- Email: [email protected]
Improvement in corrosion resistance of a nodular cast iron surface modified by plasma beam treatment
Applied Surface Science 286 (2013) 334–343
Contents lists available at ScienceDirect
Applied Surface Science journal homepage: www.elsevier.com/locate/apsusc
Improvement in corrosion resistance of a nodular cast iron surface modified by plasma beam treatment Xiu Cheng, Shubing Hu ∗ , Wulin Song, Xuesong Xiong State Key Laboratory of Material Processing and Die and Mould Technology, Huazhong University of Science and Technology, Wuhan, Hubei 430074, China
a r t i c l e
i n f o
Article history: Received 4 July 2013 Received in revised form 12 September 2013 Accepted 14 September 2013 Available online 21 September 2013 Keywords: Nodular cast iron Surface modification Plasma beam Tempering Corrosion resistance
a b s t r a c t Nodular cast iron (NCI) specimens with corrosion-resistant surfaces were fabricated by plasma beam treatment and tempering (400 ◦ C, 1 h), which consisted of plasma surface melting, plasma surface melting + tempering, plasma surface alloying and plasma surface alloying + tempering. In this manner, near-surface graphite nodules were eliminated, and inter-dendrites and eutectics with a hyper-eutectic structure appeared on the modified surfaces, as indicated by SEM. The corrosion behaviour of treated specimens in 3.5 wt% NaCl was characterised by electrochemical methods and compared with that of an untreated NCI specimen at 25 ◦ C. The corrosion resistance ranked as follows: surface-alloyed and tempered specimen > surface-alloyed specimen ≈ surface-melted and tempered specimen > surface-melted specimen > the untreated NCI specimen. Metallographic as well as electrochemical corrosion studies illustrate the beneficial effects of surface modification in refining the microstructure and in enhancing the corrosion resistance of NCI. © 2013 Elsevier B.V. All rights reserved.
1. Introduction Nodular cast iron (NCI) has been widely used in various industrial fields, such as in automotive and machine parts, tubes, drawing moulds and even nuclear waste containers, due to their high strength, high yield limit, toughness and relatively low price, in addition to their excellent castability and machinability [1–4]. The microstructure of NCI consists of graphite phases within and iron matrix [5], where graphite plays the role of a cathode and the adjacent iron that of an anode, which accelerates the anodic dissolution of iron; specifically, during graphitic corrosion, only the network structure of graphite is left behind [6]. From the perspective of surface engineering, the presence of graphite in NCI is a potential problem. Therefore, it is imperative to prevent the formation of nodular graphite immediately beneath the surface of NCI [5], and near-surface graphite phases must be eliminated to improve the corrosion resistance of NCI. Recently, high-energy beams such as electron beams [7,8], laser beams [1,2,9] and plasma beams [10,11] have been used extensively to eliminate superficial graphite nodules in NCI by rapid melting and cooling/solidification processes. Rapid cooling during solidification also leads to the formation of large amount of hypereutectic cementite instead of soft graphite [4,12]. However, because of the high cost of operating lasers and the need for a vacuum in
∗ Corresponding author. Tel.: +86 13995667466; fax: +86 27 87540057. E-mail address: [email protected] (S. Hu). 0169-4332/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.apsusc.2013.09.083
electron beam, attention has been focussed on using inexpensive, flexible and easy-to-operate plasma beam apparatus for the surface treatments of cast iron. A plasma beam is an extremely high-temperature flow usually possessing a power density of approximately 109 W/m2 , which allows for the rapid heating of almost every type of solid material to its melting or evaporating point via a type of rapid, non-equilibrium metallurgical process. The heating efficiency of plasma beam (85%) could be much higher than that of a laser beam (30%) in heating material surfaces [13]. Plasma beams provide significant advantages in treating iron-carbon alloys, including selective hardening, minimum part distortion, controllable case depth and structural refinement [14,15]. As high-energy-intensity heating sources, plasma beams are widely used for surface hardening to improve the corrosion resistance of surface-modified ferrous workpieces [16]. Plasma surface treatments, including plasma surface melting and plasma surface alloying, can be used as cost-effective alternative approaches for the elimination of near-surface graphite in cast iron. Surface melting can enhance the surface properties of ferrous materials, such as their hardness, wear resistance and corrosion resistance. Such enhancement is possible because a high cooling rate can be achieved due to localised melting that results in nonequilibrium phases as well as refined microstructures [17,18]. In general, surface melting can lead to an improvement in the corrosion resistance of all analysed materials [19].
X. Cheng et al. / Applied Surface Science 286 (2013) 334–343
335
Table 2 The composition of alloy powders (wt%). WC
Ni
TiC
Cr
Si
La2 O3
Fe
3
15
20
15
2
0.2
Balance
in Table 2. The powders were mechanically mixed and dispersed in kerosene to form alloy coatings. 2.2. Plasma surface treatment
Fig. 1. Microstructure of as-received NCI, F: ferrite, G: graphite and P: pearlite.
Surface alloying offers a versatile approach to the production of surface layers with a wide range of structures and compositions on a variety of substrates; moreover, the structures may consist of fine grains as a result of the relatively rapid cooling rates that can be achieved from the melt. Consequently, an improvement in corrosion resistance may also be achieved by surface alloying via microstructural homogenisation and refinement and the formation of new alloys on surfaces [20]. The micro-alloying of suitable elements on surfaces may also help enhance surface properties such as corrosion resistance [7]. There have been many studies that have investigated the mechanical properties of NCI [8,9,20] submitted to laser and electron beam treatments, but there are few published works on the corrosion characteristics of NCI surfaces modified by plasma beam treatment. In this study, a plasma beam was used as a heating source to modify the surface of NCI, eliminate superficial graphite nodules and hence improve the corrosion resistance of NCI. The surface treatments included plasma surface melting (M method), plasma surface melting + tempering (MT method), plasma surface alloying (A method) and plasma surface alloying + tempering (AT method). Furthermore, the corrosion resistance of the treated specimens compared with that of ferrite and pearlite NCI substrate (S specimen) was tested by potentiodynamic polarisation and electrochemical impedance spectroscopy (EIS).
Plasma surface melting (M method), plasma surface melting + tempering (MT method), plasma surface alloying (A method) and plasma surface alloying + tempering (AT method) were used to modify the surfaces of the NCI specimens. Plasma surface melting and alloying was carried out using a homemade set-up for combined transferred and non-transferred arc plasma beam treatment, as shown in Fig. 2. An atmospheric, high-density plasma was generated by the equipment, and the ionisation degree was less than 0.1%. The plasma pressure was greater than 0.1 MPa. The specimen and nozzle served as the anode, and a tungsten needle served as the cathode (Fig. 2). The nozzle diameter selected for testing was 2 mm, and the distance between the nozzle and the specimen was 4 mm. The plasma torch was controlled by a small variable-speed DC motor; thus, the speed of the torch could be adjusted and held constant throughout the test. The argon used in the process served as both a plasma gas and shielding gas, the flow rates of which were both 1 L/min. In the surface-melting process, a plasma beam with a Gaussian energy density distribution was defocussed on the surface of the NCI substrates without a pre-layer. For surface alloying, a homemade alloy coating was sprayed to form a pre-layer with a thickness of 200 m using an ejection gun on the NCI substrates before the plasma surface alloying process. The scanning speed during the plasma surface-alloying process should be lower due to the pre-layer. In the plasma surface-alloying process, the plasma beam energy melted the alloying coatings and substrate surface, and the alloy elements then seeped into the substrate to form a new alloy surface.
2. Experimental 2.1. Materials The starting material NCI (QT600-3) was produced by Dongfeng Company. The NCI was machined into plates with dimensions of 120 mm × 80 mm × 20 mm for plasma surface treatment. Fig. 1 shows a micrograph of an NCI specimen (S specimen), which consists of graphite nodules with an average diameter of 20 m surrounded by ferrite and pearlite. The chemical composition of the specimen is presented in Table 1. Alloy powders with a particle size of 1–3 m were used as an alloying layer material, the chemical composition of which is listed Table 1 Chemical composition of NCI (wt%). C
Si
Mn
S
P
Mg
Cu
Fe
3.41
2.23
0.42
0.033
0.045
0.048
0.37
Balance Fig. 2. Schematic diagram of the plasma beam heating of the nodular cast iron.
336
X. Cheng et al. / Applied Surface Science 286 (2013) 334–343
Table 3 The parameters of plasma surface melting and plasma alloying.
The nozzle diameter The distance between the nozzle and specimen Plasma torch moving speed Working current Argon of flow rate for plasma gas Argon of flow rate for shielding gas Overlapping on the former trace Arc voltage
Plasma surface melting
Plasma surface alloying
2 mm 4 mm
2 mm 4 mm
640 mm/min 50 ± 0.5 A 1 L/min 1 L/min 50% 30–48 V
320 mm/min 50 ± 0.5 A 1 L/min 1 L/min 50% 30–48 V
The plasma surface-melting and surface-alloying parameters used to produce the melted and alloyed surfaces are presented in Table 3, which were determined after a series of pre-tests. The specimens treated by plasma surface melting and surface alloying were denoted M specimens and A specimens, respectively. Coronado et al. [21] studied the effects of the tempering temperature on the abrasive wear of mottled cast iron. Their results demonstrated that no significant microstructural changes were observed after tempering at 300 ◦ C and 400 ◦ C. Therefore, the tempering temperature should be below 400 ◦ C. MT and AT specimens were obtained from the M and A specimens after tempering, respectively. The tempering treatment was carried out as follows: (1) heat to 400 ◦ C for 1 h in an air furnace and (2) furnace cool to room temperature. 2.3. Microstructure analysis After plasma surface treatment and tempering, the treated surfaces were ground by approximately 20 m to obtain an oxide-free surface. Then, the treated specimens were sectioned, mounted, polished and etched with a 4 vol% nitric acid alcohol solution. Scanning electron microscopy (SEM, Quanta 200) was used to examine the cross-sectional microstructure as well as the corroded surface. The phase structures of the modified surfaces were analysed using an X’ ˚ operPert PRO diffractometer with Cu-K␣ radiation ( = 1.5406 A) ated at 40 kV and 40 mA. The grain size and porosity of the modified surfaces were measured using the imaging tool Image-Pro Plus version 7.0 [22]. 2.4. Corrosion test The corrosion properties of the specimens were investigated by polarisation curve and EIS measurements in this study. Specimens used in electrochemical testing were machined into 1-cm cylindrical electrodes. Prior to the start of the corrosion tests, the specimens were wet ground using SiC paper with a grit number of up to 1000. Then, the specimens were washed in distilled water and ethanol and finally dried in warm air. All of the electrochemical tests were conducted in a conventional three-electrode system (CS350 electrochemical corrosion workshop) with the specimens as the working electrode (WE), Pt as the counter electrode (CE) and a saturated calomel electrode (SCE) as the reference (RE). The contact area in all cases was 0.785 cm2 , and the tests were carried out at 25 ◦ C. Polarisation tests were conducted vs. OCP (open circuit potential) after 1 h of immersion and performed in 3.5 wt% NaCl solution at a scan rate of 0.5 mV/s. The electrode potential was raised from −1000 mV to −500 mV. The corrosion behaviour of the NCI specimens was evaluated by anodic potentiodynamic polarisation tests [3]. Therefore, the corrosion potentials as well as the corrosion current densities were extracted from the plots by fitting the anode polarisation curve using C-view software.
EIS measurements were also carried out using the aforementioned equipment. After the working electrode was immersed in the electrolyte solution of the cell and its OCP became stable, its EIS was measured at the OCP with applied sinusoidal perturbations of 5 mV (peak-to-peak) at frequencies ranging from 10 kHz to 0.05 Hz, 10 steps per decade. EIS measurements were performed every 2 h of immersion up to a total of 48 h, and the same frequency range was used. The spectra were analysed using Z-view software. All experiments were repeated three times, and the results were reproducible; thus, representative data are presented in this paper. After 48 h of immersion, the surface and cross-sections of the corroded specimens were prepared and examined by SEM. 3. Results and discussion 3.1. The characterisation of surface-modified specimens Cross-sectional SEM micrographs of surface-modified specimens are shown in Fig. 3a and b. It was observed that the molten zone (MZ) and alloyed zone (AZ) were well bonded to the bulk region of the specimens with no distinctive irregular interface. In addition, it was evident from these micrographs that the graphite nodules had dissolved in the molten ferrous material during treatment, and the re-formation of graphite nodules had been suppressed by rapid self-quenching in favour of the formation of white cast iron, which enhanced the corrosion properties of the surface [7]. The molten zone on the surface-melted specimens (Fig. 3a) was approximately 105 m deep, whereas the depth of the AZ was approximately 300 m, although the pre-layer of the alloy coating was approximately 200 m. Thus, there was less epitaxial growth during the plasma alloying process when alloying particles were melted and penetrated into the substrate surface. The lower scanning speed and improvement of the hardenability owing to the alloy elements led to the deeper AZ. Fig. 4 shows high-magnification images of the MZ and AZ before and after tempering. The microstructure of the surfaces consisted of dendrites and superfine grains. This dendritic structure is similar to that of white cast iron observed by many researchers during the laser- or electron-beam melting of NCI [1,7,9]. The interdendrites were characterised by long primary and secondary arms; the secondary arm spacing was less than 0.5 m, which indicates a considerably high cooling rate during solidification [1,9]. Moreover, the fact that no cracks were observed in the micrographs indicates that the severe quenching of the plasma-beam-irradiated specimens did not cause noticeable microdefects. The molten zone showed a very fine structure that mainly consisted of austenite dendrites with an interdendritic network of carbides, as shown in Fig. 4a. In addition, some of the dendrites contained martensite [2], which indicated a high cooling rate during solidification. As reported previously, the microstructure of the alloyed zone consisted of eutectic ledeburite, alloy carbide and martensite [9]; in this case, dendrites along with a ledeburite structure were observed in the AZ, as shown in Fig. 4c. Due to the fast cooling after plasma surface alloying, a fine, cellular dendritic structure was formed and the ledeburite structure was similar to that of white cast iron. Compared with that of the specimens that were not tempered, the microstructure of the tempered specimens was more uniform and compact, which can be attributed to the formation of a hard phase and solid solution in the modified surfaces that were sufficiently high in concentration and homogeneous after tempering [23–25]. Moreover, the quenching stress may have relaxed during the tempering process to reduce the specimens’ susceptibility to stress corrosion.
X. Cheng et al. / Applied Surface Science 286 (2013) 334–343
337
Fig. 3. Cross-sectional SEM micrographs of surface-modified specimens (a) M specimen, (b) A specimen.
The average grain size in the modified surfaces was superfine, as indicated in Figs. 4 and 5. Because the treatment temperature of 400 ◦ C was not sufficiently high to make the grains grow to a great extent, the grain size remained superfine after tempering. In the plasma surface-alloying process, by shifting the plasma beam, the alloy coating and a thin surface layer of substrate material were melted simultaneously, and some compounds as well as TiC, Ni, Cr, WC and Si formed particles by heterogeneous nucleation. Thus, the number of crystal nuclei was increased during the crystallisation of the coatings, which was beneficial in increasing the nucleation rate and promoted non-spontaneous nucleation. The greater the number of crystal nuclei there are, the finer the crystal grains of the crystal become [16]. Therefore, the grain size in the molten zone appeared to be coarser than that in the alloyed zone, as shown in Figs. 4 and 5.
Plasma surface treatment is a technique that is employed under nonhomogeneous and nonequilibrium conditions, which results in multi-phase formation. The XRD results of the M specimen and MT specimen are shown in Fig. 6a and b. Both of the patterns reveal the presence of ␣-Fe (ferrite), ␥-Fe (austenite), martensite and Fe3 C. The formation of Fe3 C in the molten surfaces is attributed to the dissolution of graphite during melting that spread throughout the iron matrix. The high cooling rate of the molten pool resulted in a super-saturated austenite with carbon atoms. The phase of the A and AT specimens was similar, as shown in Fig. 6c and d. The alloyed surfaces revealed the existence of several types of phases, i.e., ␣-Fe (ferrite), ␥-(Fe, Ni) (austenite), martensite, Fe3 C, carbides TiC, Cr23 C6 , Cr7 C3 and W2 C, representing various types of carbides compared with the molten surface due to solution strengthening and chemical reaction with C in the rapid heating
Fig. 4. High-magnification images of the MZ and AZ before and after tempering (a) M (b) MT (c) A (d) AT.
338
X. Cheng et al. / Applied Surface Science 286 (2013) 334–343
Fig. 5. The average grain size in the surface of the treated and untreated specimens. Fig. 7. Polarization curves of S, M, MT, A and AT specimens in 3.5 wt% NaCl solution.
process induced by the plasma beam. The patterns demonstrated that Ni, Cr, TiC and WC could dissolve in the iron matrix of nodular cast iron. The formation of the ␥-(Fe, Ni) phase was likely due to the interaction of Ni with Fe during the solidification of the molten pool [7]. 3.2. Electrochemical test 3.2.1. Polarisation curves The polarisation behaviour of the NCI treated by different processing methods (S, M, MT, A, AT specimens) in 3.5 wt% NaCl solution at 25 ◦ C is shown in Fig. 7. The values of the corrosion potential (Ecorr ), corrosion current density (Icorr ) and polarisation resistance (R) were extracted from the polarisation curves by fitting the anode polarisation curve and are shown in Table 4. Tafel extrapolation was executed in the anodic potential region −150 mV away from Ecorr over one-half decade of current density. It is well known that the corrosion current density (Icorr ) is a good measure of a material’s corrosion resistance in a particular environment. The corrosion current is used to estimate a material’s mass loss as per Faraday’s first law. Hence, the higher the corrosion current is, the higher the mass loss and rate of corrosion become [26]. As shown in Fig. 7 and Table 4, the Ecorr of the modified surfaces was more positive than that of the untreated NCI specimen, which indicates that the modified surfaces were nobler than the untreated
Fig. 6. XRD patterns of the modified surfaces (a) M, (b) MT, (c) A, (d) AT.
NCI, and thus, the corrosion of the untreated NCI specimen could be favoured over that of the modified surfaces. In addition, the tempered specimens showed a higher potential than the specimens treated only by the plasma beam. On the other hand, the corrosion current density of the untreated NCI specimen was higher than that of the modified surfaces. Based on the estimated corrosion current densities, the protection efficiency () of different specimens can be calculated according to the following equation: (%) =
0 Icorr − Icorr 0 Icorr
× 100%
(1)
0 where Icorr and Icorr represent the corrosion current densities of untreated and treated NCI specimens, respectively, in 3.5 wt% NaCl solution. The estimated Icorr values and calculated protection efficiencies shown in Table 4 clearly indicate that the corrosion resistances of different specimens follow the order AT > A ≈ MT > M > S. The protection efficiency of the AT specimen was the highest (54.5%), indicating that the modified surface of NCI treated by plasma surface alloying and tempering had a moderate protective effect. The polarisation resistance of the AT specimen reached up to 9036 cm2 , which was 8.2 times that of S specimen (1098 cm2 ). The polarisation resistances of the A, MT and M specimens were 7671 cm2 , 7459 cm2 and 5457 cm2 , respectively. It can be concluded that the corrosion resistance of the NCI surfaces modified by plasma beam treatment was better than that of the untreated NCI specimen based on the abovementioned analysis. Moreover, the corrosion resistance of the alloyed surface was better than that of the molten surface, and the tempered specimens showed better corrosion resistance.
3.2.2. EIS Electrochemical impedance spectroscopy is well suited for monitoring any perturbation by an inhibitor in situ according to the electrochemical processes that occur at the interface of a metal and corrodent. To further understand the protective behaviour of the modified surfaces, EIS was employed to investigate the effects of different surface treatment processes on the corrosion of NCI. Fig. 8 shows the variation in the Nyquist plots of different specimens immersed in 3.5 wt% NaCl solution at various times. The resistance decreased with increasing immersion time. However, the S specimens showed a quicker speed of reduction than the other treated specimens due to the galvanic corrosion between graphite and iron.
X. Cheng et al. / Applied Surface Science 286 (2013) 334–343
339
Table 4 Electrochemical polarization parameters for various ductile irons in 3.5 wt% NaCl solution at 25 ◦ C. Specimens
Ecorr (V/SCE)
Icorr (A/cm2 )
R ( cm2 )
Protection efficiency (%)
Repetition rate (ı%)
S M MT A AT
−0.875 −0.839 −0.751 −0.781 −0.728
6.07 4.38 3.98 3.99 2.76
1098 5457 7459 7671 9036
– 27.8 34.4 34.3 54.5
95 96 95 98 97
The equivalent circuit shown in Fig. 9 can be used to analyse the EIS results, where Rs is the solution resistance; CPEdl is a constant phase element corresponding to the double-layer capacitance of the metal/solution interface; and Rp is the charge transfer resistance of the interface. The initial values were fed into the Z-view program to further fit the EIS data using the equivalent circuit model (Fig. 9) to obtain
the final Rs and Rp results. In all cases, good conformity between the curve-fitted and experimental results was obtained within an average error of only 5%. Fig. 10 shows the fitting parameters obtained for the different specimens in the NaCl solution using the equivalent circuit depicted in Fig. 9. The Rp values of different specimens in NaCl solution decreased with immersion time. However, the Rp values of
Fig. 8. EIS of different specimens in 3.5 wt% NaCl solution (a) S, (b) M, (c) MT, (d) A, (e) AT.
340
X. Cheng et al. / Applied Surface Science 286 (2013) 334–343
Fig. 9. Equivalent circuit for the modified surface in NaCl solution.
the specimens treated by plasma surface modification were greater than the Rp of the untreated NCI specimen in the NaCl solution. As the resistance increased, the corrosion protection was enhanced and fewer defects were formed on the modified surfaces [27], which suggests that the modified NCI surfaces retarded the corrosion at the metal-solution interface. The results of EIS are in agreement with the potentiodynamic polarisation results. One of the reasons for this phenomenon could be the formation of superfine grains in the modified surfaces, shown in Figs. 4 and 5, as reported by Kwok [28]. Ralston et al. [29] suggested a type of Hall–Petch correlation between Icorr and d (grain size) Icorr = a + bd−1/2
(2)
where a and b are constants whose values depend on the material being studied and on the surrounding corrosive environment. When the grain sizes are in the range of 75–1150 m, the decrease in corrosion resistance is attributed to the higher density of grain boundaries and other defects induced by refinement [26]. The introduction of grain boundaries, the high-energy areas in a microstructure, can lead to an increase in the corrosion rate by the formation of a galvanic couple between relatively anodic grain boundaries and cathodic grain interiors [30]. However, this behaviour is expected to change again when the level of grain refinement is very high while at the transition grain size, with a
Fig. 10. The Rp of different specimens as functions of immersion time in 3.5 wt% NaCl solution.
critical value of approximately 75 m, although this has not yet been established [30]. The grain sizes in the modified surfaces followed the order S (93.7 m) > MT (17.8 m) > M (15.4 m) > AT (11.8 m) > A (9.7 m), as indicated in Fig. 1, Fig. 4 and Fig. 5. Therefore, the corrosion current density of the modified surface would not obey the Hall–Petch type of correlation because the mean grain size of the substrate was greater than 75 m, whereas that of the modified surface was less than 75 m. The alloyed specimens showed better corrosion resistance than the melted specimens due to the finer grains of the alloyed specimens. In the ultrafine-grained regime (an average grain size below 70 m) [30], the very high fraction of grain boundaries is likely to reduce corrosion rates in two ways: (1) by accelerating the passivation kinetics and (2) by reducing the intensity of galvanic coupling between grain interiors and grain boundaries. Therefore, an ultrafine microstructure would lead to the formation of closely spaced electrochemical cells between anodic and cathodic regions, and the difference between the rates of anodic and cathodic reactions is expected to vary significantly, which would lead to more uniform attack and lower corrosion rates. This phenomenon explains why the modified surfaces with a superfine grain size exhibited noble corrosion properties. 3.2.3. The microstructure of the specimens after corrosion NCI usually consists of a graphite phase and a base phase (Fe). The graphite phase is the cathodic phase, which could accelerate the corrosion of Fe. As indicated in Fig. 11, which shows the S specimen immersed in 3.5 wt% NaCl solution for 48 h, there were many corrosion products on the surface of the untreated NCI specimen; the graphite was projected relative to the matrix and showed features typical of graphitic corrosion. The results of the EDS analysis of the corrosion product shown in Fig. 11 indicate that the product contained mainly elemental Fe, Cl, O and Si. In other words, the graphite acted as the cathode and the Fe matrix as the anode in the solution, which caused the dissolution and corrosion of the anode around the graphite, leaving only a projection of the graphite, as observed by Jeong et al. [6] and Zeng et al. [31]. Fig. 12 shows typical SEM images of the corroded surface of a treated specimen after 48 h of immersion in 3.5 wt% NaCl solution. The plasma-beam-treated specimens presented slight corrosion compared to the untreated NCI specimen (Fig. 11). Additionally, in this case, the near-surface graphite nodules were dissolved, dense surfaces were formed and no graphitic corrosion occurred. The dense structures indicate that the modified surfaces possessed low porosity. From the micrographs shown in Figs. 1 and 4, the overall porosities were estimated to be S (3%) > M (0.7%) > MT (0.3%) = A (0.3%) > AT (0.1%). It is known that pores and inclusions play important roles in the corrosion behaviour of materials: a high porosity allows for the penetration of solution, whereas large inclusions lead to local corrosion [32]. The ledeburite structure presented on the modified surfaces is similar to that of white cast iron, which can improve the corrosion properties of the surfaces [7]. Moreover, all of the modified surfaces revealed a diffraction peak attributed to austenite, as shown in Fig. 6, which acted as a corrosion inhibitor [3,15]. These factors led to the improvement of the corrosion resistance of the NCI surfaces modified by plasma beam treatment and tempering. Based on the results of the electrochemical tests and corroded surface analysis, the alloyed surfaces exhibited better corrosion resistance than the molten surfaces, and tempering also improved the corrosion resistance. A finer and more densely alloyed surface was formed to improve corrosion resistance after plasma surface alloying. Ni and Cr occurred on the alloyed surfaces, as shown in Fig. 6c and d.
X. Cheng et al. / Applied Surface Science 286 (2013) 334–343
341
Fig. 11. Corroded morphology on the surface of the untreated NCI specimen after 48 h of immersion in 3.5 wt% NaCl solution and EDX of the corrosion product.
Ni is referred to as an austenite-stabilising element [3,33], and austenite is a good inhibitor [3]. Moreover, Ni and Cr distributed on a surface could form a dense oxide film to act as a barrier to the internal diffusion of aggressive media [34]. For these reasons, the alloyed surfaces presented better corrosion resistance. During tempering, hard phases and solid solutions may be distributed more uniformly over a modified surface due to the reheating effect of the tempering, and quenching stress may also be released [23–25]. Furthermore, the porosity of modified surfaces can be reduced after heat treatment [35]. Hence, tempering can reduce the surfaces’ susceptibility to corrosion.
Cross-sectional images of the M specimen as well as the untreated NCI specimen after 48 h of immersion are shown in Fig. 13a and b. No distinct cracks or pitting corrosion can be observed because the immersion time was not sufficiently long. Verdian et al. [34] stated that no obvious indication of blistering or delimitation could be observed in SEM micrographs of the cross-section of an air plasma spraying (APS) coating after 63 days of immersion. In this study, there were no changes in the crosssectional structure of the M specimen before and after corrosion (4a and 13b, respectively). To summarise, it was determined that there was no obvious corrosion of the MT, A and AT specimens along their depth.
Fig. 12. Corroded morphology on the surface of the treated specimens after 48 h of immersion in 3.5 wt% NaCl solution (a) M, (b) MT, (c) A, (d) AT.
342
X. Cheng et al. / Applied Surface Science 286 (2013) 334–343
Fig. 13. Corroded morphology of transverse after the 48 h immersion in 3.5 wt% NaCl solution (a) S specimen, (b) M specimen.
4. Conclusions From the present investigation, the following conclusions can be drawn: (1) Plasma surface melting and plasma surface alloying could eliminate the near-surface graphite nodules of NCI, and a dense and fine eutectic dendritic microstructure was obtained to enhance the corrosion resistance. (2) Specimens showed good corrosion resistance if submitted to plasma surface treatment and tempering. The better corrosion resistance of the alloyed surface in 3.5 wt% NaCl was due to the development of a denser structure and finer grain size. (3) The corrosion resistances of the specimens ranked as follows: surface-alloyed and tempered specimen > surface-alloyed specimen ≈ surface-melted and tempered specimen > surfacemelted specimen > untreated NCI specimen. Acknowledgements This work was supported by the Province and Ministry cooperation of university-industry collaboration (No. 2010090200047) and the national natural science foundation of China (No. 51375005). The authors express their sincere thanks for the help of analytical and testing center of HUST. References [1] K.Y. Benyounis, O.M.A. Fakron, J.H. Abboud, A.G. Olabi, M.J.S. Hashmi, Surface melting of nodular cast iron by Nd-YAG laser and TIG, J. Mater. Process. Technol. 170 (2005) 127–132. [2] K.F. Alabeedi, J.H. Abboud, K.Y. Benyounis, Microstructure and erosion resistance enhancement of nodular cast iron by laser melting, Wear 266 (2009) 925–933. [3] C.H. Hsu, M.L. Chen, Corrosion behavior of nickel alloyed and austempered ductile irons in 3.5% sodium chloride, Corros. Sci. 52 (2010) 2945–2949. [4] J.H. Abboud, Microstructure and erosion characteristic of nodular cast iron surface modified by tungsten inert gas, Mater. Des. 35 (2012) 677–684. [5] M.B. Karamıs¸, K. Yıldzlı, Surface modification of nodular cast iron: a comparative study on graphite elimination, Mater. Sci. Eng., A 527 (2010) 5225–5229. [6] B.Y. Jeong, M.H. Kim, Corrosion characteristics of duplex surfacetreated spheroidal graphite cast iron, Surf. Coat. Technol. 141 (2001) 262–268. [7] A. Gulzar, J.I. Akhter, M. Ahmadb, G. Ali, M. Mahmoodd, M. Ajmal, Microstructure evolution during surface alloying of ductile iron and austempered ductile iron by electron beam melting, Appl. Surf. Sci. 255 (2009) 8527–8532. [8] M.D. Jean, Y.F. Tzeng, Optimisation of electron-beam surface hardening of cast iron for high wear resistance using the Taguchi method, Int. J. Adv. Manuf. Technol. 24 (2004) 190–198.
[9] H. Yan, A.H. Wang, Z.T. Xiong, K.D. Xu, Z.W. Huang, Microstructure and wear resistance of composite layers on a ductile iron with multicarbide by laser surface alloying, Appl. Surf. Sci. 256 (2010) 7001–7009. [10] W.S. Dai, L.H. Chen, T.S. Lui, SiO2 particle erosion of spheroidal graphite cast iron after surface remelting by the plasma transferred arc process, Wear 248 (2001) 201–210. [11] T. Ishida, Local melting of nodular cast iron by plasma arc, J. Mater. Sci. 18 (1983) 1773–1784. [12] M. Shamanian, S.M.R. Mousavi Abarghouie, S.R. Mousavi Pour, Effects of surface alloying on microstructure and wear behavior of ductile iron, Mater. Des. 31 (2010) 2760–2766. [13] W.X. Pan, X. Meng, G. Li, Q.X. Fei, C.K. Wu, Feasibility of laminar plasma-jet hardening of cast iron surface, Surf. Coat. Technol. 197 (2005) 345–350. [14] L. Giordano, A. Tiziani, A. Zambon, Characterization of surface chromium and molybdenum alloying on gray cast iron obtained by the plasma-transferred arc technique, Mater. Sci. Eng., A 140 (1991) 727–732. [15] H. Chen, H.Q. Li, Microstructure and wear resistance of Fe-based coatings formed by plasma jet surface metallurgy, Mater. Lett. 60 (2006) 1311–1314. [16] H. Chen, H. Li, Y.Z. Sun, M. Li, Microstructure and properties of coatings with rare earth formed by DC-plasma jet surface metallurgy, Surf. Coat. Technol. 200 (2006) 4741–4745. [17] M.H. Sohi, G. Karshenas, S.M.A. Boutorabi, Electron beam surface melting of as cast and austempered ductile irons, J. Mater. Process. Technol. 153–154 (2004) 199–202. [18] M.Y. Li, Y. Wang, B. Han, W.M. Zhao, T. Han, Microstructure and properties of high chrome steel roller after laser surface melting, Appl. Surf. Sci. 255 (2009) 7574–7579. [19] A. Conde, R. Colac¸o, R. Vilar, J.D. Damborenea, Corrosion behaviour of steels after laser surface melting, Mater. Des. 21 (2000) 441–445. [20] G.F. Sun, R. Zhou, P. Li, A.X. Feng, Y.K. Zhang, Laser surface alloying of CB-W-Cr powders on nodular cast iron rolls, Surf. Coat. Technol. 205 (2011) 2747–2754. [21] J.J. Coronado, A. Gómez, A. Sinatora, Tempering temperature effects on abrasive wear of mottled cast iron, wear 267 (2009) 2070–2076. [22] K.M. Pedersen, N.S. Tiedje, Graphite nodule count and size distribution in thinwalled ductile cast iron, Mater. Charact. 59 (2008) 1111–1121. [23] H.X. Liu, C.Q. Wang, X.W. Zhang, Y.H. Jiang, C.X. Cai, S.J. Tang, Improving the corrosion resistance and mechanical property of 45 steel surface by laser cladding with Ni60CuMoW alloy powder, Surf. Coat. Technol. 228 (2013) S296–S300. [24] P.D. Bilmes, C.L. Llorente, L.S. Huamán, L.M. Gassa, C.A. Gervasi, Microstructure and pitting corrosion of 13CrNiMo weld metals, Corros. Sci. 48 (2006) 3261–3270. [25] R.A. Carnério, R.C. Ratnapuli, V.F.C. Lins, The influence of chemical composition and microstructure of API linepipe steels on hydrogen induced cracking and sulfide stress corrosion cracking, Mater. Sci. Eng., A 357 (2003) 104–110. [26] S. Gollapudi, Grain size distribution effects on the corrosion behaviour of materials, Corros. Sci. 62 (2012) 90–94. ˇ [27] D.K. Merl, P. Panjan, M. Cekada, M. Maˇcek, The corrosion behavior of Cr-(C,N) PVD hard coatings deposited on various substrates, Electrochim. Acta 49 (2004) 1527–1533. [28] C.T. Kwok, F.T. Cheng, H.C. Man, Microstructure and corrosion behavior of laser surface-melted high-speed steels, Surf. Coat. Technol. 202 (2007) 336–348. [29] K.D. Ralston, N. Birbilis, C.H.J. Davies, Revealing the relationship between grain size and corrosion rate of metals, Scr. Mater. 63 (2010) 1201–1204. [30] G.R. Argade, S.K. Panigrahi, R.S. Mishra, Effects of grain size on the corrosion resistance of wrought magnesium alloys containing neodymium, Corros. Sci. 58 (2012) 145–151.
X. Cheng et al. / Applied Surface Science 286 (2013) 334–343 [31] D.W. Zeng, K.C. Yung, C.S. Xie, Investigation of corrosion behavior of high nickel ductile iron by laser surface alloying with copper, Scr. Mater. 44 (2001) 2747–2752. [32] J.H. Chang, J.M. Chou, R.I. Hsieh, J.L. Lee, Corrosion behaviour of vacuum induction-melted Ni-based alloy in sulphuric acid, Corros. Sci. 52 (2010) 2323–2330. [33] ASTM A439, Annual Book of ASTM Standards, 01.02, Beuth Publications, Berlin, Germany, 1989, pp. 238–242.
343
[34] M.M. Verdian, K. Raeissi, M. Salehi, Corrosion performance of HVOF and APS thermally sprayed NiTi intermetallic coatings in 3.5% NaCl solution, Corros. Sci. 52 (2010) 1052–1059. [35] X.B. Zhao, D.Y. Yan, S. Li, C.G. Lu, The effect of heat treatment on the electrochemical corrosion behavior of reactive plasma-sprayed TiN coatings, Appl. Surf. Sci. 257 (2011) 10078–10083.
Contents lists available at ScienceDirect
Applied Surface Science journal homepage: www.elsevier.com/locate/apsusc
Improvement in corrosion resistance of a nodular cast iron surface modified by plasma beam treatment Xiu Cheng, Shubing Hu ∗ , Wulin Song, Xuesong Xiong State Key Laboratory of Material Processing and Die and Mould Technology, Huazhong University of Science and Technology, Wuhan, Hubei 430074, China
a r t i c l e
i n f o
Article history: Received 4 July 2013 Received in revised form 12 September 2013 Accepted 14 September 2013 Available online 21 September 2013 Keywords: Nodular cast iron Surface modification Plasma beam Tempering Corrosion resistance
a b s t r a c t Nodular cast iron (NCI) specimens with corrosion-resistant surfaces were fabricated by plasma beam treatment and tempering (400 ◦ C, 1 h), which consisted of plasma surface melting, plasma surface melting + tempering, plasma surface alloying and plasma surface alloying + tempering. In this manner, near-surface graphite nodules were eliminated, and inter-dendrites and eutectics with a hyper-eutectic structure appeared on the modified surfaces, as indicated by SEM. The corrosion behaviour of treated specimens in 3.5 wt% NaCl was characterised by electrochemical methods and compared with that of an untreated NCI specimen at 25 ◦ C. The corrosion resistance ranked as follows: surface-alloyed and tempered specimen > surface-alloyed specimen ≈ surface-melted and tempered specimen > surface-melted specimen > the untreated NCI specimen. Metallographic as well as electrochemical corrosion studies illustrate the beneficial effects of surface modification in refining the microstructure and in enhancing the corrosion resistance of NCI. © 2013 Elsevier B.V. All rights reserved.
1. Introduction Nodular cast iron (NCI) has been widely used in various industrial fields, such as in automotive and machine parts, tubes, drawing moulds and even nuclear waste containers, due to their high strength, high yield limit, toughness and relatively low price, in addition to their excellent castability and machinability [1–4]. The microstructure of NCI consists of graphite phases within and iron matrix [5], where graphite plays the role of a cathode and the adjacent iron that of an anode, which accelerates the anodic dissolution of iron; specifically, during graphitic corrosion, only the network structure of graphite is left behind [6]. From the perspective of surface engineering, the presence of graphite in NCI is a potential problem. Therefore, it is imperative to prevent the formation of nodular graphite immediately beneath the surface of NCI [5], and near-surface graphite phases must be eliminated to improve the corrosion resistance of NCI. Recently, high-energy beams such as electron beams [7,8], laser beams [1,2,9] and plasma beams [10,11] have been used extensively to eliminate superficial graphite nodules in NCI by rapid melting and cooling/solidification processes. Rapid cooling during solidification also leads to the formation of large amount of hypereutectic cementite instead of soft graphite [4,12]. However, because of the high cost of operating lasers and the need for a vacuum in
∗ Corresponding author. Tel.: +86 13995667466; fax: +86 27 87540057. E-mail address: [email protected] (S. Hu). 0169-4332/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.apsusc.2013.09.083
electron beam, attention has been focussed on using inexpensive, flexible and easy-to-operate plasma beam apparatus for the surface treatments of cast iron. A plasma beam is an extremely high-temperature flow usually possessing a power density of approximately 109 W/m2 , which allows for the rapid heating of almost every type of solid material to its melting or evaporating point via a type of rapid, non-equilibrium metallurgical process. The heating efficiency of plasma beam (85%) could be much higher than that of a laser beam (30%) in heating material surfaces [13]. Plasma beams provide significant advantages in treating iron-carbon alloys, including selective hardening, minimum part distortion, controllable case depth and structural refinement [14,15]. As high-energy-intensity heating sources, plasma beams are widely used for surface hardening to improve the corrosion resistance of surface-modified ferrous workpieces [16]. Plasma surface treatments, including plasma surface melting and plasma surface alloying, can be used as cost-effective alternative approaches for the elimination of near-surface graphite in cast iron. Surface melting can enhance the surface properties of ferrous materials, such as their hardness, wear resistance and corrosion resistance. Such enhancement is possible because a high cooling rate can be achieved due to localised melting that results in nonequilibrium phases as well as refined microstructures [17,18]. In general, surface melting can lead to an improvement in the corrosion resistance of all analysed materials [19].
X. Cheng et al. / Applied Surface Science 286 (2013) 334–343
335
Table 2 The composition of alloy powders (wt%). WC
Ni
TiC
Cr
Si
La2 O3
Fe
3
15
20
15
2
0.2
Balance
in Table 2. The powders were mechanically mixed and dispersed in kerosene to form alloy coatings. 2.2. Plasma surface treatment
Fig. 1. Microstructure of as-received NCI, F: ferrite, G: graphite and P: pearlite.
Surface alloying offers a versatile approach to the production of surface layers with a wide range of structures and compositions on a variety of substrates; moreover, the structures may consist of fine grains as a result of the relatively rapid cooling rates that can be achieved from the melt. Consequently, an improvement in corrosion resistance may also be achieved by surface alloying via microstructural homogenisation and refinement and the formation of new alloys on surfaces [20]. The micro-alloying of suitable elements on surfaces may also help enhance surface properties such as corrosion resistance [7]. There have been many studies that have investigated the mechanical properties of NCI [8,9,20] submitted to laser and electron beam treatments, but there are few published works on the corrosion characteristics of NCI surfaces modified by plasma beam treatment. In this study, a plasma beam was used as a heating source to modify the surface of NCI, eliminate superficial graphite nodules and hence improve the corrosion resistance of NCI. The surface treatments included plasma surface melting (M method), plasma surface melting + tempering (MT method), plasma surface alloying (A method) and plasma surface alloying + tempering (AT method). Furthermore, the corrosion resistance of the treated specimens compared with that of ferrite and pearlite NCI substrate (S specimen) was tested by potentiodynamic polarisation and electrochemical impedance spectroscopy (EIS).
Plasma surface melting (M method), plasma surface melting + tempering (MT method), plasma surface alloying (A method) and plasma surface alloying + tempering (AT method) were used to modify the surfaces of the NCI specimens. Plasma surface melting and alloying was carried out using a homemade set-up for combined transferred and non-transferred arc plasma beam treatment, as shown in Fig. 2. An atmospheric, high-density plasma was generated by the equipment, and the ionisation degree was less than 0.1%. The plasma pressure was greater than 0.1 MPa. The specimen and nozzle served as the anode, and a tungsten needle served as the cathode (Fig. 2). The nozzle diameter selected for testing was 2 mm, and the distance between the nozzle and the specimen was 4 mm. The plasma torch was controlled by a small variable-speed DC motor; thus, the speed of the torch could be adjusted and held constant throughout the test. The argon used in the process served as both a plasma gas and shielding gas, the flow rates of which were both 1 L/min. In the surface-melting process, a plasma beam with a Gaussian energy density distribution was defocussed on the surface of the NCI substrates without a pre-layer. For surface alloying, a homemade alloy coating was sprayed to form a pre-layer with a thickness of 200 m using an ejection gun on the NCI substrates before the plasma surface alloying process. The scanning speed during the plasma surface-alloying process should be lower due to the pre-layer. In the plasma surface-alloying process, the plasma beam energy melted the alloying coatings and substrate surface, and the alloy elements then seeped into the substrate to form a new alloy surface.
2. Experimental 2.1. Materials The starting material NCI (QT600-3) was produced by Dongfeng Company. The NCI was machined into plates with dimensions of 120 mm × 80 mm × 20 mm for plasma surface treatment. Fig. 1 shows a micrograph of an NCI specimen (S specimen), which consists of graphite nodules with an average diameter of 20 m surrounded by ferrite and pearlite. The chemical composition of the specimen is presented in Table 1. Alloy powders with a particle size of 1–3 m were used as an alloying layer material, the chemical composition of which is listed Table 1 Chemical composition of NCI (wt%). C
Si
Mn
S
P
Mg
Cu
Fe
3.41
2.23
0.42
0.033
0.045
0.048
0.37
Balance Fig. 2. Schematic diagram of the plasma beam heating of the nodular cast iron.
336
X. Cheng et al. / Applied Surface Science 286 (2013) 334–343
Table 3 The parameters of plasma surface melting and plasma alloying.
The nozzle diameter The distance between the nozzle and specimen Plasma torch moving speed Working current Argon of flow rate for plasma gas Argon of flow rate for shielding gas Overlapping on the former trace Arc voltage
Plasma surface melting
Plasma surface alloying
2 mm 4 mm
2 mm 4 mm
640 mm/min 50 ± 0.5 A 1 L/min 1 L/min 50% 30–48 V
320 mm/min 50 ± 0.5 A 1 L/min 1 L/min 50% 30–48 V
The plasma surface-melting and surface-alloying parameters used to produce the melted and alloyed surfaces are presented in Table 3, which were determined after a series of pre-tests. The specimens treated by plasma surface melting and surface alloying were denoted M specimens and A specimens, respectively. Coronado et al. [21] studied the effects of the tempering temperature on the abrasive wear of mottled cast iron. Their results demonstrated that no significant microstructural changes were observed after tempering at 300 ◦ C and 400 ◦ C. Therefore, the tempering temperature should be below 400 ◦ C. MT and AT specimens were obtained from the M and A specimens after tempering, respectively. The tempering treatment was carried out as follows: (1) heat to 400 ◦ C for 1 h in an air furnace and (2) furnace cool to room temperature. 2.3. Microstructure analysis After plasma surface treatment and tempering, the treated surfaces were ground by approximately 20 m to obtain an oxide-free surface. Then, the treated specimens were sectioned, mounted, polished and etched with a 4 vol% nitric acid alcohol solution. Scanning electron microscopy (SEM, Quanta 200) was used to examine the cross-sectional microstructure as well as the corroded surface. The phase structures of the modified surfaces were analysed using an X’ ˚ operPert PRO diffractometer with Cu-K␣ radiation ( = 1.5406 A) ated at 40 kV and 40 mA. The grain size and porosity of the modified surfaces were measured using the imaging tool Image-Pro Plus version 7.0 [22]. 2.4. Corrosion test The corrosion properties of the specimens were investigated by polarisation curve and EIS measurements in this study. Specimens used in electrochemical testing were machined into 1-cm cylindrical electrodes. Prior to the start of the corrosion tests, the specimens were wet ground using SiC paper with a grit number of up to 1000. Then, the specimens were washed in distilled water and ethanol and finally dried in warm air. All of the electrochemical tests were conducted in a conventional three-electrode system (CS350 electrochemical corrosion workshop) with the specimens as the working electrode (WE), Pt as the counter electrode (CE) and a saturated calomel electrode (SCE) as the reference (RE). The contact area in all cases was 0.785 cm2 , and the tests were carried out at 25 ◦ C. Polarisation tests were conducted vs. OCP (open circuit potential) after 1 h of immersion and performed in 3.5 wt% NaCl solution at a scan rate of 0.5 mV/s. The electrode potential was raised from −1000 mV to −500 mV. The corrosion behaviour of the NCI specimens was evaluated by anodic potentiodynamic polarisation tests [3]. Therefore, the corrosion potentials as well as the corrosion current densities were extracted from the plots by fitting the anode polarisation curve using C-view software.
EIS measurements were also carried out using the aforementioned equipment. After the working electrode was immersed in the electrolyte solution of the cell and its OCP became stable, its EIS was measured at the OCP with applied sinusoidal perturbations of 5 mV (peak-to-peak) at frequencies ranging from 10 kHz to 0.05 Hz, 10 steps per decade. EIS measurements were performed every 2 h of immersion up to a total of 48 h, and the same frequency range was used. The spectra were analysed using Z-view software. All experiments were repeated three times, and the results were reproducible; thus, representative data are presented in this paper. After 48 h of immersion, the surface and cross-sections of the corroded specimens were prepared and examined by SEM. 3. Results and discussion 3.1. The characterisation of surface-modified specimens Cross-sectional SEM micrographs of surface-modified specimens are shown in Fig. 3a and b. It was observed that the molten zone (MZ) and alloyed zone (AZ) were well bonded to the bulk region of the specimens with no distinctive irregular interface. In addition, it was evident from these micrographs that the graphite nodules had dissolved in the molten ferrous material during treatment, and the re-formation of graphite nodules had been suppressed by rapid self-quenching in favour of the formation of white cast iron, which enhanced the corrosion properties of the surface [7]. The molten zone on the surface-melted specimens (Fig. 3a) was approximately 105 m deep, whereas the depth of the AZ was approximately 300 m, although the pre-layer of the alloy coating was approximately 200 m. Thus, there was less epitaxial growth during the plasma alloying process when alloying particles were melted and penetrated into the substrate surface. The lower scanning speed and improvement of the hardenability owing to the alloy elements led to the deeper AZ. Fig. 4 shows high-magnification images of the MZ and AZ before and after tempering. The microstructure of the surfaces consisted of dendrites and superfine grains. This dendritic structure is similar to that of white cast iron observed by many researchers during the laser- or electron-beam melting of NCI [1,7,9]. The interdendrites were characterised by long primary and secondary arms; the secondary arm spacing was less than 0.5 m, which indicates a considerably high cooling rate during solidification [1,9]. Moreover, the fact that no cracks were observed in the micrographs indicates that the severe quenching of the plasma-beam-irradiated specimens did not cause noticeable microdefects. The molten zone showed a very fine structure that mainly consisted of austenite dendrites with an interdendritic network of carbides, as shown in Fig. 4a. In addition, some of the dendrites contained martensite [2], which indicated a high cooling rate during solidification. As reported previously, the microstructure of the alloyed zone consisted of eutectic ledeburite, alloy carbide and martensite [9]; in this case, dendrites along with a ledeburite structure were observed in the AZ, as shown in Fig. 4c. Due to the fast cooling after plasma surface alloying, a fine, cellular dendritic structure was formed and the ledeburite structure was similar to that of white cast iron. Compared with that of the specimens that were not tempered, the microstructure of the tempered specimens was more uniform and compact, which can be attributed to the formation of a hard phase and solid solution in the modified surfaces that were sufficiently high in concentration and homogeneous after tempering [23–25]. Moreover, the quenching stress may have relaxed during the tempering process to reduce the specimens’ susceptibility to stress corrosion.
X. Cheng et al. / Applied Surface Science 286 (2013) 334–343
337
Fig. 3. Cross-sectional SEM micrographs of surface-modified specimens (a) M specimen, (b) A specimen.
The average grain size in the modified surfaces was superfine, as indicated in Figs. 4 and 5. Because the treatment temperature of 400 ◦ C was not sufficiently high to make the grains grow to a great extent, the grain size remained superfine after tempering. In the plasma surface-alloying process, by shifting the plasma beam, the alloy coating and a thin surface layer of substrate material were melted simultaneously, and some compounds as well as TiC, Ni, Cr, WC and Si formed particles by heterogeneous nucleation. Thus, the number of crystal nuclei was increased during the crystallisation of the coatings, which was beneficial in increasing the nucleation rate and promoted non-spontaneous nucleation. The greater the number of crystal nuclei there are, the finer the crystal grains of the crystal become [16]. Therefore, the grain size in the molten zone appeared to be coarser than that in the alloyed zone, as shown in Figs. 4 and 5.
Plasma surface treatment is a technique that is employed under nonhomogeneous and nonequilibrium conditions, which results in multi-phase formation. The XRD results of the M specimen and MT specimen are shown in Fig. 6a and b. Both of the patterns reveal the presence of ␣-Fe (ferrite), ␥-Fe (austenite), martensite and Fe3 C. The formation of Fe3 C in the molten surfaces is attributed to the dissolution of graphite during melting that spread throughout the iron matrix. The high cooling rate of the molten pool resulted in a super-saturated austenite with carbon atoms. The phase of the A and AT specimens was similar, as shown in Fig. 6c and d. The alloyed surfaces revealed the existence of several types of phases, i.e., ␣-Fe (ferrite), ␥-(Fe, Ni) (austenite), martensite, Fe3 C, carbides TiC, Cr23 C6 , Cr7 C3 and W2 C, representing various types of carbides compared with the molten surface due to solution strengthening and chemical reaction with C in the rapid heating
Fig. 4. High-magnification images of the MZ and AZ before and after tempering (a) M (b) MT (c) A (d) AT.
338
X. Cheng et al. / Applied Surface Science 286 (2013) 334–343
Fig. 5. The average grain size in the surface of the treated and untreated specimens. Fig. 7. Polarization curves of S, M, MT, A and AT specimens in 3.5 wt% NaCl solution.
process induced by the plasma beam. The patterns demonstrated that Ni, Cr, TiC and WC could dissolve in the iron matrix of nodular cast iron. The formation of the ␥-(Fe, Ni) phase was likely due to the interaction of Ni with Fe during the solidification of the molten pool [7]. 3.2. Electrochemical test 3.2.1. Polarisation curves The polarisation behaviour of the NCI treated by different processing methods (S, M, MT, A, AT specimens) in 3.5 wt% NaCl solution at 25 ◦ C is shown in Fig. 7. The values of the corrosion potential (Ecorr ), corrosion current density (Icorr ) and polarisation resistance (R) were extracted from the polarisation curves by fitting the anode polarisation curve and are shown in Table 4. Tafel extrapolation was executed in the anodic potential region −150 mV away from Ecorr over one-half decade of current density. It is well known that the corrosion current density (Icorr ) is a good measure of a material’s corrosion resistance in a particular environment. The corrosion current is used to estimate a material’s mass loss as per Faraday’s first law. Hence, the higher the corrosion current is, the higher the mass loss and rate of corrosion become [26]. As shown in Fig. 7 and Table 4, the Ecorr of the modified surfaces was more positive than that of the untreated NCI specimen, which indicates that the modified surfaces were nobler than the untreated
Fig. 6. XRD patterns of the modified surfaces (a) M, (b) MT, (c) A, (d) AT.
NCI, and thus, the corrosion of the untreated NCI specimen could be favoured over that of the modified surfaces. In addition, the tempered specimens showed a higher potential than the specimens treated only by the plasma beam. On the other hand, the corrosion current density of the untreated NCI specimen was higher than that of the modified surfaces. Based on the estimated corrosion current densities, the protection efficiency () of different specimens can be calculated according to the following equation: (%) =
0 Icorr − Icorr 0 Icorr
× 100%
(1)
0 where Icorr and Icorr represent the corrosion current densities of untreated and treated NCI specimens, respectively, in 3.5 wt% NaCl solution. The estimated Icorr values and calculated protection efficiencies shown in Table 4 clearly indicate that the corrosion resistances of different specimens follow the order AT > A ≈ MT > M > S. The protection efficiency of the AT specimen was the highest (54.5%), indicating that the modified surface of NCI treated by plasma surface alloying and tempering had a moderate protective effect. The polarisation resistance of the AT specimen reached up to 9036 cm2 , which was 8.2 times that of S specimen (1098 cm2 ). The polarisation resistances of the A, MT and M specimens were 7671 cm2 , 7459 cm2 and 5457 cm2 , respectively. It can be concluded that the corrosion resistance of the NCI surfaces modified by plasma beam treatment was better than that of the untreated NCI specimen based on the abovementioned analysis. Moreover, the corrosion resistance of the alloyed surface was better than that of the molten surface, and the tempered specimens showed better corrosion resistance.
3.2.2. EIS Electrochemical impedance spectroscopy is well suited for monitoring any perturbation by an inhibitor in situ according to the electrochemical processes that occur at the interface of a metal and corrodent. To further understand the protective behaviour of the modified surfaces, EIS was employed to investigate the effects of different surface treatment processes on the corrosion of NCI. Fig. 8 shows the variation in the Nyquist plots of different specimens immersed in 3.5 wt% NaCl solution at various times. The resistance decreased with increasing immersion time. However, the S specimens showed a quicker speed of reduction than the other treated specimens due to the galvanic corrosion between graphite and iron.
X. Cheng et al. / Applied Surface Science 286 (2013) 334–343
339
Table 4 Electrochemical polarization parameters for various ductile irons in 3.5 wt% NaCl solution at 25 ◦ C. Specimens
Ecorr (V/SCE)
Icorr (A/cm2 )
R ( cm2 )
Protection efficiency (%)
Repetition rate (ı%)
S M MT A AT
−0.875 −0.839 −0.751 −0.781 −0.728
6.07 4.38 3.98 3.99 2.76
1098 5457 7459 7671 9036
– 27.8 34.4 34.3 54.5
95 96 95 98 97
The equivalent circuit shown in Fig. 9 can be used to analyse the EIS results, where Rs is the solution resistance; CPEdl is a constant phase element corresponding to the double-layer capacitance of the metal/solution interface; and Rp is the charge transfer resistance of the interface. The initial values were fed into the Z-view program to further fit the EIS data using the equivalent circuit model (Fig. 9) to obtain
the final Rs and Rp results. In all cases, good conformity between the curve-fitted and experimental results was obtained within an average error of only 5%. Fig. 10 shows the fitting parameters obtained for the different specimens in the NaCl solution using the equivalent circuit depicted in Fig. 9. The Rp values of different specimens in NaCl solution decreased with immersion time. However, the Rp values of
Fig. 8. EIS of different specimens in 3.5 wt% NaCl solution (a) S, (b) M, (c) MT, (d) A, (e) AT.
340
X. Cheng et al. / Applied Surface Science 286 (2013) 334–343
Fig. 9. Equivalent circuit for the modified surface in NaCl solution.
the specimens treated by plasma surface modification were greater than the Rp of the untreated NCI specimen in the NaCl solution. As the resistance increased, the corrosion protection was enhanced and fewer defects were formed on the modified surfaces [27], which suggests that the modified NCI surfaces retarded the corrosion at the metal-solution interface. The results of EIS are in agreement with the potentiodynamic polarisation results. One of the reasons for this phenomenon could be the formation of superfine grains in the modified surfaces, shown in Figs. 4 and 5, as reported by Kwok [28]. Ralston et al. [29] suggested a type of Hall–Petch correlation between Icorr and d (grain size) Icorr = a + bd−1/2
(2)
where a and b are constants whose values depend on the material being studied and on the surrounding corrosive environment. When the grain sizes are in the range of 75–1150 m, the decrease in corrosion resistance is attributed to the higher density of grain boundaries and other defects induced by refinement [26]. The introduction of grain boundaries, the high-energy areas in a microstructure, can lead to an increase in the corrosion rate by the formation of a galvanic couple between relatively anodic grain boundaries and cathodic grain interiors [30]. However, this behaviour is expected to change again when the level of grain refinement is very high while at the transition grain size, with a
Fig. 10. The Rp of different specimens as functions of immersion time in 3.5 wt% NaCl solution.
critical value of approximately 75 m, although this has not yet been established [30]. The grain sizes in the modified surfaces followed the order S (93.7 m) > MT (17.8 m) > M (15.4 m) > AT (11.8 m) > A (9.7 m), as indicated in Fig. 1, Fig. 4 and Fig. 5. Therefore, the corrosion current density of the modified surface would not obey the Hall–Petch type of correlation because the mean grain size of the substrate was greater than 75 m, whereas that of the modified surface was less than 75 m. The alloyed specimens showed better corrosion resistance than the melted specimens due to the finer grains of the alloyed specimens. In the ultrafine-grained regime (an average grain size below 70 m) [30], the very high fraction of grain boundaries is likely to reduce corrosion rates in two ways: (1) by accelerating the passivation kinetics and (2) by reducing the intensity of galvanic coupling between grain interiors and grain boundaries. Therefore, an ultrafine microstructure would lead to the formation of closely spaced electrochemical cells between anodic and cathodic regions, and the difference between the rates of anodic and cathodic reactions is expected to vary significantly, which would lead to more uniform attack and lower corrosion rates. This phenomenon explains why the modified surfaces with a superfine grain size exhibited noble corrosion properties. 3.2.3. The microstructure of the specimens after corrosion NCI usually consists of a graphite phase and a base phase (Fe). The graphite phase is the cathodic phase, which could accelerate the corrosion of Fe. As indicated in Fig. 11, which shows the S specimen immersed in 3.5 wt% NaCl solution for 48 h, there were many corrosion products on the surface of the untreated NCI specimen; the graphite was projected relative to the matrix and showed features typical of graphitic corrosion. The results of the EDS analysis of the corrosion product shown in Fig. 11 indicate that the product contained mainly elemental Fe, Cl, O and Si. In other words, the graphite acted as the cathode and the Fe matrix as the anode in the solution, which caused the dissolution and corrosion of the anode around the graphite, leaving only a projection of the graphite, as observed by Jeong et al. [6] and Zeng et al. [31]. Fig. 12 shows typical SEM images of the corroded surface of a treated specimen after 48 h of immersion in 3.5 wt% NaCl solution. The plasma-beam-treated specimens presented slight corrosion compared to the untreated NCI specimen (Fig. 11). Additionally, in this case, the near-surface graphite nodules were dissolved, dense surfaces were formed and no graphitic corrosion occurred. The dense structures indicate that the modified surfaces possessed low porosity. From the micrographs shown in Figs. 1 and 4, the overall porosities were estimated to be S (3%) > M (0.7%) > MT (0.3%) = A (0.3%) > AT (0.1%). It is known that pores and inclusions play important roles in the corrosion behaviour of materials: a high porosity allows for the penetration of solution, whereas large inclusions lead to local corrosion [32]. The ledeburite structure presented on the modified surfaces is similar to that of white cast iron, which can improve the corrosion properties of the surfaces [7]. Moreover, all of the modified surfaces revealed a diffraction peak attributed to austenite, as shown in Fig. 6, which acted as a corrosion inhibitor [3,15]. These factors led to the improvement of the corrosion resistance of the NCI surfaces modified by plasma beam treatment and tempering. Based on the results of the electrochemical tests and corroded surface analysis, the alloyed surfaces exhibited better corrosion resistance than the molten surfaces, and tempering also improved the corrosion resistance. A finer and more densely alloyed surface was formed to improve corrosion resistance after plasma surface alloying. Ni and Cr occurred on the alloyed surfaces, as shown in Fig. 6c and d.
X. Cheng et al. / Applied Surface Science 286 (2013) 334–343
341
Fig. 11. Corroded morphology on the surface of the untreated NCI specimen after 48 h of immersion in 3.5 wt% NaCl solution and EDX of the corrosion product.
Ni is referred to as an austenite-stabilising element [3,33], and austenite is a good inhibitor [3]. Moreover, Ni and Cr distributed on a surface could form a dense oxide film to act as a barrier to the internal diffusion of aggressive media [34]. For these reasons, the alloyed surfaces presented better corrosion resistance. During tempering, hard phases and solid solutions may be distributed more uniformly over a modified surface due to the reheating effect of the tempering, and quenching stress may also be released [23–25]. Furthermore, the porosity of modified surfaces can be reduced after heat treatment [35]. Hence, tempering can reduce the surfaces’ susceptibility to corrosion.
Cross-sectional images of the M specimen as well as the untreated NCI specimen after 48 h of immersion are shown in Fig. 13a and b. No distinct cracks or pitting corrosion can be observed because the immersion time was not sufficiently long. Verdian et al. [34] stated that no obvious indication of blistering or delimitation could be observed in SEM micrographs of the cross-section of an air plasma spraying (APS) coating after 63 days of immersion. In this study, there were no changes in the crosssectional structure of the M specimen before and after corrosion (4a and 13b, respectively). To summarise, it was determined that there was no obvious corrosion of the MT, A and AT specimens along their depth.
Fig. 12. Corroded morphology on the surface of the treated specimens after 48 h of immersion in 3.5 wt% NaCl solution (a) M, (b) MT, (c) A, (d) AT.
342
X. Cheng et al. / Applied Surface Science 286 (2013) 334–343
Fig. 13. Corroded morphology of transverse after the 48 h immersion in 3.5 wt% NaCl solution (a) S specimen, (b) M specimen.
4. Conclusions From the present investigation, the following conclusions can be drawn: (1) Plasma surface melting and plasma surface alloying could eliminate the near-surface graphite nodules of NCI, and a dense and fine eutectic dendritic microstructure was obtained to enhance the corrosion resistance. (2) Specimens showed good corrosion resistance if submitted to plasma surface treatment and tempering. The better corrosion resistance of the alloyed surface in 3.5 wt% NaCl was due to the development of a denser structure and finer grain size. (3) The corrosion resistances of the specimens ranked as follows: surface-alloyed and tempered specimen > surface-alloyed specimen ≈ surface-melted and tempered specimen > surfacemelted specimen > untreated NCI specimen. Acknowledgements This work was supported by the Province and Ministry cooperation of university-industry collaboration (No. 2010090200047) and the national natural science foundation of China (No. 51375005). The authors express their sincere thanks for the help of analytical and testing center of HUST. References [1] K.Y. Benyounis, O.M.A. Fakron, J.H. Abboud, A.G. Olabi, M.J.S. Hashmi, Surface melting of nodular cast iron by Nd-YAG laser and TIG, J. Mater. Process. Technol. 170 (2005) 127–132. [2] K.F. Alabeedi, J.H. Abboud, K.Y. Benyounis, Microstructure and erosion resistance enhancement of nodular cast iron by laser melting, Wear 266 (2009) 925–933. [3] C.H. Hsu, M.L. Chen, Corrosion behavior of nickel alloyed and austempered ductile irons in 3.5% sodium chloride, Corros. Sci. 52 (2010) 2945–2949. [4] J.H. Abboud, Microstructure and erosion characteristic of nodular cast iron surface modified by tungsten inert gas, Mater. Des. 35 (2012) 677–684. [5] M.B. Karamıs¸, K. Yıldzlı, Surface modification of nodular cast iron: a comparative study on graphite elimination, Mater. Sci. Eng., A 527 (2010) 5225–5229. [6] B.Y. Jeong, M.H. Kim, Corrosion characteristics of duplex surfacetreated spheroidal graphite cast iron, Surf. Coat. Technol. 141 (2001) 262–268. [7] A. Gulzar, J.I. Akhter, M. Ahmadb, G. Ali, M. Mahmoodd, M. Ajmal, Microstructure evolution during surface alloying of ductile iron and austempered ductile iron by electron beam melting, Appl. Surf. Sci. 255 (2009) 8527–8532. [8] M.D. Jean, Y.F. Tzeng, Optimisation of electron-beam surface hardening of cast iron for high wear resistance using the Taguchi method, Int. J. Adv. Manuf. Technol. 24 (2004) 190–198.
[9] H. Yan, A.H. Wang, Z.T. Xiong, K.D. Xu, Z.W. Huang, Microstructure and wear resistance of composite layers on a ductile iron with multicarbide by laser surface alloying, Appl. Surf. Sci. 256 (2010) 7001–7009. [10] W.S. Dai, L.H. Chen, T.S. Lui, SiO2 particle erosion of spheroidal graphite cast iron after surface remelting by the plasma transferred arc process, Wear 248 (2001) 201–210. [11] T. Ishida, Local melting of nodular cast iron by plasma arc, J. Mater. Sci. 18 (1983) 1773–1784. [12] M. Shamanian, S.M.R. Mousavi Abarghouie, S.R. Mousavi Pour, Effects of surface alloying on microstructure and wear behavior of ductile iron, Mater. Des. 31 (2010) 2760–2766. [13] W.X. Pan, X. Meng, G. Li, Q.X. Fei, C.K. Wu, Feasibility of laminar plasma-jet hardening of cast iron surface, Surf. Coat. Technol. 197 (2005) 345–350. [14] L. Giordano, A. Tiziani, A. Zambon, Characterization of surface chromium and molybdenum alloying on gray cast iron obtained by the plasma-transferred arc technique, Mater. Sci. Eng., A 140 (1991) 727–732. [15] H. Chen, H.Q. Li, Microstructure and wear resistance of Fe-based coatings formed by plasma jet surface metallurgy, Mater. Lett. 60 (2006) 1311–1314. [16] H. Chen, H. Li, Y.Z. Sun, M. Li, Microstructure and properties of coatings with rare earth formed by DC-plasma jet surface metallurgy, Surf. Coat. Technol. 200 (2006) 4741–4745. [17] M.H. Sohi, G. Karshenas, S.M.A. Boutorabi, Electron beam surface melting of as cast and austempered ductile irons, J. Mater. Process. Technol. 153–154 (2004) 199–202. [18] M.Y. Li, Y. Wang, B. Han, W.M. Zhao, T. Han, Microstructure and properties of high chrome steel roller after laser surface melting, Appl. Surf. Sci. 255 (2009) 7574–7579. [19] A. Conde, R. Colac¸o, R. Vilar, J.D. Damborenea, Corrosion behaviour of steels after laser surface melting, Mater. Des. 21 (2000) 441–445. [20] G.F. Sun, R. Zhou, P. Li, A.X. Feng, Y.K. Zhang, Laser surface alloying of CB-W-Cr powders on nodular cast iron rolls, Surf. Coat. Technol. 205 (2011) 2747–2754. [21] J.J. Coronado, A. Gómez, A. Sinatora, Tempering temperature effects on abrasive wear of mottled cast iron, wear 267 (2009) 2070–2076. [22] K.M. Pedersen, N.S. Tiedje, Graphite nodule count and size distribution in thinwalled ductile cast iron, Mater. Charact. 59 (2008) 1111–1121. [23] H.X. Liu, C.Q. Wang, X.W. Zhang, Y.H. Jiang, C.X. Cai, S.J. Tang, Improving the corrosion resistance and mechanical property of 45 steel surface by laser cladding with Ni60CuMoW alloy powder, Surf. Coat. Technol. 228 (2013) S296–S300. [24] P.D. Bilmes, C.L. Llorente, L.S. Huamán, L.M. Gassa, C.A. Gervasi, Microstructure and pitting corrosion of 13CrNiMo weld metals, Corros. Sci. 48 (2006) 3261–3270. [25] R.A. Carnério, R.C. Ratnapuli, V.F.C. Lins, The influence of chemical composition and microstructure of API linepipe steels on hydrogen induced cracking and sulfide stress corrosion cracking, Mater. Sci. Eng., A 357 (2003) 104–110. [26] S. Gollapudi, Grain size distribution effects on the corrosion behaviour of materials, Corros. Sci. 62 (2012) 90–94. ˇ [27] D.K. Merl, P. Panjan, M. Cekada, M. Maˇcek, The corrosion behavior of Cr-(C,N) PVD hard coatings deposited on various substrates, Electrochim. Acta 49 (2004) 1527–1533. [28] C.T. Kwok, F.T. Cheng, H.C. Man, Microstructure and corrosion behavior of laser surface-melted high-speed steels, Surf. Coat. Technol. 202 (2007) 336–348. [29] K.D. Ralston, N. Birbilis, C.H.J. Davies, Revealing the relationship between grain size and corrosion rate of metals, Scr. Mater. 63 (2010) 1201–1204. [30] G.R. Argade, S.K. Panigrahi, R.S. Mishra, Effects of grain size on the corrosion resistance of wrought magnesium alloys containing neodymium, Corros. Sci. 58 (2012) 145–151.
X. Cheng et al. / Applied Surface Science 286 (2013) 334–343 [31] D.W. Zeng, K.C. Yung, C.S. Xie, Investigation of corrosion behavior of high nickel ductile iron by laser surface alloying with copper, Scr. Mater. 44 (2001) 2747–2752. [32] J.H. Chang, J.M. Chou, R.I. Hsieh, J.L. Lee, Corrosion behaviour of vacuum induction-melted Ni-based alloy in sulphuric acid, Corros. Sci. 52 (2010) 2323–2330. [33] ASTM A439, Annual Book of ASTM Standards, 01.02, Beuth Publications, Berlin, Germany, 1989, pp. 238–242.
343
[34] M.M. Verdian, K. Raeissi, M. Salehi, Corrosion performance of HVOF and APS thermally sprayed NiTi intermetallic coatings in 3.5% NaCl solution, Corros. Sci. 52 (2010) 1052–1059. [35] X.B. Zhao, D.Y. Yan, S. Li, C.G. Lu, The effect of heat treatment on the electrochemical corrosion behavior of reactive plasma-sprayed TiN coatings, Appl. Surf. Sci. 257 (2011) 10078–10083.