructures via optical and electron microscopy to infer processing history and mechanical properties. This iterative approach is time-consuming and resource intensive. To accelerate optimization and discovery, we present a quantitative method for extracting key microstructural features—such as phase constituents, grain size and grain shape—from digital micrographs. Quantifying these structural attributes shows improvements in the accuracy of mechanical property predictions by establishing robust process-structure-property (P-S-P) relationships. By linking structural data with both mechanical performance and processing parameters, this method enables more predictive, data-driven alloy development. When integrated into automated workflows, it can significantly reduce iteration time and speed up the transition from innovation to commercialization. As a proof of concept, we apply this approach to high-strength low-alloy (HSLA) steels, highlighting the role of image segmentation in P-S-P linkage. We further demonstrate extensibility by applying microstructural learning to predict properties in microstructurally complex advanced high strength steels. |