Search

International audience; The design of the dissimilar metal weld investigated here is aimed at applications in the steam generator of a sodium-cooled nuclear reactor, with a multi-decade lifespan in demanding operational conditions. It consists in a narrow-gap joint between 2.25Cr-1Mo low-alloy steel and an austenitic alloy using a nickel-based alloy 82 as filler material. This study focuses on understanding the microstructural and micromechanical evolution in the near fusion boundary region between the low-alloy steel and the nickel alloy filler metal during post-weld heat treatment, using notably electron probe microanalysis and nano-indentation. The difference in matrix phase and chemical composition between the two alloys leads to a large difference in chemical potential for carbon, which is mobile at the post-weld heat treatment temperature. A number of fine-scale characterization techniques were used to assess the gradient of composition, hardness and microstructures across the fusion boundary, both as welded and after post-weld heat treatment. This complete analysis permits to highlight and understand the main microstructural and micromechanical changes occurring during post-weld heat treatment and opens the way to their long term study in service conditions.

International audience; In ferrite/austenite dissimilar metal welds, carbon can diffuse from the ferritic low-alloy steel to the austenitic weld metal during heat treatment and thermal aging due to the large carbon and chromium content differences between the two alloys. This carbon influx leads to the precipitation of carbides in the high-alloy weld metal, near the fusion boundary, which can affect the properties of the joint. This precipitation was investigated using multiple characterization techniques: electron probe microanalysis, small angle X-ray scattering and atom probe tomography. The studied joint consisted of low-alloy steel and stainless steel base metals that were arc welded using a nickel-based filler metal. Given the heterogeneous microstructure of such joints, a microstructure navigation strategy was implemented to ensure all characterizations were carried out in the same regions of interest and the associated data could later be correlated. The precipitates were found to be mostly M23C6. The precipitation was quantified using two different methods that gave comparable results, whose differences highlight the capabilities and limitations of each technique. Precipitate fractions are higher following thermal aging but remain far from equilibrium in all cases, indicating that precipitation kinetics should be accounted for to model the microstructural evolution of such dissimilar metal welds more accurately.

International audience; Detailed characterization of the near fusion-boundary region of 18MND5 / Alloy 52 dissimilar metal weld joints was performed to investigate the effect of thermal aging on the microstructure and mechanical properties and to quantify the diffusion of carbon into the weld metal. It was shown that the microstructural features of dissimilar metal welds, such as the size of the partially mixed zone in the weld and the adjacency to micro-segregations in the low-alloy steel have an effect on the peak carbon concentration in the enriched region of the weld. Aging conditions (time and temperature) influence the position of this concentration peak in the weld. Those trends were supported by thermodynamic calculations.

International audience; The industrial importance of carbonitriding is owed to the exceptional wear and fatigue resistance it imparts to treated steel parts. This resistance is related to microstructural changes occurring during the enrichment treatment and upon cooling. Here, the effects of interstitial contents, in particular nitrogen, and transformation temperature were investigated in 23MnCrMo5 steel. Samples were homogeneously enriched in the austenitic phase and the isothermal transformation of the enriched austenite between 750 °C and 600 °C was studied. CrN was found to precipitate during the enrichment treatment. During subsequent isothermal holding, CrN precipitate as fine platelets in nitrogen containing samples. At equal carbon content, ferrite formed faster and in finer grains in presence of nitrogen. Preexisting CrN facilitate ferrite nucleation resulting in more numerous ferrite grains. The intense nitride precipitation is the main origin for enhanced hardness in nitrogen-enriched alloys. The exact mechanism leading to the observed microstructures could not be determined and remains under investigation. In particular, the high nitrogen supersaturation of ferrite required to produce the observed fraction of CrN has to be explained.

The growth kinetics of ferrite during intercritical annealing of steel has been investigated using in situ high energy X-ray diffraction with a specially designed furnace allowing highly quantitative measurements of phase volume fractions. Kinetics have been obtained at 730 ∘ C, 750 ∘ C and 775 ∘ C in different ternary Fe-C-X (where X : Mn, Ni, Mo and Cr) alloys and in a quaternary Fe-C-1Mn-1Cr (wt.%) alloy. The obtained results were compared with the predictions of classical Local Equilibrium and Para-Equilibrium models as well as an improved version of a three-jump solute drag model, where the interactions between the solute elements and the moving interface are described comprehensively. Good agreement was obtained between the measured ferrite growth kinetics and the predictions of the solute drag model for the different Fe-C-X systems using only one fitting parameter, namely the Fe-X interaction parameter at the interface. The interactions found here qualitatively match those reported in literature for all solutes except for Ni, which displayed an attractive interaction with the interface. For the quaternary Fe-C-Mn-Cr system, the solute drag model succeeded in predicting ferrite growth kinetics using the same interaction values as used for the ternary Fe-C-Mn and Fe-C-Cr systems.

Journal of alloys and compounds 658, 832 - 838(2016). doi:10.1016/j.jallcom.2015.11.007, The effect on the bainitic transformation of carburizing, nitriding and carbonitriding of a 23MnCrMo5 low-alloyed steel in the austenitic field was examined by in situ high-energy synchrotron X-ray diffraction (HEXRD) and transmission electron microscopy. CrN nitrides precipitated in austenite during the enrichment in nitrogen and were quantified. Enrichment in nitrogen leads to a fine bainitic ferrite microstructure, which contains the CrN nitrides as well as some AlN, VN and MnSiN2 precipitates, with much smaller number density. The precipitates are frequently gathered into aggregates. By using HEXRD, it was possible to monitor ferrite, austenite, cementite and CrN mass fraction during the isothermal bainitic transformation, whereas HEXRD experiments in previous studies focused mostly on carbide-free bainite. Ferrite formed first, followed by cementite precipitation and, in the case of nitrided steels, increase of CrN mass fraction. Enrichment in carbon and/or nitrogen tends to slow down the bainitic transformation compared to the initial steel, as expected from their austenite-stabilizing character. However, comparison between enriched steels shows that kinetics is not simply related to the total interstitial content when both carbon and nitrogen are simultaneously involved. Possible role of CrN precipitated in austenite on ferrite formation acceleration is discussed., Published by Elsevier, Lausanne

International audience; The isothermal decomposition of austenite between BS and MS was investigated after carburizing and carbonitriding in low-alloy steel 23MnCrMo5. The microstructure of the products differed significantly between carburizing and carbonitriding. The carburized alloy produced an expected bainitic microstructure consisting mainly of bainitic ferrite with some cementite and retained austenite. In contrast, the microstructure of the nitrogen-containing sample showed finer, entangled ferrite plates, retained austenite and nitrides. The ferrite plates contain numerous transformation twins, which can be as thin as only a few nanometers. This original result proves that abundant twinning can be found in bainite as well as in martensite.

The effect on the bainite formation of carburizing (0.6 wt.% C), nitriding (0.12 wt.% C-0.25 wt.% N) and carbonitriding (0.7 wt.% C-0.25 wt.% N) of a 23MnCrMo5 low-alloyed steel in the austenitic field was examined by in situ high-energy synchrotron X-ray diffraction (HEXRD) and transmission electron microscopy. The enrichment in nitrogen induces strong acceleration of the bainite transformation kinetics in carbonitrided steel compared to carburized steel, despite the γ-stabilizing character of nitrogen. This is attributed to the nucleation of ferrite on CrN nitrides, which precipitated during the enrichment, either at γ grain boundaries or intragranularly. AlN, VN and MnSiN2 nitrides were observed as well, with much smaller number density. They formed frequently aggregates with the CrN nitrides. The bainite microstructure is much finer than in initial or carburized steel. It shares some common features with intragranularly nucleated bainite, i. e. acicular ferrite. From HEXRD, the chronology of the phase formation (ferrite and precipitates) during bainite formation as well as cell parameter evolutions are analyzed.

Knowledge of solute interaction with the interface during the transformation of austenite into ferrite is fundamental in predicting its kinetics in multicomponent steel. This interaction notably translates in segregation, or depletion, of the solutes at the transformation interface. Here, this segregation was successfully quantified by atom probe tomography (APT) in four ternary Fe-X-C systems involving substitutional solutes commonly found in modern steel grades (X = Cr, Mn, Ni, Mo). Controlled decarburization was used to grow a uniform, planar and incoherent ferrite layer at the surface of fully austenitic samples. In the case of Fe-Cr-C and Fe-Mo-C, the interfacial concentrations permitted the evaluation of the binding energy of each substitutional solute to the interface, which was found to be comparable to its respective grain boundary binding energy. In the case of Fe-Mn-C and Fe-Ni-C, undesirable motion of the interface during the quench of the samples could not be avoided, preventing a reliable estimation of their binding energy since temperature and interface velocity were unknown.

Annealing conditions leading to the crystallization of amorphous silicon nitride occurring upon nitriding of Fe–Si alloys have been found. A brief thermodynamic explanation for the occurrence of this unexpected amorphous-to-crystal transition is given. The crystalline precipitates were found to be α-Si3N4 and follow the orientation relationship given by ( 1 1 ¯ 0 ) α -Fe / / ( 10 1 ¯ 0 ) α - Si 3 N 4 with [ 111 ] α -Fe / / [ 0001 ] α - Si 3 N 4 . This transition happens in precipitates with a constant Si3N4 composition. This structural change also leads to a morphological change from the initial cube-like shape to a hexagonal prism.

International audience; Atom probe tomography (APT) was employed to investigate the precipitation process of Si3N4 particles after plasma assisted nitriding and subsequent annealing of Fe-Si alloy. The nitriding was done using 15N2 medium to enable quantitative analysis of Si and N by APT. Al, as a minor alloying element, is found to play an important role in the precipitation sequence. It is shown that Al forms Al-N clusters before Si3N4 precipitation takes place, and provides nucleation sites for the precipitation of Si3N4. Composition measurements show that silicon nitrides are enriched in Al, Ti and Mn as compared to the matrix.

International audience; The binary B2 (ordered bcc-type structure) FeAl40 Grade 3 aluminide containing 40 at.% Al was nitrided at 600 °C for 15 min with a pulsed DC plasma in a 95% N2/5% H2 gas mixture. The nitrided layer was described at the micro-, nano- and atomic scales. The nitrided layer consists of a thin outer sublayer (∼1 μm in thickness) containing γ′-Fe4N and having a moderate hardness (870 Hv) below which a thicker (∼36 μm in thickness) and harder (1400 Hv) inner sublayer is formed. While the outer sublayer is completely depleted in Al, the inner sublayer is made of a mixture of hex. AlN and α-Fe phases regularly structured at different levels: i) At the micrometre scale, observations showed a network of α-Fe “wave-like” veins (∼2 μm in length and ∼100 nm in width) located at the grain boundaries and preferentially aligned parallel to the nitrided surface. ii) At the nanometre scale, observations evidenced that such veins are embedded in a “lamellar-like” morphology matrix alternately composed of α-Fe and α-Fe/hex. AlN lamellae (∼300 nm in length and ∼10 nm in width). iii) At the atomic scale, analysis pointed out that each α-Fe/hex. AlN lamella presents a finer structural arrangement in which hex. AlN-rich clusters (∼4 nm in diameter) are trapped between thin α-Fe walls (∼2 nm in thickness). The diversity of the morphologies evidenced in the present study highlights the complexity of the nitriding mechanisms of iron-based aluminides. The observations support however a mechanism of discontinuous precipitation in the inner nitrided sublayer which seems to be strongly promoted by the high amount of Al in the FeAl40 grade 3 alloy (40 at.%).

International audience; The segregation of solutes to austenite/ferrite transformation interfaces during decarburization/denitriding of Fe-Mn-C, Fe-Mn-N, and Fe-Si-C ternary alloys was studied by using atom probe tomography. Manganese was found to segregate noticeably to the transformation interface in the presence of carbon, while no segregation could be detected in the presence of nitrogen. This result might indicate that manganese interacts little with the interface itself and that its interaction with the interstitial controls its segregation behavior. In the case of Fe-Si-C, the experiments were complicated by interface motion during quenching. Preliminary results suggest that silicon was depleted at the interface in contrast to the commonly observed segregation behavior of silicon at grain boundaries of ferrite and austenite. This observation could be explained by taking into account the repulsive interaction between silicon and carbon along with the intense segregation of carbon to the interface. This would lead to a net repulsive interaction of silicon with the interface even when considering the intrinsic tendency of silicon to segregate to the boundary in the absence of carbon. The results presented here emphasize the need to account for the interaction of all solutes present at the interface in ferrite growth models.

International audience; A theoretical approach is proposed to take into the account nucleation, growth and coarsening of nitrides in alloys featuring heterogeneous nitrogen content. It is based on physical considerations and accounts for both bulk nitrogen diffusion, which results from the nitriding process and nitrides precipitation kinetics. It predicts local information such as average particle density, radius and volume fraction of nitrides as a function of depth. The work presented in this paper leads to two important conclusions. First, the precipitation rate is not so high that precipitation can be considered as infinitely faster than the diffusion of nitrogen into the bulk. Second, the precipitation state at a given depth depends on the local interaction between nucleation, growth and coarsening phenomena, themselves depending on the local nitrogen content. Finally, the precipitation in alloys with heterogeneous nitrogen content induced by nitriding is radically different from classical precipitation in a single phase where the driving force for precipitation is consumed.

The limiting conditions for non-partitioned ferrite growth were investigated under controlled decarburization conditions. An abrupt change in the growth kinetics and morphology of ferrite was observed when the temperature increased above the limit defined by the local equilibrium no-partitioning (LENP) model. The ferrite layer formed below the LENP limit consisted of columnar grains and showed continuous, parabolic growth kinetics. In contrast, ferrite growth above the LENP limit showed two distinct stages. Initially, a thin layer of columnar ferrite grains formed and grew rapidly for a short period of time. In the second stage, the ferrite layer appeared to increase in thickness at a very low rate. The slow growth of the ferrite layer was associated with a strong depletion of carbon content in austenite as well as the formation of new ferrite grains as opposed to the growth of existing grains. The evolution of the carbon concentration in austenite was used to infer the operating interfacial contact conditions as a function of time. The evolution of the interfacial austenite conditions is suggested to penetrate deep into the two-phase region as predicted by ferrite growth theories that account for free-energy dissipation during the austenite-to-ferrite phase transformation.

International audience; We present a detailed analysis based on both experimental and modeling approaches of the unique silicon nitride precipitation sequence recently observed in ferritic Fe–Si alloys upon nitriding. At 570 °C, silicon nitride forms as an amorphous phase of size-dependent cuboidal morphology which results from the symmetry of matrix crystal structure. These amorphous precipitates are stable over remarkably long treatment durations. However, we demonstrate here that it is possible to trigger a transition to the precipitation of a crystalline modification of Si3N4 by switching to a denitriding medium at the same temperature. This change in structure is associated with a change from the cube-like morphology to a hexagonal prism shape. This unique phenomenon, driven by a classical nucleation and growth process, can be explained by the shift of stress-relief mechanism related to the change of atmosphere. The plentiful availability of nitrogen atoms during nitriding allows a strong relaxation of the precipitation-induced stress, which is energetically more favorable to the amorphous phase. Upon annealing in a low nitrogen activity medium, nitrogen in solid solution diffuses outward and no longer relieves the precipitation-induced stress. In that configuration, the crystalline modification of Si3N4 becomes more stable owing to its lower associated stress. It precipitates in a hexagonal prism shape, which is the equilibrium shape dictated by the crystal symmetries of both the matrix and the precipitates.