Investigation of spark plasma sintered oxide-dispersion strengthened steels by means of small-angle neutron scattering

Highlights

• SPS to consolidate ODS samples, milling parameters varied.

• Suitable milling parameters identified.

• Particles of radii between 0.5 and 3 nm are dominant in ODS samples.

• Dominant type of these particles consistent with Y₂TiO₅ and/or Y₂Ti₂O₇.

Abstract

Spark plasma sintering (SPS) is an advanced consolidation technique, which particularly allows coarsening of the microstructure to be limited. The basis for the characteristics of the strengthening nanofeatures is already set in the milling process preceding SPS. The present study is focused on the dependence of the size distribution and nature of the nanofeatures in an ODS Fe-14Cr-1W-0.4Ti alloy as a function of the applied milling parameters and amount of added Y₂O₃ while keeping the SPS parameters constant. Statistically reliable averages of the particle characteristics representative of a macroscopic sample volume have been obtained by means of small-angle neutron scattering (SANS). The measured magnetic-to-nuclear scattering ratios have been critically compared to values calculated on the basis of structures and compositions reported in the literature. Milling parameters suitable to completely transform the added Y₂O₃ into Ti-containing nm-sized oxide particles have been identified. Two size ranges of particles have been analyzed separately: 0.5–3 nm and 3–15 nm (radius). The former size range is dominant in all ODS samples, the magnetic-to-nuclear scattering ratio indicates these particles to be predominantly Y₂TiO₅ and/or Y₂Ti₂O₇.

Introduction

Oxide dispersion strengthened (ODS) steels are considered as candidate materials for components of generation IV (Gen IV) fission and fusion nuclear reactors [1], [2]. Ferritic Fe-14Cr alloys exhibit good resistance to corrosion and swelling compared to austenitic steels [3]. They are creep resistant up to 550–600 °C [4], but in order to apply these materials at higher operation temperatures (e.g. 650 °C in Gen IV fission reactors [5]), the creep properties have to be improved. The addition of ODS particles (e.g. Y₂O₃) was found to be a good option [2], [6], [7]. Besides refining the grain size of the matrix, the particles themselves have an influence on the mechanical behavior at high temperatures. Moreover, the interface between small particles (few nm) and the matrix is assumed to act as a sink for radiation-induced primary defects, such as vacancies and interstitials, and as a nucleation site for small He-agglomerations which are uncritical as opposed to large pressurized He bubbles [8], [9]. Along with the grain size, the oxide particle size distribution is an important factor with regard to the radiation resistance. Here, two factors play a role: the mean free diffusion path towards sinks and the sink capacity. The finer the dispersion of ODS particles (at constant volume fraction) is, the larger is their specific area and, simultaneously, the smaller is the mean free diffusion path towards sinks/nucleation sites.

Usually, ODS steels are produced applying a powder-metallurgy route which consists of mechanical alloying (MA) with subsequent consolidation via hot isostatic pressing (HIP) [10], [11], [12] or hot extrusion [12], [13], [14], [15] and thermal/thermomechanical treatments. Recently, MA followed by spark plasma sintering (SPS) was shown to be suitable to achieve oxide particle sizes in the lower nm range in ODS Fe-Cr alloys [12], [16], [17], [18], [19].

It has been observed that changes in the structural properties of the Y₂O₃ take place during the MA process. After milling for several hours, the intensities of the Y₂O₃ peaks in powder X-ray diffraction (XRD) patterns decrease and finally disappear [20], [21], [22], [23], [24], [25], [26]. This may be due to either complete dissolution of Y and O [23], [24] or, alternatively, to fragmentation of Y₂O₃ down to the sub-nm scale [21], [22]. The details are still under debate. Finally, the milling parameters govern the size distribution, spatial distribution and structure of oxide particles after consolidation. During sintering of the milled powders, Y₂O₃ reappears [17] or, if the matrix contains Ti, Y–Ti complex oxides form [27]. The variation of the Y₂O₃ content allows the Y/Ti ratio to be adjusted.

Consolidation by means of SPS is based on heat generation directly at the powder particle contacts. The main advantage of SPS is the shorter holding time and a limitation of grain growth [28] compared to conventional techniques such as HIP.

The nature of nanoclusters in ODS steels is still a matter of debate [18], [29], [30], [31]. Yttrium-containing oxide particles have been extensively studied by means of TEM, atom probe tomography (APT) and XRD [32], [33], [34], [35]. Zhang et al. applied TEM and APT to identify the particles in a spark plasma sintered material to be Y₂Ti₂O₇, which is known to form stable particles [36], [37]. Nagini et al. investigated the influence of milling time on the microstructure, including the particle size and composition, of a 9Cr ODS steel by means of TEM. However, the volume analyzed using these techniques is very small, which can cause problems in samples with an inhomogeneous spatial distribution of particles. In contrast, small-angle neutron scattering (SANS) provides information about the size distribution averaged over a statistically representative number of oxide particles randomly sampled from a macroscopic volume (∼50 mm³ in the present case) [38]. In existing SANS studies, several aspects of the ODS particle distribution and its role in the fabrication process are considered [39], [40].

This work is focused on the influence of milling parameters on the size distribution and composition of oxide particles in ODS Fe-14Cr alloys consolidated by means of SPS. For this purpose, SANS is applied to ODS and non-ODS samples. The SANS analysis allows bimodal particle size distributions to be explored and both particle populations to be analyzed individually. The main aim is to identify a set of milling parameters suitable to obtain a high number density of nm-size stable strengthening particles in the steel matrix. The suitability of the non-ODS samples as reference samples is examined in order to distinguish the desired strengthening particles from contamination oxides. Moreover, the ratio between the magnetic and nuclear scattering derived from SANS is critically compared to values calculated on the basis of the structure and composition of oxide particles reported in the literature.