- Email: [email protected]
Effect of surface topography on the dynamics of the abrasive particles during micro-abrasion
Wear 324-325 (2015) 129–139
Contents lists available at ScienceDirect
Wear journal homepage: www.elsevier.com/locate/wear
Effect of surface topography on the dynamics of the abrasive particles during micro-abrasion$ H.L. Costa n, M.A.N. Ardila, W.S. Labiapari, W.M. Silva, J.D.B. de Mello Laboratório de Tribologia e Materiais, Universidade Federal de Uberlândia, Uberlandia, MG Brazil
art ic l e i nf o
a b s t r a c t
Article history: Received 12 September 2014 Received in revised form 4 December 2014 Accepted 6 December 2014 Available online 19 December 2014
During abrasive wear, the wear mechanism has been shown to be associated with the movement of the active particles present at the wear interface: rolling, evidenced by indentations on the worn surface, and sliding, which produces scratching and/or ploughing. Particle dynamics can vary with tribological parameters such as different combinations of ball and specimen materials, applied load, slurry concentration, abrasive material, ball condition and equipment configuration (fixed or free-ball). In this article, the effect of surface topography of both the ball and the specimen on the dynamics of the abrasive particles and micro-abrasion wear is investigated for SiO2 abrasive particles. The effect of the ball surface topography was investigated using a fixed-ball rig, zirconia balls (Sa ¼ 0.06, 0.34, and 0.54 mm) and stainless steel specimens (Sa ¼0.10 mm). When the roughness of the ball increased, the wear mechanism changed from sliding to mixed and then to rolling and the micro abrasion coefficient k increased substantially, the difference between the smoothest and the roughest ball being around 510%. The effect of the specimen surface topography was investigated using a free-ball rig, AISI 52100 steel balls (Sa ¼ 0.82 mm) and tool steel specimens (Sq ¼ 0.025 mm and 0.414 mm). The influence of the directionality of the specimen surface finish was also analyzed by conducting tests parallel and perpendicular to the grinding marks using three slurry concentrations. The effect of the topography of the specimens on wear coefficients and mechanisms was much less pronounced than that found for the ball topography. For the highest slurry concentration (20 wt%), k increased for the rougher specimens (around 23%) and a slight change in mechanism occurred from mixed (sliding in the center and rolling at borders of craters), to sliding. This effect was less significant for lower concentrations. The influence of surface directionality on abrasive wear was negligible. & 2014 Elsevier B.V. All rights reserved.
Keywords: Microabrasion Surface topography Wear severity Wear mechanisms
1. Introduction Micro-abrasion tests, also known as ball-cratering tests, have been used for over twenty years. They were originally proposed to characterize the intrinsic mechanical quality of thin coatings when a commercial dimple grinder, conventionally used in thinning of transmission electron microscopy (TEM) samples, was adapted to grind craters into the surfaces of coating/substrate composites [1]. Later, a rotating sphere apparatus [2] was proposed, which is today commercially available and widely used. In the free-ball version, the ball rotates freely driven by friction at the contact with a notched drive shaft. The normal load results from the weight of the ball resting on the test sample, which can be varied by changing the dimensions and material of the ball and/or by altering the angle of the sample holding plate, and depends on $ This article was presented at the 2nd International Conference on Abrasive Processes. n Corresponding author. Tel.: þ 55 34 3239 4036. E-mail address: [email protected] (H.L. Costa).
http://dx.doi.org/10.1016/j.wear.2014.12.011 0043-1648/& 2014 Elsevier B.V. All rights reserved.
the friction coefficient between the ball and the test surface. In the fixed-ball version, the ball is driven directly by clamping the ball in a split drive shaft and the sample is loaded against the ball with the desired normal force by a lever arm arrangement. This version allows a much wider range of normal loads to be used [3]. Microabrasion wear mechanisms and wear coefficients can vary significantly with the operating conditions used in the tests [4–7]. Grooving wear mechanism occurs when a significant proportion of the abrasive particles embed in the surface of the ball and slide acting as fixed indenters, producing a series of fine parallel grooves in the specimen surface. A rolling mechanism is produced when the abrasive particles do not embed, but roll between the two surfaces, producing a heavily deformed, multiply indented surface, with no surface directionality. A mixed mechanism may also occur, producing grooving in the center and three-body rolling at the edges of the wear craters [4,5]. A model proposed by Williams and Hyncica [8] shows that an abrasive particle between two surfaces undergoes a transition from rolling to grooving at a critical value D/h, where D is the particle major axis and h is the separation of the surfaces. When D/h is only
130
H.L. Costa et al. / Wear 324-325 (2015) 129–139
Fig. 1. Test rig details: (a) General view; (b) setup for force measurements.
Table 1 Chemical composition and Vickers hardness (4.9 N, 30 s) of the specimens to evaluate effect of ball topography. Steel
AISI 304
Chemical Composition [wt%]
Hardness [GPa]
C
Mn
Cr
Ni
Nb
0.055
1.15
18.28
8.01
0.01
1.99
slightly larger than 1, the particle indents both surfaces. As there is a similar indentation occurring at the opposite corner of the particle, the two forces acting on the particle, which will not generally be
collinear, form a couple tending to rotate the particle. This produces multiple indentations on the surfaces. However, when D/h increases, the particle initially rotates until eventually the two forces acting on it become collinear. When this happens, there is no impetus to cause further rotation and so the particle will tend to remain at this inclination, producing grooves on the surface. Other authors have adapted this model to associate the transition between the two mechanisms to changes in the severity of the contact. An initial model by Trezona et al. [4] suggests the transition to occur for a certain ratio of normal load to volume fraction of abrasives in the slurry, which varies depending on the type of abrasive. If the contact contains many particles under a low load,
H.L. Costa et al. / Wear 324-325 (2015) 129–139
131
Fig. 2. 3D surface topography of the balls after form removal: (a) fresh, Sa ¼ 0.06 mm; (b) ground ball, Sa ¼ 0.34 mm; ground ball, Sa ¼0.54 mm.
Table 2 Surface conditioning using abrasive papers and final roughness of the zirconia balls. Surface conditioning
Sa (mm)
Fresh (as received) Grit size # 220 for 20 min Grit size # 220 for 20 minþ Grit size # 120 for 10 minþ Grit size # 120 for 10 min
0.067 0.01 0.34 7 0.06 0.54 7 0.04
each particle will indent the surfaces only lightly, and so rolling prevails. Contrarily, grooving wear occurs when a few heavily loaded particles indent the surface more deeply and therefore slide. Consequently, sliding of abrasive particles is generally dominant at high loads and/or low slurry concentrations, whereas rolling tends to prevail at low loads and/or high slurry concentrations. However, sliding of the abrasive particles can occur even at very high slurry concentrations if a sufficiently high load is used. A later model [5] suggests the transition to occur for a constant ratio between the severity of the contact (S) and the relative hardness between specimen (HS) and ball (HB), where: S¼W/(A.ϑ.H´), W is the normal load, ϑ is the volume fraction of abrasive in the slurry, A is the region over which the separation of the ball and the specimen is less than the diameter of the abrasive particles (D), given by A¼ π (a2 þ 2RD), R is the radius of the ball, a is the hertzian elastic radius, and 1/H´¼1/HB þ1/HS. Another relevant point regarding the transition in particle dynamics is that because sliding of particles normally occurs at lower slurry concentrations in microabrasion tests, the wear rate, or wear severity, associated with grooving wear, is not necessarily any higher than that produced by rolling abrasion, although the literature normally associates grooving two-body wear with higher wear rates than three-body rolling wear [9]. From a more general point of view, the transition in particle dynamics depends on the amount of abrasive particles that entrain the contact between ball and specimen. Rutherford and Hutchings [2] proposed a model to analyze the conditions under which a spherical abrasive particle can be transported into the gap between the sliding sphere and a plane surface. They show that for a particle to be drawn into the contact the coefficients of friction between the particle and
the sphere and between the particle and the specimen must not be equal. They also show that for larger balls and smaller abrasive particles, the difference in friction coefficients to ensure particle entrainment is greatly reduced. Following this rationale, an experimental work investigated the influence of the surface finish of the ball [10] during micro-scale abrasion tests with a fixed-ball rig. Tests were carried out with fresh, as-received 52100 steel balls and also with roughened and slightly corroded balls due to previous use. They tested both plain carbon silver steel and aluminum specimens using a slurry of SiC angular abrasive particles in distilled water. All tests with the new balls showed much greater variability than the tests with the used balls. Evidence was shown that abrasive particle entrainment only occurred after the ball surface presented a large number of irregularities due to ball use during previous tests. They showed that the irregularities in the ball surface helped to engage abrasive particles and drag them into the contact. [10]. The influence of ball topography was also suggested by Shipway [11] during microabrasion tests of glass specimens. When the author used alumina balls, apparently the abrasive did not travel parallel to the sliding direction but instead moved sideways out of the contact leaving the tail end of the wear scar starved of abrasive particles. The author associated this to the lack of mechanical key of the alumina balls. The use of ceramic balls (mainly zirconia and alumina) is of particular relevance when microabrasion tests are combined with corrosion tests to investigate the synergy between corrosion and abrasion of corrosion-resistant soft materials such as stainless steels [12]. Entrainment of abrasive particles could be a particular issue in these systems due to the difficulty of the abrasive particles
132
H.L. Costa et al. / Wear 324-325 (2015) 129–139
10 9
8
Frequency (%)
7 6 5 4 3 2 1
0
0.2 0.2 0.3 0.3 0.4 0.5 0.6 0.7 0.9 1.0 1.3 1.5 1.8 2.2 2.6 3.1 3.8 4.5 5.5 6.6 7.9 9.5 11.4 13.7 16.4 19.7 23.7 28.5 34.2
Diameter (µm)
Fig. 3. Size of the abrasive particles: (a) size distribution histogram by laser granulometry; (b) SEM iamge.
Table 3 Surface topography conditions of the specimens. Nomenclature
Surface finish
Sa (mm)
#80 #4000
Abrasive paper, grit size 80 Abrasive paper, grit size 4000
0.414 0.025
to embed into the hard ceramic balls. Another aggravate to the problem is that in many situations involving abrasion and corrosion, including the bio-fuel and sugar industry [13], the abrasive particles present in the real application are relatively soft, such as silica sand. This article aims to investigate the role of the surface topography of hard balls within a wide range of values of Sa (from 0.06 to 0.54 mm) on the microabrasion of soft stainless steel. On the other hand, besides the surface topography of the ball, the surface topography of the specimen could also influence particle entrainment and therefore abrasive wear. The literature recognizes this possibility by generally using polished specimens in the tests, although this option mainly aims to help identifying the limits of the crater. In some cases
original surface grooves present in the specimen can be quite similar to grooves produced during the microabrasion tests, in the case of parallel motion. Therefore, in addition to the effect of the ball topography, an investigation of the effect of the topography of the specimens on microabrasion is also carried out. The experimental approach used in this article aims to generalize the role of surface topography (ball and specimen) on microabrasion mechanisms and wear coefficients. 2. Experimental procedure To evaluate the effect of the ball surface topography, microabrasion wear tests were carried out using a recently developed rig [14]. It consists of a fixed-ball micro abrasion tester where the specimen is supported by a three-axis load cell, which measures simultaneously three forces and three moments acting on the specimen. This rig is also able to carry out abrasion-corrosion tests (Fig. 1.a). A very flexible protective membrane has been designed to connect the load cell, located outside the test chamber, to the specimen (Fig. 1.b). This membrane poses negligible mechanical interference
H.L. Costa et al. / Wear 324-325 (2015) 129–139
133
Fig. 4. Scheme of the relative orientation between the direction of particle entrainment (thick arrow) and the direction of the grinding marks (thin arrows): (a) parallel; and (b) perpendicular. 140 120
k [ m3.N-1.m 1 ].10-15
100 80 60 Sa=0.34
40
Sa=0.54
20 0
0
5
10
15
20
25
30
35
Test time [ min ]
Fig. 5. Evolution of abrasive wear coefficients with test time for the roughened balls; each point represents an average of three different repetitions.
140
k [10-15.m3/N.m]
120 100 80 60 40 20 0
0.06
0.37
0.54
Sa of the ball (µm) Fig. 6. Average wear coefficients within the steady state regime for the different ball conditions.
to the load measurements. The force is transmitted from its application point to the sample by means of a lever arm system, which has been designed to have a 1:1 ratio between the applied force and the normal force on the contact area. A full description of this new test rig has been submitted for publication [14]. AISI 304 austenitic stainless steel specimens (Table 1) were cut into coupons of 35 25 5 mm, and then sanded with 220 and 600 grit sandpapers, which gave a final Sq of 0.1 mm, and then ultrasonically cleaned in acetone for 5 min before the tests. The counter body was a zirconia ball with 2R¼25.4 mm. To vary the surface topography of the ball in a controlled way, a simple
Fig. 7. Optical microscopy of typical wear craters when zirconia balls with different surface topographies are used as counterbodies: (a) Sa ¼0.06 mm; (b) Sa ¼0.34 mm.
apparatus was developed and adapted to a commercial polisher, which allowed the ball to rotate over an abrasive paper. The ball surface topography was varied by using different abrasive grit sizes
134
H.L. Costa et al. / Wear 324-325 (2015) 129–139
Fig. 8. SEM of the wear craters produced using zirconia balls with different surface topographies as counterbodies: (a) Sa ¼ 0.06 mm; (b) Sa ¼ 0.34 mm; (c) Sa ¼ 0.54 mm.
350
350
300
300
250
250
200
200
150
150
100
0
2
4
6
8
10
12
14
100
16
0
2
4
6
Test time [ min ]
8
10
12
14
16
Test time [ min ] 350
300
250
200
150
100
0
2
4
6
8
10
12
14
16
Test time [ min ] Fig. 9. Evolution of abrasive wear coefficients with test time for different slurry concentrations; each point represents an average of three different tests: (a) 5 wt%.; (b) 10 wt%; (c) 20 wt%.
and grinding times. For comparison, a fresh ball was also investigated. The surface finish of the balls was evaluated using a 3D laser interferometer, model UBM MESSTECHNIK MicroFocus. The measuring rate was 300 points/s, using continuous measurement
mode. The resolution was chosen so that the distance between neighboring sampling points was 1 mm in both measurement directions (x and y). An example of the 3D surface topography of the balls after form removal is shown for a fresh ball (Fig. 2.a) and for ground
H.L. Costa et al. / Wear 324-325 (2015) 129–139
135
300 20 wt% 250
10 wt%
k [10-14.m2/N]
5 wt% 200 150 100 50
0 par
per #80
Par
per #4000
Fig. 10. Average wear coefficients for specimens with different surface roughness and relative orientation between topography and direction of particle entrainment.
balls (Fig. 2.b and c). From those measurements, average values of Sa were calculated for each condition (Table 2). SiO2 particles were used as abrasives and their sizes were measured by laser (Fig. 3), giving a mean particle diameter of 2.25 mm. The tests were performed using a mixture of distilled water with 10 wt% of abrasives. The rotary speed was set at 150 rpm and normal load of 1.42 N. In order to identify the steady state wear regime the tests were interrupted at every 3 min. to measure the crater diameter and to calculate the microabrasion wear coefficient (k) [2]. At least three repetitions were carried out for each test condition. The tests to investigate the effect of the specimen surface topography on microabrasion were carried out using a free-ball commercial rig, whereas the tests to evaluate the effect of the surface topography of the ball used a fixed-ball rig. AISI 52100 steel bearing balls (25.4 mm in diameter) were used as counter bodies in all the tests. Laser interferometry evaluation of the surface topography of the ball indicated Sa ¼0.83 mm. This ball size and the inclination of the apparatus induced a normal load of 0.133 N, monitored by a load cell. The ball rotation was kept constant at 150 rpm. The tests were performed using a mixture of distilled water with SiO2 at three concentrations (5, 10 and 20 wt%). Abrasive particle size was measured by laser as described for the tests with different ball topographies, giving a mean particle diameter of 2.25 mm. Microabrasion wear coefficients (k) were also calculated for interruptions at every 3 min. Tool steel coupons were used as specimens. Their hardness was 8.2470.28 GPa, measured using a Vickers indenter under an applied load of 49 N during 30 s. Their surface topography was varied by grinding using abrasive papers with different grit sizes (Table 3). Grinding of the specimens with abrasive papers resulted in a surface topography composed of parallel grooves. In order to assess the effect of the relative orientation between the grooves in the specimen and the direction of particle entrainment during microabrasion, tests were carried out both for parallel relative orientation (Fig. 4.a) and for perpendicular orientation (Fig. 4.b). After the tests, scanning electron microscopy (SEM) and 3D laser interferometry were used to assess the surfaces within the wear craters.
3. Results 3.1. Effect of the surface topography of the balls The evolution with time of abrasive wear coefficients (k) calculated from tests interrupted at every 3 min is shown in
Fig. 5 for the conditioned balls. In this figure, each point corresponds to the average value of the three repetitions for each tested condition. The wear coefficient (k) tends to a constant value as the test time increases. Steady state wear was considered to be achieved when the variation of k was inferior to 5%, that is, considering the last five points. A similar procedure was used for the fresh balls. The comparison of the average wear coefficients within the steadystate region for the different ball topographies is summarized in Fig. 6. As the roughness of the ball increases, the abrasive wear coefficient increases substantially. The difference between the values of k found for the fresh and the roughest ball was around 510%. Typical wear craters produced in the stainless steel samples due to the use of balls with the different surface topographies are presented in Fig. 7. Detail of the wear mechanisms within the wear craters were detected by SEM (Fig. 8). The variation of the ball topography clearly induced changes in the wear micromechanisms. For the smoothest ball (Fig. 8.a), grooving is clear within the wear craters, indicating sliding of the abrasive particles. The use of the roughest ball led to the occurrence of multiple indentations (Fig. 8.c), indicating rolling of the abrasives. A mixed regime was observed for intermediate surface roughness of the ball (Fig. 8.b). 3.2. Effect of the surface topography of the specimens The evolution of k with time for specimens with different surface topographies and relative orientations between topography and direction of particle entrainment is presented in Fig. 9 for different slurry concentrations, where each point represents the average of three repetitions. Average values of k within the steady state region were calculated using the last four points in Fig. 9, and are presented in Fig. 10. The results summarized in Fig. 10 did not identify any significant effect of the relative orientation between the grinding marks in the sample and the direction of particle entrainment in the test. The increase of slurry concentration increased the wear coefficient. However, for the roughest specimen (#80), no significant difference was detected between 10 wt% and 20 wt%. The effect of the roughness of the specimen was negligible for a slurry concentration of 5 wt%. When the slurry concentration increased to 10 wt%, a slight reduction in k (around 12%) was observed for the rougher specimen when compared with the smoother specimen. For the highest slurry concentration (20 wt%), k of the rougher specimen was around 23% lower than for the smoother specimen.
136
H.L. Costa et al. / Wear 324-325 (2015) 129–139
Fig. 11. SEM of wear craters evidencing the effect of specimen topography on wear mechanisms: (a) 20 wt%, #80; (b) 20 wt%., #4000; (c) 5 wt%, #80; (d) 5 wt% #4000; (e) 5 wt%, #80, perpendicular.
SEM was used to identify the effect of the surface topography of the specimens on the wear micromechanisms within the craters (Fig. 11). A first common feature to all images is that a completely new surface topography was generated on the samples due to wear, with no evident correlation with the original surface topography, either for parallel orientation (Fig. 11.a to d) or for perpendicular orientation (Fig. 11.e). For the highest slurry
concentration of 20 wt%, a slight change in mechanism was observed, from mixed (sliding in the center and rolling at borders of craters, Fig. 11.b), to pure sliding (Fig. 11.a). Another interesting feature is that apparently the spacing between the grooves formed in the specimen seems wider when the original surface topography was rougher. When the slurry concentration decreased, there was no change in mechanism, which is grooving due to sliding of
H.L. Costa et al. / Wear 324-325 (2015) 129–139
137
Fig. 12. 3D laser interferometry of the craters: (a) # 80, 20 wt%, parallel, (b) #80, 5 wt%, perpendicular; (c) # 80, 20 wt%, parallel, after form removal; (d) # 4000, 20 wt%, parallel, after form removal.
the abrasive particles for both specimen topographies (Fig. 11.c to e). However, the difference in the spacing of the grooves remains. The quantification of the surface topography within the wear scars by laser interferometry confirmed the qualitative tendencies provided by SEM. Fig. 12.a shows an example of a crater produced in the roughest specimen (#80) and the largest slurry concentration (20 wt%) with the surface topography oriented in the direction of entrainment of the particles. An example of crater produced with perpendicular orientation is shown in Fig. 12.b for the roughest specimen and the lowest slurry concentration. After form removal, an area of approximately 600 600 mm2 was chosen in the middle of each crater. Comparing Fig. 12.c and d, it is possible to visualize the effect of the specimen surface topography for the most concentrated slurry (20 wt%). The rougher specimen (Fig. 12.c) resulted in craters with apparently fewer grooves than the smoother specimen (Fig. 12.d). The parameters chosen to quantify the surface topography inside the craters were Sq, an amplitude parameter, SSk, a height distribution parameter and Sdq, the average inclination of the irregularities (Table 4). This table shows that when the specimens were rougher, the values of SSk are negative, whereas for the smoothest specimens, SSk was positive. This indicates that the topography of the craters in the rough specimens is of the type plateau-valley, but of the type plateau-peak in the smooth specimen [15]. Also, with the increase of the roughness of the specimens, the average height of the irregularities in the craters increases and the average slope of the irregularities increases. These values support the hypothesis that fewer grooves are produced for the rough specimen, leaving a topography composed of alternate relatively smooth regions and deep grooves.
4. Discussion The results show that the topography of both the ball and specimen can influence microabrasion, but to different extents.
The surface topography of the ball was changed for a hard zirconia ball and great variations were found for the wear coefficients, which were accompanied by transitions in the wear micromechanisms found in the craters. Those results agree with the findings by Allsopp et al. [10], who found that for smoother balls, particle entrainment was more difficult, reducing wear. They suggest a mechanism similar to that schematized in Fig. 12.a and b. For abrasive wear to occur, abrasive particles must become entrained into the contact area between the ball and the specimen. If the ball is very smooth (Fig. 12.a), abrasive particles could embed in the specimen and remain stationary at the edge of the contact zone. However, if the ball is relatively rough (Fig. 12.b), the abrasive particles can engage with the ball surface irregularities. For the tests in this work, the abrasive particles were softer and rounder than those used by Allsopp et al. [10], which probably makes engagement with the ball surface more difficult. Also, the specimen is very soft, which must help the abrasive particles to embed in the sample and remain stationary outside the contact, which agrees with their findings that the influence of ball surface topography was greater for the softer aluminum specimen. The severity of the contact was estimated using the model proposed by Adachi and Hutchings [5], although the tests in the present study were carried out with a much softer abrasive, which must not be undeformable within the contact. This calculation gave a value of 0.059 (W¼ 1.46 N, ϑ ¼ 0.065 vol.%, a¼49.5 mm, E0 ¼228 GPa, R¼12.7 mm, D¼2.24 mm). For HB ¼13.24 GPa (nominal) and HS ¼1.99 GPa (measured), the ratio HS / HB becomes 0.15. According to their model, these conditions are close to the transition between sliding and rolling of abrasive particles. They used a relatively rough ball in their tests and did not account for surface topography. The present results show that it is possible to vary from the mixed regime found for balls with intermediate roughness to rolling if a rougher ball is used or to sliding if a smoother ball is used. No account for change in wear mechanism as a consequence of ball surface topography was given in the work by Allsopp et al. [10]. However, other works have shown transitions in wear
138
H.L. Costa et al. / Wear 324-325 (2015) 129–139
mechanism as a function of the hardness of the ball, which apparently changes embedment of the abrasive particles in the ball and therefore the amount of particle entrainment [7,11,16]. In another work, transition in wear mechanisms was identified as a function of the hardness of the abrasive particles, which also probably influences the amount of particle embedment in the ball [17]. They found that rolling prevailed for harder abrasive particles, whereas sliding occurred for softer abrasive particles. An interesting point of the present work is the quantification of the ball surface topography. A simple methodology to vary the surface topography of the ball in a controlled way was proposed. This should lead in future works to a more precise control of the ball topography in microabrasion tests. The results here presented confirm this aspect to be of paramount importance to obtain reproducible results. As an example, a zirconia ball that had been previously used in many abrasion-corrosion tests was used in microabrasion tests. The surface topography for this condition was found to be Sa ¼0.2 mm. Three repetitions were carried out and the value of k within the steady state regime was 40 (10 15.m2/N). This value was well below to those obtained when balls conditioned in a controlled way are used. When the surface topography of the specimens was varied, the first important point is that the effect on wear mechanisms and k was much less pronounced than when the topography of the balls was varied. This was not surprising, since the severity of abrasion tends to produce completely new topographies on the specimen,
Table 4 3D Surface roughness within craters for different specimen topographies. Parameters
Sq (mm) SSk Sdq
5 wt%
20 wt%
#80
#4000
#80
#4000
0.37 1.07 0.462
0.193 0.382 0.271
0.835 0.211 0.826
0.478 0.305 0.401
as confirmed by the observation of the craters (Fig. 11). However, in some cases, despite this fact, a small effect of the surface topography of the specimen was not negligible. This was the case particularly for the highest slurry concentration. However, the effect was opposite to that of the ball, i.e., the increase of the surface roughness of the specimens decreased wear rates. In the system investigated, the hardness of the ball and specimen were very similar, which should make particle entrainment more difficult, since the friction coefficients between ball and abrasive and between specimen and abrasive will be very similar [2]. This is particularly relevant for rounder abrasive particles, which is the case of the present work. SEM of the craters showed that the amount of grooves is smaller for the rougher specimens (Fig. 11.a and c) than for the smoother specimens (Fig. 11.b and d). It is proposed that although a new surface topography is generated during abrasion, the original surface roughness of the specimens influences the amount of abrasive particles that entrain into the contact. When the specimen is rougher, the particles will embed more easily in the specimen and may remain stationary outside the contact (Fig. 12.d), as opposed to a smoother specimen (Fig. 12.c) When the slurry concentration was small, the number of abrasive particles that entrain into the contact was probably reduced, as normally reported in the literature [4,5]. Therefore, sliding of abrasive particles prevailed independently of the surface roughness of the specimen. The severity of the contact was again estimated using the model in [5], which gave values of 0.0047, 0.0019 and 0.001 for ϑ ¼ 0.032, 0.065 and 0.129 vol.%, respectively (W¼0.133 N, a ¼45 mm, E0 ¼ 210 GPa, R¼12.7 mm, D¼ 2.6 mm), which were much lower than those for the tests with different zirconia balls. For HS / HB 1, multiple indentations produced by rolling of the abrasive particles were expected to occur, but most tests showed grooving by sliding of abrasive particles. This suggests that when much softer abrasive particles are used, combined with a small difference in friction coefficient between ball and abrasive and between specimen and abrasive, particle entrainment becomes difficult. The results for the highest slurry
Fig. 13. Scheme suggesting the effect of surface topography on particle entrainment: (a) smooth hard ball and soft specimen; (b) rough hard ball and soft specimen; (c) ball and smooth sample of similar hardness; (d) ball and rough sample of similar hardness.
H.L. Costa et al. / Wear 324-325 (2015) 129–139
concentration indicates that probably the use of smoother specimens could help particle entrainment Fig. 13. Another important aspect to be considered is the possibility of fragmentation of the abrasive particles, which must be particularly relevant for the relatively soft but brittle (K1c o1 MPa.m1/2 [18]) abrasive used in this work. If particle fragmentation occurs, it is likely that the number of particles within the contact increases, which could also be responsible for the observed phenomena. Laser granulometry was used to measure the size distribution of the abrasive particles. There was a tendency for the smoother specimens inducing more fragmentation of the particles. However, results for the particle size distribution for different tests under similar conditions were not very repeatable. The effect of roughness of the specimens on particle fragmentation needs to be further investigated. 5. Conclusions The effect of the surface topography of the ball and of the specimen on the dynamics of the abrasive particles and therefore on microabrasion coefficients was investigated. For the microabrasion of hard zirconia balls and soft specimens using relatively soft abrasives (SiO2) and a fixed-ball rig, the increase of the roughness of the ball changed the particle dynamics from grooving to rolling and a large increase in k was observed (510%). It is believed that this occurs because a rougher ball helps the abrasive particles to engage with the ball, increasing the number of particles entrained into the contact. For the microabrasion of ball and specimens with similar hardness using SiO2 in a free-ball rig, the effect of the topography of the specimens was much less pronounced. However, the use of rougher specimens under high slurry concentrations changed the mechanisms from mixed to grooving and reduced k (23%). It is believed that rougher specimens make particle entrainment more difficult when they embed into the valleys of the surface topography. Acknowledgments The authors are grateful to CNPq, Capes and Fapemig (Brazil) for financial support.
139
References [1] A. Kassman, S. Jacobson, L. Erickson, P. Hedenqvist, M. Olsson, A new test method for the intrinsic abrasion resistance of thin coatings, Surf. Coat. Technol. 50 (1991) 75–84. [2] K.L. Rutherford, I.M. Hutchings, A micro-abrasive wear test, with particular application to coated systems, Surf. Coat. Technol. 79 (1996) 231–239. [3] M.G. Gee, A. Gant, I. Hutchings, R. Bethke, K. Schiffman, K. Van Acker, S. Poulat, Y. Gachon, J. von Stebut, Progress towards standardisation of ball cratering, Wear 255 (2003) 1–13. [4] R.I. Trezona, D.N. Allsopp, I.M. Hutchings, Transitions between two-body and three-body abrasive wear: influence of test conditions in the microscale abrasive wear test, Wear 225 (1999) 205–214. [5] K. Adachi, I.M. Hutchings, Wear-mode mapping for the micro-scale abrasion test, Wear 255 (2003) 23–29. [6] M.M. Stack, M. Mathew, Micro-abrasion transitions of metallic materials (vol 255, pg 14, 2003), Wear 256 (2004) (854-854). [7] R.C. Cozza, J.D.B. de Mello, D.K. Tanaka, R.M. Souza, Relationship between test severity and wear mode transition in micro-abrasive wear tests, Wear 263 (2007) 111–116. [8] J.A. Williams, A.M. Hyncica, Mechanisms of abrasive wear in lubricated contacts, Wear 152 (1992) 57–74. [9] J.D. Gates, Two-body and three-body abrasion: a critical discussion, Wear 214 (1998) 139–146. [10] D.N. Allsopp, R.I. Trezona, I.M. Hutchings, The effects of ball surface condition in the micro-scale abrasive wear test, Tribol. Lett. 5 (1998) 259–264. [11] P.H. Shipway, The role of test conditions on the microabrasive wear behaviour of soda-lime glass, Wear 233 (1999) 191–199. [12] R.J.K. Wood, D. Sun, M.R. Thakare, A.D. Rozas, J.A. Wharton, Interpretation of electrochemical measurements made during micro-scale abrasion-corrosion, Tribol. Int. 43 (2010) 1218–1227. [13] W.S. Labiapari, C.M. Alcantara, H.L. Costa, J.D.B. de Mello, Stainless steel as an antiwear material for the bio-fuel industry, Wear 302 (2013) 1536–1545. [14] M.B. Santos, W.S. Labiapari, M.A.N. Ardilla, Abrasion-corrosion: new insights from force measurements, Submitted to the Twentyth International Conference on Wear of Materials, 2015. [15] W.P. Dong, P.J. Sullivan, K.J. Stout, Comprehensive study of parameters for characterizing 3-dimensional surface-topography.3. Parameters for characterizing amplitude and some functional-properties, Wear 178 (1994) 29–43. [16] P.H. Shipway, C.J.B. Hodge, Microabrasion of glass - the critical role of ridge formation, Wear 237 (2000) 90–97. [17] W.M. da Silva, R. Binder, J.D.B. de Mello, Abrasive wear of steam-treated sintered iron, Wear 258 (2005) 166–177. [18] A.C. Bozzi, J.D.B. De Mello, Wear mechanism and wear resistance of WC-12%Co thermal sprayed coatings in three body abrasion, Wear 233 (1999) 575–587.
Contents lists available at ScienceDirect
Wear journal homepage: www.elsevier.com/locate/wear
Effect of surface topography on the dynamics of the abrasive particles during micro-abrasion$ H.L. Costa n, M.A.N. Ardila, W.S. Labiapari, W.M. Silva, J.D.B. de Mello Laboratório de Tribologia e Materiais, Universidade Federal de Uberlândia, Uberlandia, MG Brazil
art ic l e i nf o
a b s t r a c t
Article history: Received 12 September 2014 Received in revised form 4 December 2014 Accepted 6 December 2014 Available online 19 December 2014
During abrasive wear, the wear mechanism has been shown to be associated with the movement of the active particles present at the wear interface: rolling, evidenced by indentations on the worn surface, and sliding, which produces scratching and/or ploughing. Particle dynamics can vary with tribological parameters such as different combinations of ball and specimen materials, applied load, slurry concentration, abrasive material, ball condition and equipment configuration (fixed or free-ball). In this article, the effect of surface topography of both the ball and the specimen on the dynamics of the abrasive particles and micro-abrasion wear is investigated for SiO2 abrasive particles. The effect of the ball surface topography was investigated using a fixed-ball rig, zirconia balls (Sa ¼ 0.06, 0.34, and 0.54 mm) and stainless steel specimens (Sa ¼0.10 mm). When the roughness of the ball increased, the wear mechanism changed from sliding to mixed and then to rolling and the micro abrasion coefficient k increased substantially, the difference between the smoothest and the roughest ball being around 510%. The effect of the specimen surface topography was investigated using a free-ball rig, AISI 52100 steel balls (Sa ¼ 0.82 mm) and tool steel specimens (Sq ¼ 0.025 mm and 0.414 mm). The influence of the directionality of the specimen surface finish was also analyzed by conducting tests parallel and perpendicular to the grinding marks using three slurry concentrations. The effect of the topography of the specimens on wear coefficients and mechanisms was much less pronounced than that found for the ball topography. For the highest slurry concentration (20 wt%), k increased for the rougher specimens (around 23%) and a slight change in mechanism occurred from mixed (sliding in the center and rolling at borders of craters), to sliding. This effect was less significant for lower concentrations. The influence of surface directionality on abrasive wear was negligible. & 2014 Elsevier B.V. All rights reserved.
Keywords: Microabrasion Surface topography Wear severity Wear mechanisms
1. Introduction Micro-abrasion tests, also known as ball-cratering tests, have been used for over twenty years. They were originally proposed to characterize the intrinsic mechanical quality of thin coatings when a commercial dimple grinder, conventionally used in thinning of transmission electron microscopy (TEM) samples, was adapted to grind craters into the surfaces of coating/substrate composites [1]. Later, a rotating sphere apparatus [2] was proposed, which is today commercially available and widely used. In the free-ball version, the ball rotates freely driven by friction at the contact with a notched drive shaft. The normal load results from the weight of the ball resting on the test sample, which can be varied by changing the dimensions and material of the ball and/or by altering the angle of the sample holding plate, and depends on $ This article was presented at the 2nd International Conference on Abrasive Processes. n Corresponding author. Tel.: þ 55 34 3239 4036. E-mail address: [email protected] (H.L. Costa).
http://dx.doi.org/10.1016/j.wear.2014.12.011 0043-1648/& 2014 Elsevier B.V. All rights reserved.
the friction coefficient between the ball and the test surface. In the fixed-ball version, the ball is driven directly by clamping the ball in a split drive shaft and the sample is loaded against the ball with the desired normal force by a lever arm arrangement. This version allows a much wider range of normal loads to be used [3]. Microabrasion wear mechanisms and wear coefficients can vary significantly with the operating conditions used in the tests [4–7]. Grooving wear mechanism occurs when a significant proportion of the abrasive particles embed in the surface of the ball and slide acting as fixed indenters, producing a series of fine parallel grooves in the specimen surface. A rolling mechanism is produced when the abrasive particles do not embed, but roll between the two surfaces, producing a heavily deformed, multiply indented surface, with no surface directionality. A mixed mechanism may also occur, producing grooving in the center and three-body rolling at the edges of the wear craters [4,5]. A model proposed by Williams and Hyncica [8] shows that an abrasive particle between two surfaces undergoes a transition from rolling to grooving at a critical value D/h, where D is the particle major axis and h is the separation of the surfaces. When D/h is only
130
H.L. Costa et al. / Wear 324-325 (2015) 129–139
Fig. 1. Test rig details: (a) General view; (b) setup for force measurements.
Table 1 Chemical composition and Vickers hardness (4.9 N, 30 s) of the specimens to evaluate effect of ball topography. Steel
AISI 304
Chemical Composition [wt%]
Hardness [GPa]
C
Mn
Cr
Ni
Nb
0.055
1.15
18.28
8.01
0.01
1.99
slightly larger than 1, the particle indents both surfaces. As there is a similar indentation occurring at the opposite corner of the particle, the two forces acting on the particle, which will not generally be
collinear, form a couple tending to rotate the particle. This produces multiple indentations on the surfaces. However, when D/h increases, the particle initially rotates until eventually the two forces acting on it become collinear. When this happens, there is no impetus to cause further rotation and so the particle will tend to remain at this inclination, producing grooves on the surface. Other authors have adapted this model to associate the transition between the two mechanisms to changes in the severity of the contact. An initial model by Trezona et al. [4] suggests the transition to occur for a certain ratio of normal load to volume fraction of abrasives in the slurry, which varies depending on the type of abrasive. If the contact contains many particles under a low load,
H.L. Costa et al. / Wear 324-325 (2015) 129–139
131
Fig. 2. 3D surface topography of the balls after form removal: (a) fresh, Sa ¼ 0.06 mm; (b) ground ball, Sa ¼ 0.34 mm; ground ball, Sa ¼0.54 mm.
Table 2 Surface conditioning using abrasive papers and final roughness of the zirconia balls. Surface conditioning
Sa (mm)
Fresh (as received) Grit size # 220 for 20 min Grit size # 220 for 20 minþ Grit size # 120 for 10 minþ Grit size # 120 for 10 min
0.067 0.01 0.34 7 0.06 0.54 7 0.04
each particle will indent the surfaces only lightly, and so rolling prevails. Contrarily, grooving wear occurs when a few heavily loaded particles indent the surface more deeply and therefore slide. Consequently, sliding of abrasive particles is generally dominant at high loads and/or low slurry concentrations, whereas rolling tends to prevail at low loads and/or high slurry concentrations. However, sliding of the abrasive particles can occur even at very high slurry concentrations if a sufficiently high load is used. A later model [5] suggests the transition to occur for a constant ratio between the severity of the contact (S) and the relative hardness between specimen (HS) and ball (HB), where: S¼W/(A.ϑ.H´), W is the normal load, ϑ is the volume fraction of abrasive in the slurry, A is the region over which the separation of the ball and the specimen is less than the diameter of the abrasive particles (D), given by A¼ π (a2 þ 2RD), R is the radius of the ball, a is the hertzian elastic radius, and 1/H´¼1/HB þ1/HS. Another relevant point regarding the transition in particle dynamics is that because sliding of particles normally occurs at lower slurry concentrations in microabrasion tests, the wear rate, or wear severity, associated with grooving wear, is not necessarily any higher than that produced by rolling abrasion, although the literature normally associates grooving two-body wear with higher wear rates than three-body rolling wear [9]. From a more general point of view, the transition in particle dynamics depends on the amount of abrasive particles that entrain the contact between ball and specimen. Rutherford and Hutchings [2] proposed a model to analyze the conditions under which a spherical abrasive particle can be transported into the gap between the sliding sphere and a plane surface. They show that for a particle to be drawn into the contact the coefficients of friction between the particle and
the sphere and between the particle and the specimen must not be equal. They also show that for larger balls and smaller abrasive particles, the difference in friction coefficients to ensure particle entrainment is greatly reduced. Following this rationale, an experimental work investigated the influence of the surface finish of the ball [10] during micro-scale abrasion tests with a fixed-ball rig. Tests were carried out with fresh, as-received 52100 steel balls and also with roughened and slightly corroded balls due to previous use. They tested both plain carbon silver steel and aluminum specimens using a slurry of SiC angular abrasive particles in distilled water. All tests with the new balls showed much greater variability than the tests with the used balls. Evidence was shown that abrasive particle entrainment only occurred after the ball surface presented a large number of irregularities due to ball use during previous tests. They showed that the irregularities in the ball surface helped to engage abrasive particles and drag them into the contact. [10]. The influence of ball topography was also suggested by Shipway [11] during microabrasion tests of glass specimens. When the author used alumina balls, apparently the abrasive did not travel parallel to the sliding direction but instead moved sideways out of the contact leaving the tail end of the wear scar starved of abrasive particles. The author associated this to the lack of mechanical key of the alumina balls. The use of ceramic balls (mainly zirconia and alumina) is of particular relevance when microabrasion tests are combined with corrosion tests to investigate the synergy between corrosion and abrasion of corrosion-resistant soft materials such as stainless steels [12]. Entrainment of abrasive particles could be a particular issue in these systems due to the difficulty of the abrasive particles
132
H.L. Costa et al. / Wear 324-325 (2015) 129–139
10 9
8
Frequency (%)
7 6 5 4 3 2 1
0
0.2 0.2 0.3 0.3 0.4 0.5 0.6 0.7 0.9 1.0 1.3 1.5 1.8 2.2 2.6 3.1 3.8 4.5 5.5 6.6 7.9 9.5 11.4 13.7 16.4 19.7 23.7 28.5 34.2
Diameter (µm)
Fig. 3. Size of the abrasive particles: (a) size distribution histogram by laser granulometry; (b) SEM iamge.
Table 3 Surface topography conditions of the specimens. Nomenclature
Surface finish
Sa (mm)
#80 #4000
Abrasive paper, grit size 80 Abrasive paper, grit size 4000
0.414 0.025
to embed into the hard ceramic balls. Another aggravate to the problem is that in many situations involving abrasion and corrosion, including the bio-fuel and sugar industry [13], the abrasive particles present in the real application are relatively soft, such as silica sand. This article aims to investigate the role of the surface topography of hard balls within a wide range of values of Sa (from 0.06 to 0.54 mm) on the microabrasion of soft stainless steel. On the other hand, besides the surface topography of the ball, the surface topography of the specimen could also influence particle entrainment and therefore abrasive wear. The literature recognizes this possibility by generally using polished specimens in the tests, although this option mainly aims to help identifying the limits of the crater. In some cases
original surface grooves present in the specimen can be quite similar to grooves produced during the microabrasion tests, in the case of parallel motion. Therefore, in addition to the effect of the ball topography, an investigation of the effect of the topography of the specimens on microabrasion is also carried out. The experimental approach used in this article aims to generalize the role of surface topography (ball and specimen) on microabrasion mechanisms and wear coefficients. 2. Experimental procedure To evaluate the effect of the ball surface topography, microabrasion wear tests were carried out using a recently developed rig [14]. It consists of a fixed-ball micro abrasion tester where the specimen is supported by a three-axis load cell, which measures simultaneously three forces and three moments acting on the specimen. This rig is also able to carry out abrasion-corrosion tests (Fig. 1.a). A very flexible protective membrane has been designed to connect the load cell, located outside the test chamber, to the specimen (Fig. 1.b). This membrane poses negligible mechanical interference
H.L. Costa et al. / Wear 324-325 (2015) 129–139
133
Fig. 4. Scheme of the relative orientation between the direction of particle entrainment (thick arrow) and the direction of the grinding marks (thin arrows): (a) parallel; and (b) perpendicular. 140 120
k [ m3.N-1.m 1 ].10-15
100 80 60 Sa=0.34
40
Sa=0.54
20 0
0
5
10
15
20
25
30
35
Test time [ min ]
Fig. 5. Evolution of abrasive wear coefficients with test time for the roughened balls; each point represents an average of three different repetitions.
140
k [10-15.m3/N.m]
120 100 80 60 40 20 0
0.06
0.37
0.54
Sa of the ball (µm) Fig. 6. Average wear coefficients within the steady state regime for the different ball conditions.
to the load measurements. The force is transmitted from its application point to the sample by means of a lever arm system, which has been designed to have a 1:1 ratio between the applied force and the normal force on the contact area. A full description of this new test rig has been submitted for publication [14]. AISI 304 austenitic stainless steel specimens (Table 1) were cut into coupons of 35 25 5 mm, and then sanded with 220 and 600 grit sandpapers, which gave a final Sq of 0.1 mm, and then ultrasonically cleaned in acetone for 5 min before the tests. The counter body was a zirconia ball with 2R¼25.4 mm. To vary the surface topography of the ball in a controlled way, a simple
Fig. 7. Optical microscopy of typical wear craters when zirconia balls with different surface topographies are used as counterbodies: (a) Sa ¼0.06 mm; (b) Sa ¼0.34 mm.
apparatus was developed and adapted to a commercial polisher, which allowed the ball to rotate over an abrasive paper. The ball surface topography was varied by using different abrasive grit sizes
134
H.L. Costa et al. / Wear 324-325 (2015) 129–139
Fig. 8. SEM of the wear craters produced using zirconia balls with different surface topographies as counterbodies: (a) Sa ¼ 0.06 mm; (b) Sa ¼ 0.34 mm; (c) Sa ¼ 0.54 mm.
350
350
300
300
250
250
200
200
150
150
100
0
2
4
6
8
10
12
14
100
16
0
2
4
6
Test time [ min ]
8
10
12
14
16
Test time [ min ] 350
300
250
200
150
100
0
2
4
6
8
10
12
14
16
Test time [ min ] Fig. 9. Evolution of abrasive wear coefficients with test time for different slurry concentrations; each point represents an average of three different tests: (a) 5 wt%.; (b) 10 wt%; (c) 20 wt%.
and grinding times. For comparison, a fresh ball was also investigated. The surface finish of the balls was evaluated using a 3D laser interferometer, model UBM MESSTECHNIK MicroFocus. The measuring rate was 300 points/s, using continuous measurement
mode. The resolution was chosen so that the distance between neighboring sampling points was 1 mm in both measurement directions (x and y). An example of the 3D surface topography of the balls after form removal is shown for a fresh ball (Fig. 2.a) and for ground
H.L. Costa et al. / Wear 324-325 (2015) 129–139
135
300 20 wt% 250
10 wt%
k [10-14.m2/N]
5 wt% 200 150 100 50
0 par
per #80
Par
per #4000
Fig. 10. Average wear coefficients for specimens with different surface roughness and relative orientation between topography and direction of particle entrainment.
balls (Fig. 2.b and c). From those measurements, average values of Sa were calculated for each condition (Table 2). SiO2 particles were used as abrasives and their sizes were measured by laser (Fig. 3), giving a mean particle diameter of 2.25 mm. The tests were performed using a mixture of distilled water with 10 wt% of abrasives. The rotary speed was set at 150 rpm and normal load of 1.42 N. In order to identify the steady state wear regime the tests were interrupted at every 3 min. to measure the crater diameter and to calculate the microabrasion wear coefficient (k) [2]. At least three repetitions were carried out for each test condition. The tests to investigate the effect of the specimen surface topography on microabrasion were carried out using a free-ball commercial rig, whereas the tests to evaluate the effect of the surface topography of the ball used a fixed-ball rig. AISI 52100 steel bearing balls (25.4 mm in diameter) were used as counter bodies in all the tests. Laser interferometry evaluation of the surface topography of the ball indicated Sa ¼0.83 mm. This ball size and the inclination of the apparatus induced a normal load of 0.133 N, monitored by a load cell. The ball rotation was kept constant at 150 rpm. The tests were performed using a mixture of distilled water with SiO2 at three concentrations (5, 10 and 20 wt%). Abrasive particle size was measured by laser as described for the tests with different ball topographies, giving a mean particle diameter of 2.25 mm. Microabrasion wear coefficients (k) were also calculated for interruptions at every 3 min. Tool steel coupons were used as specimens. Their hardness was 8.2470.28 GPa, measured using a Vickers indenter under an applied load of 49 N during 30 s. Their surface topography was varied by grinding using abrasive papers with different grit sizes (Table 3). Grinding of the specimens with abrasive papers resulted in a surface topography composed of parallel grooves. In order to assess the effect of the relative orientation between the grooves in the specimen and the direction of particle entrainment during microabrasion, tests were carried out both for parallel relative orientation (Fig. 4.a) and for perpendicular orientation (Fig. 4.b). After the tests, scanning electron microscopy (SEM) and 3D laser interferometry were used to assess the surfaces within the wear craters.
3. Results 3.1. Effect of the surface topography of the balls The evolution with time of abrasive wear coefficients (k) calculated from tests interrupted at every 3 min is shown in
Fig. 5 for the conditioned balls. In this figure, each point corresponds to the average value of the three repetitions for each tested condition. The wear coefficient (k) tends to a constant value as the test time increases. Steady state wear was considered to be achieved when the variation of k was inferior to 5%, that is, considering the last five points. A similar procedure was used for the fresh balls. The comparison of the average wear coefficients within the steadystate region for the different ball topographies is summarized in Fig. 6. As the roughness of the ball increases, the abrasive wear coefficient increases substantially. The difference between the values of k found for the fresh and the roughest ball was around 510%. Typical wear craters produced in the stainless steel samples due to the use of balls with the different surface topographies are presented in Fig. 7. Detail of the wear mechanisms within the wear craters were detected by SEM (Fig. 8). The variation of the ball topography clearly induced changes in the wear micromechanisms. For the smoothest ball (Fig. 8.a), grooving is clear within the wear craters, indicating sliding of the abrasive particles. The use of the roughest ball led to the occurrence of multiple indentations (Fig. 8.c), indicating rolling of the abrasives. A mixed regime was observed for intermediate surface roughness of the ball (Fig. 8.b). 3.2. Effect of the surface topography of the specimens The evolution of k with time for specimens with different surface topographies and relative orientations between topography and direction of particle entrainment is presented in Fig. 9 for different slurry concentrations, where each point represents the average of three repetitions. Average values of k within the steady state region were calculated using the last four points in Fig. 9, and are presented in Fig. 10. The results summarized in Fig. 10 did not identify any significant effect of the relative orientation between the grinding marks in the sample and the direction of particle entrainment in the test. The increase of slurry concentration increased the wear coefficient. However, for the roughest specimen (#80), no significant difference was detected between 10 wt% and 20 wt%. The effect of the roughness of the specimen was negligible for a slurry concentration of 5 wt%. When the slurry concentration increased to 10 wt%, a slight reduction in k (around 12%) was observed for the rougher specimen when compared with the smoother specimen. For the highest slurry concentration (20 wt%), k of the rougher specimen was around 23% lower than for the smoother specimen.
136
H.L. Costa et al. / Wear 324-325 (2015) 129–139
Fig. 11. SEM of wear craters evidencing the effect of specimen topography on wear mechanisms: (a) 20 wt%, #80; (b) 20 wt%., #4000; (c) 5 wt%, #80; (d) 5 wt% #4000; (e) 5 wt%, #80, perpendicular.
SEM was used to identify the effect of the surface topography of the specimens on the wear micromechanisms within the craters (Fig. 11). A first common feature to all images is that a completely new surface topography was generated on the samples due to wear, with no evident correlation with the original surface topography, either for parallel orientation (Fig. 11.a to d) or for perpendicular orientation (Fig. 11.e). For the highest slurry
concentration of 20 wt%, a slight change in mechanism was observed, from mixed (sliding in the center and rolling at borders of craters, Fig. 11.b), to pure sliding (Fig. 11.a). Another interesting feature is that apparently the spacing between the grooves formed in the specimen seems wider when the original surface topography was rougher. When the slurry concentration decreased, there was no change in mechanism, which is grooving due to sliding of
H.L. Costa et al. / Wear 324-325 (2015) 129–139
137
Fig. 12. 3D laser interferometry of the craters: (a) # 80, 20 wt%, parallel, (b) #80, 5 wt%, perpendicular; (c) # 80, 20 wt%, parallel, after form removal; (d) # 4000, 20 wt%, parallel, after form removal.
the abrasive particles for both specimen topographies (Fig. 11.c to e). However, the difference in the spacing of the grooves remains. The quantification of the surface topography within the wear scars by laser interferometry confirmed the qualitative tendencies provided by SEM. Fig. 12.a shows an example of a crater produced in the roughest specimen (#80) and the largest slurry concentration (20 wt%) with the surface topography oriented in the direction of entrainment of the particles. An example of crater produced with perpendicular orientation is shown in Fig. 12.b for the roughest specimen and the lowest slurry concentration. After form removal, an area of approximately 600 600 mm2 was chosen in the middle of each crater. Comparing Fig. 12.c and d, it is possible to visualize the effect of the specimen surface topography for the most concentrated slurry (20 wt%). The rougher specimen (Fig. 12.c) resulted in craters with apparently fewer grooves than the smoother specimen (Fig. 12.d). The parameters chosen to quantify the surface topography inside the craters were Sq, an amplitude parameter, SSk, a height distribution parameter and Sdq, the average inclination of the irregularities (Table 4). This table shows that when the specimens were rougher, the values of SSk are negative, whereas for the smoothest specimens, SSk was positive. This indicates that the topography of the craters in the rough specimens is of the type plateau-valley, but of the type plateau-peak in the smooth specimen [15]. Also, with the increase of the roughness of the specimens, the average height of the irregularities in the craters increases and the average slope of the irregularities increases. These values support the hypothesis that fewer grooves are produced for the rough specimen, leaving a topography composed of alternate relatively smooth regions and deep grooves.
4. Discussion The results show that the topography of both the ball and specimen can influence microabrasion, but to different extents.
The surface topography of the ball was changed for a hard zirconia ball and great variations were found for the wear coefficients, which were accompanied by transitions in the wear micromechanisms found in the craters. Those results agree with the findings by Allsopp et al. [10], who found that for smoother balls, particle entrainment was more difficult, reducing wear. They suggest a mechanism similar to that schematized in Fig. 12.a and b. For abrasive wear to occur, abrasive particles must become entrained into the contact area between the ball and the specimen. If the ball is very smooth (Fig. 12.a), abrasive particles could embed in the specimen and remain stationary at the edge of the contact zone. However, if the ball is relatively rough (Fig. 12.b), the abrasive particles can engage with the ball surface irregularities. For the tests in this work, the abrasive particles were softer and rounder than those used by Allsopp et al. [10], which probably makes engagement with the ball surface more difficult. Also, the specimen is very soft, which must help the abrasive particles to embed in the sample and remain stationary outside the contact, which agrees with their findings that the influence of ball surface topography was greater for the softer aluminum specimen. The severity of the contact was estimated using the model proposed by Adachi and Hutchings [5], although the tests in the present study were carried out with a much softer abrasive, which must not be undeformable within the contact. This calculation gave a value of 0.059 (W¼ 1.46 N, ϑ ¼ 0.065 vol.%, a¼49.5 mm, E0 ¼228 GPa, R¼12.7 mm, D¼2.24 mm). For HB ¼13.24 GPa (nominal) and HS ¼1.99 GPa (measured), the ratio HS / HB becomes 0.15. According to their model, these conditions are close to the transition between sliding and rolling of abrasive particles. They used a relatively rough ball in their tests and did not account for surface topography. The present results show that it is possible to vary from the mixed regime found for balls with intermediate roughness to rolling if a rougher ball is used or to sliding if a smoother ball is used. No account for change in wear mechanism as a consequence of ball surface topography was given in the work by Allsopp et al. [10]. However, other works have shown transitions in wear
138
H.L. Costa et al. / Wear 324-325 (2015) 129–139
mechanism as a function of the hardness of the ball, which apparently changes embedment of the abrasive particles in the ball and therefore the amount of particle entrainment [7,11,16]. In another work, transition in wear mechanisms was identified as a function of the hardness of the abrasive particles, which also probably influences the amount of particle embedment in the ball [17]. They found that rolling prevailed for harder abrasive particles, whereas sliding occurred for softer abrasive particles. An interesting point of the present work is the quantification of the ball surface topography. A simple methodology to vary the surface topography of the ball in a controlled way was proposed. This should lead in future works to a more precise control of the ball topography in microabrasion tests. The results here presented confirm this aspect to be of paramount importance to obtain reproducible results. As an example, a zirconia ball that had been previously used in many abrasion-corrosion tests was used in microabrasion tests. The surface topography for this condition was found to be Sa ¼0.2 mm. Three repetitions were carried out and the value of k within the steady state regime was 40 (10 15.m2/N). This value was well below to those obtained when balls conditioned in a controlled way are used. When the surface topography of the specimens was varied, the first important point is that the effect on wear mechanisms and k was much less pronounced than when the topography of the balls was varied. This was not surprising, since the severity of abrasion tends to produce completely new topographies on the specimen,
Table 4 3D Surface roughness within craters for different specimen topographies. Parameters
Sq (mm) SSk Sdq
5 wt%
20 wt%
#80
#4000
#80
#4000
0.37 1.07 0.462
0.193 0.382 0.271
0.835 0.211 0.826
0.478 0.305 0.401
as confirmed by the observation of the craters (Fig. 11). However, in some cases, despite this fact, a small effect of the surface topography of the specimen was not negligible. This was the case particularly for the highest slurry concentration. However, the effect was opposite to that of the ball, i.e., the increase of the surface roughness of the specimens decreased wear rates. In the system investigated, the hardness of the ball and specimen were very similar, which should make particle entrainment more difficult, since the friction coefficients between ball and abrasive and between specimen and abrasive will be very similar [2]. This is particularly relevant for rounder abrasive particles, which is the case of the present work. SEM of the craters showed that the amount of grooves is smaller for the rougher specimens (Fig. 11.a and c) than for the smoother specimens (Fig. 11.b and d). It is proposed that although a new surface topography is generated during abrasion, the original surface roughness of the specimens influences the amount of abrasive particles that entrain into the contact. When the specimen is rougher, the particles will embed more easily in the specimen and may remain stationary outside the contact (Fig. 12.d), as opposed to a smoother specimen (Fig. 12.c) When the slurry concentration was small, the number of abrasive particles that entrain into the contact was probably reduced, as normally reported in the literature [4,5]. Therefore, sliding of abrasive particles prevailed independently of the surface roughness of the specimen. The severity of the contact was again estimated using the model in [5], which gave values of 0.0047, 0.0019 and 0.001 for ϑ ¼ 0.032, 0.065 and 0.129 vol.%, respectively (W¼0.133 N, a ¼45 mm, E0 ¼ 210 GPa, R¼12.7 mm, D¼ 2.6 mm), which were much lower than those for the tests with different zirconia balls. For HS / HB 1, multiple indentations produced by rolling of the abrasive particles were expected to occur, but most tests showed grooving by sliding of abrasive particles. This suggests that when much softer abrasive particles are used, combined with a small difference in friction coefficient between ball and abrasive and between specimen and abrasive, particle entrainment becomes difficult. The results for the highest slurry
Fig. 13. Scheme suggesting the effect of surface topography on particle entrainment: (a) smooth hard ball and soft specimen; (b) rough hard ball and soft specimen; (c) ball and smooth sample of similar hardness; (d) ball and rough sample of similar hardness.
H.L. Costa et al. / Wear 324-325 (2015) 129–139
concentration indicates that probably the use of smoother specimens could help particle entrainment Fig. 13. Another important aspect to be considered is the possibility of fragmentation of the abrasive particles, which must be particularly relevant for the relatively soft but brittle (K1c o1 MPa.m1/2 [18]) abrasive used in this work. If particle fragmentation occurs, it is likely that the number of particles within the contact increases, which could also be responsible for the observed phenomena. Laser granulometry was used to measure the size distribution of the abrasive particles. There was a tendency for the smoother specimens inducing more fragmentation of the particles. However, results for the particle size distribution for different tests under similar conditions were not very repeatable. The effect of roughness of the specimens on particle fragmentation needs to be further investigated. 5. Conclusions The effect of the surface topography of the ball and of the specimen on the dynamics of the abrasive particles and therefore on microabrasion coefficients was investigated. For the microabrasion of hard zirconia balls and soft specimens using relatively soft abrasives (SiO2) and a fixed-ball rig, the increase of the roughness of the ball changed the particle dynamics from grooving to rolling and a large increase in k was observed (510%). It is believed that this occurs because a rougher ball helps the abrasive particles to engage with the ball, increasing the number of particles entrained into the contact. For the microabrasion of ball and specimens with similar hardness using SiO2 in a free-ball rig, the effect of the topography of the specimens was much less pronounced. However, the use of rougher specimens under high slurry concentrations changed the mechanisms from mixed to grooving and reduced k (23%). It is believed that rougher specimens make particle entrainment more difficult when they embed into the valleys of the surface topography. Acknowledgments The authors are grateful to CNPq, Capes and Fapemig (Brazil) for financial support.
139
References [1] A. Kassman, S. Jacobson, L. Erickson, P. Hedenqvist, M. Olsson, A new test method for the intrinsic abrasion resistance of thin coatings, Surf. Coat. Technol. 50 (1991) 75–84. [2] K.L. Rutherford, I.M. Hutchings, A micro-abrasive wear test, with particular application to coated systems, Surf. Coat. Technol. 79 (1996) 231–239. [3] M.G. Gee, A. Gant, I. Hutchings, R. Bethke, K. Schiffman, K. Van Acker, S. Poulat, Y. Gachon, J. von Stebut, Progress towards standardisation of ball cratering, Wear 255 (2003) 1–13. [4] R.I. Trezona, D.N. Allsopp, I.M. Hutchings, Transitions between two-body and three-body abrasive wear: influence of test conditions in the microscale abrasive wear test, Wear 225 (1999) 205–214. [5] K. Adachi, I.M. Hutchings, Wear-mode mapping for the micro-scale abrasion test, Wear 255 (2003) 23–29. [6] M.M. Stack, M. Mathew, Micro-abrasion transitions of metallic materials (vol 255, pg 14, 2003), Wear 256 (2004) (854-854). [7] R.C. Cozza, J.D.B. de Mello, D.K. Tanaka, R.M. Souza, Relationship between test severity and wear mode transition in micro-abrasive wear tests, Wear 263 (2007) 111–116. [8] J.A. Williams, A.M. Hyncica, Mechanisms of abrasive wear in lubricated contacts, Wear 152 (1992) 57–74. [9] J.D. Gates, Two-body and three-body abrasion: a critical discussion, Wear 214 (1998) 139–146. [10] D.N. Allsopp, R.I. Trezona, I.M. Hutchings, The effects of ball surface condition in the micro-scale abrasive wear test, Tribol. Lett. 5 (1998) 259–264. [11] P.H. Shipway, The role of test conditions on the microabrasive wear behaviour of soda-lime glass, Wear 233 (1999) 191–199. [12] R.J.K. Wood, D. Sun, M.R. Thakare, A.D. Rozas, J.A. Wharton, Interpretation of electrochemical measurements made during micro-scale abrasion-corrosion, Tribol. Int. 43 (2010) 1218–1227. [13] W.S. Labiapari, C.M. Alcantara, H.L. Costa, J.D.B. de Mello, Stainless steel as an antiwear material for the bio-fuel industry, Wear 302 (2013) 1536–1545. [14] M.B. Santos, W.S. Labiapari, M.A.N. Ardilla, Abrasion-corrosion: new insights from force measurements, Submitted to the Twentyth International Conference on Wear of Materials, 2015. [15] W.P. Dong, P.J. Sullivan, K.J. Stout, Comprehensive study of parameters for characterizing 3-dimensional surface-topography.3. Parameters for characterizing amplitude and some functional-properties, Wear 178 (1994) 29–43. [16] P.H. Shipway, C.J.B. Hodge, Microabrasion of glass - the critical role of ridge formation, Wear 237 (2000) 90–97. [17] W.M. da Silva, R. Binder, J.D.B. de Mello, Abrasive wear of steam-treated sintered iron, Wear 258 (2005) 166–177. [18] A.C. Bozzi, J.D.B. De Mello, Wear mechanism and wear resistance of WC-12%Co thermal sprayed coatings in three body abrasion, Wear 233 (1999) 575–587.