analytical method for temporal changes in repeated sliding phenomena

Procedia Engineering 68 ( 2013 ) 213 – 218

Available online at www.sciencedirect.com

1877-7058 © 2013 The Authors. Published by Elsevier Ltd.Selection and peer-review under responsibility of The Malaysian Tribology Society (MYTRIBOS), Department of Mechanical Engineering, Universiti Malaya, 50603 Kuala Lumpur, Malaysiadoi: 10.1016/j.proeng.2013.12.170

ScienceDirect

The Malaysian International Tribology Conference 2013, MITC2013

Analytical Method for Temporal Changes in Repeated Sliding Phenomena

Kanao Fukudaa,b,c,*, Takehiro Moritad aMalaysia-Japan International Institute of Technology, Universiti Teknologi Malaysia, Jalan Semarak, Kuala Lumpur 54100, Malaysia

bInternational Research Centert for Hydrogen Energy, Kyushu University, 744 Motooka, Nishi-ku, Fukuoka 819-0395, Japan cInternational Institute for Carbon-Neutral Energy Research, Kyushu University, 744 Motooka, Nishi-ku, Fukuoka 819-0395, Japan

dFaculty of Mechanical Engineering, Kyushu University, 744 Motooka, Nishi-ku, Fukuoka 819-0395,Japan

Abstract

Understanding the sliding phenomena based on fundamental mechanisms is indispensable to improve the stability and expected life time of sliding parts which work in industrial field. In this paper, adhesive wear in a repeated sliding system was analyzed with a newly devised pin-on-disk apparatus. The analysis on time series data showed some correlational relationship between wear particle sizes and specimen displacement perpendicular to the sliding surface. © 2013 The Authors. Published by Elsevier Ltd. Selection and peer-review under responsibility of The Malaysian Tribology Society (MYTRIBOS), Department of Mechanical Engineering, Universiti Malaya, 50603 Kuala Lumpur, Malaysia. Keywords: Wear, friction, analysis, temporal changes, running-in.

Nomenclature

d Pin displacement differential ( m)

1. Introduction

A tribological phenomenon, especially siding is widely recognized as a time-dependent phenomenon and a running-in process [1-3] and a severe-mild wear transition [4, 5] are typical examples. In most cases, a large temporal change is found in the early stage of sliding phenomena. The change often varies and results in various tribological subsequences in a steady state which follows the early stage. Thus the running-in period of newly introduced equipment holds a major significance for the condition of the equipment in long term usage. Usually, moderate operational conditions i.e., lighter load, lower velocity and temperature, etc. are applied for the running-in period than those for regular operation and some optimized conditions of sliding surface such as the surface covered with protective oxide film are expected to be generated in the running-in period. The operational conditions of running-in processes have been determined empirically because principles that enable rational selection and logical improvement of the conditions have not been established. For establishing generalized rules to determine the running-in conditions, fundamental understanding on the mechanisms of friction force generation and wear especially in the early stage of sliding is indispensable.

© 2013 The Authors. Published by Elsevier Ltd.Selection and peer-review under responsibility of The Malaysian Tribology Society (MYTRIBOS), Department of Mechanical Engineering, Universiti Malaya, 50603 Kuala Lumpur, Malaysia

214 Kanao Fukuda and Takehiro Morita / Procedia Engineering 68 ( 2013 ) 213 – 218

Adhesion, abrasion, corrosion and fatigue are widely known as wear mechanisms and an adhesive wear mechanism is recognized as predominant one for the sliding mechanical parts in industrial machinery. To understand wear mechanism, an adhesive wear model has been originated and developed by many researchers such as Holm [6], Rabinowicz [7] and Archard [8] and a compelling model has been proposed by Sasada [9]. Sasada’s model explains the generation of wear particles as the process accompanied by mutual transfer of adhered stuff between sliding surfaces and growth and discharge of the stuff as a wear particle. The model successfully explains phenomena which appear in adhesive wear processes including the generation of wear particles with variety of sizes. Though the effectiveness as described above, the model is not able to explain the phenomena especially the time dependent changes quantitatively in detail.

Novel analysis method was devised to study repeated sliding phenomena and the method enabled the visual mapping of dynamic information e.g. friction force on a plane which employs two axes at normal to each other of sliding position and the number of repeated sliding [10-13]. One of the current authors further devised the data collection and analysis method using a computer to compile data as shown in Fig.1 and enabled a correlative analysis on different kinds of dynamic information (e.g. information A, B, C in Fig.1) based on the sliding position [14]. The devised method is effective to avoid mistakes of researchers such as wrong correlational analysis between plural kinds of information obtained at different sliding position. Fig. 2 shows a distributional analysis of friction force and pin displacement differential which were obtained using pin-on-disk apparatus. The effectiveness of developed analysis system was verified by comparison on the experimental data obtained for typical adhesive wear and abrasive wear and their temporal changes.

The final objective of this study is improving an adhesive wear model which is available as the basis for optimizing operational conditions of running-in process. Quantitative improvement of the wear model is necessary for the objective. In this paper, the relationship between the size of wear particle and other information was studied and the other information was obtained at the sliding position where the wear particle was generated. Images of wear particles which had been regarded as static data i.e. information D shown in Fig.1was captured as time dependent information and the sizes of wear particles were compared with other dynamic information.

Fig. 1. Devised analysis method enables compilation of plural kinds of dynamic data which obtained at same sliding position [14].


Fig. 2. Distribution analysis on the relationship between friction force and pin displacement differential

2. Experimental procedure

2.1. Sliding test

Fig. 3 shows (a) a schematic illustration of newly developed pin-on-disk apparatus and (b) shapes and dimensions of specimens. A ball of 8mm diameter was utilized as a pin specimen. The ball was grasped by a holder to prevent rotating while performing a sliding test. Both disk and ball specimens were made of austenitic stainless steel (JIS SUS316) because it is known that SUS316 tends to show relatively large scattering in wear test results and clarifying the cause of the scattering is one of the long term purposes of this study. The surfaces of the specimens to be tested were polished using a 3 micrometer diamond slurry, and the typical surface roughness of the specimens was 0.005 m Ra. Then the specimens were twice cleaned ultrasonically using a mixture of acetone and hexane for 10 minutes each and then set in the apparatus. The weight of the specimens was determined before and after the sliding tests using an electrical balance. The difference in the readings was converted into a specific wear rate of wear volume for each unit load and unit sliding distance. Disk rotated around the horizontal axis to allow generated wear particles to fall and then be collected by a wear debris collector shown in Fig. 3(a). The wear debris collector was driven synchronously with disk rotation to enable correlational analysis between collected wear debris, friction force and pin displacement perpendicular to the sliding surface.

Fig. 3. (a) Pin-on-disk apparatus and (b) the shapes and dimensions of specimens.

Measurement interval time, t

Pin displacement differential, d

Pin movement

Friction force


2.2. Experimental conditions

Table 1 sets out the experimental conditions of the sliding test in this study. The test was conducted in air with controlled relative humidity at 49% to avoid the fluctuation in influence of relative humidity throughout the test period.

Table 1. Experimental conditions

(N) 10

Sliding speed (m/s) 0.063

Sliding distance (m) 126

Total disk rotations 2000

Lubrication Non

Atmosphere Air with relative humidity 49% at normal pressure,

R.T.

3. Results and discussion

Specific wear rates of pin and disk specimens were 3.7x10-7 mm2/N and 6.7x10-7 mm2/N, respectively. These values indicate that the sliding in the test was governed by typical adhesive wear mechanism and it was in the sever regime [4]. Fig. 4 shows temporal changes of the values for average and standard deviation of the pin displacement perpendicular to the sliding surface per disk rotation, respectively. Fig. 4 (a), (b) and (c) show the optical micrographs of wear debris collected by wear debris collector at 0-25, 975-1,025 and 1,975 to 2,000 disk rotations, respectively. From Fig. 4, dynamic and static information, i.e. the pin displacement data and the micrographs of wear particles can be compared.

All observed wear particles have metallic shine and supports that the sliding was in the severe regime. Wear particle can be categorized by their size into 3 groups namely large, medium and small. The sizes of large particles are around 200 m, those of medium particles are ranged from 10 to 100 m and those of small particles are less than 10 m. In the beginning of sliding that is Fig. 4(a), wear particles mainly consist of large and medium but small ones. Then the constitution of the wear particle sizes in (b) changed and contained some small particles. After 2,000 disk rotations, (c) shows no large particles but the number of small ones increased considerably.

In Fig. 4, charts of average and standard deviation values of the pin displacement show large and frequent variations in the beginning of the sliding. The frequency of those variations continuously reduced from the beginning until 2,000 rotations and the frequency was supposed to correspond with that of the generation of large wear particles. Though tendency of the charts corresponds with the optical micrograph observation of wear particles, it is difficult to imagine that large particles as bigger than 100 m could be produced. The discrepancy between the charts of the pin displacement and the observation of the wear particles should be caused by statistical treatment of the pin displacement data and the flattened shape of the large wear particles. If the sizes of wear particles distribute as normal distribution, around 0.2% of fluctuation can be bigger than 3 times of S.D. and this can reach around 80

m at largest. The shapes of the large wear particles are not round or cubic as shown in Fig. 4 (a) and (b) and somewhat flattened. The smallest reading of the wear particle size should contribute to the measured value of the pin displacement and the size estimated above namely 80 m looks reasonable.

The correlational analysis on the statistical data of the pin displacement and the sizes of the wear particle could provide some explanation on the relationship between those data. However further precise and detailed correlational analysis is necessary to improve the adhesive wear models. The correlational analysis should employ the raw data of pin displacement and the size of individual wear particle.


Fig. 4. Temporal changes of average and standard deviation values of pin displacement perpendicular to the sliding surface and optical micrographs of wear particles obtained at (a) 0-25, (b) 975-1,025 and (c) 1,975-2,000 rotations.

4. Conclusions

Adhesive wear in the repeated sliding system of austenitic stainless steel JIS SUS316 was analyzed with a newly devised pin-on-disk apparatus. Correlational analysis on temporal changes along the sliding test time was carried out on the size of wear particles and the data of pin displacement perpendicular to the siding surface. Some relationship between the size of wear particles and the statistical values of the pin displacement data was recognized for the generation of large wear particles. Further study on the detailed relationship between data is necessary to improve the adhesive wear models.

Acknowledgements

This work was supported by JSPS KAKENHI Grant Number 23560163.

References

[1] Burwell J., T., Strang C., D., 1952, Metallic Wear, Proceedings of the Royal Society of London A 212, p. 470. [2] Ostvik R., Christensen H., 1968, Changes in Surface Topography with Running-In, Proceedings of the Institution of Mechanical Engineers

183, p. 57. [3] Montgomery, R., S., 1969, RUN-IN AND GLAZE FORMATION ON GRAY CAST IRON SURFACES, Wear 14, p. 99. [4] Archard J. F., Hirst W., 1956, The wear of metals under unlubricated conditions, Proceedings of Royal Society A236, p. 397. [5] Lancaster, J. K., 1963, The formation of surface films at the transition between mild and severe metallic wear, Proceedings of Royal Society

A273, p. 466.

(a) 0-25 rotations (b) 975-1,025 rotations (c) 1,975-2,000 rotations


[6] Holm R., 1946. Electric Contacts, Stockholm. [7] Rabinowicz E., 1951, Journal of Applied Physics 22, p. 1373. [8] Archard J., F., 1953, Contact and rubbing of flat surfaces, Journal of Applied Physics 24, p. 981. [9] Sasada T., Norose S., 1996, The Formation and Growth of Wear Particles through Mutual Material Transfer, Proceedings of JSLE-ASLE

Lubrication Conference, P.82. [10] Fukuda K., Ueki M., 1992, A method and a device to measure friction force, Japanese patent, H4-208949. [11] Belin M., Martin J. M., 1992, Triboscopy, a new approach to surface degradations of thin films, Wear 156, p. 151. [12] Fukuda K., 1998, Friction force distribution and its alternation with repeated sliding, Japanese Journal of Tribologists 43, p. 1143. [13] Fukuda K., 2004, Analysis of specimen displacement in repeated sliding system, Japanese Journal of Tribologists 49, p. 738. [14] Fukuda K., 2008, Combinational analysis of multi-data obtained in a repeated sliding system, Wear 264, p. 499.

analytical method for temporal changes in repeated sliding phenomena

Documents