- Email: [email protected]
Microstructural comparison of Oxide Dispersion Strengthened Fe-14Cr steels produced by HIP and SPS
Fusion Engineering and Design 146 (2019) 2328–2333
Contents lists available at ScienceDirect
Fusion Engineering and Design journal homepage: www.elsevier.com/locate/fusengdes
Microstructural comparison of Oxide Dispersion Strengthened Fe-14Cr steels produced by HIP and SPS
T
David Pazosa,b, Marta Suárezc, Adolfo Fernándezc, Pilar Fernándezd, Iñigo Iturrizaa,b, ⁎ Nerea Ordása,b, a
Ceit-IK4, Paseo Manuel Lardizabal 15, 20018, Donostia-San Sebastián, Spain UNAV (Tecnun), Paseo Manuel Lardizabal 13, 20018, Donostia-San Sebastian, Spain c Nanomaterials and Nanotechnology Research Center (CINN-CSIC), Universidad de Oviedo (UO), Principado de Asturias (PA), Avda. de la Vega, 4-6, 33940, El Entrego, Spain d Ciemat, National Fusion Laboratory. Technology Division. Avda. Complutense, 40, 28040, Madrid, Spain b
ARTICLE INFO
ABSTRACT
Keywords: ODS steels Y-Ti-O oxides STARS route HIP Hot deformation SPS
Once the feasibility of the STARS route (Surface Treatment of gas Atomized powder followed by Reactive Synthesis) has been demonstrated to produce ODS ferritic steels by internal oxidation avoiding mechanical alloying, two strategies are proposed in this work to enhance precipitation of oxide nanoparticles. Firstly, consolidation by Spark Plasma Sintering (SPS) is used to produce finer microstructures compared to HIP consolidation. SPS parameters like temperature, heating rate, pressure and time of exposure under pressure were explored, and dense samples (up to 99,5%RD) were obtained. Secondly, precipitation of nanoparticles is achieved through a combination of HIP consolidation at low temperature (700–900 °C) followed by hot deformation under the presence of metastable oxides, to increase the density of dislocations, preferential nucleation sites for the precipitation of nanometric Y-Ti-O oxides. Deformation dilatometry and plane strain compression tests were performed to simulate hot deformation under different hot rolling schedules. Total deformation, number of passes, time between passes or initial and final rolling temperatures are some of the parameters explored. Microstructural characterization by Scanning Electron Microscopy (SEM) of consolidated materials is presented.
1. Introduction The extraordinary creep properties, high temperature strength, and good resistance to helium embrittlement and swelling make Oxide Dispersion Strengthened (ODS) steels an excellent option as structural materials in future fission and fusion reactors [1–5]. The reason is a very fine grain size microstructure and the presence of ultrafine (< 5 nm) complex oxides. ODS steels are commonly produced following a route that includes MA of metal powders to dissolve Y in the steel matrix, followed by hot extrusion or hot isostatic pressing (HIP) and final thermomechanical treatment. Nevertheless, MA has some drawbacks, like brittleness, long processing times, increased cost and contamination with C, N and O from grinding media and atmosphere [6,7]. The innovation of the STARS route (Surface Treatments of gas Atomized powder followed by Reactive Synthesis) proposed by Ceit to obtain ODS ferritic steels (FS) [8–12] and based in the GARS route
⁎
developed by Rieken [13,14], resides precisely in avoiding the MA. Powder with the target composition Fe-14Cr-2W-0.3Ti-0.24Y is produced by gas (Ar) atomization. During oxidation at low temperature (< 450 °C), a metastable Cr- and Fe- oxide layer grows at the surface of powder particles. During HIPping, these oxides dissociate, the oxygen available diffuses inside grains and reacts with the metallic Y and Ti, resulting in Y-Ti-O nanometric particles (NPs) precipitate. In this work, spark plasma sintering (SPS) and hot rolling are proposed as two alternatives to promote the diffusion of oxygen from prior particle boundaries (PPBs), increase the density of dislocations, preferential nucleation sites and, hence, improve the density and distribution of NPs. 2. Experimental The chemical composition of powders atomized under Ar (Table 1) was analyzed by inductively coupled plasma optical emission
Corresponding author at: Ceit-IK4, Paseo Manuel Lardizábal 15, 20018, Donostia-San Sebastián, Madrid. E-mail address: [email protected] (N. Ordás).
https://doi.org/10.1016/j.fusengdes.2019.03.182 Received 8 October 2018; Received in revised form 20 March 2019; Accepted 27 March 2019 Available online 10 April 2019 0920-3796/ © 2019 ASOCIACION CENTRO TECNOLOGICO CEIT-IK4. Published by Elsevier B.V. All rights reserved.
Fusion Engineering and Design 146 (2019) 2328–2333
D. Pazos, et al.
Table 1 Composition of as-atomized powder. Composition (wt.%)
Table 3 Samples obtained by SPS. Contaminants (ppm)
Fe
Cr
W
Ti
Y
Al
C
O
N
Bal.
14,0
2,0
0,3
0,3
150
119
169
2
spectroscopy (ICP-OES). The concentration of light elements (O, N and C, S) was obtained by the inert gas fusion principle and combustion infrared detection technique. Powders between 20 and 45 μm were oxidized to adjust the oxygen concentration in the range 1000–2000 ppm. Powders with 1000, 1500 and 2000 ppm of oxygen were consolidated by HIP at 700–900 °C and 140 MPa during 4–6 hours. Experimental simulation of hot rolling was conducted by plain strain compression on prismatic samples (60 ×30 × 13 mm3) machined from HIPped samples. Each specimen was deformed to a 70% reduction in area (RA), using 7% increments (10 passes, with 10 s as interpass time). Table 2 shows the temperature range during deformation. Powders with 2000 ppm of oxygen were consolidated by SPS, using BN as diffusion barrier to reduce carbon diffusion from graphite punches and die to the ferritic powder. Table 3 shows the parameters followed. The microstructure of powders and consolidated samples was analyzed with SEM and TEM equipped with EDS analyzers. Samples for metallographic analysis were extracted from hot deformed materials in the transverse plane (RD-ND plane). Hardness was measured with Vickers microhardness (HV1, 1 kg load) tests.
Code
Oxygen
Temp (ºC)
Press. (MPa)
Time (min)
Relative density (%)
Hardness (HV1)
1050-5 1000-5 1000-1 1000-10
2000 2000 2000 2000
1050 1000 1000 1000
50 80 80 80
5 5 1 10
95.5 98.0 98.0 98.0
142 160 167 165
Metastable Cr-rich oxides remained at PPBs, especially after HIPping at 700 °C (Fig. 2b). Inside grains, Ti remained in solid solution [11,12], and submicrometric (Y,W) intermetallic particles precipitated (see bright particles inside grains in Fig. 2a, c and d). With increasing HIP temperature up to 900 °C, dissociation of Cr2O3 from PPBs, and diffusion of Ti and incipient precipitation of Y- and Ti-rich nanometric oxides became more evident at the vicinities of PPBs (Fig. 2d). The low values of hardness of as-HIPped samples (Table 4) are associated to the low density of oxide nanoparticles and a microstructure with relatively large grain size. Fig. 3 shows the stress-strain curves recorded during plain strain compression tests. The steady increase in the stress observed during the initial passes, is associated to the temperature reduction between successive passes. From the 5th pass (above ε˜0,6) the additional increase in the stress is caused by change from recovery to work hardening behavior. As expected, deformation performed at lower temperatures (starting at 800 °C) led to higher stress values, compared to samples deformed at higher temperature (starting at 1000 °C). The stress values of sample C10-9-8, not included in Fig. 3, are significantly lower, due to the lower number of passes and lower strain rate. An increase in the microhardness measurements of 33–47% after hot deformation, compared to the as-HIPped condition (Table 4), evidenced the work hardening behavior. The values obtained, are slightly lower to those measured by Rieken in similar ODS ferritic steels produced with the GARS route after cold rolling to a slightly higher RA (80%), 271–282 HV [13]. Higher level of work hardening needs to be introduced in the HIPped material to reach common microhardness values measured on conventional ODS ferritic steels (340–400 HV) [16–18]. The microstructures developed after plain strain compression tests are significantly finer than those observed after HIP (compare Figs. 2a and 4 a). A narrower grain size distribution was measured, and the mean grain size (D50) decreased from 6,6 and 6,9 μm after HIP at 700 and 900 °C, respectively, to 2–3 μm after hot deformation (Table 4). In addition, almost complete dissolution of the oxides at PPBs was achieved after hot deformation. Only few submicrometric Ti (and Y)-
3. Results 3.1. Atomization and oxidation of powder Particles of atomized powders are spherical, with low presence of satellites (Fig. 1a). After oxidation, a metastable and continuous Fe2O3 layer develops at the surface, and Cr2O3 isolated particles grow preferentially at grain and subgrain boundaries, as previous works showed [10]. Although Y is slightly oxidized at the surface, it remains metallic in solid solution, or precipitated at grain boundaries inside larger particles (Fig. 1b and c). 3.2. HIP and hot deformation After HIP consolidation at low temperature (≤900 °C), porous samples with relative density < 95% were obtained (Fig. 2a).
Table 2 Samples obtained by HIP and hot deformation. Code
C10-9-8 C15-7-8 C15-7-10 C20-8-8 C20-8-10
O (ppm)
1000 1500 1500 2000 2000
HIP T.(ºC)
900 700 700 800 800
Hot deformation Ti (ºC)
Tf (ºC)
passes
strain
strain rate
tpass (s)
800 800 1000 800 1000
700 600 600 600 600
3 10 10 10 10
1 1,4 1,4 1,4 1,4
1,4 5 5 5 5
10 10 10 10 10
2329
Fusion Engineering and Design 146 (2019) 2328–2333
D. Pazos, et al.
Fig. 1. SEM micrographs of (a) surface of as atomized powder, (b) interior of oxidized powder with 1000 ppm of O and (c) surface of oxidized powder with 2000 ppm of O.
Fig. 2. SEM detailed microstructure of samples consolidated at low temperature. (a) and (c): powder oxidized with 1000 ppm of O and consolidated by HIP at 900 °C and 140 M P during 4 h. (b): powder oxidized with 1000 ppm of O and consolidated by HIP at 700 °C and 140 M P during 6 h. (d): powder oxidized with 1500 ppm of O and consolidated by HIP at 800 °C and 140 M P during 6 h. (e) and (f) EDS line scans from (c) and (d), respectively.
rich oxides below 200 nm remained. (Fig. 4b). Inside most grains, a large number density of fine nanoparticles (< 20 nm) were found (Fig. 4b, region inside dotted line). Their precipitation, together with the grain refinement, and the presence of a fine scale dislocation substructure [12] could explain the additional increase in yield strength in the last passes during plain strain compression. Finally, very few
submicrometric rounded particles rich in Y were found inside some grains. These particles, partly enriched in O and Ti, and in some cases also in W, correspond to the bright (Y, W) particles found after HIP at low temperature. TEM analysis after plain strain compression tests confirmed the presence of a bimodal particle size distribution (Fig. 5a). Larger 2330
Fusion Engineering and Design 146 (2019) 2328–2333
D. Pazos, et al.
Table 4 Hardness and grain size after HIP and hot deformation. Code
HIP
Plain strain compression
Grain size (μm)
C15-7-8 C15-7-10 C20-9-8 C20-9-10
HV1
D10
D50
D90
2,1
6,9
14,8
174
2,0
6,6
14,8
167
Grain size (μm)
HV1
D10
D50
D90
0,6 0,7 0,6 0,8
1,9 2,2 1,8 2,9
6,2 6 4,4 9,8
232 247 245 245
Fig. 3. Stress-strain curves of plain strain compression tests simulating hot rolling.
particles (> 50 nm), most of them round or elongated, are preferentially located at PPBs and grain boundaries, GBs. Smaller particles (< 50 nm), located inside grains, are mostly round. Most of them are < 10 nm and are not uniformly distributed in the matrix as can be derived from Fig. 5c and d. Most of the analyzed larger particles (> 50 nm) were identified as Y2O3, whereas approximately half of the finest NPs are Y, Ti-rich oxides, with variable Ti/Y ratios. No clear evidences of the effect of initial oxygen content, HIP temperature or rolling temperatures in the final microstructure were found from the parameters under study in this work. Although previous works [9] suggested that temperatures as high as 1100 °C were required to decompose the Y reservoirs and allow the precipitation of Y-Ti-O NPs, this work demonstrates that a high density of dislocations can destabilize (Y, W) particles at lower temperature by enhancing diffusion of oxygen towards the interior of the PPBs. Nucleation of oxide nanoparticles at pre-existing dislocations networks led to high densities of finely dispersed of Y-Ti-O nanoparticles. Fig. 6 shows how a high density of dislocations is associated with a higher density of NPs, while ferritic grains with few dislocations or depleted of them show a lower population of NPs. This observation suggests that dislocations (together with grain refinement) enhance the precipitation of a fine network of oxide nanoparticles.
Fig. 4. SEM micrographs after C15-7-8 compression test.
minimize grain size while ensuring full densification. The precipitation of oxide particles after SPS, with no need for additional thermal or thermomechanical treatment, was also pursued. Almost dense samples (RD ≥ 98%) were obtained by SPS. The absence of a MA step led to a lower density of dislocations, compared to SPS samples obtained with MA powder and, hence, larger ferritic grains (see Fig. 7a) with lower hardness, ranging between 145 and 160 ppm [19]. In addition, BN was effective as C diffusion barrier only for short times at highest temperature (≤5 min), and promoted N uptake. Consequently, part of the Ti in the alloy did not react with O and Y, but with C and N, forming submicrometric Ti(C,N) at PPBs (Fig. 7b). However, SEM analysis at high magnification also revealed the presence of a high density of very fine NPs (< 20 nm) (Fig. 7b and c, dotted regions). It is
3.3. SPS The aim of the preliminary activities performed on SPS, was to explore the feasibility to obtain ODS steels by SPS from powder that has not undergone mechanical alloying. SPS parameters were selected to
2331
Fusion Engineering and Design 146 (2019) 2328–2333
D. Pazos, et al.
Fig. 5. Representative STEM micrographs of sample C15-7-10 showing a) the microstructure observing the elongated grains, b) distribution of the oxides within the grains, and c–d) density of particles variation.
expected that by a proper selection of an inert C diffusion barrier more Ti will be available during SPS consolidation to form Y-Ti-O NPs. 4. Conclusions ODS Ferritic Steels with composition Fe-14Cr-2W-0.3Ti-0.3Y2O3 (wt.%) have been produced with the STARS route, using HIP and hot rolling or SPS to consolidate the powders. The effect of processing parameters has been investigated to understand the mechanisms involved in the development of the dispersion of Y-Ti-O NPs. The STARS route enables a reduction of the HIP temperature, compared to the conventional route to produce ODS steels. The metastable Fe2O3 and Cr2O3 nanolayer grown at powder particles surface acts as O reservoir and dissociates at high temperature to form a nanometric oxide dispersion. The combination of low temperature consolidation (< 900 °C) followed by TMTs to increase the density of nucleation sites for NPs in the STARS route, is effective in refining the microstructure and enhance the precipitation of Y-Ti-O NPs. SPS can be a promising tool to consolidate ODS steels from powder already containing Y provided contamination that consumes ODS formers is avoided.
Fig. 6. STEM micrograph of sample C10-9-8.
2332
Fusion Engineering and Design 146 (2019) 2328–2333
D. Pazos, et al.
Acknowledgments Part of this work has been carried out within the framework of the EUROfusion Consortium and has received funding from the Euratom research and training programme 2014-2018 under grant agreement No. 633053. The views and opinions expressed herein do not necessarily reflect those of the European Commission. Authors acknowledge Dr. Uranga, Dr. Isasti and J. Urbieta for their help with hot deformation experiments. References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [16] [17] [18] [19]
Fig. 7. SEM micrographs of 1000-5 sample obtained by SPS.
2333
D.T. Hoelzer, et al., J. Nucl. Mater. 367–370 (2007) 166. N. Baluc, et al., J. Nucl. Mater. 417 (2011) 149. G.R. Odette, D.T. Hoelzer, J. Miner. Met. Mater. Soc. 66 (2014) 2427. S.J. Zinkle, et al., Nucl. Eng. 57 (2017) 092005. E. Aydogan, et al., J. Nucl. Mater. 486 (2017) 86. C. Suryanarayana, Progress in Mats, Science 46 (2001) 1. G.R. Romanoski, et al., J. Nucl. Mater. 283–287 (2000) 642. E. Gil, et al., Fusion Eng. Des. 98–99 (2015) 1973. E. Gil, et al., Powder Metall. 59 (2016) 359. E. Gil, et al., Appl. Surf. Sci. 427 (2018) 182. N. Ordás, et al., J. Nucl. Mater. 504 (2018) 8. D. Pazos, et al., Nucl. Mater. Energy 17 (2018) 1. J.R. Rieken, Ph. D. Thesis, Iowa State University (2011). J.R. Rieken, et al., J. Nucl. Mater. 428 (2012) 65. Z. Oksiuta, et al., J. Nucl. Mater. 386–388 (2009) 26. Z. Oksiuta, et al., Fusion Eng. Des. 89 (2014) 137. P. He, et al., J. Nucl. Mater 428 (2012) 131. X. Boulnat, et al., Powder Metall. 57 (2014) 204.
Contents lists available at ScienceDirect
Fusion Engineering and Design journal homepage: www.elsevier.com/locate/fusengdes
Microstructural comparison of Oxide Dispersion Strengthened Fe-14Cr steels produced by HIP and SPS
T
David Pazosa,b, Marta Suárezc, Adolfo Fernándezc, Pilar Fernándezd, Iñigo Iturrizaa,b, ⁎ Nerea Ordása,b, a
Ceit-IK4, Paseo Manuel Lardizabal 15, 20018, Donostia-San Sebastián, Spain UNAV (Tecnun), Paseo Manuel Lardizabal 13, 20018, Donostia-San Sebastian, Spain c Nanomaterials and Nanotechnology Research Center (CINN-CSIC), Universidad de Oviedo (UO), Principado de Asturias (PA), Avda. de la Vega, 4-6, 33940, El Entrego, Spain d Ciemat, National Fusion Laboratory. Technology Division. Avda. Complutense, 40, 28040, Madrid, Spain b
ARTICLE INFO
ABSTRACT
Keywords: ODS steels Y-Ti-O oxides STARS route HIP Hot deformation SPS
Once the feasibility of the STARS route (Surface Treatment of gas Atomized powder followed by Reactive Synthesis) has been demonstrated to produce ODS ferritic steels by internal oxidation avoiding mechanical alloying, two strategies are proposed in this work to enhance precipitation of oxide nanoparticles. Firstly, consolidation by Spark Plasma Sintering (SPS) is used to produce finer microstructures compared to HIP consolidation. SPS parameters like temperature, heating rate, pressure and time of exposure under pressure were explored, and dense samples (up to 99,5%RD) were obtained. Secondly, precipitation of nanoparticles is achieved through a combination of HIP consolidation at low temperature (700–900 °C) followed by hot deformation under the presence of metastable oxides, to increase the density of dislocations, preferential nucleation sites for the precipitation of nanometric Y-Ti-O oxides. Deformation dilatometry and plane strain compression tests were performed to simulate hot deformation under different hot rolling schedules. Total deformation, number of passes, time between passes or initial and final rolling temperatures are some of the parameters explored. Microstructural characterization by Scanning Electron Microscopy (SEM) of consolidated materials is presented.
1. Introduction The extraordinary creep properties, high temperature strength, and good resistance to helium embrittlement and swelling make Oxide Dispersion Strengthened (ODS) steels an excellent option as structural materials in future fission and fusion reactors [1–5]. The reason is a very fine grain size microstructure and the presence of ultrafine (< 5 nm) complex oxides. ODS steels are commonly produced following a route that includes MA of metal powders to dissolve Y in the steel matrix, followed by hot extrusion or hot isostatic pressing (HIP) and final thermomechanical treatment. Nevertheless, MA has some drawbacks, like brittleness, long processing times, increased cost and contamination with C, N and O from grinding media and atmosphere [6,7]. The innovation of the STARS route (Surface Treatments of gas Atomized powder followed by Reactive Synthesis) proposed by Ceit to obtain ODS ferritic steels (FS) [8–12] and based in the GARS route
⁎
developed by Rieken [13,14], resides precisely in avoiding the MA. Powder with the target composition Fe-14Cr-2W-0.3Ti-0.24Y is produced by gas (Ar) atomization. During oxidation at low temperature (< 450 °C), a metastable Cr- and Fe- oxide layer grows at the surface of powder particles. During HIPping, these oxides dissociate, the oxygen available diffuses inside grains and reacts with the metallic Y and Ti, resulting in Y-Ti-O nanometric particles (NPs) precipitate. In this work, spark plasma sintering (SPS) and hot rolling are proposed as two alternatives to promote the diffusion of oxygen from prior particle boundaries (PPBs), increase the density of dislocations, preferential nucleation sites and, hence, improve the density and distribution of NPs. 2. Experimental The chemical composition of powders atomized under Ar (Table 1) was analyzed by inductively coupled plasma optical emission
Corresponding author at: Ceit-IK4, Paseo Manuel Lardizábal 15, 20018, Donostia-San Sebastián, Madrid. E-mail address: [email protected] (N. Ordás).
https://doi.org/10.1016/j.fusengdes.2019.03.182 Received 8 October 2018; Received in revised form 20 March 2019; Accepted 27 March 2019 Available online 10 April 2019 0920-3796/ © 2019 ASOCIACION CENTRO TECNOLOGICO CEIT-IK4. Published by Elsevier B.V. All rights reserved.
Fusion Engineering and Design 146 (2019) 2328–2333
D. Pazos, et al.
Table 1 Composition of as-atomized powder. Composition (wt.%)
Table 3 Samples obtained by SPS. Contaminants (ppm)
Fe
Cr
W
Ti
Y
Al
C
O
N
Bal.
14,0
2,0
0,3
0,3
150
119
169
2
spectroscopy (ICP-OES). The concentration of light elements (O, N and C, S) was obtained by the inert gas fusion principle and combustion infrared detection technique. Powders between 20 and 45 μm were oxidized to adjust the oxygen concentration in the range 1000–2000 ppm. Powders with 1000, 1500 and 2000 ppm of oxygen were consolidated by HIP at 700–900 °C and 140 MPa during 4–6 hours. Experimental simulation of hot rolling was conducted by plain strain compression on prismatic samples (60 ×30 × 13 mm3) machined from HIPped samples. Each specimen was deformed to a 70% reduction in area (RA), using 7% increments (10 passes, with 10 s as interpass time). Table 2 shows the temperature range during deformation. Powders with 2000 ppm of oxygen were consolidated by SPS, using BN as diffusion barrier to reduce carbon diffusion from graphite punches and die to the ferritic powder. Table 3 shows the parameters followed. The microstructure of powders and consolidated samples was analyzed with SEM and TEM equipped with EDS analyzers. Samples for metallographic analysis were extracted from hot deformed materials in the transverse plane (RD-ND plane). Hardness was measured with Vickers microhardness (HV1, 1 kg load) tests.
Code
Oxygen
Temp (ºC)
Press. (MPa)
Time (min)
Relative density (%)
Hardness (HV1)
1050-5 1000-5 1000-1 1000-10
2000 2000 2000 2000
1050 1000 1000 1000
50 80 80 80
5 5 1 10
95.5 98.0 98.0 98.0
142 160 167 165
Metastable Cr-rich oxides remained at PPBs, especially after HIPping at 700 °C (Fig. 2b). Inside grains, Ti remained in solid solution [11,12], and submicrometric (Y,W) intermetallic particles precipitated (see bright particles inside grains in Fig. 2a, c and d). With increasing HIP temperature up to 900 °C, dissociation of Cr2O3 from PPBs, and diffusion of Ti and incipient precipitation of Y- and Ti-rich nanometric oxides became more evident at the vicinities of PPBs (Fig. 2d). The low values of hardness of as-HIPped samples (Table 4) are associated to the low density of oxide nanoparticles and a microstructure with relatively large grain size. Fig. 3 shows the stress-strain curves recorded during plain strain compression tests. The steady increase in the stress observed during the initial passes, is associated to the temperature reduction between successive passes. From the 5th pass (above ε˜0,6) the additional increase in the stress is caused by change from recovery to work hardening behavior. As expected, deformation performed at lower temperatures (starting at 800 °C) led to higher stress values, compared to samples deformed at higher temperature (starting at 1000 °C). The stress values of sample C10-9-8, not included in Fig. 3, are significantly lower, due to the lower number of passes and lower strain rate. An increase in the microhardness measurements of 33–47% after hot deformation, compared to the as-HIPped condition (Table 4), evidenced the work hardening behavior. The values obtained, are slightly lower to those measured by Rieken in similar ODS ferritic steels produced with the GARS route after cold rolling to a slightly higher RA (80%), 271–282 HV [13]. Higher level of work hardening needs to be introduced in the HIPped material to reach common microhardness values measured on conventional ODS ferritic steels (340–400 HV) [16–18]. The microstructures developed after plain strain compression tests are significantly finer than those observed after HIP (compare Figs. 2a and 4 a). A narrower grain size distribution was measured, and the mean grain size (D50) decreased from 6,6 and 6,9 μm after HIP at 700 and 900 °C, respectively, to 2–3 μm after hot deformation (Table 4). In addition, almost complete dissolution of the oxides at PPBs was achieved after hot deformation. Only few submicrometric Ti (and Y)-
3. Results 3.1. Atomization and oxidation of powder Particles of atomized powders are spherical, with low presence of satellites (Fig. 1a). After oxidation, a metastable and continuous Fe2O3 layer develops at the surface, and Cr2O3 isolated particles grow preferentially at grain and subgrain boundaries, as previous works showed [10]. Although Y is slightly oxidized at the surface, it remains metallic in solid solution, or precipitated at grain boundaries inside larger particles (Fig. 1b and c). 3.2. HIP and hot deformation After HIP consolidation at low temperature (≤900 °C), porous samples with relative density < 95% were obtained (Fig. 2a).
Table 2 Samples obtained by HIP and hot deformation. Code
C10-9-8 C15-7-8 C15-7-10 C20-8-8 C20-8-10
O (ppm)
1000 1500 1500 2000 2000
HIP T.(ºC)
900 700 700 800 800
Hot deformation Ti (ºC)
Tf (ºC)
passes
strain
strain rate
tpass (s)
800 800 1000 800 1000
700 600 600 600 600
3 10 10 10 10
1 1,4 1,4 1,4 1,4
1,4 5 5 5 5
10 10 10 10 10
2329
Fusion Engineering and Design 146 (2019) 2328–2333
D. Pazos, et al.
Fig. 1. SEM micrographs of (a) surface of as atomized powder, (b) interior of oxidized powder with 1000 ppm of O and (c) surface of oxidized powder with 2000 ppm of O.
Fig. 2. SEM detailed microstructure of samples consolidated at low temperature. (a) and (c): powder oxidized with 1000 ppm of O and consolidated by HIP at 900 °C and 140 M P during 4 h. (b): powder oxidized with 1000 ppm of O and consolidated by HIP at 700 °C and 140 M P during 6 h. (d): powder oxidized with 1500 ppm of O and consolidated by HIP at 800 °C and 140 M P during 6 h. (e) and (f) EDS line scans from (c) and (d), respectively.
rich oxides below 200 nm remained. (Fig. 4b). Inside most grains, a large number density of fine nanoparticles (< 20 nm) were found (Fig. 4b, region inside dotted line). Their precipitation, together with the grain refinement, and the presence of a fine scale dislocation substructure [12] could explain the additional increase in yield strength in the last passes during plain strain compression. Finally, very few
submicrometric rounded particles rich in Y were found inside some grains. These particles, partly enriched in O and Ti, and in some cases also in W, correspond to the bright (Y, W) particles found after HIP at low temperature. TEM analysis after plain strain compression tests confirmed the presence of a bimodal particle size distribution (Fig. 5a). Larger 2330
Fusion Engineering and Design 146 (2019) 2328–2333
D. Pazos, et al.
Table 4 Hardness and grain size after HIP and hot deformation. Code
HIP
Plain strain compression
Grain size (μm)
C15-7-8 C15-7-10 C20-9-8 C20-9-10
HV1
D10
D50
D90
2,1
6,9
14,8
174
2,0
6,6
14,8
167
Grain size (μm)
HV1
D10
D50
D90
0,6 0,7 0,6 0,8
1,9 2,2 1,8 2,9
6,2 6 4,4 9,8
232 247 245 245
Fig. 3. Stress-strain curves of plain strain compression tests simulating hot rolling.
particles (> 50 nm), most of them round or elongated, are preferentially located at PPBs and grain boundaries, GBs. Smaller particles (< 50 nm), located inside grains, are mostly round. Most of them are < 10 nm and are not uniformly distributed in the matrix as can be derived from Fig. 5c and d. Most of the analyzed larger particles (> 50 nm) were identified as Y2O3, whereas approximately half of the finest NPs are Y, Ti-rich oxides, with variable Ti/Y ratios. No clear evidences of the effect of initial oxygen content, HIP temperature or rolling temperatures in the final microstructure were found from the parameters under study in this work. Although previous works [9] suggested that temperatures as high as 1100 °C were required to decompose the Y reservoirs and allow the precipitation of Y-Ti-O NPs, this work demonstrates that a high density of dislocations can destabilize (Y, W) particles at lower temperature by enhancing diffusion of oxygen towards the interior of the PPBs. Nucleation of oxide nanoparticles at pre-existing dislocations networks led to high densities of finely dispersed of Y-Ti-O nanoparticles. Fig. 6 shows how a high density of dislocations is associated with a higher density of NPs, while ferritic grains with few dislocations or depleted of them show a lower population of NPs. This observation suggests that dislocations (together with grain refinement) enhance the precipitation of a fine network of oxide nanoparticles.
Fig. 4. SEM micrographs after C15-7-8 compression test.
minimize grain size while ensuring full densification. The precipitation of oxide particles after SPS, with no need for additional thermal or thermomechanical treatment, was also pursued. Almost dense samples (RD ≥ 98%) were obtained by SPS. The absence of a MA step led to a lower density of dislocations, compared to SPS samples obtained with MA powder and, hence, larger ferritic grains (see Fig. 7a) with lower hardness, ranging between 145 and 160 ppm [19]. In addition, BN was effective as C diffusion barrier only for short times at highest temperature (≤5 min), and promoted N uptake. Consequently, part of the Ti in the alloy did not react with O and Y, but with C and N, forming submicrometric Ti(C,N) at PPBs (Fig. 7b). However, SEM analysis at high magnification also revealed the presence of a high density of very fine NPs (< 20 nm) (Fig. 7b and c, dotted regions). It is
3.3. SPS The aim of the preliminary activities performed on SPS, was to explore the feasibility to obtain ODS steels by SPS from powder that has not undergone mechanical alloying. SPS parameters were selected to
2331
Fusion Engineering and Design 146 (2019) 2328–2333
D. Pazos, et al.
Fig. 5. Representative STEM micrographs of sample C15-7-10 showing a) the microstructure observing the elongated grains, b) distribution of the oxides within the grains, and c–d) density of particles variation.
expected that by a proper selection of an inert C diffusion barrier more Ti will be available during SPS consolidation to form Y-Ti-O NPs. 4. Conclusions ODS Ferritic Steels with composition Fe-14Cr-2W-0.3Ti-0.3Y2O3 (wt.%) have been produced with the STARS route, using HIP and hot rolling or SPS to consolidate the powders. The effect of processing parameters has been investigated to understand the mechanisms involved in the development of the dispersion of Y-Ti-O NPs. The STARS route enables a reduction of the HIP temperature, compared to the conventional route to produce ODS steels. The metastable Fe2O3 and Cr2O3 nanolayer grown at powder particles surface acts as O reservoir and dissociates at high temperature to form a nanometric oxide dispersion. The combination of low temperature consolidation (< 900 °C) followed by TMTs to increase the density of nucleation sites for NPs in the STARS route, is effective in refining the microstructure and enhance the precipitation of Y-Ti-O NPs. SPS can be a promising tool to consolidate ODS steels from powder already containing Y provided contamination that consumes ODS formers is avoided.
Fig. 6. STEM micrograph of sample C10-9-8.
2332
Fusion Engineering and Design 146 (2019) 2328–2333
D. Pazos, et al.
Acknowledgments Part of this work has been carried out within the framework of the EUROfusion Consortium and has received funding from the Euratom research and training programme 2014-2018 under grant agreement No. 633053. The views and opinions expressed herein do not necessarily reflect those of the European Commission. Authors acknowledge Dr. Uranga, Dr. Isasti and J. Urbieta for their help with hot deformation experiments. References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [16] [17] [18] [19]
Fig. 7. SEM micrographs of 1000-5 sample obtained by SPS.
2333
D.T. Hoelzer, et al., J. Nucl. Mater. 367–370 (2007) 166. N. Baluc, et al., J. Nucl. Mater. 417 (2011) 149. G.R. Odette, D.T. Hoelzer, J. Miner. Met. Mater. Soc. 66 (2014) 2427. S.J. Zinkle, et al., Nucl. Eng. 57 (2017) 092005. E. Aydogan, et al., J. Nucl. Mater. 486 (2017) 86. C. Suryanarayana, Progress in Mats, Science 46 (2001) 1. G.R. Romanoski, et al., J. Nucl. Mater. 283–287 (2000) 642. E. Gil, et al., Fusion Eng. Des. 98–99 (2015) 1973. E. Gil, et al., Powder Metall. 59 (2016) 359. E. Gil, et al., Appl. Surf. Sci. 427 (2018) 182. N. Ordás, et al., J. Nucl. Mater. 504 (2018) 8. D. Pazos, et al., Nucl. Mater. Energy 17 (2018) 1. J.R. Rieken, Ph. D. Thesis, Iowa State University (2011). J.R. Rieken, et al., J. Nucl. Mater. 428 (2012) 65. Z. Oksiuta, et al., J. Nucl. Mater. 386–388 (2009) 26. Z. Oksiuta, et al., Fusion Eng. Des. 89 (2014) 137. P. He, et al., J. Nucl. Mater 428 (2012) 131. X. Boulnat, et al., Powder Metall. 57 (2014) 204.