Microstructure and mechanical behavior of neutron irradiated ultrafinegrained ferritic steel

Ahmad Alsabbagh a,n, Apu Sarkar a, Brandon Miller b, Jatuporn Burns c, Leah Squires b,Douglas Porter b, James I. Cole b, K.L. Murty a

a Department of Nuclear Engineering, North Carolina State University, Raleigh, NC 27695, USAb ATR National Scientific User Facility, Idaho National Laboratory, Idaho Falls, ID 83415, USAc Center for Advanced Energy Studies, Idaho Falls, ID 83401, USA

a r t i c l e i n f o

Article history:Received 9 May 2014Received in revised form3 July 2014Accepted 22 July 2014Available online 30 July 2014

Keywords:Neutron irradiation effectsUltrafine-grained materialsLow carbon steelPrecipitate strengtheningDislocation forest hardening

a b s t r a c t

Neutron irradiation effects on ultra-fine grain (UFG) low carbon steel prepared by equal channel angularpressing (ECAP) have been examined. Counterpart samples with conventional grain (CG) sizes have beenirradiated alongside with the UFG ones for comparison. Samples were irradiated in the Advanced TestReactor (ATR) at Idaho National Laboratory (INL) to 1.37 dpa. Atom probe tomography revealedmanganese and silicon-enriched clusters in both UFG and CG steel after neutron irradiation. Mechanicalproperties were characterized using microhardness and tensile tests, and irradiation of UFG carbon steelrevealed minute radiation effects in contrast to the distinct radiation hardening and reduction ofductility in its CG counterpart. After irradiation, micro hardness indicated increases of around 9% for UFGversus 62% for CG steel. Similarly, tensile strength revealed increases of 8% and 94% respectively for UFGand CG steels while corresponding decreases in ductility were 56% versus 82%. X-ray quantitativeanalysis showed that dislocation density in CG increased after irradiation while no significant changewas observed in UFG steel, revealing better radiation tolerance. Quantitative correlations betweenexperimental results and modeling were demonstrated based on irradiation induced precipitatestrengthening and dislocation forest hardening mechanisms.

Published by Elsevier B.V.

1. Introduction

Advanced nuclear reactors employ higher temperature environ-ments and require higher irradiation exposures for core structuralcomponents and fuel cladding compared to the current operating lightwater reactors (LWR) [1,2]. Thus, development of new materials withsuperior properties which can withstand such severe conditions isrequired to fill these needs [3–5]. Ferritic steels have been used widelyas structural materials in light water reactors, but they suffer fromirradiation hardening and embrittlement accompanied by increasedductile to brittle transition temperature (DBTT) and decreased uppershelf energy [6]. Previous studies have shown that major factorsaffecting low carbon steels under neutron irradiation are neutronfluence, irradiation temperature and chemical composition [7]. How-ever, designing materials with tailored response that can sustain highamounts of radiation damage while maintaining their mechanicalproperties is a grand challenge in materials research [8]. Duringirradiation, point defects (vacancies and interstitials) are produced as

a result of displacement cascade [9–11]. These point defects can clusterto form other types of defects that will alter the mechanical propertiesof irradiated materials. One method to suppress accumulation ofthese point defects is by annihilating them at interfaces such asgrain boundaries. Ultra-fine grained materials are expected to be moreradiation tolerant since grain refinement increases the area of grainboundaries which can act as sinks for radiation induced pointdefects.

It has been theorized that the large amount of grain boundary areawill help to prevent accumulation of defects which can adverselyaffect mechanical properties [12–20]. Singh [13] showed that voidswelling in electron irradiated helium doped stainless steel decreasesas the grain size decreases. Rose et al. [21] illustrated using TEM thatthe defect density in ZrO2 irradiated by Kr ions reduces as the grainsize decreases. Matsouka et al. [22] studied the effects of neutronirradiation on UFG SUS316L stainless steel and their TEM observationsrevealed defect-free zones along grain boundaries suggesting that thegrain boundaries are acting as sinks for radiation induced defects.Kurishita et al. [23] showed that the density of voids in UFG W-0.5 wt%TiC is much lower than that in CG tungsten after neutron irradiationat 600 1C. Sun et al. [24] found that dislocation loops and He bubbledensities in UFG Fe–Cr–Ni alloy after He ion irradiation are less than

Contents lists available at ScienceDirect

journal homepage: www.elsevier.com/locate/msea

Materials Science & Engineering A

http://dx.doi.org/10.1016/j.msea.2014.07.0700921-5093/Published by Elsevier B.V.

n Corresponding author.E-mail address: ahalsabb@ncsu.edu (A. Alsabbagh).

Materials Science & Engineering A 615 (2014) 128–138

those in its CG counterpart. Computer simulation studies [14,25–27]demonstrated that materials with large surface area of interfaces orgrain boundaries have a potential to increase irradiation resistance.The effect of a small grain size (large grain boundary area density) onradiation tolerance of a low-carbon ferritic steel is assessed inthis study.

2. Materials and methods

2.1. Materials

Two material cases are considered in this study; ultra-fine grainlow carbon steel with a composition of 0.1 C, 0.5 Mn, 0.27 Si withbalance of Fe in wt% processed through ECAP [28–30] and theirconventional coarse grain counterparts that were produced byannealing the UFG material at 800 1C for one hour [12]. UFG steelwas made by ECAP using Bc route [31] with four passes where thematerial is rotated 901 in the same direction after each pass. Thisprocessing route ensures eventual restoration of the material cubicelement [32]. More details about ECAP routes can be foundelsewhere [33,34].

2.2. Irradiation experiment

The irradiation experiment was carried out in the E-7 positionof the East Flux Trap (Fig. 1b [35]) in the Advanced Test Reactor(ATR) at Idaho National Laboratory. The materials were irradiatedto a neutron fluence of 1.78�1025 neutrons/m2 (E40.1 MeV)corresponding to a dose of 1.37 dpa. Fig. 1a is a schematic of theirradiation test assembly consisting of the experimental basket,support rod and capsule assemblies. The support rod was inserted

at the bottom of the experimental basket to ensure that the testcapsules are at the maximum flux location. The experimentalbasket of the test assembly is an aluminum tube designed forinsertion in the capsule assembly in the ATR. The basket wasdesigned such that there is adequate coolant circulation to preventtemperature distortions or mechanical effects, and that there isadequate mechanical support to secure the test capsule through-out the reactor insertion, irradiation, and removal from the reactor.Samples were cut from the bulk ECAP material and were preparedby grinding with a series of silicon carbide papers (600, 800 and1200 grits) to optical flatness and then polished in colloidal silicaresulting in deformation free surfaces. The prepared samples (UFGand CG steels) were loaded in sample holders (Fig. 2a) and a thinaluminum disc was tack-welded to the open end of each holder toposition the samples inside it. Each group of sample holders wasstrung together using aluminum rods and was assembled into asample train which was designed to hold the samples for easierremoval following irradiation and also to help keeping the samplesat the desired irradiation temperature (Fig. 2b, c and d). Finally, thesample trains were sealed in a stainless steel containment capsulethat is essentially a sealed pressure vessel filled with helium(Fig. 2e). The integrity of the capsules was ensured using heliumleak testing, dye penetrant testing and visual inspection.

Thermal analysis was performed using a detailed finite elementmodel of the experiment using ABAQUS code [36]. MCNP code [37]was used to calculate the heat generation rate for each part of theexperiment which was then used as an input to the finite elementmodel (Fig. 3). The specimen temperature during irradiation wasfound to vary between 70 1C to less than and 100 1C depending onthe sample position in the irradiation capsule. The low irradiationtemperature is mainly due to the small helium gas gaps betweenspecimens and holder, and between holder and capsule.

Fig. 1. Irradiation Test Assembly for ATR East Flux Trap Position (a) and Radial Cross Section View of the ATR Reactor Core, E-7 Irradiation Test Position (b) [35].

A. Alsabbagh et al. / Materials Science & Engineering A 615 (2014) 128–138 129

2.3. Post irradiation examination

The grain size distributions for UFG steel were measured usingtransmission electron microscopy (TEM) while electron back scattereddiffraction (EBSD) technique was employed for CG steel; differenttechniques were used due to the differences in the mean grain sizesand an average of about 400 grains was used to estimate the averagegrain size. TEM studies were performed using a Tecnai TF30-FEGscanning transmission electron microscope operating at 300 keV. TheEBSD measurements were conducted on a Quanta 3D FEG scanningelectron microscope, and the crystallographic orientation mappingswere made for the CG steel using a step size of 0.5 μm in an area of100 μm�100 μm. In addition, the microstructures of the low carbonsteel for both grain sizes (CG and UFG) were characterized via X-raydiffraction (XRD) and atom probe tomography (APT) before and afterirradiation. XRD was performed using PANalytical diffractometer withCuKα radiation (wavelength 0.15406 nm). All the diffraction profileswere obtained by varying 2θ from 401 to 1201 with a scan step of0.0261 and the samples were rotated at 4 s/rev. APT was conductedusing CAMECA LEAP 4000X HR instrument. The analyses were madeusing voltage pulsing (20% pulse fraction) at a specimen temperatureof 50 K. The atom maps were analyzed employing IVAS v3.6.6(CAMECA Instrument) software. The APT specimens were needleshaped with an apex radius of �30 nm using focused ion beam(FIB), which was carried out using a Quanta 3D field emission gun(FEG) with a Ga ion source. Due to the irradiation embrittlement, someatom probe specimens fractured under the high stresses applied duringthe test. Thus, six tips were prepared from each sample and 3 tips fromeach sample were analyzed to maintain good statistics.

Irradiation induced hardening of the steel samples was quantifiedusing both microhardness and tensile tests. Micro hardness measure-ments were carried out using Leco LM247AT Vickers micro hardnesstester at a load of 500 g-f (�5 N). Twelve indents were made on eachsample with a dwell time of 13 s and their average value is reported.The indents were 300 mm apart to avoid any influence from theprevious indents on the hardness value. Tensile tests were performedon the unirradiated and irradiated samples using a closed loop Instron5967 machine with 5 kN load cell. The tensile tests were carried out ata cross-head speed of 2�10�3 mm/s (strain rate of 1�10�3 s�1) and

three tests were performed for each condition. In order to minimizeinduced radioactivity, relatively small tensile specimens (2 mm gagelength) were used in this study, and special tensile grips weredesigned for testing the miniature tensile samples (Fig. 4).

3. Results

Fig. 5 shows the microstructures of both UFG and CG steels beforeirradiation and the mean grain sizes for the UFG and CG steels are0.3570.18 and 4.471.8 μm, respectively. The XRD peaks (Fig. 6) showthat there are no phase changes due to irradiation in both steels. Allthe X-ray diffraction peaks were fitted simultaneously using ModifiedRietveld technique with suitable weightage by a pseudo-Voigt (pV)function using LS1 program [38]. This program uses the ModifiedRietveld technique in its analysis and includes the simultaneousrefinement of the crystal structure and the microstructural parameterslike the domain size and the microstrain within the domain. Themethod involves Fourier analysis of the broadened peaks. Consideringan isotropic model, the lattice parameters (a), surface weightedaverage domain size (Ds) and the average microstrain ⟨ε2L ⟩

1=2 wereused simultaneously as the fitting parameters to obtain the best fit[39]. Fig. 7 represents a typical whole pattern fit using LS1 program forirradiated CG steel. Using the results obtained from the XRD analysis,the dislocation density, ρ has been estimated from the followingrelations [40],

ρ¼ ðρDρSÞ1=2ρD ¼ 3

D2s

ρS ¼ k⟨ε2L ⟩=b2

9>>>=>>>;; ð1Þ

where ρD and ρS are the dislocation densities due to domain size andstrain respectively, k is material constant equal to 14.4 for bcc metals[40]. X-ray analysis results are summarized in Table 1.

The reconstructed three dimensional atom maps of Si, Mn andC are shown in Figs. 8 and 9 for UFG and CG steels, respectively. InUFG steel, the grain boundaries were observed in both unirra-diated and irradiated samples by solute enrichment that can onlybe attributed to segregation at grain boundaries. In CG steel, the

Fig. 3. Sample irradiation temperature calculated by a detailed finite element model of the experiment using ABAQUS code.

Fig. 2. Sample loading into the aluminum holders (a). Loaded sample holders before being strung on the Al wire and loaded into the stainless steel capsules (b). Sampleholders strung together with two fine aluminum wires (c). Sample holder stack in the test train assembly before loading into the stainless steel capsules; the Al wires haveend beads to secure the stack (d). Test train assembly loaded into the experimental capsule tube (e).

A. Alsabbagh et al. / Materials Science & Engineering A 615 (2014) 128–138130

average total length of the matter analyzed is about 150 nm. Sincethe mean grain size for CG steel is about 4.4 μm, no grain boundarywas observed in these volumes.

Solute atom analysis was performed using statistical methods todetect deviations from a random solid solution. A maximum separa-tion method was used to identify the clusters in the examinedmaterials [41]. This process defines a cluster based on a concept ofnearest neighbor analysis. The solute atoms are identified as precipi-tate solute atoms (not solute atoms in a random solid solution) if theyare within a maximum separation distance, Dmax from one another.Minimum size of clusters in terms of solute atoms that constitutes a

significant cluster (Nmin) is used to eliminate clusters of atoms that arenot delineated as precipitates, so that any cluster with less number ofsolute atoms will not be considered. In order to include non-soluteatoms in the defined cluster, all the matrix atoms within a distance (L)from the solute atoms are taken as being part of the same cluster.However, this process results in a shell of matrix atoms being includedaround each cluster. Thus, an erosion distance (E) is used to removethe shell of matrix atoms that lies within a distance less than E fromthe nearest atom not defined as being part of the cluster [41–43].

Based on the deviation of the solute distribution in the testedspecimens from random distribution, the maximum separation

Fig. 4. Photographs of low carbon steel mini tensile sample (a) and tensile test grips (b).

Fig. 5. Transmitted electron microscopy (TEM) micrograph showing the grain size for UFG steel (a) and Electron back scattered diffraction (EBSD) micrograph showing thegrain size for CG Steel (b).

A. Alsabbagh et al. / Materials Science & Engineering A 615 (2014) 128–138 131

between solute atoms (Dmax) was chosen to be 0.6–0.7 nm. Boththe maximum separation of additional elements (L) and theerosion distance for removal of atoms near the cluster matrixinterface (E) are equal to Dmax. The minimum size of the cluster interms of solute atoms that constitutes a significant cluster (Nmin) istaken as 11 atoms. APT characterizations of the irradiated steelrevealed formation of Mn–Si-enriched nanoclusters (Fig. 10). How-ever, the matrices of un-irradiated UFG and CG steel are found tobe solid solutions with no solute clustering (the boxes in Fig. 10

Fig. 6. XRD profiles for unirradiated and irradiated (1.37 dpa) CG (top) and UFG (bottom) steels.

Fig. 7. Whole XRD pattern fit for irradiated CG steel at dose 1.37 dpa using LS1 program.

Table 1Values of domain size (Ds), microstrain ð⟨ε2L ⟩1=2Þ and dislocation density (ρ) fordifferent samples obtained by Modified Rietveld analysis.

Sample Ds (Å) ⟨ε2L ⟩1=2 ρ (m�2)

Unirr CG 1850 7.40�10�4 1.06 (70.13)�1014

Irr CG 622 1.00�10�3 4.26 (70.56)�1014

Unirr UFG 385 1.38�10�3 9.50 (71.24)�1014

Irr UFG 366 1.24�10�3 8.98 (71.39)�1014

A. Alsabbagh et al. / Materials Science & Engineering A 615 (2014) 128–138132

were empty). The fact that these small clusters were not observedbefore irradiation confirms that their formation was radiation-induced. The cluster number density is estimated using:

Nv¼ npξ

nV; ð2Þ

where np and n are the number of clusters in the analyzed volumeand the total number of atoms in the same volume, respectively. Vis the atomic volume and ξ is the detection efficiency which is�37% for local electrode atome probe (LEAP) instruments [43].The cluster size is estimated by finding the Guinier radius forspherical precipitates using [44,45]:

rG ¼ffiffiffi53

rlg ; ð3Þ

where lg is the standard radius of gyration. The average size(radius) of gyration of the nanoclusters is found to be 0.75 and0.7 nm for the UFG and CG steels, respectively. This yields anaverage Guinier radius of 0.9770.23 and 0.970.16 nm, respec-tively (Fig. 11). The number density of clusters is1.270.4�1024 m�3 and 6.771.1�1023 m�3 for the UFG and CGsteels, respectively.

Radiation hardening and embrittlement are of primary concernin neutron irradiated ferritic steels. The change in the microhardness post irradiation was measured for both UFG and CGsteels. Vickers microhardness data are included in Fig. 12. Theaverage micro hardness values for CG steel increased after irradia-tion by 62% compared to only 8.6% for the UFG steel. Mechanicalcharacteristics were also evaluated from tensile tests. Fig. 13includes stress–strain curves for CG and UFG steel before and

after neutron irradiation and the appropriate results are tabulatedin Table 2. Tensile test results revealed that CG yield strengthincreased after irradiation by 132% and its ductility decreased by82%. On the other hand, UFG steel yield strength increased by 30%and the ductility reduced by 56%.

4. Discussion

Significant changes were observed in the values of the domainsize and micro strain of the CG steel after irradiation to 1.37 dpadue to the increased defect concentration (Table 1). As shown inEq. (1), dislocation density is proportional to the micro strain butinversely proportional to the domain size and thus the dislocationdensity for irradiated CG steel is found to be about 4 times higherthan before irradiation. On the other hand, UFG steel did not showsignificant change in its dislocation densities (values are withinthe error bars). This can be attributed to the high density of grainboundaries in UFG steel that act as sinks for irradiation-induceddefects.

Irradiation can cause local enrichment or depletion of solutes. Ifthe local solute concentration exceeded the solubility limit, soluteclusters or precipitates are formed. As shown earlier, whileunirradiated steels show homogenous matrixes with no clustersdetected, after irradiation both steels revealed formation of nanomanganese–silicon enriched clusters. Although the sizes of clus-ters in both irradiated steels were very similar (mean cluster radiiare 0.9 nm and 0.97 nm in CG and UFG steels, respectively), theirradiated UFG steel revealed higher number density of clusterscompared to CG steel. The irradiation temperature was relatively

Fig. 8. Representation of three-dimensional reconstruction of UFG steel, (a) before and (b) after neutron irradiation by three-dimensional (3D) atom-probe microscopy.

A. Alsabbagh et al. / Materials Science & Engineering A 615 (2014) 128–138 133

low (�80 1C) and thus defect mobility will be relatively lowthereby increasing the defect recombination rate and limitingthe number of defects available to annihilate at sinks. However,the large grain boundary area per unit volume and the highdislocation density (resulting from the severe plastic deformationthrough ECAP processing) in UFG steel, increase the probabilitythat the defects will migrate to the sink before being recombined.Thus, significant participation of solutes in the defect fluxes resultsin pronounced segregation at sinks, raising the local concentrationabove the solubility limit and thus having higher number densityof clusters in UFG steel compared to CG steel that were irradiatedat the same conditions.

Higher strength of the UFG steel compared to the CG steelbefore irradiation is due to the grain refinement (Hall–Petchrelation). After irradiation, the CG steel exhibits increased hard-ness and strength accompanied by decreased ductility as per thecommonly observed radiation hardening and embrittlement.However, the UFG steel clearly indicates less significant changes.Tensile strength in CG steel increased after irradiation by 94%compared to 8% in UFG steel (Table 2). Fig. 13 shows that yield

point phenomena is observed only in the unirradiated CG steeldue to non-negligible source hardening attributable to the pinningof dislocations by impurity atoms (principally C) [46]. Before aFrank-Read (FR) source can be operated by the applied stress,dislocations have to be unpinned from the impurity atoms.However, during irradiation the impurity atoms get attracted toradiation produced defects thereby decreasing source hardening[47] resulting in reduced yield points which disappear followinghigher radiation dose (1.37 dpa). On the contrary, no yield pointphenomena were observed in the UFG steel presumably becauseimpurity atoms (principally carbon) tend to migrate to the grainboundaries thereby not being available for pinning thedislocations.

Although irradiation hardening was minute in the UFG steelcompared to the CG counterparts as delineated by Fig. 13, theirradiation induced embrittlement is clear in the UFG steel afterirradiation. However, the % decrease in the ductility of UFG steel isquite less than that of CG counterparts. According to Odette andLucas [48], the primary mechanism of embrittlement in ferriticsteels is the hardening produced by nanometer size features that

Fig. 9. Representation of three-dimensional reconstruction of CG low carbon steel, (a) before and (b) after neutron irradiation by three-dimensional (3D) atom-probemicroscopy.

A. Alsabbagh et al. / Materials Science & Engineering A 615 (2014) 128–138134

develop as a consequence of radiation exposure. However, sinceresults showed that there is no significant change in dislocationdensity in UFG steel after irradiation, the high density of irradia-tion induced Mn–Si-enriched clusters found in UFG steel isconsidered to be responsible for the observed embrittlement inUFG steel.

To achieve better understanding of the effect of irradiationmicrostructural changes on the steel's mechanical properties, theincrease in yield stress is related to different strengthening

mechanisms. After irradiation, four major mechanisms may beresponsible for increased strength as shown in the following:

Δσirr ¼ΔσGSþΔσSSþΔσCluþΔσDis; ð4Þwhere the subscripts GS, SS, Clu and Dis correspond to strengthen-ing due to grain size, solid solution, clusters and dislocations,respectively. However, no significant change in grain size isobserved for both UFG and CG steels after irradiation (Table 3).Thus, strengthening due to Hall–Petch effect is neglected

Fig. 10. Si–Mn-enriched cluster distribution post neutron irradiation for (a) CG and (b) UFG steel. The dimension of the interior (colored) boxes is 20�20�100 nm3.

A. Alsabbagh et al. / Materials Science & Engineering A 615 (2014) 128–138 135

(ΔσGS¼0). APT results showed that the percentage of solute atomsthat left the matrix to produce clusters is very small (mean atomicconcentration of solutes reduces by 0.05% after irradiation). There-fore, the change in the concentration of solutes before and afterirradiation does not cause any significant change in the hardeningdue to solid solution and the solid solution strengthening isignored. As a result, the remaining two strengthening mechanismsdominate and they cause the increase in the strength afterirradiation: cluster strengthening (Orowan precipitate hardening)and strengthening by dislocations (forest strengthening). Thus, thechange in the yield strength due to irradiation can be representedas:

Δσy ¼ σaf ter irr�σbef ore irr ¼ΔσCluþΔσDis: ð5Þ

As discussed earlier, APT analyses of the samples revealedpresence of Mn–Si-enriched clusters in the irradiated UFG andCG steels. However, no clusters were observed in both the un-irradiated steels. Hence, the strength increase due to irradiation-induced clusters can be described using Orowan–Ashby model[49,50] shown in Eq. (6). It is important to note that Orowan–Ashby model is commonly used for incoherent precipitates;however, previous studies [51,52] on different alloys showed thathardening due to nano clusters modeled by Orowan model iscapable of rendering explanation for observed yield stress. Thus,

ΔσOr�Ash ¼M 0:83Gb2πð1�υÞ0:5

1λ ln

2rs4b

� �rs ¼

ffiffi23

qr

9>=>;; ð6Þ

where G is the shear modulus of the α-Fe matrix (78 GPa), b is themagnitude of the matrix Burgers vector (0.248 nm), ν is thePoisson's ratio (0.33) [53], M is Taylor factor used for convertingshear strength to an equivalent uniaxial yield strength (3.06) [54],rs is the average radius of cross section of clusters on the slipplanes, r is the mean radius of the clusters and λ is inter-particlespacing which describes the basic characteristics of the clustersinside the irradiated material in terms of their volume fraction andthe average cluster radius [49,55].

λ¼ffiffiffiπ

f

r�2

� �rs; ð7Þ

where f is the volume fraction of the clusters.To evaluate the role of dislocations to strengthening in the CG

and UFG steels, yield strength increment due to dislocation foreststrengthening is given by [56]:

Δσdis ¼ αMGbð ffiffiffiffiffiffiffiρirr

p � ffiffiffiffiffiffiffiffiffiffiffiffiffiffiρun� irr

p Þ; ð8Þwhere ρirr and ρunirr are dislocation densities in irradiated andunirradiated materials respectively, and α is the correspondingstrengthening coefficient; for body-centered cubic metals, α¼0.4[57]. The change in the material yield strength due to irradiationcan then be found through the following equation:

Δσy ¼ΔσOr�AshþΔσDis ¼M0:83Gb

2πð1�υÞ0:51λln

2rs4b

� �þαMGbð ffiffiffiffiffiffiffi

ρirrp � ffiffiffiffiffiffiffiffiffiffiffi

ρunirrp Þ;

ð9ÞUsing the results obtained from both X-ray and APT analyses,

the increase in yield strength due to neutron irradiation ðΔσyÞ iscalculated using Eq. (9). Error calculations were made using errorpropagation method (Eq. (10)) starting from the standard devia-tion of the cluster size distribution (Fig. 11) and propagatingforward by finding the error in each equation leading to Eq. (9).Calculated results for different strengthening mechanisms andtheir contributions to yield strength with their correspondingerrors (uncertainty) are tabulated in Table 4.

σf ðx;yÞ ¼ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiddx

f ðx; yÞ� �2

ðσxÞ2þddy

f ðx; yÞ� �2

ðσyÞ2s

; ð10Þ

where σf ðx;yÞ is the standard deviation or error in the function f(x,y).Both tensile yield stress measurements and strengtheningmechanisms calculations illustrate similar results. Agreementbetween the two methods indicates that the Orowan–Ashby alongwith Taylor strengthening model is sufficient and can be useful forexplaining the contribution of nano-cluster strengthening anddislocation forest hardening to the overall strength of the ECAPultra-fine grained and the CG steels. The results show that while

Fig. 11. Size distribution of Si–Mn- enriched precipitates in the neutron irradiated (a) UFG and (b) CG steel.

Fig. 12. Micro hardness before and after irradiation for both UFG and CG lowcarbon steel.

A. Alsabbagh et al. / Materials Science & Engineering A 615 (2014) 128–138136

irradiation induced dislocation density has a high influence on thetotal yield stress increase in neutron irradiated CG steel, theirradiation hardening in the UFG steel was mainly due to theirradiation induced clusters.

5. Summary and conclusions

Neutron irradiation effects on ECAP steel were investigatedafter irradiation to 1.37 dpa dose in the ATR reactor at IdahoNational Laboratory. Changes in both microstructural and mechan-ical properties due to irradiation were analyzed using atom probetomography, X-ray diffraction, micro-hardness and tensile testing.The following conclusions are drawn:

– High number densities of nano Mn–Si-enriched precipitateswere observed in both CG and UFG steels after irradiation.However, the number density and the radius of the clusters

were larger in the case of the UFG steels due to the shorter paththat the defects need to diffuse before they reach the grainboundary and hence less defect recombination probability inthe matrix. The fact that these small clusters were not observedbefore irradiation confirms that their formation was radiation-induced.

– Both micro hardness and tensile tests revealed radiation hard-ening in the conventional grain sized steel as expected. How-ever, UFG steel hardness/strength showed minute changes afterirradiation indicating less radiation hardening effect.

– Irradiation induced precipitate strengthening and dislocationforest strengthening were evaluated and correlated to theexperimental measurements. Irradiation induced dislocationdensity in the UFG steel is found to be negligible and thusthe change in UFG steel strength after irradiation is consideredto be due mainly to the cluster hardening.

– As the area of grain boundaries (which act as sinks forradiation-induced defects) is significantly increased by grainrefinement, UFG steel revealed better irradiation tolerance.However, irradiation induced solute clustering in UFG alloysneeds to be carefully considered.

Acknowledgments

This research was supported by the Advanced Test ReactorNational Scientific User Facility (ATR-NSUF) project 13-450_RTE_ATR.Acknowledgments are due to Prof. Ruslan Valiev of Ufa State AviationTechnical University for providing the UFG steel samples. The authorsalso acknowledge the beneficial discussions with Mr. Peter Wells ofUniversity of California, Santa Barbara regarding the atom probetomography analysis.

References

[1] S.J. Zinkle, G.S. Was, Acta Mater. 61 (2013) 735.[2] K.L. Murty, J. Nucl. Energy Sci. Power Gener. Technol. 1 (2012) 1.[3] S.J. Zinkle, J.T. Busby, Mater. Today 12 (2009) 12.[4] J. Wadsworth, G. Crabtree, R. Hemley, Basic Research Needs for Materials

Under Extreme Environments: Report of the Basic Energy Sciences Workshopfor Materials Under Extreme Environments, 2007.

[5] P. Yvon, F. Carré, J. Nucl. Mater. 385 (2009) 217.[6] K.L. Murty, C.S. Seok, JOM 53 (2001) 23.[7] K.L. Murty, JOM 37 (1985) 34.[8] A. Misra, L. Thilly, MRS Bull. 35 (2010) 1.[9] D. Olander, Fundamental Aspects of Nuclear Reactor Fuel Elements, US Dept.

of Energy, Tennessee, 1976.

Table 4Estimated strength increment for different strengthening mechanisms for both CGand UFG steels.

ΔσOro�Ash (MPa) ΔσDis (MPa) ΔσCalculated (MPa) ΔσMeasured (Mpa)

CG Steel 137744 245735 382757 391727UFG Steel 230790 �20773 2107115 234750

Fig. 13. Engineering stress–strain curves for UFG (a) and CG (b) before and after irradiation to 1.37 dpa.

Table 2Mechanical properties obtained from both tensile and Vickers micro hardness testsfor both steels (CG and UFG) before and after irradiation.

Dose(dpa)

Yield strength(MPa)

Tensilestrength (MPa)

Vickershardness (MPa)

Ductility(%)

CG 0 296717 390718 1284713 63731.37 687720 754720 2081751 1171

UFG 0 77579 980711 3088754 18721.37 1009749 1060743 3353768 871

Table 3Mean grain size (μm) for both CG and UFG steels pre and post irradiation.

Mean grain size (μm) Unirradiated 1.37 dpa

CG Steel 4.4271.82 4.5171.78UFG Steel 0.3570.18 0.3470.15

A. Alsabbagh et al. / Materials Science & Engineering A 615 (2014) 128–138 137

[10] G.S. Was, Fundamentals of Radiation Materials Science, Springer, New York,2007.

[11] K.L. Murty, I. Charit, An Introduction to Nuclear Materials: Fundamentals andApplications, Wiley, Weinheim, Germany, 2013.

[12] A. Alsabbagh, R.Z. Valiev, K.L. Murty, J. Nucl. Mater. 443 (2013) 302.[13] B.N. Singh, Philos. Mag. 29 (1) (1974) 25.[14] W. Mohamed, Influence of Fast Neutron Irradiation on Mechanical Properties

and Microstructure of Nanocrystalline Copper (Ph.D. Dissertation), NorthCarolina State University, 2012.

[15] N. Nita, R. Schaeublin, M. Victoria, R.Z. Valiev, Philos. Mag. 85 (2005) 723.[16] S. Wurster, R. Pippan, Scr. Mater. 60 (2009) 1083.[17] Y. Estrin, A. Vinogradov, Acta Mater. 61 (2013) 782.[18] Y. Chimi, A. Iwase, N. Ishikawa, M. Kobiyama, T. Inami, S. Okuda, J. Nucl. Mater.

297 (2001) 355.[19] T.D. Shen, S. Feng, M. Tang, J. Valdez, Y. Wang, K.E. Sickafus, Appl. Phys. Lett. 90

(2007) 263115.[20] A. Etienne, B. Radiguet, N.J. Cunningham, G.R. Odette, R. Valiev, P. Pareige,

Ultramicroscopy 111 (2011) 659.[21] M. Rose, Nucl. Instrum. Methods Phys. Res. 128 (1997) 119.[22] H. Matsuoka, T. Yamasaki, Y.J. Zheng, T. Mitamura, M. Terasawa, T. Fukami,

Mater. Sci. Eng. A 449–451 (2007) 790.[23] H. Kurishita, S. Kobayashi, K. Nakai, T. Ogawa, A. Hasegawa, K. Abe, H. Arakawa,

S. Matsuo, T. Takida, K. Takebe, M. Kawai, N. Yoshida, J. Nucl. Mater. 377 (2008) 34.[24] C. Sun, K.Y. Yu, J.H. Lee, Y. Liu, H. Wang, L. Shao, S.A. Maloy, K.T. Hartwig,

X. Zhang, J. Nucl. Mater. 420 (2012) 235.[25] Y. Chang, Q. Guo, J. Zhang, L. Chen, Y. Long, F. Wan, Front. Mater. Sci. 7 (2013) 143.[26] M. Samaras, P. Derlet, H. Van Swygenhoven, M. Victoria, J. Nucl. Mater. 351

(2006) 47.[27] X. Bai, A.F. Voter, R.G. Hoagland, M. Nastasi, B.P. Uberuaga, Science 327 (2010) 1631.[28] R. Valiev, T.G. Langdon, Rev. Adv. Mater. 13 (2006) 15.[29] R.Z. Valiev, Mater. Sci. Forum 584–586 (2008) 22.[30] K.T. Park, Y.S. Kim, D.H. Shin, Metall. Mater. Trans. A 32 (2001) 2373.[31] V. Stolyarov, Y. Zhu, I. Alexandrov, T. Lowe, R. Valiev, Mater. Sci. Eng. A 299

(2001) 59.

[32] Y.T. Zhu, T.C. Lowe, Mater. Sci. Eng. A 291 (2000) 46.[33] T.G. Langdon, M. Furukawa, M. Nemoto, Z. Horita, JOM 52 (2000) 30.[34] M. Furukawa, Z. Horita, M. Nemoto, T.G. Langdon, J. Mater. Sci. 36 (2001) 2835.[35] F.M. Marshall, Joint TRTR – IGORR Meeting, 2005.[36] H. D. Hibbitt, B. I. Karlsson, E. P. Sorensen, Abaqus, Abaqus User's Manual, 2011.[37] L.A.N.L. X-5 Monte Carlo Team, MCNP, 2003.[38] L. Lutterotti, P. Scardi, J. Appl. Crystallogr. 23 (1990) 246.[39] P.S. Chowdhury, A. Sarkar, P. Mukherjee, N. Gayathri, M. Bhattacharya, P. Barat,

Mater. Charact. 61 (2010) 1061.[40] G.K. Williamson, R.E. Smallman, Philos. Mag. 1 (1956) 34.[41] J.M. Hyde, E.A. Marquis, K.B. Wilford, T.J. Williams, Ultramicroscopy 111 (2011) 440.[42] D. Vaumousse, A. Cerezo, P.J. Warren, Ultramicroscopy 95 (2003) 215.[43] E. Marquis, J.M. Hyde, Mater. Sci. Eng. R 69 (2010) 37.[44] M.K. Miller, K.F. Russell, J. Nucl. Mater. 371 (2007) 145.[45] M.K. Miller, Atom Probe Tomography: Analysis at the Atomic Level, Kluwer

Academic, New York, 2000.[46] K.L. Murty, D.J. Oh, Scr. Metall. 17 (1983) 317.[47] I. Charit, C.S. Seok, K.L. Murty, J. Nucl. Mater. 361 (2007) 262.[48] G.R. Odette, G.E. Lucas, JOM 53 (2001) 18.[49] L. Anand, J. Gurland, Metall. Trans. A 7 (1976) 191.[50] J. Martin, Micromechanisms in Particle-Hardened Alloys, Cambridge Univer-

sity Press, Cambridge, 1980.[51] M. Lambrecht, E. Meslin, L. Malerba, M. Hernández-Mayoral, F. Bergner,

P. Pareige, B. Radiguet, A. Almazouzi, J. Nucl. Mater. 406 (2010) 84.[52] F. Bergner, C. Pareige, M. Hernández-Mayoral, L. Malerba, C. Heintze, J. Nucl.

Mater. 448 (2014) 96.[53] K.-T. Park, Y.-S. Kim, J.G. Lee, D.H. Shin, Mater. Sci. Eng. A 293 (2000) 165.[54] R.E. Stoller, S.J. Zinkle, J. Nucl. Mater. 283–287 (2000) 349.[55] C.J. Hsu, C.Y. Chang, P.W. Kao, N.J. Ho, C.P. Chang, Acta Mater. 54 (2006) 5241.[56] M. Ashby, Strengthening methods in crystals, in: A. Kelly, R. Nicholson (Eds.),

Halsted Press Division, Wiley, New York, 1971.[57] T. Courtney, Mechanical Behavior of Materials, second, McGraw-Hill, New

York, 2000.

A. Alsabbagh et al. / Materials Science & Engineering A 615 (2014) 128–138138