Ion nitrided AISI H13 tool steel Part I – Microstructural aspects M. R. Cruz*1, L. Nachez2, B. J. Gomez2, L. Nosei2, J. N. Feugeas2 and M. H. Staia1 In the present paper, five different operating conditions of a plasma nitriding process on annealed AISI H13 were investigated. A systematic variation of processing parameters such as gas mixture, time of processing and current density has been carried out in order to study their effect on microstructure. X-ray diffraction, microhardness testing and scanning electron microscopy techniques coupled with semiquantitative energy dispersive X-ray analysis were used to characterise the nitrided samples. The results revealed that the current density during plasma processing has a considerable influence on both the compound layer formation and depth of the diffusion zone. Also, it has been shown that the geometrical configuration of the workload may affect the uniformity of the diffusion zone, giving rise to the presence of heterogeneous nitrogen distribution zones along it. Keywords: Ion nitriding, Optical spectroscopy, H13 tool steel, Current density, Diffusion layer, Compound layer, Nitrided tool steel, Microhardness, X-ray diffraction, Nitrogen diffusion, Chromium nitride precipitates, Case depth
Introduction In the last four decades, a considerable amount of research has been focused on different aspects of plasma nitriding of iron base alloys, since this process has demonstrated to be a very important technique with many industrial applications. Gru¨n et al.1 have compiled several industrial experiences related to handling of a plasma nitriding plant and have found that variables such as processing temperature, duration of the process, chemical composition and flow of the gas mixture, pressure in the vacuum vessel, applied plasma voltage and current density and their interaction have an utmost importance on the product quality. The generation of N2 active species during plasma nitriding has been studied using optical emission spectroscopy (OES), which allowed the modelling of the reactions on the substrate surface2–4 and the analysis of all diffusive species during the DC pulse discharge and in the afterglow.5 Few results on the influence of current density have been reported,1,6–8 mainly related to its influence on the hardness of the nitrided samples. Some authors9–11 have determined that both microstructural configuration and uniformity of the nitrided layer are influenced by the sample geometry and its size, as well as its relative position inside the vacuum chamber. However, the relationships between the current density, the concentration of N2 active species and
workpiece geometry with the microstructural characteristics are not well established. The effect of nitriding on the mechanical behaviour of steels is well known and the substantial increase in fatigue life and wear is mainly due to the formation of two different layers during processing: the compound layer and the diffusion layer. Therefore, in order to maximise the influence of these layers in tailoring the mechanical properties of the products with possible applications in service, a series of combinations of such layers were obtained and studied in the present investigation. Thus, the main objective of the present work is to offer information on the specific plasma nitriding processing parameters of an AISI H13 tool steel, such as current density, concentration of N2 active species and workpiece geometry, in order to be able to obtain a precise microstructural configuration and to study the response of these different nitrided systems under sliding wear conditions. The present research encompasses two different aspects. The first one, presented in the present paper, is related to the use of OES of glow discharge to control in situ the nitriding process. Additional X-ray diffraction (XRD) analysis, morphological characterisation by using scanning electron microscopy (SEM) and hardness measurements were conducted to relate the operational parameters to the different microstructural features obtained. The second part of the present research, which will be presented in a subsequent paper, will focus on the high temperature sliding wear resistance behaviour of these systems.
1
School of Metallurgical Engineering and Materials Science, Universidad Central de Venezuela, Apartado Postal 49141, Caracas 1042-A, Venezuela 2 Instituto de Fı´sica de Rosario, Bv. 27 de Febrero 210 Bis, Rosario, Argentina *Corresponding author, email
[email protected]
ß 2006 Institute of Materials, Minerals and Mining Published by Maney on behalf of the Institute Received 30 January 2006; accepted 5 January 2006 DOI 10.1179/174329406X126663
Experimental Discs of 4 mm thickness were machined from a cylindrical H13 annealed steel bar of W531.4 mm.
Surface Engineering
2006
VOL
22
NO
5
359
Cruz et al.
Ion nitrided AISI H13 tool steel: Part 1
Before nitriding, the samples were polished to a mirror like finish with 1 mm alumina powder leading to an average roughness of Ra50.025¡0.002 mm. Subsequently, these were degreased and cleaned with acetone before being introduced in the process chamber. The nitriding treatments were carried out in an experimental nitriding plasma reactor with a cathode planar disc of y167 cm2. Before nitriding, plasma cleaning was carried out using a hydrogen atmosphere at 1 T for 10 min. Subsequently, a flow of a (N2zH2) gas mixture was used at a total pressure of 3 T, which was kept constant under all the nitriding conditions. During the process, the applied voltage was controlled to heat the samples up to the treatment temperature of 380uC. Owing to relatively fast heating time (y12 min), no auxiliary heating system was used. Nitriding times of 3 and 10 h, with two different gas mixtures [25 vol.-%N2 and 75 vol.-%(N2zH2)] and two different cathode areas (182.44 and 238.11 cm2) were used. Table 1 summarises the plasma nitriding conditions employed and the sample nomenclature. Figure 1 shows a schematic representation of the applied process. When the plasma nitriding process is generated by a DC source, it depends on a duty factor, which is characterised by the on-off relation.12 The duty factors used during processing are shown in Table 2. A Princeton applied research (PAR) optical spectrometer (focal length 275 mm; grating width 1024 grooves mm21) and an optical multichannel analyser III (PAR) were employed for the detection of plasma near the cathode. The data acquisition of the optical spectrum was done in a continuum manner along the pulsed glow discharge. X-ray diffraction was carried out by means of a Cu Ka radiation source, using BraggBrentano geometry. The morphology and microstructure of the nitrided samples were studied by SEM coupled with semiquantitative energy dispersive X-ray analysis (EDS). Nitrogen profiles were determined by means of semiquantitative EDS on the samples cross-section. Hardness profiles were determined by conducting standard Vickers indentations on the samples crosssections by employing a motorised stage microhardness tester. A load of 0.05 kg was used and it was maintained for 15 s. The results reported represent the average value of five indentations carried out in each case. Image analysis techniques were used in the order to estimate from SEM images, the area (statistical
1 Schematic diagram of plasma nitriding process
percentage) in the diffusion zone which corresponded to the CrN enriched precipitates.
Results and discussion Optical spectroscopy Figure 2 presents the optical emission spectra from the different processing conditions, which were analysed. Sample 25B1 shows the same spectrum as sample 25EB1, since the only difference between them is related to the nitriding duration. The spectra obtained allow the visualisation of the radiatives transitions with interest to plasma diagnostics. Optical diagnostic is a non-destructive method that can be easily set up on any plasma reactor. However, in industrials applications, the plasma is often treated like a ‘black box’ whose macroscopic characteristics corresponding to the desired metallurgic effect are found as a result of many experiments. The spectral lines detected by the optical spectrometer are indicated in Table 3. The emission from the (N2zH2) discharge was produced principally in the negative glow zone around of steel sample (cathode). The dominant species were the first negative bands from the N2z ion and the first and second positive bands form N2. The spectra shown in Fig. 2 indicate that the first negative bands from the N2z ion (at 391.4 nm) are more intense. This results are characteristic for negative glow discharges, as it has been previously reported.5,12,13 The relative intensities which correspond to the excited species of N2z (391.4 nm) and
Table 1 Treatment conditions Batch number Sample ID Nitriding time, h Temperature, uC N2/(H2zN2), % Cathode area, cm2 Current density, mA cm22 B1 B2 B1 B2 B1
360
75B1 75B2 25B1 25B2 25EB1
3 3 3 3 10
380 380 380 380 380
182.44 238.11 182.44 238.11 182.44
75 75 25 25 25
5.60¡0.27 3.22¡0.87 4.17¡0.19 3.12¡0.93 4.17¡0.19
Table 2 Duty factor for pulse DC source of nitriding process, ms
Table 3 Spectral lines detected by means of optical spectrometry
Batch 1 (B1 samples)
Batch 2 (B2 samples)
Specie
Wavelength l, nm
ON: 3.6 OFF: 6.4
ON: 2.8 OFF: 7.2
N2 N2z
371.05/375.54/380.49/394.30/399.84/405.94 391.44
Surface Engineering
2006
VOL
22
NO
5
Cruz et al.
Ion nitrided AISI H13 tool steel: Part 1
a 75B1; b 75B2; c 25B1; d 25B2 2 Optical emission spectra for (N2zH2) plasma under each condition under study
N2 (380.5 nm of C 3 Pu ?B3 Pg transition) are shown in Fig. 3, indicating that under the same current density the N2z species have similar intensities. This results show that, under the processing conditions employed, the N2 concentration does not depend on the generation of N2z species inside of plasma. Another important aspect to be observed in Fig. 3 is the relative intensities of the N2 excited states, which varied under different conditions. The way in which this variation affects the compound zone generation will be shown in the next section.
3 Spectral intensities corresponding to different processing conditions
X-ray diffraction and microstructure Figure 4 shows the XRD pattern of the untreated AISI H13, which presents well defined peaks corresponding to a-Fe. However, in the literature, it was reported that AISI H13 steel under annealed conditions contained a fine dispersion of M7C3 carbides,14 which could not be detected during the X-ray analysis in the present work. Figure 5 shows the XRD pattern of all the samples treated by plasma nitriding grouped according to the N2 content in the gas mixture. As observed, all the samples which were nitrided employing 75%N2 show the presence of a compound layer. This compound layer has a predominant single nitride structure (e-Fe2–3N), which corresponds to the 75B1 condition, and a mixture
4 X-ray pattern corresponding to AISI H13
Surface Engineering
2006
VOL
22
NO
5
361
Cruz et al.
Ion nitrided AISI H13 tool steel: Part 1
erosion rate can be expressed as17 R~62:3Jd YMa =r
5 X-ray diffractions patterns under each plasma nitriding condition (Bragg-Brentano geometry)
of e-Fe2–3N/c-Fe4N which corresponds to the 75B2 condition. When the nitriding processes was performed by using a gas mixture containing only 25%N2, i.e. conditions 25B1 and 25EB1, only the diffusion zone was observed. This diffusion zone, which was present in all samples, is characterised by the existence of similar diffraction peaks that correspond to a-Fe, but with a typical peak shifting and broadening, as consequence of the increase in the lattice constant owing to the saturation of a-Fe lattice with nitrogen.15,16 According to the literature,15,16 in all nitriding processes it is well recognised that under each fixed processing condition there is a critical nitrogen potential, called the threshold nitriding potential, below which an iron nitride layer does not form on the surface during nitriding. However, although its dependence on the main processing parameters such as nitriding time, nitrogen gas concentration and temperature has been already indicated, the influence of the current density has not been reported previously. As it will be shown below, it was found that the current density affects the formation of the compound layer and this effect is clearly observable from the XRD pattern which corresponds to the 25B2 processing condition (see Fig. 5). For the latter, the formation of a predominant c-Fe4N single nitride structure is observed, although it is characterised by the same nitrogen potential which corresponds to both 25B1 and 25EB1 conditions. The current density inside the plasma affects the kinetics of nitride formation owing to the erosion rates R associated to the sputtering component over the sample surface under plasma nitriding processing. This
362
Surface Engineering
2006
VOL
22
NO
5
where Jd is the current density, Y is the sputtering yield and Ma and r are the atomic weight and density of the material under bombardment respectively. Under conditions with higher current density (5.6 and . 4 17 mA cm22) the sputtering yield was lower and therefore, the formation of a compound layer was favoured. The results obtained by optical spectroscopy allow the corroboration of the XRD findings, since as it was indicated in Fig. 3, samples which have shown the presence of a compound layer have higher emission intensity of N2 (380.5 nm) bands than those for which the compound layer was absent. These results agree with those reported by Marchand et al.3 who represented the phenomena at the plasma/solid interface, based on the mass balance of nitrogen in solution at the solid surface, which accounts for the reaction of the nitrogen active species from both plasma and the cathodic sputtering of the surface by the heavy high energy particles (ions and fast neutral species). In the present study, therefore, higher emission intensities from N2 (380.5 nm) indicated that near the cathodic surface (surface under treatment) a higher density of molecules exited to a low energy level causing the generation of a compound layer. The nitriding potential (%N), however, will determine the microstructure of the compound layer, i.e. if it will be composed of a single or two phase layer. In order to corroborate the results obtained by doing XRD analysis, a series of SEM images, in backscattering mode (BSE), of the cross-sections of the samples produced according to the processing conditions indicated in the Table 1 are shown in Fig. 6. As it is observed, samples 75B1, 75B2 and 25B2 present a uniform compound layer with the microstructural characteristics indicated previously. However, samples 25B1 and 25EB1, as expected, show the presence of the diffusion zone only. Figure 7 presents a detail of the diffusion zone with the corresponding elemental mapping for Cr, N, Mo and V. As it is observed, Mo and V indicate the presence of globular carbides, characteristic of annealed tool steels.14 The presence of zones around the ferritic grains, which are rich in chromium, coupled with the information from the BSE micrograph (darker zones) are indicative of the presence of CrN enriched precipitates, which were also detected by XRD (see Fig. 5). It was shown that the presence of CrN precipitates contributes considerably to the increase in the hardness values, from y600 HV, which corresponds to pure iron nitrided,18 to nearly 1200 HV for the nitrided sample were chromium was present.19 In case of the plasma nitrided H13 steel, the presence of CrN precipitates has been also confirmed by using TEM techniques.20 Table 4 shows the thickness of the compound layer corresponding to 75B1, 75B2 and 25B2 processing conditions. It is clearly observable that the highest compound layer thickness was obtained under the 75B2 condition, which is characterised by the lowest current density and highest nitrogen potential. Since the extension of the diffusion zone is difficult to measure by means of standard metallographic techniques, the following criterion was used in order to further analyse the evolution of the case depth.
Cruz et al.
Ion nitrided AISI H13 tool steel: Part 1
a 75B1; b 75B2; c 25B1; d 25B2; e 25EB1 6 Images (SEM) (BSE mode) of nitrided samples cross-sections
7 Energy dispersive X-ray analysis mapping of diffusion zone Table 4 Microstructural characteristics of samples under study Compound Thickness Thickness diffusion zone Thickness diffusion zone % area in diffusion zone Sample layer compound layer, mm up to 900 HV0.05, mm up to 600 HV0.05, mm of CrN enriched precipitated 75B1 75B2 25B1 25B2 25EB1
ezc ezc – c –
6 12 – 3 –
69 128 30 87 65
76 133 35 102 75
4.77 7.63 3.91 5.69 3.31
Surface Engineering
2006
VOL
22
NO
5
363
Cruz et al.
Ion nitrided AISI H13 tool steel: Part 1
8 Hardness profiles nitrided samples
obtained
from
cross-sections
of
The case depth was measured between the points where the hardness reached 600 and 900 HV0.05 respectively. 600 HV0.05, corresponds to the standard hardness value of a hardened tool steel and 900 HV0.05 to an arbitrary value related to the zone where the hardness begin to decrease. The area percentage of the diffusion zone of CrN enriched precipitated (see Table 4) was measured from SEM images at a magnification indicated in Fig. 7. Each result corresponds to the mean value obtained from five fields which were taken at y5 mm below the surface. These results show that a decrease in the current density value favoured the compound layer formation owing to a lower sputtering rate. In this case, the diffusion of nitrogen allows the formation of rich precipitates of CrN, as consequence of the decomposition of the globular carbides rich in chromium.21
Hardness profiles Figure 8 presents a comparison between the changes in microhardness as a function of the distance from the sample surface under the different plasma nitriding processing conditions used. In these graphs, each point represents the mean value from five indentations with a maximum standard deviation of 4%. The effect of the different nitriding conditions is clearly observable in Fig. 8. As it could be noticed, the values of both current density and nitrogen potential appear to be the principal variables that affected the extent of the diffusion zone. It is also found, that the nitriding time is the variable with less impact in the generation of a deeper diffusion zone, as it can be observed for the samples which were nitrided by using 25%N2 in the gas mixture, where an increase in duration of the nitriding process from 3 to 10 h (conditions 25B1 and 2EB1 respectively) produced an increase of 200% in the diffusion zone depth. However, the decrease in the current density value from 4.17 to 3.12 mA cm22, maintaining the duration of the nitriding process for 3 h (25B1 and 25B2 respectively), had a more pronounced effect on the diffusion zone extension, i.e. producing an increase of almost 300%. Figure 9 shows the hardness profiles and the corresponding nitrogen profiles measured. Each nitrogen profile was measured from the beginning of diffusion zone inside the ferrite grains by EDS techniques.
364
Surface Engineering
2006
VOL
22
NO
5
As expected, the nitrogen decay corresponds to the hardness decay across the cross-section of the sample. What it is interesting to notice is the localised decrease in the nitrogen concentration inside the diffusion zone, which also affected the hardness profile. This behaviour was associated to the ‘erosion effect’, which has been explained recently by Alves et al.11 It was shown that this phenomenon could produce zones which are characterised by a small amount of nitrogen within zones with high nitrogen content. This localised decrease in hardness is more evident for the sample 25B2. This can be related to the contribution of CrN precipitates to the bulk hardness, which is lower in the sample 25B2 (see Table 4). Therefore, the diffusion zone is more sensible to hardness changes related to the decrease in the nitrogen concentration of the ferrite grains. The geometrical configuration of the cathode, probably could also affect the uniformity of the diffusion zone owing to a localised heating and the response to this phenomenon has been detected from hardness measurements across the sample cross-section.11 The experimental conditions tested in the present investigation have allowed obtaining different combinations between the products, i.e. the presence or not of the compound layer and the diffusion layer during plasma nitriding of this steel. Owing to numerous industrial applications of these systems, any condition which provides a specific microstructural feature could be of interest. Research related to ductility, abrasion and resistance to friction of the deep drawn products have shown22 that the role of the compound layer depends principally of two characteristics such as homogeneity and thickness. It was found that the mechanical properties improve considerably if the compound layer is composed mainly of one phase, i.e. either Fe4N or Fe2–3N. In the case of the existence of both phases, high residual stresses are present which are inherent to the region where the phase transition occurs between them, as a result of their different structures (fcc and hexagonal respectively), which will generate microcracks even for smaller applied external stresses. This phenomenon could have a deleterious effect on the tribological response of this material, since it will promote the generation of a high amount of debris which will act as third body abrasive particles in the contact. Therefore, the nitrided steel produced under condition 75B2 could not be used as obtained in service unless prior removal of the biphasic compound layer is performed, which will be much difficult for components with complex geometry. As regards ductility, it is well known that this property decreases with the increase in the thickness of the compound layer.22 As a consequence, the optimum tribological properties will be achieved when the compound zone will be constituted by one phase only and will have a thickness sufficient to insure both a good abrasion and corrosion resistance. In this case, the experimental conditions corresponding to 75B1 and 25B2 could eventually offer this behaviour especially in industrial applications related to their use for synchroniser components for transmissions (components with close dimensional tolerance) and screws and cylinders for the extrusion of plastic materials respectively. The experimental conditions which provided only the production of the diffusion layer (25B1 and 25EB1)
Cruz et al.
Ion nitrided AISI H13 tool steel: Part 1
a 75B1; b 75B2; c 25B1; d 25B2; e 25EB1 9 Hardness and nitrogen profiles
could be of importance for the production of the reduction gears for marine steam turbines, where the ion nitriding practice makes possible the selective nitriding of the gears teeth. Good performance of components such as deep drawing punches, where high compressive stresses are present, could also be achieved by employing materials obtained from the experimental conditions 25B1 and 25EB1, the later offering an extended diffusional zone allowing, thus, a higher load carrying capacity of the system, with a homogeneous distribution of rich CrN precipitates.
Conclusions The following conclusions can be derived from the results obtained in the present work:
1. Optical spectroscopy was a useful technique to determine the relationships among the actives species generated during processing and the microstructural configuration of the treated samples. Two spectral intensities were analysed: N2z (391.4 nm) and N2 (380.5 nm). The relative intensity of N2z was determined by the current density applied during the process, whereas the relative intensity of N2 offered valuable information which could be related to the compound layer generation. 2. The nitrogen potential is an important parameter on the generation of a compound layer, but it is not the only one to be taken in account. In the present work, it has been demonstrated that the current density value has also an important influence on both compound layer formation and the depth of the diffusion zone.
Surface Engineering
2006
VOL
22
NO
5
365
Cruz et al.
Ion nitrided AISI H13 tool steel: Part 1
3. The geometrical configuration of the workload may affect the uniformity of the diffusion zone, generating an ‘erosion effect’, which gives rise to the appearance of heterogeneous nitrogen distribution zones along the diffusion zone. In the case of alloyed steels such as H13 tool steel employed in the present work, this effect could be masked by the precipitation of CrN intermetallic compounds. Further investigation will be conducted to characterise the microstructural configuration of ion nitrided specimens obtained for a wide range of H2/N2 ratios and total pressure inside the chamber. Special consideration will be given to the geometrical configuration and its relation with the plasma kinetics.
Acknowledgements The authors wish to acknowledge the financial support received from Fondo Nacional de Ciencia, Tecnologı´ a e Innovacio´n (FONACIT) through the project S12001000759, project UCV F-2001000600 and to CDCH-UCV PI 08-17-5120-2003.
References 1. R. Gru¨n and H.-J. Gu¨nther: Mater. Sci. Eng. A, 1991, 140A, 435– 441. 2. K. T. Rie: Surf. Coat. Technol., 1999, 112, 56–62. 3. H. Michel, T. Czerwiec, M. Gantois, D. Ablitzer and A. Ricard: Surf. Coat. Technol., 1995, 72, 103–111.
366
Surface Engineering
2006
VOL
22
NO
5
4. T. Czerwiec, H. Michel and E. Bergmann: Surf. Coat. Technol., 1998, 108–109, 182–190. 5. S. P. Bru¨hl, M. W. Russell, B. J. Go´mez, G. M. Grigioni, J. N. Feugeas and A. Ricard: J. Phys. D, 1997, 30D, 2917–2922. 6. L. F. Zagonel, C. A. Figueroa, R. Droppa, Jr and F. Alvarez: Surf. Coat. Technol., 2006, 201 (1–2), 452–457. 7. L. H. Corredor, B. Chornik and K. Ishizaki: Scr. Metall., 1981, 15, 195. 8. I. Alphonsa, A. Chainami, P. M. Raole, B. Ganguli and P. I. John: Surf. Coat. Technol., 2002, 263, 146–150. 9. G. Nayal, D. B. Lewis, M. Lembke, W.-D. Mu¨nz and J. E. Cockrem: Surf. Coat. Technol., 1999, 111, (2–3), 148–157. 10. A. R. P. de Ataı´de, C. Alves, Jr., V. Hajek and J. P. Leite: Surf. Coat. Technol., 2003, 167, 52–58. 11. C. Alves, Jr., E. F. Da Silva and A. E. Martinelli: Surf. Coat. Technol., 2001, 139, 1–5. 12. B.-Y. Jeong and M.-Ho Kim: Surf. Coat. Technol., 2001, 141, 182– 186. 13. Y. M. Kim, J. U. Kim and J. G. Han: Surf. Coat. Technol., 2002, 151–152, 227–232. 14. G. Krauss: ‘Steels: heat treatment and processing principles’, 406– 410; 1989, Materials Park, OH, ASM International. 15. A. Alsaran and A. Celik: Mater. Charact., 2001, 47, 207–213. 16. Y. Sun and T. Bell: Mater. Sci. Eng. A, 1991, 140A, 419–434. 17. N. A. G. Ahmed: ‘Ion plating technology, developments and applications’, 5–8; 1987, New York, John Wiley and Sons. 18. E. J. Miola, S. D. de Souza and M. Olzon-Dionysio: Surf. Coat. Technol., 2003, 167, 33–40. 19. B.-Y. Jeong and M.-Ho Kim: Surf. Coat. Technol., 2001, 137, 249– 254. 20. O. Salas, J. Oseguera, N. Garcı´a and U. Figueroa: JMEPEG, 2001, 10, 649–655. 21. B. Edenhofer: Heat Treat. Met., 1974, 1, 23–28. 22. B. Edenhofer: Heat Treat. Met., 1974, 2, 59–67.
