Simulation of nitriding processes in dilatometric tests
Simulation of nitriding process at 400°C for 48 h carried out in dilatometer didn’t revealed any significant changes in the phase composition of treated nanobainitic steel – on dilatometric curve any changes in sample length, which could indicate the transformation of austenite or precipitation of carbides weren’t observed (Fig. 1), indicating high thermal stability of the microstructure. A small change in sample length, visible on dilatometric curve (Fig. 1) is a result of change of the temperature in the laboratory during long-term test.
However it should be taking to account, that the temperature of 400°C can be too low to ensure optimal conditions of diffusion of nitrogen during nitriding process, and extension of process duration to 48 h can be unfavorable from economic point of view. Thus in order to determine the highest limit temperature of thermal stability of the nanobainitic microstructure, subsequent simulations of nitriding processes were carried out – at a temperature of 425°C for 24 h, at 450°C for 12 h and at 480°C for 8 h (Fig. 1, 2). Analysis of dilatograms allowed to conclude, that when the temperature exceed 425°C, the changes in sample length occur at different stages of the nitriding simulation. On the dilatometric curve from simulation of nitriding at 450°C clearly visible decrease of sample’s length during annealing was observed (Fig. 1). This effect is most probably related to the carbides precipitation process, similar as it was reported by H.K.D.H. Bhadeshia and C. Garcia-Mateo et. al [7–9]. At the same time, after simulation of nitriding at 450°C increase of sample’s length and effects corresponding to the martensitic transformation during cooling to the room temperature weren’t observed, which indicate, that the retained austenite remained stable (Fig. 2). Analysis of dilatometric curve coming from simulation of nitriding at 480°C has shown a strong shrinkage at the beginning, with a subsequent increase in length of the sample (Fig. 1). In this case, reduction of sample length is most probably associated with intense carbides precipitation, leading to a significant reduction of carbon content in austenite, which may causes a restart of bainitic transformation in austenite blocks, manifested by growth of sample length. Changes that occur at the temperature of 480°C are significant and cause strong destabilization of austenite and, consequently, martensitic transformation during final cooling to the room temperature (Fig. 2) [10]. Some changes in sample length (initial decrease and sudden increase) can be also observed on the dilatometric curve for annealing at 400 ° C for 48 h (Fig. 1). This is related to temperature changes in the laboratory during the test – this effect was confirmed by the graphs of temperature changes in the dilatometer chamber.
The changes in the phase composition, observed on dilatograms are partially reflected in the results of mechanical tests. After simulation of nitriding at 400°C mechanical parameters an ductility slightly increase which indicate some changes in the microstructure. It can be result of two different phenomenon. First, it is possible, that due to a relatively small concentration of carbon in blocky austenite, when the tempering temperature is in between Ms’ and Bs’ (new martensitic and bainitic start temperature, respectively, determined in the austenite remained after austempering process), bainitic transformation is restarted and continued in blocky austenite [11]. In this case most of the film austenite, which is more stable, remain unchanged. The amount of residual austenite after austempering amounts 25.7 ± 2% and after austempering and tempering 21.2 ± 2, what suggest small microstructural changes during annealing at a temperature of 400°C. At the same time, in this temperature a small amount of very fine carbides can precipitate. Bainitic transformation results an increase in sample length and carbide precipitation – decrease in sample length, thus dilatometric effect is invisible. The changes in mechanical properties of the sample may be also related to higher and more uniform concentration of carbon in large blocks which in turn causes that TRIP effect occurs at higher stress. In novel high strength steels it’s TRIP effect that is usually responsible for increasing mechanical properties. It means that in many bainitic steels excellent strength and elongation are related not only to a grain size of bainite but also amount, size and distribution of blocks of retained austenite. It has been proved, that during early stage of deformation, large blocks of austenite undergo martensitic transformation [12] [13] - austenite grain boundaries decrease Ms temperature by limiting growth of martensite and increasing a driving force of martensitic transformation [13]. Higher stabilization of austenite can results TRIP effect at higher stress, which is manifested by greater tensile, yield strength and elongation. Thus improvement of mechanical properties can be also associated with redistribution of alloying elements, especially carbon, that occurs during tempering. Concentration of carbon inside blocks of retained austenite after austempering heat treatment can be non-uniform, and annealing during tempering allows redistribution and homogenization of chemical composition in blocky austenite. Carbon can also diffuse from bainite to untransformed austenite and stabilize it - similar as during Q&P processes [14]. During redistribution and partitioning process, volume fraction of bainitic ferrite and austenite should remain unchanged, thus this effect cannot be observed at dilatometric curve as well. Earlier investigation [15, 16] suggested that during tempering carbon do not diffuse from bainitic ferrite to austenite, however microscopic observations didn’t revealed any significant changes in the microstructure of the steel and VSM (Vibrating Sample Magnetometer) analysis indicates only slight changes in the phase composition, which suggest that partitioning from ferrite to austenite may occur instead of precipitation of carbides and continuation of bainitic transformation. Thus it can be assumed that improvement of mechanical properties may be a result of a combination of both described phenomena – continuation of bainitic transformation and partitioning and redistribution of alloying elements, especially carbon. A significant drop of tensile strength and ductility as a function of nitriding simulation temperature was observed (Tab. III). This confirm degradation of the microstructure during annealing at higher temperature – carbides precipitation and retained austenite transformation [14, 17].
Table III. Mechanical properties of X37CrMoV5-1after nanobainitization and various simulations of nitriding processes.
| Process parameters | R0.2 [MPa] | Rm [MPa] | A [%] | Charpy impact KV [J/cm2] |
| Austempering at 300°C for 19h | 734 | 1763 | 16 | 15.1 |
| Austempering at 300°C for 19h + simulation of nitriding process at 400 °C for 8h | 827 | 1827 | 19.3 | 17 |
| Austempering at 300°C for 19h + simulation of nitriding process at 450 °C for 8h | 698 | 1290 | 3.9 | 9.7 |
On the basis of dilatometric and mechanical tests the temperature of low temperature glow discharge nitriding was selected at 420°C. A typical time of glow discharge nitriding of 6 h was assumed.
Microstructure on the cross-section of steel after glow discharge nitriding
Observations of the nitrided layers produced at 420°C using LM didn’t revealed the microstructure and the thickness of the layer on both, quenched and tempered and austempered steel (Fig. 3a, c). Only Nomarski contrast has shown some differences in the microstructure (Fig. 3b, d), however such differences can be also related to the preparation of the cross section of the samples – due to a different hardness between steel and mounting resin, near surface zone can polished in different way than the core of the sample, thus some contrast may appear. Glow discharge nitriding at higher temperature allows to produce much thicker layers, with the near-surface zone of nitrides, so called white layer (Fig. 4a, b), very clearly visible on Nomarski contrast (Fig. 4b). At the same time, it is worth noting that after high temperature nitriding, the diffusion layer is much thinner, and the layer is distinctly separated from the substrate on austempered steel (Fig. 4c, d) as compared to the martensitic quenched and tempered steel (Fig. 4a, b). The thickness of the nitrided layers produced at 520 ° C measured on LM images amounts 147 ± 4 µm on quenched and tempered substrate, and 93 ± 4 µm on austempered substrate.
LM observations were confirmed by SEM studies. As a result of a low temperature nitriding, a diffusion layers weren’t observed on both analyzed substrates (Fig. 5). After nitriding at higher temperatures, layers are much thicker and consist of the nitrides zone (Fig. 6). Observations at higher magnification allowed to estimate the thickness of nitrides zone on all investigated samples. The thickness of the white layer after low temperature nitriding was very low and amount about 500 nm on both, quenched and tempered and austempered samples (Fig. 5b, d). The nitrides zone on substrates treated at higher temperature were much thicker, about 4 µm on quenched and tempered sample and around 1,5 µm on austempered steel (Fig. 6b. d). In all analyzed samples prior austenite grain boundaries in nitride layer are clearly visible. In case of nitriding at lower temperature, austenite grains are etched. After nitriding at higher temperature, some nitrides precipitation at grain boundaries are visible.
Nanobainitic microstructure is composed of nanometric plates of carbide-free bainitic ferrite, with high amount of thin films of residual austenite and some blocky austenite. On the one hand, high density of grain boundaries, which are easy diffusion path, and high content of residual austenite, in which the solubility of N is greater than in ferrite [10], should facilitate forming of nitride layers. On the other hand, diffusion coefficient of nitrogen in austenite is lower than in ferrite [18] and nitrogen can be trapped in austenite at octahedral interstitial sites or chromium atoms. At the same time bainitic ferrite is supersaturated with carbon, strongly dislocated and strained [19–22], up to tetragonal distortion [22] which can significantly hinder nitrogen diffusion through ferritic grains. The results of microscopic observations suggest, that this supersaturation of ferrite and high amount of residual austenite inhibits diffusion of the nitrogen into the material and the high density of grain boundaries doesn’t accelerate the diffusion of the nitrogen as it was supposed.
Microstructure of the core of steel after glow discharge nitriding
Observations of the core of steel at low magnification confirmed dilatometric tests - low temperature nitriding did not affect the microstructure of steel (Fig. 7a-d, 8a-d). Also at higher nitriding temperatures, the microstructure of the quenched and tempered steel remains unchanged (Fig. 7e, f) - the temperature of last tempering was higher than the temperature of nitriding process. In case of austempered steel nitrided at 520°C, effect of decomposition of nanobainite was clearly visible (Fig. 8e, f). This is due to the fact, that the temperature of –nitriding process was significantly higher than the temperature of thermal stability of nanobainitic microstructure in investigated steel. Despite of very high temperature stability of nanobainite in X37CrMoV5-1 steel, when the temperature exceeds 425°C, nanobainite decompose into ferrite and carbides.
Quantitative chemical and phase composition of nitrided layers
Chemical composition analysis on the cross section revealed that due to a low temperature glow discharge nitriding nitrogen diffuse to a relatively small depth (Fig. 9a-d). Diffusion depth of nitrogen in steel after conventional quenching and tempering is slightly higher (55 µm) and the nitrogen profile is more gradual as compared to the steel after austempering (40 µm). The greatest concentration of nitrogen occurs at the depth below 5 µm in both samples, however in case of austempered steel, concentration of nitrogen in the near surface zone was slightly higher.
Higher temperature of nitriding provides similar concentration and much deeper nitrogen diffusion into the material as compared to low tempereature nitriding – at a depth 60–80 µm nitrogen was still observed (Fig. 9c, d). It’s obvious, because with increasing temperature of the nitriding, diffusion coefficient of nitrogen also increases. The highest concentration of nitrogen was observed near the surface of the material, up to the depth 8 µm and 6 µm for quenched and tempered and austempered steel respectively. In case of quenched and tempered sample nitrogen profile shows two slops of nitrogen profiles, indicating two zones of nitride layer – compound zone and diffusion layer (Fig. 9c). The concentration profiles of nitrogen on the cross section of samples after nitriding at higher temperature are more flat as compared to low temperature nitriding, regardless of the microstructure of the core. According to Marchev [22] this is due to different nitriding mechanism at various temperatures. In long term nitriding processes (about 100 hours) nitrogen profile is more flat at low temperature [22]. However some authors revealed, that in short term processes (about 4 hours) with increasing temperature of the nitriding, nitride profile is less sharp [23, 24]. They attributed it to nitrides precipitation – nitrides increase activation energy during nitriding, therefore nitrogen, which already exist in the layer can diffuse into the material. During low temperature nitriding processes nitrides form as well, and with increasing time of the process, nitride zone become thicker. At the same time lower concentration of nitrogen in the diffusion layer and lower supersaturation of the substrate with nitrogen (which results lower stresses) facilitates diffusion into the material with the process time.
Internal stresses in material can act on the diffusion of the layer-forming element in a similar way as supersaturation with nitride. Tempering process removes stresses from martensite, and after high tempering, the microstructure consist of carbides in a ferritic matrix. The microstructure of tempered martensite is not supersaturated with carbon and is stress-relieved, therefore the diffusion of nitrogen should occur more easily as compared to nanobainitic microstructure, so that the layer produced on quenched and tempered substrate is thicker than the layer produced on the substrate austempered stteel.
In all analyzed case an increase of carbon at some depth was observed. This suggest that nitrogen during diffusion pushes carbon, increasing its [25–27]concentration in the near nitride layer zone. According to mechanism proposed by Tier et. al. [25] carbon in carbides may be substituted by nitrogen and diffuse into the core of material. Our studies confirm this theory - in investigated materials, higher concentration of carbon was observed in samples after quenching and tempering heat treatment, which produced a microstructure of tempered martensite with carbides.
In all analyzed cases higher concentration of chromium in the near surface zone suggest chromium diffusion to the surface however without chromium nitrides formation - XRD studies didn’t reveal the presence of chromium nitrides, only ε - Fe3N iron nitrides on steel subjected to LTN. Higher temperature of nitriding produced the mixture of Fe4N and Fe3N nitrides on both substrates. Due to a relatively low concentration of chromium in X37CrMoV5-1 steel (lower than 5.6%), chromium nitrides probably didn’t form. It is also possible, that in the near surface zone some chromium nitrides precipitate, however their amount and size are too low to be detected by XRD. The presence of fine-grained nitrides in the nitrided layer produced at a temperature higher than 450 °C was confirmed by Marchev [22]. At lower temperature diffusion of chromium is very low, thus chromium nitrides shouldn’t precipitate.
Formation and composition of a white layer during nitriding strongly depends on time, temperature and chemical composition of nitriding atmosphere, which determines nitriding potential. At a temperature 400–550°C a mixture of both, γ’ and ε nitrides form in the nitride layer, regardless of the chemical composition of the gas mixture and temperature [26–28]. However, the content of individual phases in the compound zone depends on process parameters. During nitriding in the atmosphere contains 75% of nitrogen, ε nitrides dominate. By reducing the nitrogen in the gas mixture up to 5%, the ε nitrides content can be significant reduced [26, 27]. Formation of a layer with a predominance of γ’ nitrides is also favored by the extension of the time, up to 15 h [28]. At the same time Pindeo showed, that the compound layer should be avoided when nitriding is carried out at temperature lower than 520°C for 4 hours with a gas mixture containing 50% N2 [29]. In this study gas mixture contains of 25% of nitrogen and 75% of hydrogen. Addition of hydrogen plays very important role during glow discharge nitriding. Hydrogen is a reducing agent and cleans the surface during first stage of the process and lowers the nitrogen potential during nitriding [30]. Lower content of nitrogen in a gas mixture (lower nitrogen potential) should avoid nitrides formation. However hydrogen can provides decarburization in the near surface zone – carbon from the material can react with hydrogen ions and form hydro-carbides. Due to a carbon gradient in the near surface zone carbon diffuse from the material towards the surface and its further removed by hydro-carbides formation [25]. Decarburization enhances the diffusion of nitrogen into the steel [31] and increases concentration of nitrogen in the layer [25, 32], thus some nitrides in the layer can appeared. Some of the authors [33] also indicate, that nitrogen-rich iron nitrides form during first hour of the process, even at very low temperature and their presence is related not only to the temperature and time of the process, but also the effect of the cathode sputtering phenomenon, occurring during the heating of the treated samples to the temperature of the nitriding process, which is influenced by the pressure of gas atmosphere in the working chamber and its chemical composition [34].
Hardness of steel after glow discharge nitriding
All nitriding processes significantly improved the hardness of investigated materials (Fig. 10a, b). The highest increase of hardness was observed after the nitriding process at higher temperature. However hardness of the layers produced at lower temperatures is only slightly lower. According to XRD analysis after LTN processes surface nitride zone was composed of ε-Fe3N nitrides, which are characterized by very high hardness. Higher temperature of the processes produced a mixture of ε-Fe3N and γ’-Fe4N nitrides zone, thus hardness should be reduced as compare to LTN processes. At the same time after nitriding at 520°C both, surface zone of iron nitrides and diffusion zone are much thicker which results relatively high hardness. This can be confirmed by a drop of hardness at 2kg load, especially for nitriding at lower temperature (Fig. 10b) – the most probable scenario is that indenter penetrate through the layer during the test at a given load, and the measured hardness is a sum of the effects of hard and thin nitrides zone, thin diffusion layer and the substrate. The hardness of the core in most of samples remains unchanged (Fig. 10c). The only exception is steel after austempering and nitriding processes, especially at higher temperature, where hardness increases. Slight increase of core hardness after nitriding at lower temperature indicates some microstructural changes, corresponds to continuation of bainitic transformation, as it was describe in “Simulation of nitriding processes in dilatometric tests” chapter. In case of high temperature nitriding hardness results confirm significant changes in the nanobainitic microstructure such a carbides precipitation, austenite decomposition and martensitic transformation [14], which was also observed on the dilatometric curve. Fresh martensite, which appears during cooling to the room temperature as a result of a carbides precipitation and depletion of austenite with carbon increases hardness of the steel. Decomposition of nanobainite starts at a temperature higher than 425°C and at a temperature 520°C this effect is very strong, and high amount of martensite appears in the core therefore an increase of the core hardness after austempering and nitriding at higher temperature is more visible than in steel in which the bainite transformation took place.
Wear resistance of nitrided layers
Wear resistance before and after glow discharge nitriding of steel strongly depends on applied load (Fig. 11). Before nitriding processes, volumetric wear of austempered steel at higher load is slightly lower as compared to the steel after conventional heat treatment. This may be due to the TRIP effect occurring in retained austenite during wear test [35–38]. At this load, low temperature glow discharge nitriding decreases wear resistance, especially in case of quenched and tempered steel. The most probably reason is a structure and the thickness of the layer and presence of hard and brittle ε - nitrides zone and a very thin diffusion layer - during the test nitrides debris may spall from the layer and become an additional abrasive. The TRIP effect occurring in nanobanitic steel compensates this effect, therefore the drop of wear resistance for nanobainitic steel is lower. After nitriding processes at higher temperatures, the diffusion zone is much thicker, resulting a higher hardness at a higher depth. The nitrides zone is composed of a mixture of ε and γ’ nitrides. It’s well known, that γ’ nitrides are responsible for better wear properties. Thus it can be assumed, that phase composition and a thickness of the nitrided layers produced at higher temperature increase the wear resistance of the material. In this case the influence of the substrate to wear resistance wasn’t observed. At lower load some relationships are reversed – quenched and tempered steel is much more resistant to friction wear than austempered sample – applied load is too low to ensure TRIP effect in nanobainitic steel. In this case, surface layers produced at low temperature provide similar wear resistance of both steels (slightly worse for quenched and tempered steel and slightly better for nanobainitic steel), however the reduction of volumetric loss as compared to the initial state is significantly higher for austempered steel. It is worth noting that the values of volumetric loss, measured at different load for both steels after high temperature nitriding are identical, regardless of the heat treatment of the substrate. It suggest, that in case of high temperature nitriding there is no influence of the substrate on the useful properties of steel. Despite of the conditions of glow discharge nitriding and applied load during the test, all produced surface layers reduce friction coefficient as compared to both steels before surface treatment (Fig. 12). Some Authors [39] suggest, that friction coefficient strongly depends on hardness of material. Higher hardness results lower deformation of the surface and thus the contact surface with the counter-sample is smaller, which results lower value of friction coefficient.