Atom probe investigations on temper embrittlement and reversible temper embrittlement in S 690 steel weld metal

Abstract

Severe embrittlement was observed in weld material of a brand new penstock of a huge hydro power plant. Temper embrittlement (TE) was found as root case of embrittlement. Reversible temper embrittlement (RTE) treatment characterised by a short-time heating at about 600°C, by which the toughness of embrittled weld material can significantly be recovered, was qualified and successfully applied in the plant. Basic investigations were performed to explain the embrittlement as well as the de-embrittlement effect. By the application of high resolution analytics as Atom Probe Tomography (APT) applied on TE as well as on the RTE-treated material, revealed phosphorus segregation in the grain boundaries as root cause of embrittlement. By application of RTE treatment the APT results revealed, that the phosphorus segregation in the grain boundaries disappeared. The mechanism of this behaviour can be explained by referring the McLean [Grain boundaries in metals. Oxford: Clarendon Press; 1957] based grain boundary equilibrium segregation of phosphorous. During RTE treatment, which occur at higher temperatures (600°C) that segregation (which starts during cooling at about 550°C), desegregation occurs. During this higher temperature, the diffusion is much faster than segregation producing the fast recovery of toughness.

Community-Driven Methods for Open and Reproducible Software Tools for Analyzing Datasets from Atom Probe Microscopy

“…The test case exemplifies how one-dimensional composition profiling at a grain boundary is typically performed. The test case is based on Cerjak et al (2017). High-throughput analyzing MPIE (Max-Planck-Institut für Eisenforschung) reports how the dataset of the Leoben test case was used to develop two new open-source tools for high-throughput analyses of APT datasets.…”

“…The test case exemplifies how one-dimensional composition profiling at a grain boundary is typically performed. The test case is based on Cerjak et al (2017).…”

“…The authors filled in the template and added an exemplar analysis to one of their specimens from the Cerjak et al (2017) study. The authors submitted the associated original datasets as POS and RRNG files, respectively.…”

On Strong-Scaling and Open-Source Tools for High-Throughput Quantification of Material Point Cloud Data: Composition Gradients, Microstructural Object Reconstruction, and Spatial Correlations

“…Furthermore, we use it to present an automated method for composition profiling to support scientists with collecting reliable statistics of composition gradients across interfaces in an efficient manner. As an example, we process a reconstructed dataset of Cerjak et al [117] that was taken from a S690 steel specimen. The dataset contains 23.8 × 10 6 ions in total.…”

“…The dataset contains 23.8 × 10 6 ions in total. We work with the original reconstructed ion positions and ranging definitions of the authors [117] for which data were made public and experimental procedures explained in previous work [21]. The dataset represents a material volume which contains a partially measured carbide located at a grain boundary segment which is decorated with phosphorus as solute atoms.…”