- Email: [email protected]
Diesel engine crankshaft journal bearings failures: Case study
Accepted Manuscript Diesel Engine Crankshaft Journal Bearings Failures: Case Study Aleksandar Vencl, Aleksandar Rac PII: DOI: Reference:
S1350-6307(14)00161-7 http://dx.doi.org/10.1016/j.engfailanal.2014.05.014 EFA 2326
To appear in:
Engineering Failure Analysis
Received Date: Revised Date: Accepted Date:
1 December 2013 14 May 2014 20 May 2014
Please cite this article as: Vencl, A., Rac, A., Diesel Engine Crankshaft Journal Bearings Failures: Case Study, Engineering Failure Analysis (2014), doi: http://dx.doi.org/10.1016/j.engfailanal.2014.05.014
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Diesel Engine Crankshaft Journal Bearings Failures: Case Study Aleksandar Vencl University of Belgrade – Faculty of Mechanical Engineering, Belgrade, Serbia, [email protected] Aleksandar Rac University of Belgrade – Faculty of Mechanical Engineering, Belgrade, Serbia, [email protected] Abstract: Wear as a tribological process has major influence on the reliability and life of engine crankshaft bearings. The importance of field examinations of bearing failures due to wear is very well known. They point to the possible causes of wear and to the necessary treatment for its reduction or elimination. The paper presents the results obtained by examining 616 crankshaft bearings, damaged by different mechanisms. The bearings were installed in high-speed diesel engines, and were gathered for two years, during general repairs of the engines (overhaul), i.e. after 3000 to 5000 working hours. After the examination of the bearings, the fault tree analysis (FTA) was performed to determine the root causes for engine bearing failures. Each type of damage that was identified was illustrated with an appropriate hiresolution photograph. The investigations show that the basic and most conspicuous types of damage which cause bearing failures are abrasive, adhesive and surface fatigue wear. The paper also considers the effects of the place of installation and type of bearing material in respect to each type of wear. Keywords: diesel engine bearings, bearing failures, wear type statistic, fault tree analysis (FTA). 1. Introduction Reliability of diesel engines and other machinery is the function of faultlessness of its moving parts, from which bearings are certainly among the most important elements. An analysis of 410 recorded defects, occurred in industrial diesel engines during one year, showed that bearing class of defect takes 7 % of total occurrence [1]. Another study of diesel engine failures, which was based on experience of over 800 cases of damages during a period of four years, shows that 12 % of all damage cases are located in engine bearings [2]. The third study shows the distribution of internal combustion engine failures and their cost by parts failed first. Bearing failures head the list in both the number of engine failures (with 24.4 % of all incidents) and cost [3]. The analysis was based on the population of 180 engines, during the period of four years. According to the ISO standard, failure is defined as “termination of the ability of an item to perform a required function” [4]. Any damage or failure of the engine bearings results in a partial or complete functional failure of the engine and can cause significant economic losses and even legal liability. Although, as a rule, the price of the plain bearing is relatively small, a bearing damage that reduces the function of the system or results in its failure can cause a lot of overhead costs. Plain bearing failure involves the following categories: catastrophic failures which result in an immediate inability of a system to achieve its function; performance failures associated with a reducing performance of the equipment; reliability failures associated with a reducing reliability of the equipment [5]. In practice, damage of a bearing may often be the result of several mechanisms operating simultaneously. It is the complex combination of many influencing parameters which often causes difficulty in establishing the primary cause of damage. There are many publications dealing with plain bearing damages and failures. They are focused on damages and failures classifications and appearances [6-8], possible causes [9-11] and appropriate corrective actions [12-14], as well as on the effect of lubricants [15] and diagnostic procedures and examination tests [16]. The relevant literature and practical experience show that the plain bearing damages are mainly caused by wear, either as a direct cause or as a result of various irregularities in design, manufacture, assembly, operation and maintenance of the engine and bearings. Wear itself is a very complex process initiated by the action of different mechanisms, and can be manifested by different wear types which are often related: adhesive wear, abrasive wear, surface fatigue wear, erosive wear, cavitation wear, fretting wear, oxidative wear and corrosive wear [17].
There are very few literature data on the frequency of certain wear types occurrence in diesel engine bearings while the engine is under working conditions (field examinations). These data show which mechanisms have dominant influence on the wear of bearings, and determine what preventive measures should be taken to reduce wear. Correct diagnosis is the first step towards improved reliability. 2. Field examination details 2.1 Background The greatest application of hydrodynamic plain journal bearings is in internal combustion engines. This was the reason why we chose diesel engines as machines to collect from the adequate sample of damaged bearings. Generally, a representative sample could be gathered in two ways: during general repairs of the engines (overhaul) or during engine bench (stand) testing. In this study the first way was used, i.e. damaged bearings were gathered during general overhaul of the engines, in the period of two years. Field examinations of the bearings types of wear and its frequency of occurrence were focused on highspeed diesel engines that were in service from 3000 to 5000 working hours (overhaul engine), which corresponds to the recommended service life of investigated bearings. In other words, the general overhaul of the engines was not performed because of the catastrophic failures of the bearings, but because of the general deterioration and deviations from the minimum required performance (performance and reliability failures). That is why only in a few cases plastic deformation and fracture are registered on the examined bearings. The main difficulty of establishing separate type of wear is that they are often related, or that wear initiated by one cause can result in another effect. However, the dominant type of wear can be approximately determined. The classification of bearing damage was done according to the present wear type, i.e. it is based primarily upon the features visible on the running surfaces. The terms are correlated with the terms used in ISO 7146 standard. 2.2 Bearing types The damaged bearings that were examined were multilayer thin wall half-bearings, with and without the flange. They were used as crankshaft bearings and were installed in high-speed heavy duty diesel engines of vehicles and agricultural machinery (machines where diesel engines have the widest application). Total number of bearings in a representative sample was 616, and there were two types of bearings: main crankshaft bearings (342 pieces, i.e. 55.5 %) and connecting rod bearings (274 pieces, i.e. 44.5 %). The positions on the crankshaft of the investigated bearings are shown in Fig. 1. Figure 1. Positions on the crankshaft of the investigated bearings 2.3 Bearing materials The materials used as linings for examined crankshaft bearings were the two widely used alloys, according to the ISO standard [18] for multilayer materials for thin-walled plain bearings: cast lead-bronze alloy (76 wt.% Cu, 22 wt.% Pb, 2 wt.% Sn 2; 55-80 HB), referred as CuPb22Sn2, and aluminium alloy (79 wt.% Al, 20 wt.% Sn, 1 wt.% Cu; 30-40 HB), referred as AlSn20Cu1. Representation of the materials is almost identical, i.e. number of lead-bronze alloy bearings was 306 (49.7 %), and aluminium alloy bearings 310 (50.3 %). The percentage share of these two materials, used as bearing materials installed in diesel engines, on market is about 40-50 % for lead-bronze alloy and 50-60 % for aluminium alloy, which correspond to the percentage share used in this investigation. The distribution of the bearing materials investigated in this study, on the basis of the place of installation, also corresponds to the situation on market. For the main crankshaft bearings 210 pieces (61.4 %) are made of CuPb22Sn2, and 132 pieces (38.6 %) are made of AlSn20Cu1. On the other hand, for the connecting rod bearings the distribution is reverse and 96 pieces (35 %) are made of CuPb22Sn2, and 178 pieces (65 %) are made of AlSn20Cu1 (Fig. 2). The fabrication technology was the same for all bearings, i.e. standard casting procedure.
Figure 2. Distribution of the investigated bearings according to the place of installation and material 3. Results and discussion The main causes that lead to damage and failure of plain bearings are numerous, and very often with several causes contributing simultaneously. In addition, the bearing housing and the counter-face against which bearing operates also contribute. All these causes can be divided into several groups that include aspects of design, material selection, imperfections of material, manufacture and after-treatment, storage, transportation, assembly, inspection, maintenance, unforeseen operating conditions (like overloading) and direct mechanical and/or chemical damage during operation. In some instances, damage may be caused by a design compromise made in the interests of economy, as well [5,10,15]. Failures and damages caused by these causes commonly manifest as wear, fracture and plastic deformation of material. The results of this study (Fig. 3) show that the types of damage that occurred most frequently were abrasive (59 %), adhesive (19 %) and surface fatigue wear (11 %), while other types of damage were much less common, i.e. different types of wear were the most dominant types of damage (of all examined bearings, different types of wear were noticed 866 times, fracture 5 times and plastic deformation 8 times). These results are consistent, with the general observation that the most common type of wear is abrasive wear. According to Eyre [19] abrasive wear occurs in industry with 50 %, adhesive wear with 15 %, erosion and fretting wear with 8 %, and corrosive (chemical) wear with 5 % [19]. A comparison of the economic impact of different wear types also shows that abrasive and adhesive wear have the most relative importance (more than 80 % of all wear types) [20], and thus are the most common type of wear types. The low frequency of the corrosive wear occurrence in our study can be explained by good corrosive wear resistance of the materials that were used (lead-bronze and aluminium alloy), and by adequate and high-quality lubricants that are used for bearing lubrication. The lead-bronze and aluminium alloy are generally known as the corrosion resistant materials. For most of the observed bearings quality of the bond between lining and bearing back is rated as good to excellent, i.e. detachment of the lining material was not observed. Figure 3. Percentage occurrence of different types of bearing damages 3.1 Appearances and manifestations of the different type of bearing damages Abrasive wear: The majority of damages due to abrasive wear were recorded in middle (axial direction) of the bearing lining. These damages are manifested by scratches (from very fine to clearly visible scratches), Fig. 4a. This primarily relates to bearings made of AlSn20Cu1 alloy, where presence of craters left by displaced particles (characteristic of embedding) are also noticed (Fig. 4b). This alloy was softer than CuPb22Sn2 alloy and posses better embedding capacity [17]. At bearings made of CuPb22Sn2 alloy scratches are present all over on the surface. Severe scratching and scoring was not noticed on any of the bearings. Figure 4. Bearing surfaces damaged by abrasive wear: (a) two AlSn20Cu1 and one CuPb22Sn2 alloy bearings and (b) detail of the bearing in the middle on Fig. 4a (AlSn20Cu1 alloy) Adhesive wear: Adhesive wear was manifested either with wiping or seizure (Fig. 5). Wiping was noticed mainly in the main loaded area of bearings (shiny areas), and was accompanied by change of colour (dark blue to black) due to the local heating (Fig. 5a). Seizure was also accompanied by change of colour, as well as with surface melting and flow of the bearing material (Fig. 5b), and in some cases presence of severe seizure is noticed (Fig. 5c). Surface melting, flow of the bearing material and seizure were mainly present at the AlSn20Cu1 alloy bearings (this material was softer than the other investigated bearing alloy). Figure 5. Bearing surfaces damaged by adhesive wear: (a) wiping accompanied by change of colour of the bearing material (from left to right: CuPb22Sn2 alloy and two AlSn20Cu1 alloy bearings), (b) seizure of the AlSn20Cu1 alloy bearings and (d) detail of the bearing on the left in Fig. 5b (severe seizure)
Surface fatigue wear: Surface fatigue wear was present only in its initial stage, and was manifested with fatigue cracks which extend from the sliding surface in the loaded zone propagating as a network (Fig. 6). There were no characteristics places were cracks occur, and there were no bond faults (detachment of the lining material). Later stage of the surface fatigue (in which cracks change direction from normal to parallel to the surface above the bonding area, leading to the lining material relief) was not noticed. The used lining materials are resistant to the surface fatigue, and it is well known that with the general introduction of stronger bearing linings and improved calculation procedures, fatigue due to design overload is uncommon [7]. Figure 6. Bearing surfaces damaged by surface fatigue (initial period) Cavitation wear: Cavitation wear is seen predominantly in diesel engine bearings [21]. Two types of cavitation wear were noticed: flow cavitation (type 1) and discharge cavitation (type 4) [10]. Flow cavitation wear is noticed around circumferential oil groves (Fig. 7a), and discharge cavitation wear is noticed around oil hole (Fig. 7b). The damage of bearing surfaces was always on the same locations and is characterized as a form of micro fatigue cracking, initiated by the collapse of vapour cavities. These types of damages are typical for bearings operating at high loads, under vibrations and at high speeds. Figure 7. Bearing surfaces damaged by cavitation wear: (a) flow cavitation wear around oil grove and (b) detail of Fig. 7a showing discharge type of cavitation Corrosive wear: Corrosive wear was noticed only on CuPb22Sn2 alloy bearings, and was very mild (Fig. 8). There was no direct contact with water, i.e. formation of the oxide films was not noticed. Corrosion of the lead-bronze lining was due to the formation of organic acids by oxidation of lubricating oil. Figure 8. Bearing surfaces damaged by corrosive wear Erosive wear: Erosive wear was noticed around oil holes, and was manifested by removal of the material due to the fatigue mechanism. Most probably the inlet speed of lubricating oil was high, causing fluid erosion of the bearings (Fig. 9). In some cases even foreign particles in the oil can also lead to fluid erosion [9]. Figure 9. Bearing surfaces damaged by erosive wear (fluid erosion) Plastic deformation: Plastic deformation is manifested all across the bearing with and on considerable length (Fig. 10). Change of colour (dark blue to black) due to the local heating is noticed. Some bearings suffer significant melting at the bearing edge and within the groove. Melting of the lining material due to loss of lubricant resulted as seizure of the bearings. Figure 10. Bearing surfaces damaged by plastic deformation due to loss of lubricant Fracture: The appearance of the bearings damaged by fracture of the lining material consists of completely detached material in larger areas with clearly defined borders (rounder with ellipse in Fig. 11). Figure 11. Bearing surfaces damaged by fracture due to improper connection of lining and bearing back (clearly defined borders of detached material can be noticed) 3.2 Fault tree analysis Possible causes of diesel engine crankshaft bearing failure, which was detected during the investigation, are used (together with the literature data and experience) for the fault tree analysis (FTA). The different types of bearing damages were correlated with the ISO standard 7146 [9,10], which can be used for the
further explanation/description of the constructed fault tree root causes. The constructed fault tree diagram for diesel engine crankshaft bearing failure can be used by engineers in practice. A fault tree diagram is a logic diagram which traces the possible events leading to a major event, almost like a family tree. The desired or, in most cases, undesired main event is placed at the top of the tree and called the “top event” [22]. In our case, the top event will be entitled “crankshaft bearing failure”. In general terms, we can categorize two basic types of failure: (1) tribological and (2) mechanical, as illustrated in Fig. 12. Figure 12. Types of crankshaft bearing failure shown as Intermediate events on fault tree Tribological failures: According to our study, the tribological failures were the dominant types of crankshaft bearing failure. Tribological failures include seven broad wear types: (1) abrasive, (2) adhesive, (3) surface fatigue, (4) cavitation, (5) corrosive, (6) erosive, and (7) fretting. The discussion of different wear type failures must be more general, since they are intermediate events on fault tree of tribological crankshaft bearing failure (Fig. 13). Figure 13. Fault tree of tribological bearing failure Abrasive wear, according to ISO 7146, is classified as contamination with particles. In crankshaft bearings abrasive wear is caused by the contamination of oil either with foreign particles or with wear products. The damage typically manifests with scratches (Fig. 4a) made by contaminants, and with craters left by displaced embedded contaminants (Fig. 4b). Adhesive wear, according to ISO 7146, is classified in two categories, as wear by friction and as insufficient lubrication – starvation. In crankshaft bearings adhesive wear occurs when the oil film thickness is insufficient to prevent direct contact between the bearing and the shaft (mixed or boundary lubrication). Improper oil film thickness could be due to the: (1) insufficient clearance (caused by poor design or mounting); (2) oil film breakdown due to excessive load or vibrations (poor balancing); (3) insufficient oil supply (starvation) caused by lubrication system malfunction; and (4) improper oil characteristics (most usually low viscosity index). The adhesive wear typically manifests with wiping (light adhesive wear) accompanied by change of colour of the bearing material (Fig. 5a), or with seizure (heavy adhesive wear). Seizure is often accompanied by change of colour, and with surface melting and flow of the bearing material (Fig. 5b). Surface fatigue wear, according to ISO 7146, is classified as dynamic overload. In crankshaft bearings surface fatigue wear is caused, first of all, by the presence of dynamic load and overload (load exceeds fatigue strength). The damage typically manifests as initial pitting (Fig. 6), or as progressive pitting. Cavitation wear, according to ISO 7146, is classified as cavitation erosion. In crankshaft bearings cavitation wear occurs when the static pressure in oil is decreased below its vapour pressure, at the working temperature, so the evaporation occurs and vapour bubbles are generated in the oil. With the increase of pressure, these bubbles collapse and cause cavitation wear. Change of oil pressure could be due to: (1) incorrect oil flow due to the improper design and exploitation; (2) oil pressure fluctuation in bearing caused by crankshaft vibrations; and (3) interrupted oil flow through holes and grooves. The cavitation wear has been classified into four types: (1) flow (Fig. 7a); (2) impact; (3) suction; (4) discharge (Fig. 7b); and (5) miscellaneous. Corrosive wear, according to ISO 7146, is classified as contamination with chemicals. In crankshaft bearings corrosive wear is mainly caused by the thermal degradation of oil (low-quality oil), i.e. the corrosive nature of the oil is developed during long periods of operation as a result of contamination by combustion residues. Contamination could also be by water, antifreeze, etc. The damage typically manifests with change of colour (Fig. 8). Erosive wear, according to ISO 7146, is classified as contamination with particles. In crankshaft bearings erosive wear is caused by the fluid erosion (Fig. 9), and by the particles erosion. In both cases erosion is more intensive when oil-flow speed is higher. Fretting wear in crankshaft bearings is caused either by radial crankshaft vibration or by low frequency bearing vibration.
Mechanical failures: Two main mechanical failures are possible in crankshaft bearings: (1) plastic deformation, and (2) fracture (Fig. 14). Plastic deformation, according to ISO 7146, is classified as insufficient lubrication – starvation. Plastic deformation is caused by the loss of lubricant, when generated friction heat melts the bearing lining (Fig. 10). Fracture, according to ISO 7146, is classified as bond failure. Fracture of the lining material manifests as completely detached material in larger areas (Fig. 11), caused by static overload or Impact loads and/or improper manufacture (poor bonding). Figure 14. Fault tree of mechanical bearing failure 3.3 Influence of the place of installation and type of bearing material Number of occurrences of damage by wear is analyzed further in relation to the place of installation and type of bearing material (Fig. 15). The percentage share of main crankshaft and connecting rod bearings in total number of bearings was similar, but not equal (see Fig. 2). In addition, the number of failures was higher than the number of bearings, because in some cases more than one failure was identified at one bearing. There were 616 bearings (342 main crankshaft bearings and 274 connecting rod bearings), and 866 failures were identified on these 616 bearings (480 failures on main crankshaft bearings and 386 on connecting rod bearings). In other words, more failures were on main crankshaft bearings (480) than on connecting rod bearings (386), but the number of main crankshaft bearings was also higher (342 compared to 274). This is the reason why correction factors (Cor1 and Cor2) were introduced. Similar situation was with percentage share in total number of bearings of two bearing materials: CuPb22Sn2 and AlSn20Cu1 alloy (see Fig. 2). There were 616 bearings (310 AlSn20Cu1 bearings and 306 CuPb22Sn2 bearings), and 866 failures were identified on these 616 bearings (500 AlSn20Cu1 bearings and 366 CuPb22Sn2 bearings). In order to compare “real” percentage shear, in both cases a correction of the results was done: MainF ConnF MainF × Conn + ConnF ×Main ; × Cor1 ×100, % ; PSConn = × Cor1 ×100, % ; Cor1 = Main Conn Total CuPbF AlSnF CuPbF × AlSn + AlSnF × CuPb , ; ; PSCuPb = × Cor2 ×100, % PSAlSn = × Cor2 ×100, % Cor2 = CuPb AlSn Total
PSMain =
where: PSMain is percentage share of main crankshaft bearings; PSConn is percentage share of connecting rod bearings; PSCuPb is percentage share of CuPb22Sn2 bearings; PSAlSn is percentage share of AlSn20Cu1 bearings; Cor1 and Cor2 are the correction factor; Total is total number of bearings (616); Main is number of main crankshaft bearings (342); Conn is number of connecting rod bearings (274); MainF is number of failures on main crankshaft bearings (480); ConnF is number of failures on connecting rod bearings (386); CuPb is number of CuPb22Sn2 bearings (306); AlSn is number of AlSn20Cu1 bearings (310); CuPbF is number of failures on CuPb22Sn2 bearings (366); and AlSnF is number of failures on AlSn20Cu1 bearings (500). Figure 15. The effect of the place of installation and type of bearing material in respect to all wear types Investigation has shown that the place of installation has no influence on the occurrence of the wear, i.e. percentage shares of wear occurrence on main crankshaft bearings and connecting rod bearings were practically the same (49.9 and 50.1 %, respectively). On the other hand, wear was more pronounced on bearings made of AlSn20Cu1 alloy than on CuPb22Sn2 bearings (57.4 and 42.6 %, respectively). The effect of the place of installation and type of bearing material in respect to each type of wear is shown in Fig. 16. In order to compare “real” percentage shear of each type of wear, a correction of the results similar to the one applied for Fig. 15 was done. Figure 16. The effect of (a) place of installation and (b) type of bearing material in respect to each type of wear (AB – abrasive wear; AD – adhesive wear; SF – surface fatigue wear; CA – cavitation wear; ER – erosive wear; CO – corrosive wear)
The occurrence of the three most prominent types of wear (abrasive, adhesive and surface fatigue wear) was reported in approximately the same proportions in main crankshaft bearings and connecting rod bearings (Fig. 16a), so we could say that the place of installation has no significant influence on wear occurrence. It is interesting to note that the cavitation wear occurred mainly at the main crankshaft bearings. This could be due to the fact that connecting rod bearings were mainly ungrooved [21] and that only flow cavitation and discharge cavitation were noticed, but further analysis is necessary for a more precise explanation. The influence of bearing material on the occurrence of different wear types was not insignificant (Fig. 16b). Only the abrasive wear occurrence was approximately the same in AlSn20Cu1 and CuPb22Sn2 bearings. For other two most prominent types of wear (adhesive and surface fatigue wear) aluminium alloy showed lover resistance. Higher adhesive wear of Al alloy can be explained with its compatibility properties, i.e. Al is more compatible with Fe (the crankshaft was made of different types of steels) than Cu (the adhesion between Al and Fe is higher) [23]. Aluminium alloy also shows lower resistance to cavitation wear. Lower resistance to surface fatigue and cavitation wear of the AlSn20Cu1 alloy is related to the hardness of this alloy, which was lower than the hardness of the CuPb22Sn2 alloy. On the other hand, erosive wear is much pronounced in CuPb22Sn2 bearings, but this should be viewed through the fact that the number of the failures due to the erosive wear was small (only 2 cases for AlSn20Cu1 alloy, and 4 cases for CuPb22Sn2 alloy). Corrosive wear is only present in CuPb22Sn2 bearings, which can be explained by good resistance to corrosion of aluminium alloy. Aluminium-based linings are completely resistant to engine oils, and to their high-temperature degradation products [7]. 4. Conclusions The results obtained by field examination of failures established the main types of wear that occur in diesel engine crankshaft bearings. Establishment of the main types of wear facilitates the determination of the measures for their reduction or elimination and thus extending the service life and reliability of the crankshaft bearings. Further help in determining the root causes for engine bearing failures is provided by the fault tree analysis (FTA), which was performed. Based on the obtained results, it can be concluded that the most prominent (and thus the most important) types of wear in diesel engine crankshaft bearings are abrasive wear (app. 60 %), adhesive wear (app. 19 %) and surface fatigue wear (app. 11 %). Other types of wear (cavitation, erosive and corrosive wear) are less pronounced, although for certain conditions they can cause significant damage of bearing. The analysis of the effect of bearing place of installation shows that there is no significant difference in the occurrence of the three most prominent types of wear (abrasive, adhesive and surface fatigue wear) between the main crankshaft and connecting rod bearings. The type of bearing material, on the other hand, has an effect on the occurrence of certain types of wear under the same or similar conditions. Adhesive wear, surface fatigue wear and cavitation wear occur more often at aluminium alloy bearings, while corrosive wear was noticed only at lead-bronze alloy. Field examinations and results of the failures analysis are the base for detail engine test stands laboratory studies of individual phenomena, in order to identify particular influencing parameters. Acknowledgments This work has been performed as a part of activities within the projects TR 34028 and TR 35021. These projects are supported by the Republic of Serbia, Ministry of Education, Science and Technological Development, whose financial help is gratefully acknowledged. References [1] Collacott, R. A.: Mechanical Fault Diagnosis and Condition Monitoring. London: Chapman and Hall 1977 [2] Allianz, Handbook of Loss Prevention. Berlin: Springer 1978 [3] Bloch, H. P. and Gettner, F. K.: Practical Machinery Management for Process Plants, Volume 2: Machinery Failure Analysis and Troubleshooting. Third Edition, Houston: Gulf Publishing Company 1999
[4] ISO 13372:2004 Condition Monitoring and Diagnostics of Machines – Vocabulary [5] Rac, A. and Vencl, A.: Sliding Bearing Metallic Materials – Mechanical and Tribological Properties. Belgrade: Faculty of Mechanical Engineering 2004 (in Serbian) [6] Paine, B. and Cambell, B.: Examples of Damage Which Can Occur in Automobile Engine Bearings. TA 100/2. London: Glacier Metal Co. 1969 [7] Evans, C. and Warriner, J. F.: Bearings and Bearing Metals, in: Challen, B. and Baranescu, R. (Eds.): Diesel Engine Reference Book. Second Edition. Oxford: Butterworth-Heinemann 1999 [8] Affonso, L. O. A.: Machinery Failure Analysis Handbook: Sustain Your Operations and Maximize Uptime. Houston: Gulf Publishing Company 2006 [9] ISO 7146-1:2008 Plain Bearings – Appearance and Characterization of Damage to Metallic Hydrodynamic Bearings – Part 1: General [10] ISO 7146-2:2008 Plain Bearings – Appearance and Characterization of Damage to Metallic Hydrodynamic Bearings – Part 2: Cavitation Erosion and its Countermeasures [11] Holingan, P. T.: Plain Bearing Failures, in: Neale M. J. (Ed.): The Tribology Handbook. Second Edition. Oxford: Butterworth-Heinemann 1995 [12] Mellish, P. G.: Failures of Automobile Plain Bearings. Tribology, 2 (1969) 2, 100 – 105 [13] Kopeliovich, D.: Engine Bearing Failures and How to Avoid Them. Cedar Grove: King Engine Bearings, available at: www.kingbearings.com/files/Engine _Bearing_Failures_and_How_to_Avoid_Them.pdf [14] Engine Bearing Failure Analysis Guide. CL77-3-402. Ann Arbor: Clevite 2002 [15] Bartz, W. J.: The Influence of Lubricants on Failures of Bearings and Gears. Tribology International, 9 (1976) 5, 213 – 224 [16] Scott, D.: Bearing Failures Diagnosis and Investigation. Wear, 25 (1973) 2, 199 – 213 [17] Rac, A.: Basics of Tribology. Belgrade: Faculty of Mechanical Engineering, University of Belgrade 1991 (in Serbian). [18] ISO 4383:2012 Plain bearings – Multilayer materials for thin-walled plain bearings [19] Eyre, T. S.: Wear Characteristics of Metals. Tribology International, 9 (1976) 5, 203 – 212 [20] Rabinowicz, E.: Friction and Wear of Materials. Second Edition. New York: John Wiley & Sons 1995 [21] Garner, D. R., James, R. D. and Warriner, J. F.: Cavitation Erosion Damage in Engine Bearings: Theory and Practice. Journal for Engineering for Power, 102 (1980) 4, 847 – 857 [22] Strauss, B.M.: Fault tree analysis of bearing failures. Lubrication Engineering, 40 (1984) 11, 674 – 680 [23] Bhushan, B.: Principles and Applications of Tribology. Second Edition. New York: John Wiley & Sons 2013
Figure 01
Figure 02
Total number 616 of bearings
Main crankshaft bearings
CuPb22Sn2 210 342
Connecting 274 rod bearings
AlSn20Cu1
132
CuPb22Sn2
96
AlSn20Cu1
178
5 9 .4
6 0
F re q u e n c y , %
Figure 03
A B −a b r a s iv e w e a r A D −a d h e s iv e w e a r S F −s u r f a c e f a t ig u e w e a r C A −c a v it a t io n w e a r C O −c o r r o s iv e w e a r E R −e r o s iv e w e a r P D −p la s t ic d e f f o r m a t io n F R −f r a c t u r e
5 0 4 0 3 0
1 8 .9
2 0 1 0 0
A B
A D
1 1 .1
S F
6 .8
C A
1 .6
C O
0 .7
0 .9
0 .6
E R P D F R T y p e o f d a m a g e
Figure 04a
Figure 04b
Figure 05a
Figure 05b
Figure 05c
Figure 06
Figure 07a
Figure 07b
Figure 08
Figure 09
Figure 10
Figure 11
Figure 12
Figure 13
Figure 14
Figure 15
Percentage share
100 90 80 70 60
Main crankshaft bearings, 49.9 % (480 failures on 342 bearings)
CuPb22Sn2, 42.6 % (366 failures on 306 bearings)
50 40 30 20 10 0
Connecting rod bearings, 50.1 % (386 failures on 274 bearings)
Place of instalation
AlSn20Cu1, 57.4 % (500 failures on 310 bearings)
Bearing material All wear types
Figure 16a
Connecting rod bearings
Main crankshaft bearings
Percentage share
100 90 80 70
50.0
41.5
28.6
39.5
60
83.9
50
100.0
40 30 20
50.0
58.5
60.5
10 0
(a)
71.4
16.1
AB
AD
SF
CA
ER CO Wear type
Figure 16b
AlSn20Cu1
Percentage share
100
10.1
90 80 70
CuPb22Sn2
48.2
36.4
32.9 67.0
60 50
89.9
40 30 20
51.8
63.6
67.1 33.0
10 0
(b)
100.0
AB
AD
SF
CA
ER CO Wear type
Highlights 1. 2. 3. 4.
The most prominent types of wear are abrasive, adhesive and surface fatigue wear. Abrasive, adhesive and surface fatigue wear occur equally in both types of bearings. Adhesive, surface fatigue and cavitation wear occur more often at Al alloy bearings. Fault tree analysis (FTA) of diesel engine crankshaft bearing failure is performed.
S1350-6307(14)00161-7 http://dx.doi.org/10.1016/j.engfailanal.2014.05.014 EFA 2326
To appear in:
Engineering Failure Analysis
Received Date: Revised Date: Accepted Date:
1 December 2013 14 May 2014 20 May 2014
Please cite this article as: Vencl, A., Rac, A., Diesel Engine Crankshaft Journal Bearings Failures: Case Study, Engineering Failure Analysis (2014), doi: http://dx.doi.org/10.1016/j.engfailanal.2014.05.014
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Diesel Engine Crankshaft Journal Bearings Failures: Case Study Aleksandar Vencl University of Belgrade – Faculty of Mechanical Engineering, Belgrade, Serbia, [email protected] Aleksandar Rac University of Belgrade – Faculty of Mechanical Engineering, Belgrade, Serbia, [email protected] Abstract: Wear as a tribological process has major influence on the reliability and life of engine crankshaft bearings. The importance of field examinations of bearing failures due to wear is very well known. They point to the possible causes of wear and to the necessary treatment for its reduction or elimination. The paper presents the results obtained by examining 616 crankshaft bearings, damaged by different mechanisms. The bearings were installed in high-speed diesel engines, and were gathered for two years, during general repairs of the engines (overhaul), i.e. after 3000 to 5000 working hours. After the examination of the bearings, the fault tree analysis (FTA) was performed to determine the root causes for engine bearing failures. Each type of damage that was identified was illustrated with an appropriate hiresolution photograph. The investigations show that the basic and most conspicuous types of damage which cause bearing failures are abrasive, adhesive and surface fatigue wear. The paper also considers the effects of the place of installation and type of bearing material in respect to each type of wear. Keywords: diesel engine bearings, bearing failures, wear type statistic, fault tree analysis (FTA). 1. Introduction Reliability of diesel engines and other machinery is the function of faultlessness of its moving parts, from which bearings are certainly among the most important elements. An analysis of 410 recorded defects, occurred in industrial diesel engines during one year, showed that bearing class of defect takes 7 % of total occurrence [1]. Another study of diesel engine failures, which was based on experience of over 800 cases of damages during a period of four years, shows that 12 % of all damage cases are located in engine bearings [2]. The third study shows the distribution of internal combustion engine failures and their cost by parts failed first. Bearing failures head the list in both the number of engine failures (with 24.4 % of all incidents) and cost [3]. The analysis was based on the population of 180 engines, during the period of four years. According to the ISO standard, failure is defined as “termination of the ability of an item to perform a required function” [4]. Any damage or failure of the engine bearings results in a partial or complete functional failure of the engine and can cause significant economic losses and even legal liability. Although, as a rule, the price of the plain bearing is relatively small, a bearing damage that reduces the function of the system or results in its failure can cause a lot of overhead costs. Plain bearing failure involves the following categories: catastrophic failures which result in an immediate inability of a system to achieve its function; performance failures associated with a reducing performance of the equipment; reliability failures associated with a reducing reliability of the equipment [5]. In practice, damage of a bearing may often be the result of several mechanisms operating simultaneously. It is the complex combination of many influencing parameters which often causes difficulty in establishing the primary cause of damage. There are many publications dealing with plain bearing damages and failures. They are focused on damages and failures classifications and appearances [6-8], possible causes [9-11] and appropriate corrective actions [12-14], as well as on the effect of lubricants [15] and diagnostic procedures and examination tests [16]. The relevant literature and practical experience show that the plain bearing damages are mainly caused by wear, either as a direct cause or as a result of various irregularities in design, manufacture, assembly, operation and maintenance of the engine and bearings. Wear itself is a very complex process initiated by the action of different mechanisms, and can be manifested by different wear types which are often related: adhesive wear, abrasive wear, surface fatigue wear, erosive wear, cavitation wear, fretting wear, oxidative wear and corrosive wear [17].
There are very few literature data on the frequency of certain wear types occurrence in diesel engine bearings while the engine is under working conditions (field examinations). These data show which mechanisms have dominant influence on the wear of bearings, and determine what preventive measures should be taken to reduce wear. Correct diagnosis is the first step towards improved reliability. 2. Field examination details 2.1 Background The greatest application of hydrodynamic plain journal bearings is in internal combustion engines. This was the reason why we chose diesel engines as machines to collect from the adequate sample of damaged bearings. Generally, a representative sample could be gathered in two ways: during general repairs of the engines (overhaul) or during engine bench (stand) testing. In this study the first way was used, i.e. damaged bearings were gathered during general overhaul of the engines, in the period of two years. Field examinations of the bearings types of wear and its frequency of occurrence were focused on highspeed diesel engines that were in service from 3000 to 5000 working hours (overhaul engine), which corresponds to the recommended service life of investigated bearings. In other words, the general overhaul of the engines was not performed because of the catastrophic failures of the bearings, but because of the general deterioration and deviations from the minimum required performance (performance and reliability failures). That is why only in a few cases plastic deformation and fracture are registered on the examined bearings. The main difficulty of establishing separate type of wear is that they are often related, or that wear initiated by one cause can result in another effect. However, the dominant type of wear can be approximately determined. The classification of bearing damage was done according to the present wear type, i.e. it is based primarily upon the features visible on the running surfaces. The terms are correlated with the terms used in ISO 7146 standard. 2.2 Bearing types The damaged bearings that were examined were multilayer thin wall half-bearings, with and without the flange. They were used as crankshaft bearings and were installed in high-speed heavy duty diesel engines of vehicles and agricultural machinery (machines where diesel engines have the widest application). Total number of bearings in a representative sample was 616, and there were two types of bearings: main crankshaft bearings (342 pieces, i.e. 55.5 %) and connecting rod bearings (274 pieces, i.e. 44.5 %). The positions on the crankshaft of the investigated bearings are shown in Fig. 1. Figure 1. Positions on the crankshaft of the investigated bearings 2.3 Bearing materials The materials used as linings for examined crankshaft bearings were the two widely used alloys, according to the ISO standard [18] for multilayer materials for thin-walled plain bearings: cast lead-bronze alloy (76 wt.% Cu, 22 wt.% Pb, 2 wt.% Sn 2; 55-80 HB), referred as CuPb22Sn2, and aluminium alloy (79 wt.% Al, 20 wt.% Sn, 1 wt.% Cu; 30-40 HB), referred as AlSn20Cu1. Representation of the materials is almost identical, i.e. number of lead-bronze alloy bearings was 306 (49.7 %), and aluminium alloy bearings 310 (50.3 %). The percentage share of these two materials, used as bearing materials installed in diesel engines, on market is about 40-50 % for lead-bronze alloy and 50-60 % for aluminium alloy, which correspond to the percentage share used in this investigation. The distribution of the bearing materials investigated in this study, on the basis of the place of installation, also corresponds to the situation on market. For the main crankshaft bearings 210 pieces (61.4 %) are made of CuPb22Sn2, and 132 pieces (38.6 %) are made of AlSn20Cu1. On the other hand, for the connecting rod bearings the distribution is reverse and 96 pieces (35 %) are made of CuPb22Sn2, and 178 pieces (65 %) are made of AlSn20Cu1 (Fig. 2). The fabrication technology was the same for all bearings, i.e. standard casting procedure.
Figure 2. Distribution of the investigated bearings according to the place of installation and material 3. Results and discussion The main causes that lead to damage and failure of plain bearings are numerous, and very often with several causes contributing simultaneously. In addition, the bearing housing and the counter-face against which bearing operates also contribute. All these causes can be divided into several groups that include aspects of design, material selection, imperfections of material, manufacture and after-treatment, storage, transportation, assembly, inspection, maintenance, unforeseen operating conditions (like overloading) and direct mechanical and/or chemical damage during operation. In some instances, damage may be caused by a design compromise made in the interests of economy, as well [5,10,15]. Failures and damages caused by these causes commonly manifest as wear, fracture and plastic deformation of material. The results of this study (Fig. 3) show that the types of damage that occurred most frequently were abrasive (59 %), adhesive (19 %) and surface fatigue wear (11 %), while other types of damage were much less common, i.e. different types of wear were the most dominant types of damage (of all examined bearings, different types of wear were noticed 866 times, fracture 5 times and plastic deformation 8 times). These results are consistent, with the general observation that the most common type of wear is abrasive wear. According to Eyre [19] abrasive wear occurs in industry with 50 %, adhesive wear with 15 %, erosion and fretting wear with 8 %, and corrosive (chemical) wear with 5 % [19]. A comparison of the economic impact of different wear types also shows that abrasive and adhesive wear have the most relative importance (more than 80 % of all wear types) [20], and thus are the most common type of wear types. The low frequency of the corrosive wear occurrence in our study can be explained by good corrosive wear resistance of the materials that were used (lead-bronze and aluminium alloy), and by adequate and high-quality lubricants that are used for bearing lubrication. The lead-bronze and aluminium alloy are generally known as the corrosion resistant materials. For most of the observed bearings quality of the bond between lining and bearing back is rated as good to excellent, i.e. detachment of the lining material was not observed. Figure 3. Percentage occurrence of different types of bearing damages 3.1 Appearances and manifestations of the different type of bearing damages Abrasive wear: The majority of damages due to abrasive wear were recorded in middle (axial direction) of the bearing lining. These damages are manifested by scratches (from very fine to clearly visible scratches), Fig. 4a. This primarily relates to bearings made of AlSn20Cu1 alloy, where presence of craters left by displaced particles (characteristic of embedding) are also noticed (Fig. 4b). This alloy was softer than CuPb22Sn2 alloy and posses better embedding capacity [17]. At bearings made of CuPb22Sn2 alloy scratches are present all over on the surface. Severe scratching and scoring was not noticed on any of the bearings. Figure 4. Bearing surfaces damaged by abrasive wear: (a) two AlSn20Cu1 and one CuPb22Sn2 alloy bearings and (b) detail of the bearing in the middle on Fig. 4a (AlSn20Cu1 alloy) Adhesive wear: Adhesive wear was manifested either with wiping or seizure (Fig. 5). Wiping was noticed mainly in the main loaded area of bearings (shiny areas), and was accompanied by change of colour (dark blue to black) due to the local heating (Fig. 5a). Seizure was also accompanied by change of colour, as well as with surface melting and flow of the bearing material (Fig. 5b), and in some cases presence of severe seizure is noticed (Fig. 5c). Surface melting, flow of the bearing material and seizure were mainly present at the AlSn20Cu1 alloy bearings (this material was softer than the other investigated bearing alloy). Figure 5. Bearing surfaces damaged by adhesive wear: (a) wiping accompanied by change of colour of the bearing material (from left to right: CuPb22Sn2 alloy and two AlSn20Cu1 alloy bearings), (b) seizure of the AlSn20Cu1 alloy bearings and (d) detail of the bearing on the left in Fig. 5b (severe seizure)
Surface fatigue wear: Surface fatigue wear was present only in its initial stage, and was manifested with fatigue cracks which extend from the sliding surface in the loaded zone propagating as a network (Fig. 6). There were no characteristics places were cracks occur, and there were no bond faults (detachment of the lining material). Later stage of the surface fatigue (in which cracks change direction from normal to parallel to the surface above the bonding area, leading to the lining material relief) was not noticed. The used lining materials are resistant to the surface fatigue, and it is well known that with the general introduction of stronger bearing linings and improved calculation procedures, fatigue due to design overload is uncommon [7]. Figure 6. Bearing surfaces damaged by surface fatigue (initial period) Cavitation wear: Cavitation wear is seen predominantly in diesel engine bearings [21]. Two types of cavitation wear were noticed: flow cavitation (type 1) and discharge cavitation (type 4) [10]. Flow cavitation wear is noticed around circumferential oil groves (Fig. 7a), and discharge cavitation wear is noticed around oil hole (Fig. 7b). The damage of bearing surfaces was always on the same locations and is characterized as a form of micro fatigue cracking, initiated by the collapse of vapour cavities. These types of damages are typical for bearings operating at high loads, under vibrations and at high speeds. Figure 7. Bearing surfaces damaged by cavitation wear: (a) flow cavitation wear around oil grove and (b) detail of Fig. 7a showing discharge type of cavitation Corrosive wear: Corrosive wear was noticed only on CuPb22Sn2 alloy bearings, and was very mild (Fig. 8). There was no direct contact with water, i.e. formation of the oxide films was not noticed. Corrosion of the lead-bronze lining was due to the formation of organic acids by oxidation of lubricating oil. Figure 8. Bearing surfaces damaged by corrosive wear Erosive wear: Erosive wear was noticed around oil holes, and was manifested by removal of the material due to the fatigue mechanism. Most probably the inlet speed of lubricating oil was high, causing fluid erosion of the bearings (Fig. 9). In some cases even foreign particles in the oil can also lead to fluid erosion [9]. Figure 9. Bearing surfaces damaged by erosive wear (fluid erosion) Plastic deformation: Plastic deformation is manifested all across the bearing with and on considerable length (Fig. 10). Change of colour (dark blue to black) due to the local heating is noticed. Some bearings suffer significant melting at the bearing edge and within the groove. Melting of the lining material due to loss of lubricant resulted as seizure of the bearings. Figure 10. Bearing surfaces damaged by plastic deformation due to loss of lubricant Fracture: The appearance of the bearings damaged by fracture of the lining material consists of completely detached material in larger areas with clearly defined borders (rounder with ellipse in Fig. 11). Figure 11. Bearing surfaces damaged by fracture due to improper connection of lining and bearing back (clearly defined borders of detached material can be noticed) 3.2 Fault tree analysis Possible causes of diesel engine crankshaft bearing failure, which was detected during the investigation, are used (together with the literature data and experience) for the fault tree analysis (FTA). The different types of bearing damages were correlated with the ISO standard 7146 [9,10], which can be used for the
further explanation/description of the constructed fault tree root causes. The constructed fault tree diagram for diesel engine crankshaft bearing failure can be used by engineers in practice. A fault tree diagram is a logic diagram which traces the possible events leading to a major event, almost like a family tree. The desired or, in most cases, undesired main event is placed at the top of the tree and called the “top event” [22]. In our case, the top event will be entitled “crankshaft bearing failure”. In general terms, we can categorize two basic types of failure: (1) tribological and (2) mechanical, as illustrated in Fig. 12. Figure 12. Types of crankshaft bearing failure shown as Intermediate events on fault tree Tribological failures: According to our study, the tribological failures were the dominant types of crankshaft bearing failure. Tribological failures include seven broad wear types: (1) abrasive, (2) adhesive, (3) surface fatigue, (4) cavitation, (5) corrosive, (6) erosive, and (7) fretting. The discussion of different wear type failures must be more general, since they are intermediate events on fault tree of tribological crankshaft bearing failure (Fig. 13). Figure 13. Fault tree of tribological bearing failure Abrasive wear, according to ISO 7146, is classified as contamination with particles. In crankshaft bearings abrasive wear is caused by the contamination of oil either with foreign particles or with wear products. The damage typically manifests with scratches (Fig. 4a) made by contaminants, and with craters left by displaced embedded contaminants (Fig. 4b). Adhesive wear, according to ISO 7146, is classified in two categories, as wear by friction and as insufficient lubrication – starvation. In crankshaft bearings adhesive wear occurs when the oil film thickness is insufficient to prevent direct contact between the bearing and the shaft (mixed or boundary lubrication). Improper oil film thickness could be due to the: (1) insufficient clearance (caused by poor design or mounting); (2) oil film breakdown due to excessive load or vibrations (poor balancing); (3) insufficient oil supply (starvation) caused by lubrication system malfunction; and (4) improper oil characteristics (most usually low viscosity index). The adhesive wear typically manifests with wiping (light adhesive wear) accompanied by change of colour of the bearing material (Fig. 5a), or with seizure (heavy adhesive wear). Seizure is often accompanied by change of colour, and with surface melting and flow of the bearing material (Fig. 5b). Surface fatigue wear, according to ISO 7146, is classified as dynamic overload. In crankshaft bearings surface fatigue wear is caused, first of all, by the presence of dynamic load and overload (load exceeds fatigue strength). The damage typically manifests as initial pitting (Fig. 6), or as progressive pitting. Cavitation wear, according to ISO 7146, is classified as cavitation erosion. In crankshaft bearings cavitation wear occurs when the static pressure in oil is decreased below its vapour pressure, at the working temperature, so the evaporation occurs and vapour bubbles are generated in the oil. With the increase of pressure, these bubbles collapse and cause cavitation wear. Change of oil pressure could be due to: (1) incorrect oil flow due to the improper design and exploitation; (2) oil pressure fluctuation in bearing caused by crankshaft vibrations; and (3) interrupted oil flow through holes and grooves. The cavitation wear has been classified into four types: (1) flow (Fig. 7a); (2) impact; (3) suction; (4) discharge (Fig. 7b); and (5) miscellaneous. Corrosive wear, according to ISO 7146, is classified as contamination with chemicals. In crankshaft bearings corrosive wear is mainly caused by the thermal degradation of oil (low-quality oil), i.e. the corrosive nature of the oil is developed during long periods of operation as a result of contamination by combustion residues. Contamination could also be by water, antifreeze, etc. The damage typically manifests with change of colour (Fig. 8). Erosive wear, according to ISO 7146, is classified as contamination with particles. In crankshaft bearings erosive wear is caused by the fluid erosion (Fig. 9), and by the particles erosion. In both cases erosion is more intensive when oil-flow speed is higher. Fretting wear in crankshaft bearings is caused either by radial crankshaft vibration or by low frequency bearing vibration.
Mechanical failures: Two main mechanical failures are possible in crankshaft bearings: (1) plastic deformation, and (2) fracture (Fig. 14). Plastic deformation, according to ISO 7146, is classified as insufficient lubrication – starvation. Plastic deformation is caused by the loss of lubricant, when generated friction heat melts the bearing lining (Fig. 10). Fracture, according to ISO 7146, is classified as bond failure. Fracture of the lining material manifests as completely detached material in larger areas (Fig. 11), caused by static overload or Impact loads and/or improper manufacture (poor bonding). Figure 14. Fault tree of mechanical bearing failure 3.3 Influence of the place of installation and type of bearing material Number of occurrences of damage by wear is analyzed further in relation to the place of installation and type of bearing material (Fig. 15). The percentage share of main crankshaft and connecting rod bearings in total number of bearings was similar, but not equal (see Fig. 2). In addition, the number of failures was higher than the number of bearings, because in some cases more than one failure was identified at one bearing. There were 616 bearings (342 main crankshaft bearings and 274 connecting rod bearings), and 866 failures were identified on these 616 bearings (480 failures on main crankshaft bearings and 386 on connecting rod bearings). In other words, more failures were on main crankshaft bearings (480) than on connecting rod bearings (386), but the number of main crankshaft bearings was also higher (342 compared to 274). This is the reason why correction factors (Cor1 and Cor2) were introduced. Similar situation was with percentage share in total number of bearings of two bearing materials: CuPb22Sn2 and AlSn20Cu1 alloy (see Fig. 2). There were 616 bearings (310 AlSn20Cu1 bearings and 306 CuPb22Sn2 bearings), and 866 failures were identified on these 616 bearings (500 AlSn20Cu1 bearings and 366 CuPb22Sn2 bearings). In order to compare “real” percentage shear, in both cases a correction of the results was done: MainF ConnF MainF × Conn + ConnF ×Main ; × Cor1 ×100, % ; PSConn = × Cor1 ×100, % ; Cor1 = Main Conn Total CuPbF AlSnF CuPbF × AlSn + AlSnF × CuPb , ; ; PSCuPb = × Cor2 ×100, % PSAlSn = × Cor2 ×100, % Cor2 = CuPb AlSn Total
PSMain =
where: PSMain is percentage share of main crankshaft bearings; PSConn is percentage share of connecting rod bearings; PSCuPb is percentage share of CuPb22Sn2 bearings; PSAlSn is percentage share of AlSn20Cu1 bearings; Cor1 and Cor2 are the correction factor; Total is total number of bearings (616); Main is number of main crankshaft bearings (342); Conn is number of connecting rod bearings (274); MainF is number of failures on main crankshaft bearings (480); ConnF is number of failures on connecting rod bearings (386); CuPb is number of CuPb22Sn2 bearings (306); AlSn is number of AlSn20Cu1 bearings (310); CuPbF is number of failures on CuPb22Sn2 bearings (366); and AlSnF is number of failures on AlSn20Cu1 bearings (500). Figure 15. The effect of the place of installation and type of bearing material in respect to all wear types Investigation has shown that the place of installation has no influence on the occurrence of the wear, i.e. percentage shares of wear occurrence on main crankshaft bearings and connecting rod bearings were practically the same (49.9 and 50.1 %, respectively). On the other hand, wear was more pronounced on bearings made of AlSn20Cu1 alloy than on CuPb22Sn2 bearings (57.4 and 42.6 %, respectively). The effect of the place of installation and type of bearing material in respect to each type of wear is shown in Fig. 16. In order to compare “real” percentage shear of each type of wear, a correction of the results similar to the one applied for Fig. 15 was done. Figure 16. The effect of (a) place of installation and (b) type of bearing material in respect to each type of wear (AB – abrasive wear; AD – adhesive wear; SF – surface fatigue wear; CA – cavitation wear; ER – erosive wear; CO – corrosive wear)
The occurrence of the three most prominent types of wear (abrasive, adhesive and surface fatigue wear) was reported in approximately the same proportions in main crankshaft bearings and connecting rod bearings (Fig. 16a), so we could say that the place of installation has no significant influence on wear occurrence. It is interesting to note that the cavitation wear occurred mainly at the main crankshaft bearings. This could be due to the fact that connecting rod bearings were mainly ungrooved [21] and that only flow cavitation and discharge cavitation were noticed, but further analysis is necessary for a more precise explanation. The influence of bearing material on the occurrence of different wear types was not insignificant (Fig. 16b). Only the abrasive wear occurrence was approximately the same in AlSn20Cu1 and CuPb22Sn2 bearings. For other two most prominent types of wear (adhesive and surface fatigue wear) aluminium alloy showed lover resistance. Higher adhesive wear of Al alloy can be explained with its compatibility properties, i.e. Al is more compatible with Fe (the crankshaft was made of different types of steels) than Cu (the adhesion between Al and Fe is higher) [23]. Aluminium alloy also shows lower resistance to cavitation wear. Lower resistance to surface fatigue and cavitation wear of the AlSn20Cu1 alloy is related to the hardness of this alloy, which was lower than the hardness of the CuPb22Sn2 alloy. On the other hand, erosive wear is much pronounced in CuPb22Sn2 bearings, but this should be viewed through the fact that the number of the failures due to the erosive wear was small (only 2 cases for AlSn20Cu1 alloy, and 4 cases for CuPb22Sn2 alloy). Corrosive wear is only present in CuPb22Sn2 bearings, which can be explained by good resistance to corrosion of aluminium alloy. Aluminium-based linings are completely resistant to engine oils, and to their high-temperature degradation products [7]. 4. Conclusions The results obtained by field examination of failures established the main types of wear that occur in diesel engine crankshaft bearings. Establishment of the main types of wear facilitates the determination of the measures for their reduction or elimination and thus extending the service life and reliability of the crankshaft bearings. Further help in determining the root causes for engine bearing failures is provided by the fault tree analysis (FTA), which was performed. Based on the obtained results, it can be concluded that the most prominent (and thus the most important) types of wear in diesel engine crankshaft bearings are abrasive wear (app. 60 %), adhesive wear (app. 19 %) and surface fatigue wear (app. 11 %). Other types of wear (cavitation, erosive and corrosive wear) are less pronounced, although for certain conditions they can cause significant damage of bearing. The analysis of the effect of bearing place of installation shows that there is no significant difference in the occurrence of the three most prominent types of wear (abrasive, adhesive and surface fatigue wear) between the main crankshaft and connecting rod bearings. The type of bearing material, on the other hand, has an effect on the occurrence of certain types of wear under the same or similar conditions. Adhesive wear, surface fatigue wear and cavitation wear occur more often at aluminium alloy bearings, while corrosive wear was noticed only at lead-bronze alloy. Field examinations and results of the failures analysis are the base for detail engine test stands laboratory studies of individual phenomena, in order to identify particular influencing parameters. Acknowledgments This work has been performed as a part of activities within the projects TR 34028 and TR 35021. These projects are supported by the Republic of Serbia, Ministry of Education, Science and Technological Development, whose financial help is gratefully acknowledged. References [1] Collacott, R. A.: Mechanical Fault Diagnosis and Condition Monitoring. London: Chapman and Hall 1977 [2] Allianz, Handbook of Loss Prevention. Berlin: Springer 1978 [3] Bloch, H. P. and Gettner, F. K.: Practical Machinery Management for Process Plants, Volume 2: Machinery Failure Analysis and Troubleshooting. Third Edition, Houston: Gulf Publishing Company 1999
[4] ISO 13372:2004 Condition Monitoring and Diagnostics of Machines – Vocabulary [5] Rac, A. and Vencl, A.: Sliding Bearing Metallic Materials – Mechanical and Tribological Properties. Belgrade: Faculty of Mechanical Engineering 2004 (in Serbian) [6] Paine, B. and Cambell, B.: Examples of Damage Which Can Occur in Automobile Engine Bearings. TA 100/2. London: Glacier Metal Co. 1969 [7] Evans, C. and Warriner, J. F.: Bearings and Bearing Metals, in: Challen, B. and Baranescu, R. (Eds.): Diesel Engine Reference Book. Second Edition. Oxford: Butterworth-Heinemann 1999 [8] Affonso, L. O. A.: Machinery Failure Analysis Handbook: Sustain Your Operations and Maximize Uptime. Houston: Gulf Publishing Company 2006 [9] ISO 7146-1:2008 Plain Bearings – Appearance and Characterization of Damage to Metallic Hydrodynamic Bearings – Part 1: General [10] ISO 7146-2:2008 Plain Bearings – Appearance and Characterization of Damage to Metallic Hydrodynamic Bearings – Part 2: Cavitation Erosion and its Countermeasures [11] Holingan, P. T.: Plain Bearing Failures, in: Neale M. J. (Ed.): The Tribology Handbook. Second Edition. Oxford: Butterworth-Heinemann 1995 [12] Mellish, P. G.: Failures of Automobile Plain Bearings. Tribology, 2 (1969) 2, 100 – 105 [13] Kopeliovich, D.: Engine Bearing Failures and How to Avoid Them. Cedar Grove: King Engine Bearings, available at: www.kingbearings.com/files/Engine _Bearing_Failures_and_How_to_Avoid_Them.pdf [14] Engine Bearing Failure Analysis Guide. CL77-3-402. Ann Arbor: Clevite 2002 [15] Bartz, W. J.: The Influence of Lubricants on Failures of Bearings and Gears. Tribology International, 9 (1976) 5, 213 – 224 [16] Scott, D.: Bearing Failures Diagnosis and Investigation. Wear, 25 (1973) 2, 199 – 213 [17] Rac, A.: Basics of Tribology. Belgrade: Faculty of Mechanical Engineering, University of Belgrade 1991 (in Serbian). [18] ISO 4383:2012 Plain bearings – Multilayer materials for thin-walled plain bearings [19] Eyre, T. S.: Wear Characteristics of Metals. Tribology International, 9 (1976) 5, 203 – 212 [20] Rabinowicz, E.: Friction and Wear of Materials. Second Edition. New York: John Wiley & Sons 1995 [21] Garner, D. R., James, R. D. and Warriner, J. F.: Cavitation Erosion Damage in Engine Bearings: Theory and Practice. Journal for Engineering for Power, 102 (1980) 4, 847 – 857 [22] Strauss, B.M.: Fault tree analysis of bearing failures. Lubrication Engineering, 40 (1984) 11, 674 – 680 [23] Bhushan, B.: Principles and Applications of Tribology. Second Edition. New York: John Wiley & Sons 2013
Figure 01
Figure 02
Total number 616 of bearings
Main crankshaft bearings
CuPb22Sn2 210 342
Connecting 274 rod bearings
AlSn20Cu1
132
CuPb22Sn2
96
AlSn20Cu1
178
5 9 .4
6 0
F re q u e n c y , %
Figure 03
A B −a b r a s iv e w e a r A D −a d h e s iv e w e a r S F −s u r f a c e f a t ig u e w e a r C A −c a v it a t io n w e a r C O −c o r r o s iv e w e a r E R −e r o s iv e w e a r P D −p la s t ic d e f f o r m a t io n F R −f r a c t u r e
5 0 4 0 3 0
1 8 .9
2 0 1 0 0
A B
A D
1 1 .1
S F
6 .8
C A
1 .6
C O
0 .7
0 .9
0 .6
E R P D F R T y p e o f d a m a g e
Figure 04a
Figure 04b
Figure 05a
Figure 05b
Figure 05c
Figure 06
Figure 07a
Figure 07b
Figure 08
Figure 09
Figure 10
Figure 11
Figure 12
Figure 13
Figure 14
Figure 15
Percentage share
100 90 80 70 60
Main crankshaft bearings, 49.9 % (480 failures on 342 bearings)
CuPb22Sn2, 42.6 % (366 failures on 306 bearings)
50 40 30 20 10 0
Connecting rod bearings, 50.1 % (386 failures on 274 bearings)
Place of instalation
AlSn20Cu1, 57.4 % (500 failures on 310 bearings)
Bearing material All wear types
Figure 16a
Connecting rod bearings
Main crankshaft bearings
Percentage share
100 90 80 70
50.0
41.5
28.6
39.5
60
83.9
50
100.0
40 30 20
50.0
58.5
60.5
10 0
(a)
71.4
16.1
AB
AD
SF
CA
ER CO Wear type
Figure 16b
AlSn20Cu1
Percentage share
100
10.1
90 80 70
CuPb22Sn2
48.2
36.4
32.9 67.0
60 50
89.9
40 30 20
51.8
63.6
67.1 33.0
10 0
(b)
100.0
AB
AD
SF
CA
ER CO Wear type
Highlights 1. 2. 3. 4.
The most prominent types of wear are abrasive, adhesive and surface fatigue wear. Abrasive, adhesive and surface fatigue wear occur equally in both types of bearings. Adhesive, surface fatigue and cavitation wear occur more often at Al alloy bearings. Fault tree analysis (FTA) of diesel engine crankshaft bearing failure is performed.