- Email: [email protected]
Improvement in erosion-corrosion resistance of high-chromium cast irons by trace boron
Wear 376-377 (2017) 578–586
Contents lists available at ScienceDirect
Wear journal homepage: www.elsevier.com/locate/wear
Improvement in erosion-corrosion resistance of high-chromium cast irons by trace boron Hao Lu a, Tingzhong Li a, Juan Cui a, Qingyang Li a,b, D.Y. Li a,n a
Department of Chemical and Materials Engineering, University of Alberta, Edmonton, Alberta, Canada T6G 1H9 State Key Laboratory of Urban Water Resource and Environment, School of Chemical Engineering and Technology, Harbin Institute of Technology, Harbin 150001, China
b
art ic l e i nf o
a b s t r a c t
Article history: Received 12 September 2016 Received in revised form 28 January 2017 Accepted 3 February 2017
Extensive work was carried out to improve the performance of cast irons used for slurry transport in the oil sands industry. Because of complex condition of the multi-phase fluid containing oil sand particles, erosion-corrosion damage has long been an issue. High chromium cast irons (HCCIs) are currently used for slurry pumps, showing high resistance to sand-containing slurry erosion. However, further enhanced HCCIs are highly desired in order to more effectively reduce the damage to oil sand slurry transport facilities caused by erosion-corrosion for prolonged service life. This study was focused on tailoring an existing HCCI used in the oil sands industry by adding trace boron. Microstructure and relevant properties of the modified cast irons were characterized and evaluated. It was demonstrated that trace boron markedly increased the resistance of the cast iron to slurry erosion. The mechanism responsible for the improvement was analyzed based on observed variations in microstructure and properties. & 2017 Elsevier B.V. All rights reserved.
Keywords: Trace Boron Erosion-corrosion resistance Sliding wear Corrosion High-chromium cast irons
1. Introduction Wear is one of the most predominant failure mechanisms for material loss in oil sands industry operations. The extreme wear conditions result from aggressive abrasion and impact caused by the large quantities of silica sand contained in the oil sand and its slurry [1]. In addition to the harsh wearing conditions, corrosion and consequently the synergistic attacks from mechanical and chemical actions considerably shorten the materials lifespan as well [2]. Synergistic erosion and corrosion or erosion-corrosion is the dominant wear mechanism for slurry handling in the oil sand industry [3]. High chromium cast irons (HCCIs) are widely used to resist abrasion and erosion in many industrial processes, such as mining, mineral processing and oil sand operations [4], due to their high hardness and appropriate balance between hardness and toughness, compared to conventional white cast irons [5], as well as high corrosion resistance [6]. The wear resistance of HCCIs results mainly from the high volume fraction of hard M7C3 carbides with high hardness and the supporting from tough metallic matrix [7]. HCCIs are classified according to ASTM A532 “Standard Specification for Abrasion-Resistant Cast Irons”. Commercially used HCCIs have their compositions usually in the range of 12–27 wt% Cr and 2.4–3.6 wt% C, which process high wear resistance n
Corresponding author. E-mail address: [email protected] (D.Y. Li).
http://dx.doi.org/10.1016/j.wear.2017.02.014 0043-1648/& 2017 Elsevier B.V. All rights reserved.
and widely used in slurry transport systems in the oil sands industry [1]. Most HCCIs used in oil sand industry for slurry pump casing and impellers fall into the composition range of 2.4–3.6 wt. % C and 25 27 wt. % Cr. However, this group of HCCIs does not always perform satisfactorily, due to the synergistic attack of oil sand particle erosion and aqueous corrosion [8]. Considerable work has been done to improve the wear resistance of HCCIs. In recent years, the oil sand mining industry intends to push the Cr and C contents to higher levels. The former increases hardness while the latter improves the corrosion resistance. A proper combination of the improvements could be suitable for specific wearing conditions. Such HCCIs with high concentrations of carbon and chromium were previously considered to be non-castable due to high rejection rate (rejected pieces/processed pieces) [9]. However, it has become possible to cast HCCIs with high concentrations of chromium and carbon with the advance in foundry techniques [1]. In our previous studies, HCCIs with 30 wt% to 45 wt% chromium were fabricated and demonstrated to be highly resistant to wear and corrosion [9,10]. Other effective approaches to enhance HCCIs wear and corrosion resistance include appropriate heat treatment and alloying with minor elements, such as boron, nickel, molybdenum, titanium, vanadium, to optimize microstructure [11]. Studies have demonstrated that adding boron is promising to improve the wear resistance of white cast irons [12]. Such improvement is ascribed to both elevated hardness and toughness, resulting from formed
H. Lu et al. / Wear 376-377 (2017) 578–586
Table 1 Nominal compositions of HCCIs under study. Sample no.
%Cr
%C
%B
#1 #2 #3 #4 #5 #6
35 35 35 35 35 35
3 3 3 3 3 3
0 0.12 0.2 0.4 0.48 0.6
boride phases and boron-doped carbide as well as microstructure modification [5,13]. However, few studies have been conducted to investigate the effect of boron on HCCIs’ wear resistance, especially on the HCCIs with their chromium concentrations closer to the high end of Cr concentration range. In this study, an existing HCCI with 35 wt% chromium and 3 wt% carbon for oil sand slurry handling was modified by adding trace boron. The objective of this study is to investigate how the trace boron affects microstructure and relevant properties, including resistance to sliding wear, high speed solid particle erosion, corrosion, erosion-corrosion, of the modified cast irons.
2. Materials and methods 2.1. Materials Six HCCIs with nominal chemical compositions of 35 wt% chromium, 3 wt% carbon, and various concentrations of boron: 0, 0.12, 0.2, 0.4, 0.48, 0.6 wt%, respectively, balanced by iron were fabricated using an induction furnace. The molten metals were poured into copper molds to produce the samples. Table 1 gives nominal compositions of samples under study. 2.2. Characterization and testing Microstructures of the samples were examined using a Scanning Electron Microscope (Zeiss Sigma 300 VP-FESEM). Worn surfaces after sliding wear, high-speed particle erosion, and slurry erosion tests were observed under a Scanning Electron Microscope (Tescan Vega 3 SEM, Czech Republic).
579
Hardness of the samples was measured using a Rockwell hardness tester (Zwick/Roell Rockwell/superficial Rockwell hardness tester, Indentec Hardness Testing Machine Limited, UK) under a load of 60 kgf with a diamond cone indenter. Sliding wear tests were performed on a pin-on-disc tribometer (CSEM Instruments, Neuchatel, Switzerland) based on ASTM G 99 “Standard Test Method for Wear Testing with a Pin-on-Disk Apparatus”. The disc was the sample under study and the pin was a silicon nitride ball with 6mm diameter. All tests were performed at a sliding speed of 1 cm/s along a circle path of 2.0 mm in diameter under a normal load of 10 N for a sliding distance of 36 m. Wear tracks and corresponding volume losses of the samples were determined using a confocal microscope (ZeGage 3D optical profilemeter, Zygo Corp.). High-speed solid particle erosion tests were carried out using a home-made air-jet erosion tester [14] based on ASTM G 76 “Standard Test Method for Conducting Erosion Tests by Solid Particle Impingement Using Gas Jets” (Fig. 1(a)). The erosion tests were performed at an impingement angle of 90 °and a dry air flow with a pressure of 40 psi was used to eject sand particles to sample surface (corresponding sand flow speed: 55 m/s). Erosion at an impingement angle of 90° usually results in more damage to less tough materials, thus it more or less reflects the toughness of the samples. AFS 50/70 sand (U.S. Silica Company, USA) was used as the erodent particle. The sample mass was measured before and after the tests using a balance with a precision of 0.1 mg. Each test was repeated at least 3 times. The erosion–corrosion tests were carried out using a homemade slurry-pot tester (Fig. 1(b)). Guidelines for erosion-corrosion tests in slurry can be found in STP946 “Slurry Erosion: Uses, Applications, and Test Methods”. The tester had a cylindrical tank of 29 cm in diameter and 22 cm in height. A slurry solution containing 1.5 L silica sand (500 μm in diameter) and 6 L water was placed in the slurry container. Samples were held by a holder, which was rotated during erosion test, thus driving the samples to move in the slurry. The angle between sample surface and the moving direction was kept at 45 °and the moving velocity of samples in the slurry was 8 m/s. For each slurry erosion test, the total distance over which the sample travelled in the slurry was 15 km. After test, corrosion products were removed by rinsing and light brushing, and the samples were finally cleaned by rinsing with distilled water. The weight loss of each sample was then
Fig. 1. Schematic illustrations of (a) air-jet erosion tester used for high-speed solid particle erosion tests, and (b) slurry-pot tester used for erosion-corrosion tests.
580
H. Lu et al. / Wear 376-377 (2017) 578–586
3. Results and discussions 3.1. Microstructure characterization
Fig. 2. Crystal structure of Fe4Cr3C2B; green, brown, blue and yellow balls represent boron, carbon, chromium and iron atoms, respectively. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
measured using a balance with a precision of 0.1 mg. More details about the home-made slurry pot erosion-corrosion tester and the experimental set-up can be found in our previous publication [15]. It should be indicated that this study is conducted to investigate effect of trace B on the performance of the sample alloys. Relevant wear tests were performed for comparison purpose. Thus, the testing conditions were selected based on what we usually used for testing many other industrial materials rather than exactly follow the standard testing conditions. This helps obtain a sense about the performance of the HCCIs under study, compared to those of materials used in industry. Open circuit potentials (OCP) of the samples in a 3.5 wt% NaCl aqueous solution were measured using a computerized Gamary Instruments Frame-work electrochemical system. A saturated calomel electrode (SCE) was used as the reference electrode and a platinum plate with an area of 1 cm2 was used as the counter electrode. The tests were performed based on ASTM G 3 “Standard Practice for Conventions Applicable to Electrochemical Measurements in Corrosion Testing” and ASTM G 59 “Standard Test Method for Conducting Potentiodynamic Polarization Resistance Measurements”.
2.3. Simulation methods Mechanical properties and electron densities of carbide and boron-doped carbide were calculated, respectively. The calculations were conducted based on the first-principles density functional theory (DFT) and implemented using the Vienna ab-initio simulation package (VASP) [16,17]. Electron interactions were treated with the projector-augmented wave method [18]. The generalized gradient approximation (GGA) of Perdew-Burke-Ernzerhof (PBE) was used for the exchange and correlation effects [19]. The cut-off energy for plane wave basis was set to 400 eV. A Monkhorst-pack [20] sampling of 7 7 9 k-point mesh was adopted to achieve self-consist field convergence, and 11 11 13 k-point mesh was adopted to compute total energy, charge density and electronic structure. In all calculations, self-consistency was achieved when the total energy difference between cycles was smaller than 0.1 meV. The geometry relaxation tolerance in all forces was below 0.01 eV/A. The moduli were calculated with Voigt–Reuss–Hill approximation [21]. Space group P63mc (#186) was applied to construct the carbide structure in the simulation. Two boron atoms replaced two carbon atoms, which made the carbide Fe4Cr3C2B comparable with Fe4Cr3C3. Crystal structure of Fe4Cr3C2B is illustrated in Fig. 2. The reason to choose Fe4Cr3C3 as a reference for comparison is that it is the hardest in the (Fe, Cr)7C3 carbide series [22].
SEM images of the samples are illustrated in Fig. 3. Alloy #1 has a nearly eutectic microstructure with strip carbides. However, other five B-doped alloys show hypereutectic microstructures consisting primary carbides and eutectic matrix, similar to the microstructure observed in other boron-added conventional white cast irons [13]. It appears that the added boron more or less spheroidized the carbides as Fig. 3 illustrates. Using sample #6 containing 0.6%B as an example, we analyzed distributions of elements in different micro-constituents through EDX compositional mapping. Maps of various elements are shown in Fig. 4. As demonstrated, B is soluble in all phases, showing its ability to substitute Fe in the matrix and C in the carbide. Segregation of boron is not observed in the EDX maps, indicating that the sample only contains boron-doped carbides (Fe, Cr) 7 (C, B)3 without formation of borides such as (Fe, Cr) 2 B. It should be mentioned that EDX is not accurate to analyze light elements such as B and C, thus the EDX mapping was carried out to demonstrate B distribution rather than quantify boron contents in carbides and the matrix. The elemental distribution of B shown in Fig. 4 indicates the formation of boron-doped carbide, (Fe, Cr) 7 (C, B) 3 and boron-substituted Fe matrix. As reported and discussed in the following sections, the doped boron strengthened the carbide phase and increased its volume fraction, leading to higher hardness and enhanced wear resistance of the boron-doped alloys. 3.2. Properties 3.2.1. Hardness Bulk hardness of the six alloys was measured and results of the measurement are given in Fig. 5. As shown, all B-doped alloys are harder than the base alloy (#1). The higher overall hardness should be attributed to the following factors: 1) the doped boron strengthened (Fe, Cr)7C3, and 2) the dope B resulted in an increase in the volume fraction of carbides (Table 2). The trace boron may not produce other second phases due to its low concentration. It has been shown that boron-doped carbide phases are harder than the carbide in HCCIs [5,23]. Our modulus mapping using AFM directly confirmed the increase in modulus of carbides with doped B (Table 2). This is also supported by our first-principles calculation, which shows that the B-doped carbide exhibits higher bulk modulus, shear modulus and elastic modulus than the un-doped carbide (see Table 3). It may need to mention that solution hardening of the ferrous matrix by added boron is expected. However, our modulus mapping does not support this expectation (see Table 2). This could be due to the fact that the B concentration is not sufficient enough to result in effective solid-solution hardening. As for the increase in the volume fraction of carbides with the doped boron, we estimated volume fractions of the alloys through image analysis using the software ImageJ. As shown in Table 2, the volume fraction of carbides increased with increasing the concentration of boron. Such a phenomenon was also noticed by others [23]. The added B could facilitate nucleation of carbides. An increase in the volume fraction of carbides should contribute to the observed increase in the mechanical strength of the high-Cr cast irons. 3.2.2. Resistance of the alloys to sliding wear and high-speed particle erosion Resistance of the alloys to sliding wear and high speed particle erosion were measured and presented in Fig. 6. As illustrated, all B-doped alloys demonstrated considerably
H. Lu et al. / Wear 376-377 (2017) 578–586
581
Fig. 3. Microstructure of the fabricated alloys (a) #1, (b) #2, (c) #3, (d) #4, (e) #5, (f) #6.
reduced volume losses during sliding wear tests (only 15% to 25% of the volume loss of the base alloy without boron), which can be explained by the elevated mechanical strength of the B-doped alloys. Fig. 7 illustrates the surface morphology of selected samples after sliding wear testing. Grooves caused by plowing and plastic deformation can be observed on the worn surface. Compared to that of base alloy #1, the worn surface of the modified alloys is smoother with shallower wear grooves, indicating that the penetration of the pin ball is much more resisted by the B-doped alloys than by the base alloy. Although the B-doped alloys perform considerably better than the base alloy during sliding wear tests, their performance is inferior to that of the base alloy when are subjected to high-speed particle erosion, for which material's toughness plays an important role especially for high-angle impingement. For the erosion tests, 90-degree impingement angle was used. Since erosion at larger impingement angles usually results in more damage to less tough materials, the erosion tests with 90-degree
impingement angle were more influenced by toughness of the alloys. As shown in Fig. 6, the B-doped alloys are inferior to the base alloy during the high-speed particle erosion testing, implying that the modified alloys are not as tough as the base alloy. Eroded surfaces were examined using SEM. As illustrated in Fig. 8, the eroded surfaces show typical features of eroded surface of high chromium cast irons as reported in previous studies [24]. The eroded surfaces are covered by overlapping craters caused by impact, and the craters have edges with extruded platelets. The rate of forming these platelets and break off controls the rate of erosion for the materials [24]. As illustrated in Fig. 8, no visible cracks are observed. This may indicate that the alloys possess reasonable toughness, although the doped boron increases hardness of the material at expense of toughness to some degree, as shown by the increased material loss caused by high-speed particle striking at 90-degree impingement angle (Fig. 6). There is no significant morphological difference among eroded surfaces of different alloys. It is known that strain-hardening of
582
H. Lu et al. / Wear 376-377 (2017) 578–586
Fig. 4. EDX compositional maps of sample #6 showing the elements distributions of Fe (blue), Cr (green), C (yellow), and B (red). The first figure (top left) is a SEM image of the area for compositional analysis. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
Table 2 Volume fractions of carbides, moduli of matrix and carbides in the samples. Sample no. Volume fraction of carbides
Modulus of matrix
Modulus of carbide
#1 #2 #3 #4 #5 #6
234.4 GPa 231.2 GPa 228.6 GPa 227.2 GPa 225.0 GPa 223.1 GPa
265.6 GPa 278.5 GPa 280.2 GPa 281.3 GPa 293.0 GPa 302.4 GPa
38.1% 38.6% 40.1% 40.3% 42.3% 47.8%
Table 3 Calculated moduli of carbides with and without boron.
Fig. 5. Bulk hardness of the fabricated alloys; modified alloys possess higher hardness than the base alloy. The higher hardness of modified alloys is attributed to the formation of boron doped carbide phase, Fe4Cr3C2B, shown in Fig. 2, and increased volume fraction of carbides.
Fe4Cr3C3 Fe4Cr3C2B
B (GPa)
G (GPa)
E (GPa)
277.1 288.2
138.4 142.1
355.6 366.1
H. Lu et al. / Wear 376-377 (2017) 578–586
583
Fig. 6. Resistance of the alloys to sliding wear and high-speed silica particle erosion (particle velocity: 55 m/s; impingement angle: 90 degrees); modified alloys are more resistant to sliding wear, but less resistant to high-speed particle erosion.
metallic materials influences their erosion rate. The HCCIs with higher volume fraction of carbides have smaller strain hardening exponent and higher erosion rate [24]. This may also be a factor responsible for the lower resistance of B-doped HCCIs having higher volume fractions of carbides to high-speed solid particle erosion. 3.2.3. Corrosion resistance Open circuit potential tests were performed to evaluate how the doped B affected the corrosion tendency of the cast iron in a 3.5 wt% NaCl aqueous solution. Fig. 9. illustrates the corrosion potential of the alloys with and without doped boron. As shown, the B-doped alloys have lower open circuit potentials, compared to the base alloy, indicating that adding boron renders the HCCI prone to corrosion, compared to the base alloy. The following are possible factors contributing to the decrease in corrosion potential: a) The corrosion resistance of HCCIs largely depends on the corrosion behaviour of the ferrous matrix, in which the concentration of free chromium is crucial [25]. Sufficient free Cr atoms result in improved passivation capability with forming a protective passive film. Since the B-doped alloys have increased volume fractions of carbides (Table 2), which consumes more chromium in the matrix, the passivation capability could be lowered. b) Compared to Cr ( ϕCr = 4.5eV ) and Fe ( ϕFe = 4.5eV ), B has a lower work function ( ϕB = 4.45eV )[26,27]. Work function reflects the surface activity and lower work function corresponds to lower corrosion potential [28,29]. To determine if the B-doped alloys were more active than the base alloy, we measured electron work functions of matrixes and carbides with and without doped boron using a multi-mode atomic force microscope (Bruker, Multimode 8, and U.S.A.). Fig. 10 illustrates a representative work function map of the HCCIs samples. Average values of work functions of carbide and matrix are given in Table 4, which shows that both the B-doped matrix and carbide show lowered work functions, though not significant, corresponding to raised surface activities.
Fig. 7. Wear track morphology of selected samples (a) #1 (d) #3 (g) #6, and corresponding surface profiles (c) #1 (f) #3 (i) #6, observed by optical confocal microscope; (b) (e) (h) are SEM images of the worn surfaces of samples #1, #3, and #6. As shown, the worn tracks of boron modified alloys are narrower and shallower than that of the base alloy.
The above two factors could be mainly responsible for the decrease in open circuit potential and corrosion resistance of sample with doped B. As for the reason why the work function of carbide is lowered by doped boron, we analysed electron densities of carbides with
584
H. Lu et al. / Wear 376-377 (2017) 578–586
Fig. 8. Surfaces eroded by high-speed particles:(a)#1, (b) #3, (c) #6.
Fig. 9. Open circuit potential curves of the alloys, the modified alloys have lower open circuit potential than the base alloy.
Fig. 10. A representative work function map of HCCI, bright regions represent carbides with higher work functions, and dark regions represent matrix with low work function. A line profile of work function across both carbide and matrix is illustrated (curve in white).
Fig. 11. Electron densities of (110) plane for (a) carbide and (b) B-dope carbide, red color represents high electron density and blue color represents low electron density; local potentials of (110) plane for (c) carbide and (d) B-doped carbide, red color represents high local potential and blue color represents low local potential. Higher potential corresponds to lower work function and lower electron density. B-doped carbide shows lower electron density and work function. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
H. Lu et al. / Wear 376-377 (2017) 578–586
585
Table 4 Work functions of matrix and carbide for selected samples. Sample
EWF of matrix (eV)
EWF of carbide (eV)
#1 (no boron) #4 (0.4% B) #6 (0.6% B)
4.56 4.54 4.52
4.67 4.57 4.56
and without doped B through first-principles calculation. Fig. 11 shows electron densities and local potentials of (110) plane for carbide and B-doped carbide. As illustrated, B-doped carbide has a higher potential, corresponding to a lower work function (difference between the vacuum potential and local potential). This is also consistent with the electron density calculation, i.e. higher electron density corresponds to higher work function [30–32]. 3.2.4. Erosion-corrosion resistance of the alloys Slurry erosion tests were performed in a neutral slurry aqueous solution with 20% silica sand particles. Weight losses of the samples after testing are presented in Fig. 12. As shown, the resistance of the HCCIs to erosion-corrosion was improved remarkably when the base alloy was modified by adding trace boron. The higher the concentration of added boron, the more resistant was the alloy to slurry erosion-corrosion. As shown, the weight loss of sample #6 with 0.6 wt% B is only about 40% of that of the base alloy (#1) without doped boron. The hard carbide in HCCIs plays a main role in resisting sliding wear and slurry erosion if the slurry is not very corrosive [9]. As discussed earlier, all modified alloys possess higher hardness and demonstrate considerably reduced volume losses during sliding wear tests. Thus, the enhanced slurry erosion resistance of the modified alloys in this sand-containing aqueous slurry with low corrosivity should be mainly attributed to the increased hardness. Fig. 13 presents eroded surfaces of sample #1 (35-3) and #4 (35-30.4B) caused by slurry erosion. The slurry-erosion test with an impingement angle of 45 degrees involved both shearing and vertical stress components, which more represented general slurry erosion conditions. In this case, the samples with doped boron, which strengthened the material, showed considerably increased to resistance to the slurry erosion. Such improvement is mainly
Fig. 12. Erosion-corrosion resistance of the fabricated alloys, the modified alloys are more resistant to erosion-corrosion than the base alloy. Slurry velocity relative to sample: 8 m/s. Impingement angle: 45 degrees.
attributed to the elevated hardness of B-doped carbide and increased volume fraction of B-doped carbides. As shown in Fig. 13, the base alloy (sample #1) with lower resistance to slurry-erosion shows lower surface integrity. The slurry erosion resistance of the alloys depends not only on their hardness but also toughness. However, the velocity of slurry for the current slurry erosion tests (8 m/s) was much lower than that for the high-speed solid particle erosion testing (55 m/s). Thus, the toughness became less crucial and consequently the harder B-doped alloys demonstrated better performance than the base alloy during the slurry erosion tests. Corrosion may not play an important role either during the sand-containing aqueous slurry erosion test due to low corrosivity of the slurry and high concentration of Cr in the alloys, which benefits the corrosion resistance. As shown earlier, the trace boron helped change the carbide morphology from strips to relatively equiaxed particles. This may help reduce stress concentration around angular edges of carbides, thus benefiting the resistance to erosion as well.
Fig. 13. Eroded surfaces of (a) sample #1, 35-3 and (b) sample #4, 35-3-0.4B caused by slurry-erosion. The eroded surface of sample #1 without doped boron shows less surface integrity, compared to that of sample #4 with doped B.
586
H. Lu et al. / Wear 376-377 (2017) 578–586
4. Conclusions In this study, we investigated effects of trace boron on the performance of a high-chromium cast iron used as a wear-resistant material in the oil sands industry. Microstructure of the modified cast irons was characterized, and corresponding mechanical, electrochemical and tribological properties were evaluated and relevant mechanisms are discussed. The following conclusions are drawn from the investigation: 1) The added trace B was detected in both carbide and matrix, resulting in increased hardness of carbide but not the matrix possibly due to its low concentration. 2) The added trace boron increased the volume fraction of carbide. 3) The increases in the hardness of carbide and the raised volume fraction of carbide led to considerably enhanced resistance to sliding wear and erosion in the sand-containing aqueous slurry. 4) The B-doped alloys showed lowered resistance to high-speed solid particle erosion at 90-degree impingement, indicating that the material had its hardness increased as expense of toughness to some degree. 5) The boron addition resulted in slightly lowered corrosion potential, which may be ascribed to elevated activities of both B-doped carbide and matrix, evidenced by their lowered electron work functions. The higher volume fraction of carbides may consume Cr in the matrix, thus negatively influencing the passivation capability.
Acknowledgement The authors are grateful for financial support from the Natural Science and Engineering Research Council of Canada, Camber Technology Corporation, Suncor Energy Inc., GIW Industries Inc., Shell Canada Ltd., Magna International Inc. and Volant Products Inc.
References [1] R. Llewellyn, S. Yick, K. Dolman, Scouring erosion resistance of metallic materials used in slurry pump service, Wear 256 (2004) 592–599. [2] M. Jones, R.J. Llewellyn, Erosion–corrosion assessment of materials for use in the resources industry, Wear 267 (2009) 2003–2009. [3] L.L. Parent, D.Y. Li, Wear of hydrotransport lines in Athabasca oil sands, Wear 301 (2013) 477–482. [4] C. Scandian, et al., Effect of molybdenum and chromium contents in sliding wear of high-chromium white cast iron: the relationship between microstructure and wear, Wear 267 (2009) 401–408. [5] Y.X. Li, Z.L. Liu, X. Chen, Development of boron white cast iron, Int. J. Cast. Met. Res. 21 (2008) 67–70. [6] E. Zumelzu, et al., Wear and corrosion behaviour of high-chromium (14–30% Cr) cast iron alloys, J. Mater. Process. Technol. 128 (2002) 250–255.
[7] Ö. Doğan, J. Hawk, G. Laird II, Solidification structure and abrasion resistance of high chromium white irons, Metall. Mater. Trans. A 28 (1997) 1315–1328. [8] X. Tang, et al., Beneficial effects of the core–shell structure of primary carbides in high-Cr (45 wt%) white cast irons on their mechanical behavior and wear resistance, Tribol. Lett. 58 (2015) 1–10. [9] X. Tang, et al., Microstructure of high (45wt.%) chromium cast irons and their resistances to wear and corrosion, Wear 271 (2011) 1426–1431. [10] X. Tang, et al., Variations in microstructure of high chromium cast irons and resultant changes in resistance to wear, corrosion and corrosive wear, Wear 267 (2009) 116–121. [11] R. Chung, et al., Microstructure refinement of hypereutectic high Cr cast irons using hard carbide-forming elements for improved wear resistance, Wear 301 (2013) 695–706. [12] X. Chen, Y. Li, Effect of heat treatment on microstructure and mechanical properties of high boron white cast iron, Mater. Sci. Eng. A 528 (2010) 770–775. [13] H.K. Zeytin, et al., Effect of boron and heat treatment on mechanical properties of white cast iron for mining application, J. Iron Steel Res. Int. 18 (2011) 31–39. [14] X. Sun, Y. Wang, D.Y. Li, Mechanical properties and erosion resistance of ceria nano-particle-doped ultrafine WC–12Co composite prepared by spark plasma sintering, Wear 301 (2013) 406–414. [15] B. Yu, D. Li, A. Grondin, Effects of the dissolved oxygen and slurry velocity on erosion–corrosion of carbon steel in aqueous slurries with carbon dioxide and silica sand, Wear 302 (2013) 1609–1614. [16] G. Kresse, J. Furthmüller, Efficiency of ab-initio total energy calculations for metals and semiconductors using a plane-wave basis set, Comp. Mater. Sci. 6 (1996) 15–50. [17] G. Kresse, J. Furthmüller, Efficient iterative schemes for ab initio total-energy calculations using a plane-wave basis set, Phys. Rev. B 54 (1996) 11169. [18] G. Kresse, D. Joubert, From ultrasoft pseudopotentials to the projector augmented-wave method, Phys. Rev. B 59 (1999) 1758. [19] J.P. Perdew, K. Burke, M. Ernzerhof, Generalized gradient approximation made simple, Phys. Rev. Lett. 77 (1996) 3865–3868. [20] H.J. Monkhorst, J.D. Pack, Special points for Brillouin-zone integrations, Phys. Rev. B 13 (1976) 5188. [21] D. Chung, W. Buessem, The Voigt‐Reuss‐Hill approximation and elastic moduli of polycrystalline MgO, CaF2, β‐ZnS, ZnSe, and CdTe, J. Appl. Phys. 38 (1967) 2535–2540. [22] P. Zhang, et al., Optimization on mechanical properties of Fe 7 xCrxC 3 carbides by first-principles investigation, J. Alloy. Compd. 560 (2013) 49–53. [23] S.M. TOMOVIC-PETROVC, S.V. MARKOVIC, S. ZEC, The effect of boron on the amount and type of carbides in chromium white irons, J. Serbian Chem. Soc. 67 (2002) 697–707. [24] T.A. Adler, Ö.N. Doğan, Erosive wear and impact damage of high-chromium white cast irons, Wear 225 (1999) 174–180. [25] B. Lu, J. Luo, S. Chiovelli, Corrosion and wear resistance of chrome white irons —A correlation to their composition and microstructure, Metall. Mater. Trans. A 37 (2006) 3029–3038. [26] H.B. Michaelson, The work function of the elements and its periodicity, J. Appl. Phys. 48 (1977) 4729–4733. [27] S. Halas, 100 years of work function, Mater. Sci.-Pol. 24 (2006) 952. [28] C. Oernek, D.L. Engelberg, Correlative EBSD and SKPFM characterisation of microstructure development to assist determination of corrosion propensity in grade 2205 duplex stainless steel, J. Mater. Sci. 51 (2016) 1931–1948. [29] X.C. Huang, H. Lu, D.Y. Li, Understanding the corrosion behavior of isomorphous Cu–Ni alloy from its electron work function, Mater. Chem. Phys. 173 (2016) 238–245. [30] S. Halas, T. Durakiewicz, Work functions of elements expressed in terms of the Fermi energy and the density of free electrons, J. Phys. Condens. Matter 10 (1999) 10815. [31] H. Lu, et al., Electron work function–a promising guiding parameter for material design, Sci. Rep. 6 (2016) 24366. [32] H. Lu, G. Hua, D. Li, Dependence of the mechanical behavior of alloys on their electron work function—An alternative parameter for materials design, Appl. Phys. Lett. 103 (2013) 261902.
Contents lists available at ScienceDirect
Wear journal homepage: www.elsevier.com/locate/wear
Improvement in erosion-corrosion resistance of high-chromium cast irons by trace boron Hao Lu a, Tingzhong Li a, Juan Cui a, Qingyang Li a,b, D.Y. Li a,n a
Department of Chemical and Materials Engineering, University of Alberta, Edmonton, Alberta, Canada T6G 1H9 State Key Laboratory of Urban Water Resource and Environment, School of Chemical Engineering and Technology, Harbin Institute of Technology, Harbin 150001, China
b
art ic l e i nf o
a b s t r a c t
Article history: Received 12 September 2016 Received in revised form 28 January 2017 Accepted 3 February 2017
Extensive work was carried out to improve the performance of cast irons used for slurry transport in the oil sands industry. Because of complex condition of the multi-phase fluid containing oil sand particles, erosion-corrosion damage has long been an issue. High chromium cast irons (HCCIs) are currently used for slurry pumps, showing high resistance to sand-containing slurry erosion. However, further enhanced HCCIs are highly desired in order to more effectively reduce the damage to oil sand slurry transport facilities caused by erosion-corrosion for prolonged service life. This study was focused on tailoring an existing HCCI used in the oil sands industry by adding trace boron. Microstructure and relevant properties of the modified cast irons were characterized and evaluated. It was demonstrated that trace boron markedly increased the resistance of the cast iron to slurry erosion. The mechanism responsible for the improvement was analyzed based on observed variations in microstructure and properties. & 2017 Elsevier B.V. All rights reserved.
Keywords: Trace Boron Erosion-corrosion resistance Sliding wear Corrosion High-chromium cast irons
1. Introduction Wear is one of the most predominant failure mechanisms for material loss in oil sands industry operations. The extreme wear conditions result from aggressive abrasion and impact caused by the large quantities of silica sand contained in the oil sand and its slurry [1]. In addition to the harsh wearing conditions, corrosion and consequently the synergistic attacks from mechanical and chemical actions considerably shorten the materials lifespan as well [2]. Synergistic erosion and corrosion or erosion-corrosion is the dominant wear mechanism for slurry handling in the oil sand industry [3]. High chromium cast irons (HCCIs) are widely used to resist abrasion and erosion in many industrial processes, such as mining, mineral processing and oil sand operations [4], due to their high hardness and appropriate balance between hardness and toughness, compared to conventional white cast irons [5], as well as high corrosion resistance [6]. The wear resistance of HCCIs results mainly from the high volume fraction of hard M7C3 carbides with high hardness and the supporting from tough metallic matrix [7]. HCCIs are classified according to ASTM A532 “Standard Specification for Abrasion-Resistant Cast Irons”. Commercially used HCCIs have their compositions usually in the range of 12–27 wt% Cr and 2.4–3.6 wt% C, which process high wear resistance n
Corresponding author. E-mail address: [email protected] (D.Y. Li).
http://dx.doi.org/10.1016/j.wear.2017.02.014 0043-1648/& 2017 Elsevier B.V. All rights reserved.
and widely used in slurry transport systems in the oil sands industry [1]. Most HCCIs used in oil sand industry for slurry pump casing and impellers fall into the composition range of 2.4–3.6 wt. % C and 25 27 wt. % Cr. However, this group of HCCIs does not always perform satisfactorily, due to the synergistic attack of oil sand particle erosion and aqueous corrosion [8]. Considerable work has been done to improve the wear resistance of HCCIs. In recent years, the oil sand mining industry intends to push the Cr and C contents to higher levels. The former increases hardness while the latter improves the corrosion resistance. A proper combination of the improvements could be suitable for specific wearing conditions. Such HCCIs with high concentrations of carbon and chromium were previously considered to be non-castable due to high rejection rate (rejected pieces/processed pieces) [9]. However, it has become possible to cast HCCIs with high concentrations of chromium and carbon with the advance in foundry techniques [1]. In our previous studies, HCCIs with 30 wt% to 45 wt% chromium were fabricated and demonstrated to be highly resistant to wear and corrosion [9,10]. Other effective approaches to enhance HCCIs wear and corrosion resistance include appropriate heat treatment and alloying with minor elements, such as boron, nickel, molybdenum, titanium, vanadium, to optimize microstructure [11]. Studies have demonstrated that adding boron is promising to improve the wear resistance of white cast irons [12]. Such improvement is ascribed to both elevated hardness and toughness, resulting from formed
H. Lu et al. / Wear 376-377 (2017) 578–586
Table 1 Nominal compositions of HCCIs under study. Sample no.
%Cr
%C
%B
#1 #2 #3 #4 #5 #6
35 35 35 35 35 35
3 3 3 3 3 3
0 0.12 0.2 0.4 0.48 0.6
boride phases and boron-doped carbide as well as microstructure modification [5,13]. However, few studies have been conducted to investigate the effect of boron on HCCIs’ wear resistance, especially on the HCCIs with their chromium concentrations closer to the high end of Cr concentration range. In this study, an existing HCCI with 35 wt% chromium and 3 wt% carbon for oil sand slurry handling was modified by adding trace boron. The objective of this study is to investigate how the trace boron affects microstructure and relevant properties, including resistance to sliding wear, high speed solid particle erosion, corrosion, erosion-corrosion, of the modified cast irons.
2. Materials and methods 2.1. Materials Six HCCIs with nominal chemical compositions of 35 wt% chromium, 3 wt% carbon, and various concentrations of boron: 0, 0.12, 0.2, 0.4, 0.48, 0.6 wt%, respectively, balanced by iron were fabricated using an induction furnace. The molten metals were poured into copper molds to produce the samples. Table 1 gives nominal compositions of samples under study. 2.2. Characterization and testing Microstructures of the samples were examined using a Scanning Electron Microscope (Zeiss Sigma 300 VP-FESEM). Worn surfaces after sliding wear, high-speed particle erosion, and slurry erosion tests were observed under a Scanning Electron Microscope (Tescan Vega 3 SEM, Czech Republic).
579
Hardness of the samples was measured using a Rockwell hardness tester (Zwick/Roell Rockwell/superficial Rockwell hardness tester, Indentec Hardness Testing Machine Limited, UK) under a load of 60 kgf with a diamond cone indenter. Sliding wear tests were performed on a pin-on-disc tribometer (CSEM Instruments, Neuchatel, Switzerland) based on ASTM G 99 “Standard Test Method for Wear Testing with a Pin-on-Disk Apparatus”. The disc was the sample under study and the pin was a silicon nitride ball with 6mm diameter. All tests were performed at a sliding speed of 1 cm/s along a circle path of 2.0 mm in diameter under a normal load of 10 N for a sliding distance of 36 m. Wear tracks and corresponding volume losses of the samples were determined using a confocal microscope (ZeGage 3D optical profilemeter, Zygo Corp.). High-speed solid particle erosion tests were carried out using a home-made air-jet erosion tester [14] based on ASTM G 76 “Standard Test Method for Conducting Erosion Tests by Solid Particle Impingement Using Gas Jets” (Fig. 1(a)). The erosion tests were performed at an impingement angle of 90 °and a dry air flow with a pressure of 40 psi was used to eject sand particles to sample surface (corresponding sand flow speed: 55 m/s). Erosion at an impingement angle of 90° usually results in more damage to less tough materials, thus it more or less reflects the toughness of the samples. AFS 50/70 sand (U.S. Silica Company, USA) was used as the erodent particle. The sample mass was measured before and after the tests using a balance with a precision of 0.1 mg. Each test was repeated at least 3 times. The erosion–corrosion tests were carried out using a homemade slurry-pot tester (Fig. 1(b)). Guidelines for erosion-corrosion tests in slurry can be found in STP946 “Slurry Erosion: Uses, Applications, and Test Methods”. The tester had a cylindrical tank of 29 cm in diameter and 22 cm in height. A slurry solution containing 1.5 L silica sand (500 μm in diameter) and 6 L water was placed in the slurry container. Samples were held by a holder, which was rotated during erosion test, thus driving the samples to move in the slurry. The angle between sample surface and the moving direction was kept at 45 °and the moving velocity of samples in the slurry was 8 m/s. For each slurry erosion test, the total distance over which the sample travelled in the slurry was 15 km. After test, corrosion products were removed by rinsing and light brushing, and the samples were finally cleaned by rinsing with distilled water. The weight loss of each sample was then
Fig. 1. Schematic illustrations of (a) air-jet erosion tester used for high-speed solid particle erosion tests, and (b) slurry-pot tester used for erosion-corrosion tests.
580
H. Lu et al. / Wear 376-377 (2017) 578–586
3. Results and discussions 3.1. Microstructure characterization
Fig. 2. Crystal structure of Fe4Cr3C2B; green, brown, blue and yellow balls represent boron, carbon, chromium and iron atoms, respectively. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
measured using a balance with a precision of 0.1 mg. More details about the home-made slurry pot erosion-corrosion tester and the experimental set-up can be found in our previous publication [15]. It should be indicated that this study is conducted to investigate effect of trace B on the performance of the sample alloys. Relevant wear tests were performed for comparison purpose. Thus, the testing conditions were selected based on what we usually used for testing many other industrial materials rather than exactly follow the standard testing conditions. This helps obtain a sense about the performance of the HCCIs under study, compared to those of materials used in industry. Open circuit potentials (OCP) of the samples in a 3.5 wt% NaCl aqueous solution were measured using a computerized Gamary Instruments Frame-work electrochemical system. A saturated calomel electrode (SCE) was used as the reference electrode and a platinum plate with an area of 1 cm2 was used as the counter electrode. The tests were performed based on ASTM G 3 “Standard Practice for Conventions Applicable to Electrochemical Measurements in Corrosion Testing” and ASTM G 59 “Standard Test Method for Conducting Potentiodynamic Polarization Resistance Measurements”.
2.3. Simulation methods Mechanical properties and electron densities of carbide and boron-doped carbide were calculated, respectively. The calculations were conducted based on the first-principles density functional theory (DFT) and implemented using the Vienna ab-initio simulation package (VASP) [16,17]. Electron interactions were treated with the projector-augmented wave method [18]. The generalized gradient approximation (GGA) of Perdew-Burke-Ernzerhof (PBE) was used for the exchange and correlation effects [19]. The cut-off energy for plane wave basis was set to 400 eV. A Monkhorst-pack [20] sampling of 7 7 9 k-point mesh was adopted to achieve self-consist field convergence, and 11 11 13 k-point mesh was adopted to compute total energy, charge density and electronic structure. In all calculations, self-consistency was achieved when the total energy difference between cycles was smaller than 0.1 meV. The geometry relaxation tolerance in all forces was below 0.01 eV/A. The moduli were calculated with Voigt–Reuss–Hill approximation [21]. Space group P63mc (#186) was applied to construct the carbide structure in the simulation. Two boron atoms replaced two carbon atoms, which made the carbide Fe4Cr3C2B comparable with Fe4Cr3C3. Crystal structure of Fe4Cr3C2B is illustrated in Fig. 2. The reason to choose Fe4Cr3C3 as a reference for comparison is that it is the hardest in the (Fe, Cr)7C3 carbide series [22].
SEM images of the samples are illustrated in Fig. 3. Alloy #1 has a nearly eutectic microstructure with strip carbides. However, other five B-doped alloys show hypereutectic microstructures consisting primary carbides and eutectic matrix, similar to the microstructure observed in other boron-added conventional white cast irons [13]. It appears that the added boron more or less spheroidized the carbides as Fig. 3 illustrates. Using sample #6 containing 0.6%B as an example, we analyzed distributions of elements in different micro-constituents through EDX compositional mapping. Maps of various elements are shown in Fig. 4. As demonstrated, B is soluble in all phases, showing its ability to substitute Fe in the matrix and C in the carbide. Segregation of boron is not observed in the EDX maps, indicating that the sample only contains boron-doped carbides (Fe, Cr) 7 (C, B)3 without formation of borides such as (Fe, Cr) 2 B. It should be mentioned that EDX is not accurate to analyze light elements such as B and C, thus the EDX mapping was carried out to demonstrate B distribution rather than quantify boron contents in carbides and the matrix. The elemental distribution of B shown in Fig. 4 indicates the formation of boron-doped carbide, (Fe, Cr) 7 (C, B) 3 and boron-substituted Fe matrix. As reported and discussed in the following sections, the doped boron strengthened the carbide phase and increased its volume fraction, leading to higher hardness and enhanced wear resistance of the boron-doped alloys. 3.2. Properties 3.2.1. Hardness Bulk hardness of the six alloys was measured and results of the measurement are given in Fig. 5. As shown, all B-doped alloys are harder than the base alloy (#1). The higher overall hardness should be attributed to the following factors: 1) the doped boron strengthened (Fe, Cr)7C3, and 2) the dope B resulted in an increase in the volume fraction of carbides (Table 2). The trace boron may not produce other second phases due to its low concentration. It has been shown that boron-doped carbide phases are harder than the carbide in HCCIs [5,23]. Our modulus mapping using AFM directly confirmed the increase in modulus of carbides with doped B (Table 2). This is also supported by our first-principles calculation, which shows that the B-doped carbide exhibits higher bulk modulus, shear modulus and elastic modulus than the un-doped carbide (see Table 3). It may need to mention that solution hardening of the ferrous matrix by added boron is expected. However, our modulus mapping does not support this expectation (see Table 2). This could be due to the fact that the B concentration is not sufficient enough to result in effective solid-solution hardening. As for the increase in the volume fraction of carbides with the doped boron, we estimated volume fractions of the alloys through image analysis using the software ImageJ. As shown in Table 2, the volume fraction of carbides increased with increasing the concentration of boron. Such a phenomenon was also noticed by others [23]. The added B could facilitate nucleation of carbides. An increase in the volume fraction of carbides should contribute to the observed increase in the mechanical strength of the high-Cr cast irons. 3.2.2. Resistance of the alloys to sliding wear and high-speed particle erosion Resistance of the alloys to sliding wear and high speed particle erosion were measured and presented in Fig. 6. As illustrated, all B-doped alloys demonstrated considerably
H. Lu et al. / Wear 376-377 (2017) 578–586
581
Fig. 3. Microstructure of the fabricated alloys (a) #1, (b) #2, (c) #3, (d) #4, (e) #5, (f) #6.
reduced volume losses during sliding wear tests (only 15% to 25% of the volume loss of the base alloy without boron), which can be explained by the elevated mechanical strength of the B-doped alloys. Fig. 7 illustrates the surface morphology of selected samples after sliding wear testing. Grooves caused by plowing and plastic deformation can be observed on the worn surface. Compared to that of base alloy #1, the worn surface of the modified alloys is smoother with shallower wear grooves, indicating that the penetration of the pin ball is much more resisted by the B-doped alloys than by the base alloy. Although the B-doped alloys perform considerably better than the base alloy during sliding wear tests, their performance is inferior to that of the base alloy when are subjected to high-speed particle erosion, for which material's toughness plays an important role especially for high-angle impingement. For the erosion tests, 90-degree impingement angle was used. Since erosion at larger impingement angles usually results in more damage to less tough materials, the erosion tests with 90-degree
impingement angle were more influenced by toughness of the alloys. As shown in Fig. 6, the B-doped alloys are inferior to the base alloy during the high-speed particle erosion testing, implying that the modified alloys are not as tough as the base alloy. Eroded surfaces were examined using SEM. As illustrated in Fig. 8, the eroded surfaces show typical features of eroded surface of high chromium cast irons as reported in previous studies [24]. The eroded surfaces are covered by overlapping craters caused by impact, and the craters have edges with extruded platelets. The rate of forming these platelets and break off controls the rate of erosion for the materials [24]. As illustrated in Fig. 8, no visible cracks are observed. This may indicate that the alloys possess reasonable toughness, although the doped boron increases hardness of the material at expense of toughness to some degree, as shown by the increased material loss caused by high-speed particle striking at 90-degree impingement angle (Fig. 6). There is no significant morphological difference among eroded surfaces of different alloys. It is known that strain-hardening of
582
H. Lu et al. / Wear 376-377 (2017) 578–586
Fig. 4. EDX compositional maps of sample #6 showing the elements distributions of Fe (blue), Cr (green), C (yellow), and B (red). The first figure (top left) is a SEM image of the area for compositional analysis. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
Table 2 Volume fractions of carbides, moduli of matrix and carbides in the samples. Sample no. Volume fraction of carbides
Modulus of matrix
Modulus of carbide
#1 #2 #3 #4 #5 #6
234.4 GPa 231.2 GPa 228.6 GPa 227.2 GPa 225.0 GPa 223.1 GPa
265.6 GPa 278.5 GPa 280.2 GPa 281.3 GPa 293.0 GPa 302.4 GPa
38.1% 38.6% 40.1% 40.3% 42.3% 47.8%
Table 3 Calculated moduli of carbides with and without boron.
Fig. 5. Bulk hardness of the fabricated alloys; modified alloys possess higher hardness than the base alloy. The higher hardness of modified alloys is attributed to the formation of boron doped carbide phase, Fe4Cr3C2B, shown in Fig. 2, and increased volume fraction of carbides.
Fe4Cr3C3 Fe4Cr3C2B
B (GPa)
G (GPa)
E (GPa)
277.1 288.2
138.4 142.1
355.6 366.1
H. Lu et al. / Wear 376-377 (2017) 578–586
583
Fig. 6. Resistance of the alloys to sliding wear and high-speed silica particle erosion (particle velocity: 55 m/s; impingement angle: 90 degrees); modified alloys are more resistant to sliding wear, but less resistant to high-speed particle erosion.
metallic materials influences their erosion rate. The HCCIs with higher volume fraction of carbides have smaller strain hardening exponent and higher erosion rate [24]. This may also be a factor responsible for the lower resistance of B-doped HCCIs having higher volume fractions of carbides to high-speed solid particle erosion. 3.2.3. Corrosion resistance Open circuit potential tests were performed to evaluate how the doped B affected the corrosion tendency of the cast iron in a 3.5 wt% NaCl aqueous solution. Fig. 9. illustrates the corrosion potential of the alloys with and without doped boron. As shown, the B-doped alloys have lower open circuit potentials, compared to the base alloy, indicating that adding boron renders the HCCI prone to corrosion, compared to the base alloy. The following are possible factors contributing to the decrease in corrosion potential: a) The corrosion resistance of HCCIs largely depends on the corrosion behaviour of the ferrous matrix, in which the concentration of free chromium is crucial [25]. Sufficient free Cr atoms result in improved passivation capability with forming a protective passive film. Since the B-doped alloys have increased volume fractions of carbides (Table 2), which consumes more chromium in the matrix, the passivation capability could be lowered. b) Compared to Cr ( ϕCr = 4.5eV ) and Fe ( ϕFe = 4.5eV ), B has a lower work function ( ϕB = 4.45eV )[26,27]. Work function reflects the surface activity and lower work function corresponds to lower corrosion potential [28,29]. To determine if the B-doped alloys were more active than the base alloy, we measured electron work functions of matrixes and carbides with and without doped boron using a multi-mode atomic force microscope (Bruker, Multimode 8, and U.S.A.). Fig. 10 illustrates a representative work function map of the HCCIs samples. Average values of work functions of carbide and matrix are given in Table 4, which shows that both the B-doped matrix and carbide show lowered work functions, though not significant, corresponding to raised surface activities.
Fig. 7. Wear track morphology of selected samples (a) #1 (d) #3 (g) #6, and corresponding surface profiles (c) #1 (f) #3 (i) #6, observed by optical confocal microscope; (b) (e) (h) are SEM images of the worn surfaces of samples #1, #3, and #6. As shown, the worn tracks of boron modified alloys are narrower and shallower than that of the base alloy.
The above two factors could be mainly responsible for the decrease in open circuit potential and corrosion resistance of sample with doped B. As for the reason why the work function of carbide is lowered by doped boron, we analysed electron densities of carbides with
584
H. Lu et al. / Wear 376-377 (2017) 578–586
Fig. 8. Surfaces eroded by high-speed particles:(a)#1, (b) #3, (c) #6.
Fig. 9. Open circuit potential curves of the alloys, the modified alloys have lower open circuit potential than the base alloy.
Fig. 10. A representative work function map of HCCI, bright regions represent carbides with higher work functions, and dark regions represent matrix with low work function. A line profile of work function across both carbide and matrix is illustrated (curve in white).
Fig. 11. Electron densities of (110) plane for (a) carbide and (b) B-dope carbide, red color represents high electron density and blue color represents low electron density; local potentials of (110) plane for (c) carbide and (d) B-doped carbide, red color represents high local potential and blue color represents low local potential. Higher potential corresponds to lower work function and lower electron density. B-doped carbide shows lower electron density and work function. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
H. Lu et al. / Wear 376-377 (2017) 578–586
585
Table 4 Work functions of matrix and carbide for selected samples. Sample
EWF of matrix (eV)
EWF of carbide (eV)
#1 (no boron) #4 (0.4% B) #6 (0.6% B)
4.56 4.54 4.52
4.67 4.57 4.56
and without doped B through first-principles calculation. Fig. 11 shows electron densities and local potentials of (110) plane for carbide and B-doped carbide. As illustrated, B-doped carbide has a higher potential, corresponding to a lower work function (difference between the vacuum potential and local potential). This is also consistent with the electron density calculation, i.e. higher electron density corresponds to higher work function [30–32]. 3.2.4. Erosion-corrosion resistance of the alloys Slurry erosion tests were performed in a neutral slurry aqueous solution with 20% silica sand particles. Weight losses of the samples after testing are presented in Fig. 12. As shown, the resistance of the HCCIs to erosion-corrosion was improved remarkably when the base alloy was modified by adding trace boron. The higher the concentration of added boron, the more resistant was the alloy to slurry erosion-corrosion. As shown, the weight loss of sample #6 with 0.6 wt% B is only about 40% of that of the base alloy (#1) without doped boron. The hard carbide in HCCIs plays a main role in resisting sliding wear and slurry erosion if the slurry is not very corrosive [9]. As discussed earlier, all modified alloys possess higher hardness and demonstrate considerably reduced volume losses during sliding wear tests. Thus, the enhanced slurry erosion resistance of the modified alloys in this sand-containing aqueous slurry with low corrosivity should be mainly attributed to the increased hardness. Fig. 13 presents eroded surfaces of sample #1 (35-3) and #4 (35-30.4B) caused by slurry erosion. The slurry-erosion test with an impingement angle of 45 degrees involved both shearing and vertical stress components, which more represented general slurry erosion conditions. In this case, the samples with doped boron, which strengthened the material, showed considerably increased to resistance to the slurry erosion. Such improvement is mainly
Fig. 12. Erosion-corrosion resistance of the fabricated alloys, the modified alloys are more resistant to erosion-corrosion than the base alloy. Slurry velocity relative to sample: 8 m/s. Impingement angle: 45 degrees.
attributed to the elevated hardness of B-doped carbide and increased volume fraction of B-doped carbides. As shown in Fig. 13, the base alloy (sample #1) with lower resistance to slurry-erosion shows lower surface integrity. The slurry erosion resistance of the alloys depends not only on their hardness but also toughness. However, the velocity of slurry for the current slurry erosion tests (8 m/s) was much lower than that for the high-speed solid particle erosion testing (55 m/s). Thus, the toughness became less crucial and consequently the harder B-doped alloys demonstrated better performance than the base alloy during the slurry erosion tests. Corrosion may not play an important role either during the sand-containing aqueous slurry erosion test due to low corrosivity of the slurry and high concentration of Cr in the alloys, which benefits the corrosion resistance. As shown earlier, the trace boron helped change the carbide morphology from strips to relatively equiaxed particles. This may help reduce stress concentration around angular edges of carbides, thus benefiting the resistance to erosion as well.
Fig. 13. Eroded surfaces of (a) sample #1, 35-3 and (b) sample #4, 35-3-0.4B caused by slurry-erosion. The eroded surface of sample #1 without doped boron shows less surface integrity, compared to that of sample #4 with doped B.
586
H. Lu et al. / Wear 376-377 (2017) 578–586
4. Conclusions In this study, we investigated effects of trace boron on the performance of a high-chromium cast iron used as a wear-resistant material in the oil sands industry. Microstructure of the modified cast irons was characterized, and corresponding mechanical, electrochemical and tribological properties were evaluated and relevant mechanisms are discussed. The following conclusions are drawn from the investigation: 1) The added trace B was detected in both carbide and matrix, resulting in increased hardness of carbide but not the matrix possibly due to its low concentration. 2) The added trace boron increased the volume fraction of carbide. 3) The increases in the hardness of carbide and the raised volume fraction of carbide led to considerably enhanced resistance to sliding wear and erosion in the sand-containing aqueous slurry. 4) The B-doped alloys showed lowered resistance to high-speed solid particle erosion at 90-degree impingement, indicating that the material had its hardness increased as expense of toughness to some degree. 5) The boron addition resulted in slightly lowered corrosion potential, which may be ascribed to elevated activities of both B-doped carbide and matrix, evidenced by their lowered electron work functions. The higher volume fraction of carbides may consume Cr in the matrix, thus negatively influencing the passivation capability.
Acknowledgement The authors are grateful for financial support from the Natural Science and Engineering Research Council of Canada, Camber Technology Corporation, Suncor Energy Inc., GIW Industries Inc., Shell Canada Ltd., Magna International Inc. and Volant Products Inc.
References [1] R. Llewellyn, S. Yick, K. Dolman, Scouring erosion resistance of metallic materials used in slurry pump service, Wear 256 (2004) 592–599. [2] M. Jones, R.J. Llewellyn, Erosion–corrosion assessment of materials for use in the resources industry, Wear 267 (2009) 2003–2009. [3] L.L. Parent, D.Y. Li, Wear of hydrotransport lines in Athabasca oil sands, Wear 301 (2013) 477–482. [4] C. Scandian, et al., Effect of molybdenum and chromium contents in sliding wear of high-chromium white cast iron: the relationship between microstructure and wear, Wear 267 (2009) 401–408. [5] Y.X. Li, Z.L. Liu, X. Chen, Development of boron white cast iron, Int. J. Cast. Met. Res. 21 (2008) 67–70. [6] E. Zumelzu, et al., Wear and corrosion behaviour of high-chromium (14–30% Cr) cast iron alloys, J. Mater. Process. Technol. 128 (2002) 250–255.
[7] Ö. Doğan, J. Hawk, G. Laird II, Solidification structure and abrasion resistance of high chromium white irons, Metall. Mater. Trans. A 28 (1997) 1315–1328. [8] X. Tang, et al., Beneficial effects of the core–shell structure of primary carbides in high-Cr (45 wt%) white cast irons on their mechanical behavior and wear resistance, Tribol. Lett. 58 (2015) 1–10. [9] X. Tang, et al., Microstructure of high (45wt.%) chromium cast irons and their resistances to wear and corrosion, Wear 271 (2011) 1426–1431. [10] X. Tang, et al., Variations in microstructure of high chromium cast irons and resultant changes in resistance to wear, corrosion and corrosive wear, Wear 267 (2009) 116–121. [11] R. Chung, et al., Microstructure refinement of hypereutectic high Cr cast irons using hard carbide-forming elements for improved wear resistance, Wear 301 (2013) 695–706. [12] X. Chen, Y. Li, Effect of heat treatment on microstructure and mechanical properties of high boron white cast iron, Mater. Sci. Eng. A 528 (2010) 770–775. [13] H.K. Zeytin, et al., Effect of boron and heat treatment on mechanical properties of white cast iron for mining application, J. Iron Steel Res. Int. 18 (2011) 31–39. [14] X. Sun, Y. Wang, D.Y. Li, Mechanical properties and erosion resistance of ceria nano-particle-doped ultrafine WC–12Co composite prepared by spark plasma sintering, Wear 301 (2013) 406–414. [15] B. Yu, D. Li, A. Grondin, Effects of the dissolved oxygen and slurry velocity on erosion–corrosion of carbon steel in aqueous slurries with carbon dioxide and silica sand, Wear 302 (2013) 1609–1614. [16] G. Kresse, J. Furthmüller, Efficiency of ab-initio total energy calculations for metals and semiconductors using a plane-wave basis set, Comp. Mater. Sci. 6 (1996) 15–50. [17] G. Kresse, J. Furthmüller, Efficient iterative schemes for ab initio total-energy calculations using a plane-wave basis set, Phys. Rev. B 54 (1996) 11169. [18] G. Kresse, D. Joubert, From ultrasoft pseudopotentials to the projector augmented-wave method, Phys. Rev. B 59 (1999) 1758. [19] J.P. Perdew, K. Burke, M. Ernzerhof, Generalized gradient approximation made simple, Phys. Rev. Lett. 77 (1996) 3865–3868. [20] H.J. Monkhorst, J.D. Pack, Special points for Brillouin-zone integrations, Phys. Rev. B 13 (1976) 5188. [21] D. Chung, W. Buessem, The Voigt‐Reuss‐Hill approximation and elastic moduli of polycrystalline MgO, CaF2, β‐ZnS, ZnSe, and CdTe, J. Appl. Phys. 38 (1967) 2535–2540. [22] P. Zhang, et al., Optimization on mechanical properties of Fe 7 xCrxC 3 carbides by first-principles investigation, J. Alloy. Compd. 560 (2013) 49–53. [23] S.M. TOMOVIC-PETROVC, S.V. MARKOVIC, S. ZEC, The effect of boron on the amount and type of carbides in chromium white irons, J. Serbian Chem. Soc. 67 (2002) 697–707. [24] T.A. Adler, Ö.N. Doğan, Erosive wear and impact damage of high-chromium white cast irons, Wear 225 (1999) 174–180. [25] B. Lu, J. Luo, S. Chiovelli, Corrosion and wear resistance of chrome white irons —A correlation to their composition and microstructure, Metall. Mater. Trans. A 37 (2006) 3029–3038. [26] H.B. Michaelson, The work function of the elements and its periodicity, J. Appl. Phys. 48 (1977) 4729–4733. [27] S. Halas, 100 years of work function, Mater. Sci.-Pol. 24 (2006) 952. [28] C. Oernek, D.L. Engelberg, Correlative EBSD and SKPFM characterisation of microstructure development to assist determination of corrosion propensity in grade 2205 duplex stainless steel, J. Mater. Sci. 51 (2016) 1931–1948. [29] X.C. Huang, H. Lu, D.Y. Li, Understanding the corrosion behavior of isomorphous Cu–Ni alloy from its electron work function, Mater. Chem. Phys. 173 (2016) 238–245. [30] S. Halas, T. Durakiewicz, Work functions of elements expressed in terms of the Fermi energy and the density of free electrons, J. Phys. Condens. Matter 10 (1999) 10815. [31] H. Lu, et al., Electron work function–a promising guiding parameter for material design, Sci. Rep. 6 (2016) 24366. [32] H. Lu, G. Hua, D. Li, Dependence of the mechanical behavior of alloys on their electron work function—An alternative parameter for materials design, Appl. Phys. Lett. 103 (2013) 261902.