Automated system for high-throughput process-structure-property dataset generation of structural materials: A γ/γ′ superalloy case study
We present an automated high-throughput method capable of gathering 2400 data points relating processing conditions, microstructure geometry and yield strength in just 13 days. An estimated 200 times faster than conventional methods using tensile testing specimens, a complete Process-Structure-Prope...
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2025-08-01
|
| Series: | Materials & Design |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S0264127525006999 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | We present an automated high-throughput method capable of gathering 2400 data points relating processing conditions, microstructure geometry and yield strength in just 13 days. An estimated 200 times faster than conventional methods using tensile testing specimens, a complete Process-Structure-Property (P-S-P) dataset is created from a single sample. The method is demonstrated by example of the aging heat treatment process of a γ/γ′ superalloy. By aging the sample in a temperature gradient, a wide range of aging process temperatures is mapped over the sample length. Structure analysis consists of fully automated, nanometer-resolution FE-SEM scanning, with precipitate fraction, size and shape distributions determined by automatic image analysis using the Python programming language. Mechanical properties are evaluated by nanoindentation inverse analysis, an approach combining instrumented indentation data with pile-up analysis to calculate stress/strain curves. While the necessary topographic data is typically acquired using atomic force microscopy, a significant speedup was achieved by automatic indent detection and scanning using Angular selective Backscatter FE-SEM analysis. As a method to rapidly assemble comprehensive and consistent P-S-P datasets, we expect it to facilitate efficient alloy design, given a vast majority of modeling approaches still heavily rely on empirical data. |
|---|---|
| ISSN: | 0264-1275 |