Bjorn Clausen

Hot Isostatic Press (HIP) post-processing is widely employed to enhance the performance of additive manufacturing components. However, its impact on 316L Electron Beam Powder Bed Fusion (EB-PBF) materials remains unexplored. This study investigated the effects of HIP temperature on 316L EB-PBF materials compared to as-built (AB) and powder metallurgy (PM) counterparts. In-situ neutron diffraction during tensile testing, performed with the loading direction parallel to the build direction, along with microstructural analysis, was conducted to assess the influence of HIP temperature on dislocation evolution, stress distribution, and the resulting mechanical behavior during deformation. The results revealed variations among the different product states. Differences in microstructural features, including grain size, twinning, lattice strain, and dislocation density, were observed and linked to the mechanical response. The PM product exhibited higher stiffness (193 GPa) than the AB EB-PBF product (120 GPa) while showing similar flow strength. HIP treatment significantly influenced the stress-strain behavior of EB-PBF products, affecting dislocation density and stress distribution amongst various grain orientations. These findings provide valuable insights into the behavior of EB-PBF products subjected to HIP post-processing and highlight the differences compared to PM products.

High-energy 2D synchrotron X-ray diffractometry provides important insights into the atomistic structure and phase evolution of materials, yet traditional analysis methods remain complex, knowledge-intensive, and computationally demanding. Deep- learning models offer a powerful alternative for automating their analysis. Institutions that hold these datasets may be unwilling to share their data due to privacy and security policies, as well as the challenges associated with large-scale data transfer. As a result, models trained on local datasets often perform well only on their own data but exhibit bias and poor generalization across different instruments or facilities. To overcome these limitations, we explore federated learning for 2D synchrotron diffractograms, enabling collaborative model training without exchanging raw data. In this study, 2D synchrotron diffractograms of Ti-6Al-4V alloy collected from two independent facilities are used to train convolutional neural networks for predicting the β-phase volume fraction. Experimental results show that federated global models significantly outperform locally trained models in terms of generalization and achieve accuracy comparable to centralized trained models. These findings demonstrate the potential of federated learning to enable secure, cross-institutional collaboration and enhance the scalability of deep-learning-based materials characterization.

The U.S. Department of Energy's Office of Scientific and Technical Information