About this Abstract |
| Meeting |
2026 TMS Annual Meeting & Exhibition
|
| Symposium
|
Mechanical Behavior at the Nanoscale VIII
|
| Presentation Title |
Combining Nanoindentation Mapping and Machine Learning to Sense Particles Below the Surface |
| Author(s) |
Megan J. Cordill |
| On-Site Speaker (Planned) |
Megan J. Cordill |
| Abstract Scope |
Nanoindentation is a widely used material characterization technique. However, the interaction of different phases in multi-phase materials and their influence on measurement results is often only insufficiently considered. The question of how, and to what degree, hard phases embedded invisibly beneath the surface of a softer matrix affect the load-displacement curves during nanoindentation has been ignored. To address this issue, nanoindentation mapping was performed on high-speed steels composed of a soft metal matrix and hard primary carbides. The indentation experiments were complemented by a parameterized two-dimensional axisymmetric finite element model that replicated load-displacement curves for indentations with hard carbide phases located at well-defined depths below the indenter and with machine learning classifications. The combination of experiments, finite element and explainable machine learning allows for matrix-only mechanical behavior to be measured, free of hidden carbide influence that generally lead to higher values. |
| Proceedings Inclusion? |
Planned: |
| Keywords |
Machine Learning, Mechanical Properties, Modeling and Simulation |