1
Experimental Study on Friction and Wear Behaviors of Bearing Material
2
under Gas Lubricated Conditions
3
4
Mohd Fadzli Bin Abdollaha,b*, Mohd Afiq Azfar Mazlanb,
5
Hilmi Amiruddina,b, Noreffendy Tamaldina,b, A.G. Jaharahc
6
7
a
Centre for Advanced Research on Energy, Universiti Teknikal Malaysia Melaka,
Hang Tuah Jaya,76100 Durian Tunggal, Melaka, Malaysia
8
b
9
Hang Tuah Jaya,76100 Durian Tunggal, Melaka, Malaysia
10
11
Faculty of Mechanical Engineering, Universiti Teknikal Malaysia Melaka,
c
Faculty of Engineering and Built Environment, Universiti Kebangsaan Malaysia,
43600 Bangi, Selangor, Malaysia.
12
13
14
15
*
Corresponding author for mohdfadzli@utem.edu.my
16
Phone number: +606 234 6805/6914
17
Fax number: +606 234 6884
18
19
20
21
22
23
24
1
25
Abstract
26
27
An advanced lubrication technology of a bearing material, which is carbon chromium steel,
28
was experimentally investigated under gas lubricated conditions using Taguchi method. The
29
test was performed over a broad range of applied loads (W), sliding velocities (v) and sliding
30
distances (L) using a modified ball-on-disc tribometer. The results found that gas blown to the
31
sliding surfaces in air effectively reduced the coefficient of friction as compared with the air
32
lubrication at higher applied load, sliding speed and sliding distance. In addition, a specific
33
wear rate is constant throughout the tests under gas lubricated conditions. However, the
34
specific wear rate decreases with increasing applied load, sliding speed and sliding distance
35
under air lubrication. By using the optimal design parameters, a confirmation test
36
successfully verify the N2-gas lubrication reduced coefficient of friction and simultaneously
37
improved wear resistance about 24% and 50%, respectively. This is in accordance with a
38
significant reduction of wear scar diameter and smoother worn surface on a ball.
39
40
Keywords: Friction; Wear; Gas lubrication; Bearing; Taguchi Method
41
42
43
44
45
46
47
48
2
49
1.
Introduction
50
51
Ball bearings are small metal balls that are used in a wide variety of machines and
52
other devices, allowing parts of them to spin freely and without friction. However, these ball
53
bearings will sometimes run into problems due to overuse, extreme vibrations or improper
54
upkeep [1]. Besides, the bearing lubricants such as grease must be replaced periodically and
55
if bearing gets too warm, grease melts and runs out of bearing.
56
Nowadays, there are a great variety of advanced lubrication technologies includes thin
57
film coatings [2-5], nanolubricants [6] and gas lubricant [7-10]. However, gas lubrication is
58
the most cost effective and has several advantages, such as high precision, small friction loss,
59
non–polluting, vibration-free, long life and attractive for high-temperature applications [11].
60
Cong et al. [7] found that HFC-134a gas significantly reduces the friction and wear of
61
all the ceramic couples (ionic ceramics Al2O3 and ZrO2, and the covalent ceramics Si3N4 and
62
SiC rubbing against an Al2O3 ball), and that the ionic ceramic pairs show lower friction and
63
wear. Oxygen has been found to lubricate SiC by the formation of silica and the release of
64
graphite-like material [8], while benzene and acetone vapors have been found to form sticky
65
reaction products, which reduce the friction and wear of ZrO2 [9].
66
From the past researches, friction and wear of materials are effectively reduced by
67
different gas lubrications. However, researches on this topic are not much explored. Thus, in
68
this study, the friction and wear behaviors of bearing material, which is carbon-chrome steel,
69
sliding in air with O2- or N2-gas blows, are investigated using a systematic approach, which is
70
Taguchi method. Additionally, the optimal design parameters are obtained by employing
71
analysis of signal-to-noise (SN) ratio. Then, a confirmation test was carried out to verify the
72
improvement of the quality characteristic using optimal levels of the design parameters.
3
73
2.
Experimental procedures
74
2.1
Design of Experiment (DoE)
75
Prior to experimental work, design of experiment (DoE) using Taguchi method was
76
employed. Four design parameters were determined (lubricant, applied load, sliding speed
77
and sliding distance) and three levels were taken for each parameter, as shown in Table 1.
78
Table 1: Design parameters at three different levels.
79
Design parameters
Level
1
2
3
Lubricant
Applied load
(W), N
Sliding speed
(v), rpm
Sliding distance
(L), km
Air
N2-gas
O2-gas
5
10
20
50
1000
1500
1
3
5
80
81
82
In this study, the L9 (34) orthogonal arrays was selected using Minitab statistical
software, as shown in Table 2.
83
Table 2: Taguchi L9 (34) orthogonal arrays.
84
Design parameters
Test
1
2
3
4
5
6
7
8
9
Lubricant
Applied load
(W), N
Sliding speed
(v), rpm
Sliding distance
(L), km
Air
Air
Air
N2-gas
N2-gas
N2-gas
O2-gas
O2-gas
O2-gas
5
10
20
5
10
20
5
10
20
500
1000
1500
500
1000
1500
500
1000
1500
1
3
5
5
1
3
3
5
1
4
85
2.2
Materials
86
The materials used in this study were carbon-chrome steel (SKF bearing) for a ball
87
and EN-31 steel for a disc. The ball has an average surface roughness (Ra) of 0.023µm. The
88
mechanical properties of materials are shown in Table 3.
89
Table 3: Mechanical properties of materials.
90
Properties
Carbon chromium steel1
EN-312
Hardness (H), HRC
61
80
Density (), g/cm3
7.79
7.81
91
1
From laboratory measurements.
92
2
From manufacturer.
93
94
2.3
Tribological testing
95
By selecting L9 Taguchi’s orthogonal arrays as in Table 2, nine sliding tests were
96
carried out using a ball-on-disc tribometer in accordance with ASTM standard G99-95a [12],
97
as illustrated in Fig. 1. Each test was repeated two times in order to reduce experimental
98
errors. Gas was blown to the sliding surfaces in air at a constant pressure of 10psi (70kPa), as
99
shown in Fig. 2. All tests were performed at room temperature. Prior to the sliding test, both
100
ball and disc were cleaned using acetone in an ultrasonic bath. The ball and disc has a
101
diameter of 11mm and 165mm (thickness of 8mm), respectively.
102
A coefficient of friction and frictional force encounter by the ball in sliding were
103
measured by a PC based data logging system. The coefficient of friction is then being
104
determined as follows:
105
F W
106
5
(1)
107
Where F is the frictional force in N and W is the applied load in N.
108
The wear at the ball was recorded by measuring the mass of the ball before and after
109
the wear test. The mass loss in mass unit is converted to the volume loss by dividing with
110
bulk density of material. The specific wear rate is then being determined as follows:
111
Vloss  mloss 
(2)
k  Vloss WL
(3)
112
113
114
115
Where Vloss is the volume loss in mm3, mloss is the mass loss in g,  is the bulk density of
116
material in g/mm3, k is the specific wear rate of material in mm3/Nmm, W is the applied load
117
in N and L is the sliding distance in mm.
118
Ball holder
Lever mechanism
Rotating disc
Load
Motor
119
120
Fig. 1: Schematic diagram of a ball-on-disc tribometer.
121
6
Gas blow
Rotating disc
Ball
122
123
Fig. 2: Photograph of a modified ball-on-disc tribometer with gas blown to the sliding
124
surfaces.
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
7
140
3.
Results and discussion
141
3.1
Effect of gas lubrication on friction and wear behaviors of bearing material
142
Generally, two surfaces of adjacent moving parts can be seperated by a thin film to
143
minimize direct contact between them and provides an interface of low shear strength, hence
144
reduce friction and wear. In this study, the presence of gas lubrication potentially created a
145
thin film and lowered the coefficient of friction at higher applied load, sliding speed and
146
sliding distance as compared with the air lubrication, as shown in Fig. 3. This may be due to
147
the shear strength increases less in proportion to the applied load, sliding speed and sliding
148
distance; this leads to a slight reduction of friction.
From Fig. 4, a specific wear rate is constant throughout the tests under gas lubricated
149
150
conditions. On the other hand, the specific wear rate decreases significantly with increasing
151
applied load, sliding speed and sliding distance under air lubrication. With further increase in
152
load, speed and distance; frictional heating may occur due to the interaction of the asperities
153
of two contact surfaces and in this case the wear process may consist of formation and
154
removal of oxide on the surface, as shown in Fig. 5, resulting in reduction of wear rate.
155
0.55
0.45
0.35
0.25
157
0.55
0.45
0.35
0.25
5
156
0.65
Specific wear rate, WR [mm3/Nmm]
0.65
Specific wear rate, WR [mm3/Nmm]
Coefficient of friction, 
0.65
10
Applied load, W [N]
20
Air
Nitrogen
0.55
Oxygen
0.45
0.35
0.25
50
100
1500
Sliding speed, v [rpm]
1
Fig. 3: Interaction plot for coefficient of friction.
158
8
3
Sliding distance, L [km]
5
7.00
6.00
5.00
4.00
3.00
2.00
1.00
0.00
7E+0
6E+0
5E+0
4E+0
3E+0
2E+0
1E+0
0E+0
5
159
8E+0
Specific wear rate, WR [mm3/Nmm]
8E+0
Specific wear rate, WR [mm3/Nmm]
Specific wear rate, k [mm3/Nmm] x 10-7
8.00
10
Applied load, W [N]
20
Air
7E+0
Nitrogen
6E+0
Oxygen
5E+0
4E+0
3E+0
2E+0
1E+0
0E+0
500
1000
1500
Sliding speed, v [rpm]
1
2
Sliding distance, L [km]
3
Fig. 4: Interaction plot for specific wear rate.
160
161
162
Fig. 5: EDX spectrum at the worn surface of a bearing material.
163
164
165
3.2
Optimal design parameters
166
According to Taguchi method studies, response variation using the signal-to-noise
167
(SN) ratio is important, because it can result in the minimization of quality characteristic
168
variation, due to uncontrollable parameters. The coefficient of friction was considered as
169
being the quality characteristic, using the concept of the “smaller-the-better”. The SN ratio
170
used for this type of response was given by:
171
9
172

y2 
SN  10 log 10   
n

(4)
173
174
Where, n is the number of measurement values in a test, in this case, n = 2, and y is the
175
measured value in the test. SN ratio values are calculated by taking into consideration Eqn. 4.
176
The coefficient of friction values measured from the test, and their corresponding SN
177
ratio values, are shown in Fig. 6. A greater SN ratio value corresponds to a better
178
performance (low coefficient of friction). A small increase as a mean of SN ratio indicating
179
that the presence of N2-gas lubrication effectively reduced coefficient of friction. In this
180
study, the optimal design parameters for a lower coefficient of friction are identified as
181
follows: lubricant = N2, W = 10N, v = 1000rpm, L = 1km.
182
10
10.0
9.5
9.5
9.0
9.0
Mean of SN ratio
Mean of SN ratio for
coefficient of frition, 
10.0
8.5
8.0
7.5
7.5
7.0
6.5
6.5
6.0
Air
Nitrogen
Lubricant
Oxygen
10.0
10.0
9.5
9.5
9.0
9.0
Mean of SN ratio
Mean of SN ratio for
coefficient of friction, µ
8.0
7.0
6.0
8.5
8.0
7.5
5
10
Applied load, W [N]
20
1
3
Sliding distance, L [km]
5
8.5
8.0
7.5
7.0
7.0
6.5
6.5
6.0
6.0
50
183
8.5
1000
Sliding speed, v [rpm]
1500
184
Fig. 6: Mean of SN ratio for coefficient of friction. The optimal design parameter is shown by
185
a red circle.
186
187
3.3
Confirmation test
188
A comparison between the optimized values in air and N2-gas lubrication is shown in
189
Fig. 7. A confirmation test can successfully verify the N2-gas lubrication reduced coefficient
190
of friction and simultaneously improved wear resistance about 24% and 50%, respectively.
191
This is in accordance with a significant reduction of wear scar diameter on a ball, as shown in
192
Fig. 8. Furthermore, Fig. 9 shows that a smoother worn surface (Ra = 0.162µm) was also
11
193
obtained under N2-gas lubricated conditions.
194
(a)
0.7
Applied load, W = 10N
Sliding speed, v = 1000rpm
Sliding distance, L = 1km
Coefficient of friction, µ
0.6
0.5
Air; µaverage = 0.384
0.4
0.3
Nitrogen; µaverage = 0.293
0.2
0.1
0
0.0
(b)
Specific wear rate, k [mm3/Nmm] x 10-8
5.0
4.5
4.0
0.2
0.4
0.6
Sliding distance, L [km]
0.8
1.0
Applied load, W = 10N
Sliding speed, v = 1000rpm
Sliding distance, L = 1km
3.5
3.0
2.5
2.0
1.5
2.6
1.0
1.3
1.3
0.5
0.0
Air
Nitrogen
Lubricant
195
196
Fig. 7: A confirmation test results by comparing (a) the coefficient of friction and (b) specific
197
wear rate of a material under air and N2-gas lubricated conditions using optimal design
198
parameters.
12
Air
Nitrogen
790µm
2630µm
916µm
(a)
1mm
1mm
199
200
Fig. 8: Scanning Electron Microscopy (SEM) of worn surfaces on a ball under air and N2-gas
201
lubricated conditions using optimal design parameters.
202
Air
Nitrogen
203
204
Fig. 9: Surface profile of worn surfaces on a ball under air and N2-gas lubricated conditions
205
using optimal design parameters.
13
206
4.
Conclusion
207
208
The following conclusions may be drawn from the present study:
209
a) As compared with the air lubrication, the presence of gas lubrication lowered the
210
coefficient of friction at higher normal load, sliding speed and sliding distance. This
211
may be due to the shear strength increases less in proportion to the applied load,
212
sliding speed and sliding distance.
213
b) A specific wear rate is constant throughout the tests under gas lubricated conditions.
214
On the other hand, the specific wear rate decreases with increasing applied load,
215
sliding speed and sliding distance under air lubrication. With further increase in load,
216
speed and distance; frictional heating may occur and wear process may consist of
217
formation and removal of oxide on the surface, resulting in reduction of wear rate.
218
c) The optimal design parameters for a lower coefficient of friction are: lubricant = N2,
219
W = 10N, v = 1000rpm, L = 1km.
220
d) By using the optimal design parameters, a confirmation test successfully verify the
221
N2-gas lubrication reduced coefficient of friction and simultaneously improved wear
222
resistance about 24% and 50%, respectively. This is in accordance with a significant
223
reduction of wear scar diameter and smoother worn surface on a ball.
224
225
226
227
228
229
14
230
5.
Acknowledgements
231
The authors gratefully acknowledge contributions from the Green Tribology and
232
Engine Performance (G-TriboE) research group members. This research is supported by the
233
grants from the Ministry of Higher Education Malaysia (MOHE) [grant numbers:
234
ERGS/1/2013/TK01/UTEM/02/04 and RAGS/2013/FKM/TK01/03/B00042].
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
15
254
References
255
256
[1]
Problems
with
ball
bearings,
Available
at
257
http://www.ehow.com/list_7252516_problems-ball-bearings.html, Accessed on April
258
2013.
259
[2]
Abdollah, M.F.B., Yamaguchi, Y., Akao, T., Inayoshi, N., Miyamoto, N., Tokoroyama,
260
T. and Umehara, N., 2012, Future developments of a deformation-wear transition map
261
of DLC coating, Tribology Online, 3, p. 107-111.
262
[3]
Abdollah, M.F.B., Yamaguchi, Y., Akao, T., Inayoshi, N., Miyamoto, N., Tokoroyama,
263
T. and Umehara, N., 2012, Deformation-wear transition map of DLC coating under
264
cyclic impact loading, Wear, 274-275, p. 435-441.
265
[4]
Abdollah, M.F.B., Yamaguchi, Y., Akao, T., Inayoshi, N., Miyamoto, N., Tokoroyama,
266
T. and Umehara, N., 2011, The effect of maximum normal impact load, absorbed
267
energy, and contact impulse, on the impact crater volume/depth of DLC coating,
268
Tribology online, 6, p. 257–264.
269
[5]
Abdollah, M.F.B., Yamaguchi, Y., Akao, T., Inayoshi, N., Miyamoto, N., Tokoroyama,
270
T. and Umehara, N., 2010, Phase transformation studies on the a-C coating under
271
repetitive impact, Surface and Coatings Technology, 205, p. 625-631.
272
[6]
Abdullah, M.I.H.C., Abdollah, M.F.B., Amiruddin, H., Tamaldin, N., Mat Nuri, N.R.,
273
2013, Optimization of tribological performance of HbN/Al2O3 nanoparticles as engine
274
oil additives, Procedia Engineering, DOI: 10.1016/j.proeng.2013.12.185.
275
276
[7]
Cong, P., Li, T., Mori, S., 2002, Friction-wear behavior and tribochemical reactions of
different ceramics in HFC-134a gas, Wear 252, p. 662.
16
277
[8]
Martin, J.M., Le Mohne, T., Gardos, M., 1990, “Friction of alpha silicon carbide under
278
oxygen partial pressure: high resolution analysis of interface films,” Proceedings of the
279
Conference of ITC, Nagoya, Japan, pp. 1407–1412.
280
[9]
Sasaki, S., 1999, Friction and wear behaviour of partially stabilised zirconia in reduced
281
pressure atmospheres containing organic compounds: potential for process lubrication
282
of ceramics, Lubric. Sci., 12, p. 67.
283
[10] Abdollah, M.F.B., Mazlan, M.A.A., Amiruddin, H., Tamaldin, N., 2013, Frictional
284
behavior of bearing material under gas lubricated conditions, Procedia Engineering,
285
DOI: 10.1016/j.proeng.2013.12.240.
286
[11] Introduction to Air Bearings, Available at
287
288
289
http://www.specialtycomponents.com/intro_air_bearings.html, Accessed April 2013.
[12] ASTM G99-95a, 2000, Standard Test Method for Wear Testing with a Pin-on-Disk
Apparatus, ASTM International, 6 pages.
17