About this Abstract |
| Meeting |
2026 TMS Annual Meeting & Exhibition
|
| Symposium
|
In-Situ Monitoring and Control of Solidification & Deformation Processes in Metal Additive Manufacturing
|
| Presentation Title |
Enabling Low-Latency Synchrotron XRD Analysis for Real-Time Insights Into Driven Microstructural Evolution |
| Author(s) |
Tingkun Liu, Vinay C Amatya, Arun Devaraj |
| On-Site Speaker (Planned) |
Tingkun Liu |
| Abstract Scope |
Understanding microstructural evolution under extreme, non-equilibrium conditions, such as rapid heating and cooling in additive manufacturing or high-speed shear in solid-phase processing, is critical for tailoring material properties. Synchrotron X-ray diffraction (SXRD) enables in-situ probing of these transformations, but its potential is limited by data bottlenecks during acquisition, transfer and analysis. Automated low-latency analysis of SXRD data is essential for capturing transient phenomena and making real-time experimental decisions. Here we explored the value of a python based synchrotron XRD data analysis tool Pydidas for analyzing data with low latency and implemented it on NVIDEA jetson, laptop and GPU cluster and compared their latency and performances. In addition, we discuss the challenges of real time transfer of data from the beamline detector to these computing platforms. This work paves the way for faster analyses, enabling to capture transient phenomena and adjust experimental setups in real time. |
| Proceedings Inclusion? |
Planned: |
| Keywords |
Machine Learning, Phase Transformations, Characterization |