Friction torque in thrust ball bearings: influence of thickener type and concentration

Faculdade de Engenharia da Universidade do Porto Departamento de Engenharia Mecˆanica e Gest˜ao Industrial Friction torque in thrust ball bearings: influence of the thickener type and concentration Samuel Pinto Pinho Master’s Degree Dissertation presented to Faculdade de Engenharia da Universidade do Porto Dissertation supervised by: Professor Doutor Jorge Humberto Oliveira Seabra, Full Professor of FEUP Doutor Armando Jos´e Vila¸ca Campos, Adjunct Professor at ISEP Porto, June of 2014 Acknowledgements I express my gratitude to my parents support and effort during all this years to assure me all the conditions. Without their help and encouragement it will not be possible. I would like also to thank to my brother Guilherme for the good moments that he proposed to me. My friends were also important on this process, specially, the friends from Viana do Castelo, from FEUP and from R.U.F.. I also want to thank to my girlfriend patient because I know that she have to have a lot. Finally I want to thank to my boss and co-workers in Foz Caff´e. I would like thank to Professor Jorge Seabra and Professor Armando Campos for their orientation, permanent support and leadership throughout the course of this work. David Gon¸calves also helped me during this work, then I want to thank for all his support, patient and time. Without his help in the theoretical subjects and in the practical work this work would not have been possible to accomplish. I wish to thank to CETRIB for all the conditions for all the conditions and opportunity and also the monetary help to this work provided by National Funds through FCT - Funda¸c˜ao para a Ciˆencia e a Tecnologia under the PTDC/EMEPME/122271/2010 and EXCL/SEM-PRO/0103/2012 projects. Finally, I would like to express my gratitude to Faculdade de Engenharia da Universidade do Porto (FEUP) for all the resources provided during my studies. iii Abstract The energy savings and environmental cares are very important and the science improve to reach a higher efficiency. The rolling bearings have a important role in this subject because they affects the equipment performance. Every year billions of rolling bearings are manufactured and almost 90% greased lubricated and sealed for life. Therefore is very important to understand the lubrication behavior in their internal contacts. A new type of greases of high performance have been developed. This greases with a polymeric thickener have the aim of increase the performance and the bearing life. As the tribological behavior of this greases is not well understood, three greases with different thickener percentage were tested in this work. The rheological properties were defined and several friction torque testes were performed with different temperatures and speeds in order to characterize their performance in a real application. To compare the results were also tested a ‘traditional’ lithium grease as reference and also a friction torque model was applied to the experimental results in order to identify the friction coefficients behavior. v Resumo As preocupa¸c˜oes com a redu¸c˜ao do consumo de energia e com os cuidados com o ambiente s˜ao muito importantes e a ciˆencia evoluiu no sentido de melhorar a eficiˆencia. Os rolamento tˆem um papel importante nestes assuntos uma vez que afetam a performance do equipamento. Todos os anos milh˜oes de rolamentos s˜ao produzidos e quase 90% s˜ao lubrificados com massa e selados para a vida. Assim ´e muito importante perceber como funciona a lubrifica¸c˜ao nos seus contactos internos. Um novo tipo de massas lubrificante de alta performance tem sido desenvolvido. Estas massas com um espessante polim´erico tˆem como objetivo aumentar a performance e o tempo de vida do rolamento. Uma vez que o comportamento tribol´ogico destas massas n˜ao ´e bem conhecido, trˆes massas com diferentes percentagens de espessante foram testadas neste trabalho. Foram definidas as propriedades reol´ogicas e foram feitas v´arias medi¸c˜oes do momento de atrito com temperaturas e velocidades diferentes para caracterizar o seu comportamento numa aplica¸c˜ao real. Para comparar os resultados foi tamb´em testada uma massa “tradicional” com espessante de l´ıtio como referˆencia e tamb´em foi aplicado um modelo do momento de atrito aos resultados experimentais para identificar o comportamento dos coeficientes de atrito. Keywords - Base oil - Bleed oil - Elastomer - Film Thickness - Friction torque - Grease - Rolling bearings - SKF R Model - Thickener Palavras-chave -´ Oleo Base - Bleed oil - Eslast´omero - Espessura de filme - Momento de atrito - Massa lubrificante - Rolamentos - Modelo da SKF R  - Espessante viii Contents Acknowledgements .............................. iii Abstract.................................... v Keywords ...................................viii Tableofcontents ............................... ix TableofFigures................................xiii ListofTables .................................xvii 1. Introduction 1 2. Physical Properties of Lubricants 3 2.1. LubricantsOils ............................. 3 2.2. Viscosity................................. 5 2.2.1. Dynamic Viscosity . . . . . . . . . . . . . . . . . . . . . . . 6 2.2.2. Kinematic Viscosity . . . . . . . . . . . . . . . . . . . . . . . 6 2.3. Termoviscosity ............................. 7 2.4. ViscosityIndex ............................. 9 2.5. Piezoviscosity .............................. 10 2.6. Viscosity-shear rate relationship . . . . . . . . . . . . . . . . . . . . 11 2.7. Density and thermal properties . . . . . . . . . . . . . . . . . . . . 13 2.7.1. SAE Viscosity classification . . . . . . . . . . . . . . . . . . 13 2.7.2. ISO Viscosity classification . . . . . . . . . . . . . . . . . . . 15 2.8. Greases ................................. 16 2.9. BaseOil................................. 17 2.10.BleedOil ................................ 17 2.11.Thickener ................................ 17 2.11.1. Used thickeners . . . . . . . . . . . . . . . . . . . . . . . . . 17 ix List of Figures 6.1. Diagram of the particle size distribution in the direct reading ferrograph.[8] ................................ 92 6.2. Lubricant being pumped across the transparent substrate to form aferrogram.[9] ............................. 94 7.1. Stribeck curves with free and controlled temperature measurements forM1...................................103 7.2. Stribeck curves with free and controlled temperature measurements forM2...................................104 7.3. Stribeck curves with free and controlled temperature measurements forM3...................................104 7.4. Stribeck curves with free and controlled temperature measurements forMLi. .................................105 A.1. M2 Bleed oil viscosity . . . . . . . . . . . . . . . . . . . . . . . . . 116 A.2. M3 Bleed oil viscosity . . . . . . . . . . . . . . . . . . . . . . . . . 117 A.3. MLi Bleed oil viscosity . . . . . . . . . . . . . . . . . . . . . . . . . 117 xvi List of Tables 2.1. Lubricant constants. . . . . . . . . . . . . . . . . . . . . . . . . . . 11 2.2. SAE classification of engine oils. . . . . . . . . . . . . . . . . . . . . 14 2.3. SAE classification of transmission oils. . . . . . . . . . . . . . . . . 15 2.4. ISO classification of industrial oils. . . . . . . . . . . . . . . . . . . 16 2.5. NLGI grease classification. . . . . . . . . . . . . . . . . . . . . . . . 19 3.1. Polymeric greases Density. . . . . . . . . . . . . . . . . . . . . . . . 25 3.2. Kinematic Viscosity [cSt] of the M1, M2, M3 and MLi bleed oil. . . 25 3.3. Tested greases properties. . . . . . . . . . . . . . . . . . . . . . . . 29 3.4. Rolling bearings that can be tested in the bearing assembly. . . . . 32 3.5. Rolling bearing characteristics. . . . . . . . . . . . . . . . . . . . . . 32 3.6. Kistler R 93339A characteristics. . . . . . . . . . . . . . . . . . . . 33 3.7. Thermocouples position. . . . . . . . . . . . . . . . . . . . . . . . . 35 3.8. Performedtests.............................. 37 4.1. Typical Composite Roughness values. . . . . . . . . . . . . . . . . . 52 4.2. Regime classification. . . . . . . . . . . . . . . . . . . . . . . . . . . 52 4.3. Parameter used in film thickness calculation. . . . . . . . . . . . . . 54 5.1. SKF R modelparameters. ....................... 58 5.2. Friction coefficients. . . . . . . . . . . . . . . . . . . . . . . . . . . . 77 5.3. Friction coefficients used in the model at free temperature. . . . . . 86 6.1. Ferrography index for low speeds. . . . . . . . . . . . . . . . . . . . 93 6.2. Ferrography index for low speeds. . . . . . . . . . . . . . . . . . . . 93 xvii List of Tables 7.1. Specific film thickness for all greases and controlled temperatures at1000rpm................................102 A.1. M2 Bleed oil viscosity at 60, 80 e 110◦C. ...............116 A.2. M3 Bleed oil viscosity at 60, 80 e 110◦C. ...............116 A.3. MLi Bleed oil viscosity at 60, 80 e 110◦C................117 xviii Nomenclature Symbol Unit Designation α Pa−1Piezoviscosity coeficient β[-] Termoviscosity coefficient µbl [-] coefficient depending on the additive package in the lubricant µEHL [-] Friction coefficient in full film conditions µsl [-] sliding friction coefficient φbl [-] weighting factor for the influence of asperity and lubricant shearing mechanisms φish [-] Inlet shear heating reduction factor φrs [-] Kinematic replenishment/starvation factor λ[-] Specific film thickness η0Pa s Dynamic viscosity environment pressure ηpPa s Dynamic viscosity at p pressure ηPa s Dynamic viscosity ˙γ s−1Shear rate ν[cSt] Kinematic viscosity ρ kg/m3density τPa Shear stress Am2contact area b [-] Constant dependent on the lubricant c [-] Constant dependent on the lubricant CPUC [-] Wear Particle Concentration d m bearing bore diameter D m bearing outside diameter dmm mean diameter xix List of Tables DL[-] index for large particles DS[-] index for small particles F N Axial Load Grr [-] Variable dependent on the type bearing, bearing mean diameter dm[mm] and on the axial load Gsl [-] variable depending on the bearing type, on the mean diameter [0.5(d+D)], on the radial load Fr [N] and on the axial load Fa[N] ISUC [-] Wear Particle Concentration K [-] Constant dependent on the lubricant Krs [-] replenishment/starvation constant (3×10−8for oil bath and oil jet lubrication; 6×10−8for grease and air lubrication M [N.mm] Total friction torque Mdrag [N.mm] Friction torque of the drag losses, churning, splashing etc. Mrr [N.mm] Rolling torque Mseal [N.mm] Friction torque of the seals M sl [N.mm] Sliding torque n rpm Rotational speed Sp[-] Modified Stribeck coefficient PLP [-] Percentage of Large Particles Q [N] normal load in the contact T [oC] Temperature VI [-] Viscosity index xx 1. Introduction The environmental issues make necessary to developed more efficient systems, then the lubrication has a main role to reach to this aim. New lubricants have been developed and efforts have been directed to predict grease performance in rolling bearings. Nowadays, nearly 90% of the rolling bearings are grease lubricated [10]. The main function of this equipment is to transmit load with low friction, therefore, the understanding of the internal friction in rolling bearings becomes relevant. The grease lubrication had a lot of attention during the last decades in order to understand its tribological behavior and structural changes after seing exposed to different speeds, temperature, shear stress etc. [11]. A new type of greases have been developed to improve the components behavior like the bearings or other components with relative motion. This new greases have a polimeric thickener instead of a metallic thickener. Such as its behavior is not well known several tests were carried out with low and high speeds and controlled and free temperature. The new SKF R model was also used with the experimental results in order to analyze the variation of the friction coefficients. The Ferrography and FTIR tests were performed to understand the changes on the grease after the work in the bearings. This thesis is composed by height chapters, including this introductory chapter. The second chapter is dedicated to lubricants main characteristics and the difference between oil and grease. In the third chapter the tested greases are presented and the test procedure is explained. In the fourth chapter the SKF R model and the film thickness calculation are described. In the fifth chapter the temperature, friction torque and film thickness results 1 1. Introduction are analyzed. It is also applied one model to the experimental results to characterize the friction coefficients. In the sixth chapter is presented the Ferrography and FTIR analyzes and the used methods. The seventh chapter present the general conclusion about this work. And finally, some future work are presented in order to complete this thesis. 2 2. Physical Properties of Lubricants The main function of the lubricants is to reduce friction and wear in a given system to increase its life-time. Another important aspect its quality, in others words, the resistance to degradation because the lubricants during the service its properties will changes and it may cause damage to the operating surfaces. Apart of the degradation, the oil may cause corrosion of the contacting surfaces. The price it is important too, because in large machinery the amount of oil could be large and in this cases the price is very important. 2.1. Lubricants Oils The first lubricants used, in the middle of the nineteenth century, had been obtained from animal or vegetable sources. Later, with the intensification of oil drilling and refining industries, the quantities of mineral oil available increased and, not in many years, mineral oil assumed the dominant position in the market place. In the last years the oil requirements increased due to news applications. Hence it was necessary developed a new base oils, known as synthetic lubricants, however, there is very disadvantages because they can be flammable, toxic or corrosive and the price can also be very expensive. Lubricants play a main role when there is relative motion between solids. They have mainly three functions: to separate the contacting surfaces, to reduce friction and to evacuate heat from the contact [12]. Rule of thumb these requirements are often contradictory so the selection of the lubricant is very difficult. Vegetable and animal oils Vegetable and animal lubricants were the first used by humans however some properties limited its application in industry. Natural 3 2. Physical Properties of Lubricants oils have a low resistance to both oxidation and high temperature, a high pour point and generally a short viscosity range. Since the 19th century natural oils have been replaced by hydrocarbon-based lubricants derived from petroleum. Nowadays, with the increasing of the price of the petroleum and with the environment concern increasing attention has again been paid to this type of lubricants. Mineral Oils Mineral oils are composed mainly by hydrocarbons, compounds of carbon and hydrogen which can be classified in three different classes: Paraffinic, Naphthenic and aromatic oils. This division depends on the oil basis major hydrocarbon constituent, which is often related to the local of the extraction. Paraffinic oils In paraffinic base oil the carbons atoms are linked to form long chains. The atoms structure may be branched but does not contain any ring, see Figure 2.1. Its general formula is CnH2n+2. Figure 2.1.: Paraffinic chain [1]. Naphthenic oils Its structure contains rings formed by carbons atoms. Aromatic oils Its structure is similar to the naphthenic oils but they are not saturated, that is why carbon ring contains alternate single and double bonds as show in Figure 2.3. This base oils are not very used because it degrade rapidly than the others base oils. Synthetic oils This lubricants are manufactured from one or more basic raw material components to obtain oils with desired properties and they have a well-defined molecular structure and weight distribution. Their properties 4 2.2. Viscosity Figure 2.2.: Naphthenic chain [1]. Figure 2.3.: Aromatic chain [1]. are even better, for example, synthetics have a broader temperature range than the mineral oils, lower evaporation rates, higher viscosity index, excellent low temperature range than the mineral oils and improved high temperature oxidation resistance. Due to the good properties, this lubricants are used in the most critical applications such as vacuum pumps, get engines or in food processing and pharmaceutical industry. To these base oils can be added additives which change its behavior. There are additives to change the viscosity, to protect surfaces from wear, to prevent oxidation etc... Unfortunately there is not any method to predict accurately the influence of mixing several additives since they mutually affect themselves as well as the base lubricant [12]. 2.2. Viscosity In tribology, the viscosity plays a main role because it depends on the base oil and it changes with the temperature and pressure. This parameter is very important 5 2. Physical Properties of Lubricants Figure 2.7.: Shear stress - shear rate relationship of a Newtonian fluid [1]. structural complexity of the lubricant. The differences between the lubricants is shown in Figure 2.8 12 2.7. Density and thermal properties Figure 2.8.: Viscosity variation with shear rate for different fluids [1]. 2.7. Density and thermal properties 2.7.1. SAE Viscosity classification The low temperature viscosity is an indicator of cold weather starting ability and the viscosity measured at 100◦C is related to the normal operating temperature of the engine (see Table 2.2). The oils without a ‘W’ suffix meet only one SAE grade and are classified as ‘monograde oils’, the oils with ‘W’ suffix, which stands for ‘winter’, have a good cold starting capabilities. The oils with a higher viscosity index, achieved by adding additives, are called ‘multigrade oils’, for example, SAE 20W/50 has a viscosity of SAE 20 at -18◦C and a viscosity of SAE 50 at 100◦C, see Figure 2.9. Multigrade oils usually shear thin, in other words, their viscosity drops significantly with increased shear rates [1]. The only difference between the SAE classification of transmitions oils (see table 2.3) and engine oils is that the winter grade is defined by the temperature at which the oil reaches the viscosity of 150000 [cP] which is the maximum oil viscosity which can be used without causing damage to gears. 13 2. Physical Properties of Lubricants Figure 2.9.: Monograde and multigrade oils [1]. Table 2.2.: SAE classification of engine oils. SAE Viscosity [cP] at Kinematic viscosity [cSt] viscosity grade temperature [◦C] max. at 100 ◦C Cranking Pumping min max 0W 3250 at -30 30000 at -35 3,8 - 5W 3500 at -25 30000 at -30 3,8 - 10W 3500 at -20 30000 at -25 4,1 - 15W 3500 at -20 30000 at -20 5,6 - 20W 4500 at -10 30000 at -15 5,6 - 25W 6000 at -5 30000 at -10 9,3 - 20 - - 5,6 <9.3 30 - - 9,3 <12.5 40 - - 12,5 <16.3 50 - - 16,3 <21.9 60 - - 21,9 <26.1 14 2.7. Density and thermal properties Table 2.3.: SAE classification of transmission oils. SAE Maximum Temperature for kinematic viscosity viscosity grade viscosity of 150000 cP [◦C] [cSt] at 100◦C Min. Max. 70W -55 4.1 - 75W -40 4.1 - 80W -26 7.0 - 85W -12 11.0 - 90 - 13.5 <24.0 140 - 24.0 <41.0 250 - 41.0 - 2.7.2. ISO Viscosity classification The ISO viscosity classification system was developed in USA and in United Kingdom and it is now commonly used throughout industry. The grade of a lubricant is designed as its kinematic viscosity measured in cSt or mm2s−1at 40◦C as it is shown in Table 2.4. 15 2. Physical Properties of Lubricants Table 2.4.: ISO classification of industrial oils. ISO Standard 3448 Midpoint viscosity Kinematic viscosity ASTM D-2422 limits [cSt (mm2/s)] at 40◦C[cSt] at 40◦C Minimum Maximum ISO VG 2 2.2 1.98 2.42 ISO VG 3 3.3 2.88 3.52 ISO VG 5 4.6 4.14 5.06 ISO VG 7 6.8 6.12 7.48 ISO VG 10 10 9.00 11.0 ISO VG 15 15 13.5 16.5 ISO VG 22 22 19.8 24.2 ISO VG 32 32 28.8 35.2 ISO VG 46 46 41.4 50.6 ISO VG 68 68 61.2 74.8 ISO VG 100 100 90.0 110 ISO VG 150 150 135 165 ISO VG 220 220 198 242 ISO VG 320 320 288 352 ISO VG 460 460 414 506 ISO VG 680 680 612 748 ISO VG 1000 1000 900 1100 ISO VG 1500 1500 1350 1650 2.8. Greases The American Society for Testing and Materials (ASTM) defines grease as a ‘solid to semi-fluid product or dispersion of a thickener agent in a liquid lubricant. Other ingredients imparting special properties may also be included’. The base oil is kept inside the thickener structure by a combination of Van der Waals and capillary forces. Interactions between the thickener molecules are dipole-dipole including hydrogen bonding or ionic and Van der Waals forces. The dispersed 16 2.9. Base Oil phase (thickener) may be a soap, solid hydrocarbons, inorganic matter or organic materials. 2.9. Base Oil The lubrication using grease is very dependent on base oil properties. Mineral oils are commonly used as base oil but synthetics are required for specific applications with extreme conditions. The greases tested with polypropylene thickener were formulated with a poly-alfa-opefin (PAO) oil and grease with lithium thickener was produced with a mixture of two different grades of PAO and some ester to facilitate the saponication reaction [15]. 2.10. Bleed Oil Certain greases under mechanical or thermal stress or during storage release oil. This is the bleed oil and its viscosity depends on the type of the grease thickener and the interaction between the thickener and the base oil. 2.11. Thickener The thickener type is very important because it changes the grease characteristics, for example, if the thickener used is water resistant the grease will be also water resistant, etc. The volume of thickener in a grease represents 4% to 20% and the base oil 75% to 90% of the total volume and the additives represent up to 5% [16,17]. There are different types of thickener in the market but in the sections below only will be described which are present in the used greases [18]. 2.11.1. Used thickeners Lithium soap Lithium greases have a excellent mechanical stability, good water resistance and reasonably high temperature performance, up to 120◦C. 17 2. Physical Properties of Lubricants The commonly lithium used is 12-hydroxy which provides to the grease low pumpability at low temperatures caused by high elastic properties. Polypropylene The polymers thickeners are a alternative to the traditional soap thickener. With this type of thickener the grease has improved bleeding properties at low temperatures, oil bleeding properties that are less temperature dependent, good lubricating abilities at low temperature, good mechanical stability and improved grease noise characteristics. With this non-polar thickener almost all additives can easily reach the metal surfaces which makes it possible to minimize the amount of additives and improve lubricating performance 2.12. Drop Point The drop point of the grease is its maximum operating temperature, in others words, is the temperature at which a grease changes from a semi-solid to a liquid behavior. 2.13. Grease Consistency Consistency is a measure of the hardness or shear strength of the grease and is measured measuring the penetration depth by a standard cone under very well defined conditions, see Figure 2.10. This a very important characteristic because it has a main role in the lubrication efficiency. If the grease is too smooth stable lumps of grease will never form during the operation which results in high operating temperatures and consequently short grease life. However if it is too hard causes grease starvation and it is very difficult pump. The NLGI grades are shown in Table 2.5. 18 2.14. Phases during the grease lubrication Figure 2.10.: NLGI grease classification. Table 2.5.: NLGI grease classification. NLGI Worked penetration grade range [x10−1mm] at 25◦C 000 445 - 475 00 400 - 430 0 355 - 385 1 310 - 340 2 265 - 295 3 220 - 250 4 175 - 205 5 130 - 160 6 85 - 115 2.14. Phases during the grease lubrication There are three phases during the grease lubrication [17]: •Churning Phase. •Bleeding Phase. •Severe film breakdown. 19 2. Physical Properties of Lubricants The phases which take more time during the grease lubrication are the churning and bleeding phases. The first typically takes from a few hours up to 24 hours, depending on the grease volume and on the working speed, the second phases take most of the time life of the bearing. Typically 30% of the bearing volume should be filled with grease to provide the bearing contacts with a fully flooded lubricant film. Part of the lubricant flows next to the running tracks or to under the cage bars or into the cage pocket where it will stay due to its consistency. The bearing geometry and the rheological properties are very important for the first phase because they will govern the flow behavior, therefore the friction torque is very dependent on the grease viscosity and the bearing temperature will rise. As more grease flows out the swept volume the friction torque will decrease which make the temperature decrease too until a steady value. The shear rate in the contact is in the order of 107and 104s−1[17] so the grease in the running tracks is severely worked and its rheological properties decrease rapidly. The film thickness changes rapidly because the lubricants properties changes too although the contacts are always fully flooded. During the bleeding phase there are several possible mechanisms for maintaining a lubricant film [17], but it is accepted that grease lubricated bearings are running under starved lubrication conditions. There has been very work published related to the lubrication conditions and there is not consensus about who the grease bleed the oil: the grease may release oil by bleeding, by breakdown of the thickener structure in the contacts or it may also simply provide a stiff grease film. 2.15. Polypropylene Greases The polymeric greases were formulated with the same base oil and thickener. The thickener is polypropylene and the concentration is different between them. The other grease is a commercial grease formulated with Lithium complex thickener and was tested only as reference. With this different greases we will analyze the influence of the thickener volume in the grease behavior. 20 2.16. Differences between grease and oil lubrication 2.15.1. Epoch technology The Epoch technology developed by Axel Christiernsson is based on non-polar polypropylene which allows the additives to reach the metal surfaces, see Figure 2.11. The conventional thickeners, due to their polarity, block the access of the additives to the surfaces [4]. Figure 2.11.: Epoch Technology [4]. 2.16. Differences between grease and oil lubrication There are differences between grease and oil lubrication. With grease the friction and corrosion are lower than with the oil since the churning phase only takes few hours. It is easier operate with grease than with oil due to its consistency it does not leak and a well designed bearing with a well selected grease requires no maintenance and is sealed for life. The grease out of the contacts protects the bearing whereas it forms a barrier against contaminants increasing its service life. The main disadvantage of using grease is its limited life [10] therefore bearing life is determined by the grease life which is determined by the parameter L10 that is the time that 10% of a population of bearings is expected to failure. With grease starvation can occur commonly due the side flow cause by the pressure difference inside the contact and next to the tracks which generates a side flow. With oil the re-flow is easy because it is not very consistent but with greases that is more 21 3. Experimental methods to the CH2and CH3deformation vibration. The absorbance band at 721,5 cm−1 is mainly due to the base oil [23]. Figure 3.6.: FTIR spectra of the bleed-oil and thickener obtained from lithium greases. Such as the polymeric greases there are peaks presented in the thickener spectrum which are not present in the grease or are visible but with smaller intensity due to the presence of the oil. The difference of the peaks between 2921 and 2853 cm−1is due to the presence of CH2and CH3and its deformation vibration. 3.2. Four-Ball machine The tests were performed with a modified Four-Ball machine where the four ball arrangement was replaced by a rolling bearing assembly as shown in Figure 3.7. This machine allows to control the rolling bearing temperature and measure other positions of the bearing house. It is also possible to force the air into the chamber to have a more stable measurements. 28 Table 3.3.: Tested greases properties. Grese Reference M1 M2 M3 MLi Units Thickener type Polypropylene Lithium complex - Thickener content 11 13 15 17.5 % Elastomer content 0 0 0 0 % Worked penetration (ISO 2137) 290 269 249 n.a. 10−1mm NLGI 2 2 3 n.a. - Before 286 264 256 n.a. 10−1mm Shell Roll Stability 80◦C/50h (mod. ASTM D1831) After 318 323 281 Difference 32 59 25 Flow pressure at -25◦C (DIN 51805) 320 390 670 - mbar Oil separation 100◦C/30h (ASTM D6184) 7.2 5.3 3.0 n.a. % Oil separation 100◦C/168h (IP121) 7.3 5.3 5.9 Oil evaporation 100◦C/30h (ASTM D6184) 0.2 0.9 0.8 Base oil viscosity (ASTM D445) 40◦C 48 178.7 mm2/s 100◦C 8 21.4 Bleed oil viscosity (ISO 12058) 40◦C 52.8 49.3 49.1 n.a. mm2/s 100◦C 8.3 8.3 8.3 3. Experimental methods (a) Overview. (b) House bearing assembly. Figure 3.7.: Four-ball machine. 3.2.1. Bearing House The rolling bearing assembly (Figure 3.8 ) is divided in two parts: the upper part which connect to the machine shaft and the lower part which transmit the torque to the torque cell. The shaft adapter (6) connect the upper bearing track (5). The lower track of the bearing (3) is supported by the lower race support (2) which transmit the friction torque to the torque cell (11). 30 3.2. Four-Ball machine Figure 3.8.: Components of assembly bearing. There is also seven thermocouples in different points of the equipment to measure the lubricant and house bearing temperature. With this information is possible to calculate the lubricant viscosity and the heat evacuated. The thermocouples I, II, III, IV and V measure the temperatures in the bearing house the others two measure the chamber, where the assembly is mounted, and the room temperature. This chamber is also under a continuous air convention forced by two fans which keep the air temperature surrounding the bearing house steady during the test. To make measurements with controlled temperature the bearing assembly is equipped with two heaters. With this tool is possible test four types of rolling bearing: thrust ball bearings, tapered roller bearing, angular contact ball bearing and cylindrical roller thrust bearings. The geometrical limitations imposed by the Four-Ball Machine and by the bearing housing, allow a maximum bearing outer diameter of 56.0 mm and a maximum width of 14.3 mm [24]. In the Table 3.4 is shown the features of the bearing types that can be tested on this bearing assembly. 31 3. Experimental methods Table 3.4.: Rolling bearings that can be tested in the bearing assembly. Dimensions [mm] Dynamic Limiting SKF Load [kN] Speed [rpm] Designation d D H C Single direction trust ball bearings 17 30 9 11.14 12000 51103 35 52 12 19.90 7500 51107 Cylindrical roller thrust bearings 17 28 9 11.20 8500 81102 TN 35 52 12 29.00 5600 81107 TN Single row angular contact ball bearings 17 40 12 11.00 22000 7203 20 47 14 13.30 18000 7204 Single row tapered roller bearings 15 42 14.25 22.40 18000 30302 J2 17 40 13.25 19.00 18000 30203 J2 Used bearing Torque measurements were all performed with a 51107 trust ball bearing [6], which the dimensions and characteristics are shown on Table 3.5 and its scheme on Figure 3.9 Table 3.5.: Rolling bearing characteristics. Principal Basic load ratings Fatigue Minimum Speed ratings Mass Designation dimensions load limit load factor Reference limiting dynamic static speed speed d D H C C0PuA [mm] [kN] [kN] [kN] [rpm] [kg] 35 52 12 19.9 51 1.86 0.013 5600 7500 0.08 51107 Grease Volume The grease volume used was 2cm3which is approximately 30% of the bearing free volume. The grease in excess remained between the bearing and the bearing house wall and could increase the friction torque. 32 3.2. Four-Ball machine Figure 3.9.: Scheme of the used bearing. [6] The grease was measured and applied with a syringe as is shown in the Figure 3.10. This method was used to the grease in the bearings for all the tests. 3.2.2. Torque cell To measure the friction torque was used a piezoelectric cell KISTLER R 93339A whose characteristics are shown in Table 3.6. This cell was selected because the piezoelectric sensors ensure high accuracy measurements even when the friction torque in the bearing is very small compared to the measurement range. In order to simplify and preserve the operations of mounting and dismounting, two circular steel plates were assemble on the top and the bottom of the cell. Table 3.6.: Kistler R 93339A characteristics. Measuring range [Nm] −10 to +10 Overload [Nm] −12/+ 12 Sensivity [pC/Nm] ≃ −460 Tensile/compression force, max. [kN] −5/+ 12 Side force, max. 1.5 Bending moment [Nm] 15 Operating temperatures [◦C] −40 to +120 33 3. Experimental methods Figure 3.10.: Bearing and syringe with grease. 3.2.3. Thermocouples As stated before, seven thermocouples were used to monitor the temperature in different locations. All the thermocouples used are K type with a measurement range among -40 ◦Cand 200 ◦Cand sensibility of 41 µV◦C−1. In the Table 3.7 are the function of each thermocouple. 3.2.4. Software To operate, monitor and control the system a program based on LabView R was used. In the upper of the interface (see Figure 3.11) are shown the temperatures and the speed rotational in the meddle is represented the friction torque. 34 3.2. Four-Ball machine Table 3.7.: Thermocouples position. Thermocouple of Figure 3.8 I Upper plate temperature II Bearing house temperature III Grease temperature Temperature of the lubricant IV deposited on the bottom of the bearing assembly V Inferior plate temperature Extra thermocouples VI Chamber temperature VII Room temperature Figure 3.11.: Software Interface. 35 3. Experimental methods 3.3. Test procedure The operating limits of the four-ball machine and the torque cell characteristics condition the test procedure. The limits imposed by the four-ball machine are 7000N for axial load and 5500rpm for the rotational speed. The temperature should be stabilized (±2◦C) so that the conditions are always the same. After the assembly and if the bearing house does not balance the following procedure is used. 1. For grease tests and before starting each test, a runningin period is always carried out, with an axial load of 1000 N and rotational speed varying from 100 rpm to 2000 rpm during 5 minutes to accommodate the grease. 2. With the machine in off position, the desired load is applied (in the example 7000 N) and the rotational speed set to the required value (100 rpm, for example); the fans are turned on to submit the rolling bearing assembly to a continuous forced air convection. 3. Turn on the machine and run the software to start the data acquisition. The operating temperatures rise continuously until stabilization is reached. 4. When the temperatures are stabilized, the machine is turned off and immediately restarted again together with the torque measurement. 5. After the torque measurement (120s), maintain the rotational speed on and wait until the temperatures stabilize. 6. Stages 4 and 5 are repeated three times to get three measurements of the friction torque in the same conditions When the friction torque measurements are at different speeds the test should always be conducted from the lowest to the highest speed. 3.4. Performed tests All the tests were performed with the same bearing type, load and the greases described in section 2.15. In Table 3.8 is shown the plan test (for some speeds was 36 3.4. Performed tests not possible performed the tests at 60◦C). Table 3.8.: Performed tests. Grease Temperature [◦C] Speed [rpm] MLi M1 Free, 60, 100, 250, 500, 1000, M2 80 and 100 1400, 1750, 2500, 3250 M3 The measurements were performed in two stages and always from the lowest to the highest speed. The first stage were carried out with the fast transmission and the friction torque was measured at high speeds, from 1000 to 3250 rpm. After that, the transmission was changed and were performed the tests from 100 to 1400 rpm. 37 4. Frictional torque in rolling bearings Figure 4.3.: Inlet shear heating factor. [7] Kinematic replenishment/starvation (φrs) The replenishment of lubricant in the raceway after a rolling element has passed is very difficult when the viscosity or the speed is very high because the lubricant has not sufficient time to flow back from the sides to the center of the raceway. This phenomenon is called kinematic starvation and the main consequence is the reduction of the rolling thickness. This factor can be estimated using the equation 4.4. φrs =1 e[Krs×υ×n×(d+D)qKZ 2(D−d)](4.4) Where: •φrs: kinematic replenishment/starvation reduction factor [-]. •e: base of natural logarithm [-]. •Krs: replenishment/starvation constant (3 ×10−8for oil bath and oil jet lubrication; 6 ×10−8for grease and air lubrication) [-] . 44 4.1. The SKF model for calculating the friction torque •KZ: bearing type related geometric constant [-]. •n: retational speed [rpm]. •d: bearing bore diameter [m]. •D: bearing outside diameter [m]. •υ: kinematic viscosity at operating temperature of the oil or the base oil viscosity of the grease [mm2/s]. 4.1.2. Sliding friction Sliding friction is always present in rolling contacts and it is separated in two different types: Macro-sliding is caused by contact conformity due to macro-geometry features (e.g. the contact between balls and curved raceways in ball bearings and spinning) Micro-sliding is caused by the geometrical distortion from elastic deformation. The slip profile (Figure 4.4) in the contact area will produce friction losses by means of lubricant shearing or/and asperity contact depending on the film thickness/roughness ratio. 45 4. Frictional torque in rolling bearings Figure 4.4.: Rolling element raceway contact with a curved contact surface. [7] The sliding friction moment is given by the following equation. Msl =Gsl ×µsl (4.5) •Msl: sliding friction momemt [N.mm]. •Gsl: variable depending on the bearing type, on the mean diameter [0.5(d+D)], on the radial load Fr[N] and on the axial load Fa[N]. •µsl: sliding friction coefficient [-]. Lubricant shearing The following equation gives the friction coefficient due to the lubricant shearing in one contact. µEHL =1 QZ A τdA (4.6) Where, 46 4.1. The SKF model for calculating the friction torque •µEHL: friction coefficient in full film conditions. [-] •Q: normal load in the contact. [N] •τ: shear stress in the lubricant. [Pa] •A: contact area. [m2] The shear stress depends on sliding speed and the lubricant rheology. Asperity contacts In the contacts some asperity interaction could exist when there is slip between the surfaces or the film thickness is not enough to completely separate the surfaces. The equation 4.7 gives the total friction coefficient coming from shearing the oil and from asperity contacts. µsl =φblµbl + (1 −φbl)µEHL (4.7) Where •µsl: sliding friction coefficient. [-] •µbl: coefficient depending on the additive package in the lubricant. [-] •φbl: weighting factor for the influence of asperity and lubricant shearing mechanisms. [-] The φbl (equation 4.8) gives us a idea about the influence of the asperity in the contact and its influenced by the additives in the lubricant. φbl =1 e2.6∗10−8(n.υ)1,4dm(4.8) Its behavior can be seen on Figure 4.5 47 4. Frictional torque in rolling bearings Figure 4.5.: Weighting factor φbl for the sliding friction coefficient. [7] In the Figure 4.5 is possible to see the influence of lubrication conditions on the weighting factor for the sliding friction. The value of φbl tends to zero for full-film lubrication. For low speed or viscosity the lubrication is in mixed film and the weighting factor tends to 1 and occasional metal-to-metal contact may occur and friction increases. 4.1.3. Seals The sliding between the lip of the seal and the moving steel counterface causes friction which represents a large percentage of the total friction. This type of friction will not be discussed since the tests were carried out without a bearing seals. 4.1.4. Drag losses The oil bath inside the bearing causes drag loses which increases the friction torque. The rotation speed, oil viscosity and oil level and also the lubricant reser48 4.2. Different Viscosity voir geometry or external factors influence the drag losses. Since only grease tests were made, this component will be neglected in the friction model. 4.2. Different Viscosity With the SKF R model is possible to plot separately the different sources of friction and understand the contribution of the several factors. For example, for an open unsealed spherical roller bearing lubricated in an very high viscosity oil bath at low speeds the sliding loses are very high due to the asperity interaction but due to the film build-up the sliding losses decrease quickly to a steady value. On the other hand, rolling torque is zero when the bearing is stopped and grows up very quickly until becomes dominant with a maximum value at about 500 rpm (Figure (4.7). For high speed the rolling torque decreases because of kinematic starvation and/or inlet shear heating. Figure 4.6.: Frictional moment versus rolling speed at constant temperature for a high viscosity oil. [7] For the oils with low viscosity is also possible demonstrate the influence of the different components of the friction. For this case, the sliding losses are dominant in almost all the time but decrease when the speed increases. This appends because the film thickness is so thin that the asperity interaction has a main role in the 49 4. Frictional torque in rolling bearings sliding torque (Figure 4.7). For this type of applications is recommended oils with higher viscosity because the bearing could fail due to the insufficient lubricant film. Figure 4.7.: Frictional torque versus rolling speed at constant temperature for a very low viscosity oil. [7] 4.3. Effect of centrifugal forces and spinning When a roller bearing is rotating there is centrifugal forces in the contact which will influence the lubrication. With this model is possible estimate the influence of each component of the friction torque. As it is shown in the Figure 4.8, for an open angular contact ball bearing lubricated with grease, the sliding losses are high at low speed because of the asperity interaction but as velocity increases the lubricant film builds up and they are reduced. However, when speed increases to high values, sliding losses increase again due to the increase of the centrifugal forces. This effect is taking in to account on this model by the variables Grr and Gsl. 50 4.4. Film Thickness Figure 4.8.: Frictional moment versus rolling speed at constant temperature for an open angular contact ball bearing lubricated with grease. [7] 4.4. Film Thickness The film thickness were calculated with the equation 4.9 [27]. It was assumed that the contact was under pure rolling condition in order to developed central film thickness formula for shear-thinning fluids. With the value of the specific film thickness (equation 4.11) is possible to classify the lubrication regime. In the Table 4.3 are the parameters and the values used in the intermediate calculus. hc= 1.098538U0.652G0.557W−0.0415RRx(4.9) R= 1+1.324280U0.69W0.264 G0.77Gcr 1.92 !−1.2(1−n)2.0 (4.10) With the equations above is possible notice that the Rfactor is 1 with n= 1 and consequently the equation 4.9 gives the film thickness for Newtonian fluids [27]. This parameter will be reduced if the power-law factor was smaller than 1. 51 4. Frictional torque in rolling bearings The specific film thickness is calculated with the equations 4.11 and 5.2. λ=hc σ(4.11) σ=σ12+σ221/2(4.12) In the Table 4.1 are the values of composite roughness for different bearings types. In this case the value selected was 0.14. Table 4.1.: Typical Composite Roughness values. Bearing Type σ Precision ball bearings 0.05 Thrust ball bearings 0.14 Ball bearings 0.18 Tapered and needle roller bearings 0.23 Spherical and cylindrical roller bearings 0.36 With the increasing of the λthe contact between metal surfaces and the wear are reduced. Table 4.2.: Regime classification. Regime λObservations Hydrodinamic λ≥30.0 Contact surfaces completely separated by a very thick lubricant film Full film λ≥3.0 Contact surfaces completely separated by a lubricant film. Mixed film 1.0≤λ≤3.0 Contact surfaces partially separated by a lubricant film, occurring at some points metal - metal contact Boundary film λ≤1.0 There is no lubricant film separating the surfaces in contact, predominating metal - metal contact The Figure 4.9 shows the relationship between the friction coefficient and the film thickness. When the lubrication regime is boundary lubrication (λ≤1.0) 52 4.4. Film Thickness Figure 4.9.: Relationship between friction coefficient and film thickness. [3] the friction coefficient is very high due to the low film thickness which can cause interaction between the surfaces. As the film thickness increases (λ > 1.0) the interactions between the surfaces are lower such as the friction coefficient. 53 5. Experimental results (a) Experimental and theoretical friction torque. (b) Components of the total friction torque. Figure 5.6.: Measurements at 60◦C with grease M2. (a) Specific film thickness λ. (b) Weighting factor φbl. Figure 5.7.: Specific film thickness and weighting factor with grease M2. 60 5.3. SKF model (a) Experimental and theoretical friction torque. (b) Components of the total friction torque. Figure 5.8.: Measurements at 60◦C with grease M3. (a) Specific film thickness λ. (b) Weighting factor φbl. Figure 5.9.: Specific film thickness and weighting factor with grease M3 at 60◦C. 61 5. Experimental results (a) Experimental and theoretical friction torque. (b) Components of the total friction torque. Figure 5.10.: Measurements at 60◦C with grease MLi. (a) Specific film thickness λ. (b) Weighting factor φbl. Figure 5.11.: Specific film thickness and weighting factor with grease MLi at 60◦C. 62 5.3. SKF model (a) Specific film thickness λ. (b) Weighting factor φbl. Figure 5.12.: Specific film thickness and weighting factor at 60◦C. The theoretical values of µsl where obtain with the following equations. The experimental values are also plotted in the Figure 5.13. M−Mrr =Msl (5.2) µsl =Msl Gsl λ=hc σ(5.3) 63 5. Experimental results Figure 5.13.: Stribeck curves for each grase at 60◦C. The Figures 5.4, 5.6, 5.8 and 5.10 show the total friction torque (a) and its sources (b). In general the model predicted the results with reduced error since the viscosity is constant (due to the constant temperature). It is also possible to see that with low speeds the Sliding torque (Msl) is the main responsible for the total friction because the film thickness has lower values as is shown in the Figure 5.12. With the increasing of the speed the Sliding torque decreases and the Rolling torque (Mrr) increases since the contact between the metal surfaces reduces which reduces the Msl and the energy spent to remove the grease in excess 64 5.3. SKF model from the tracks increase which makes the Mrr increasing. The polymeric greases have almost the same value of specific film thickness because only the thickener content is different between them but the lithium grease has the highest due to the highest viscosity. In the Figures 5.5, 5.7, 5.9 and 5.12 is shown the film thickness (λ) (a) and the weighting factor (φbl) (b). The speeds with lower specific film thickness have the φbl closer to 1 as predicted in the section 4.1.2. In others words, for the rotational speed of the rolling bearing with φbl ≈1 the regime is boundary lubrication. The polymeric greases have the highest φbl which correspond to the lower specific film thickness values. The Stribeck curves (Figure 5.13) also confirms this behavior because with the increasing of the speed, and consequently the specific film thickness, the value of µsl decreases and the curve has the same behavior as the curve in the Figure 4.9. With controlled temperature, only the speed changes in the modified Stribeck parameter, so is possible to compare with the value of λ. The grease M1 has the lowest value of thickener and the highest value of friction coefficient whereas the grease M3 has the lowest value of the µsl which means that the thickener content influences the friction coefficient. 65 5. Experimental results Tests performed at 80◦C (a) Experimental and theoretical friction torque. (b) Components of the total friction torque. Figure 5.14.: Measurements at 80◦C with grease M1. The friction coefficient for measurements at 80◦Cused in the model to calculate the theoretical values are in the Table 5.2. (a) Specific film thickness λ. (b) Weighting factor φbl. Figure 5.15.: Film thickness and weighting factor with grease M1 at 80◦C. 66 5.3. SKF model (a) Experimental and theoretical friction torque. (b) Components of the total friction torque. Figure 5.16.: Measurements at 80◦C with grease M2. (a) Specific film thickness lambda. (b) Weighting factor φbl. Figure 5.17.: Film thickness and weighting factor with grease M2 at 80◦C. 67 5. Experimental results (a) Experimental and theoretical friction torque. (b) Components of the total friction torque. Figure 5.18.: Measurements at 80◦C with grease M3. (a) Specific film thickness. (b) Weighting factor φbl. Figure 5.19.: Specific film thickness and weighting factor with grease M3 at 80◦C. 68 5.3. SKF model (a) Experimental and theoretical friction torque. (b) Components of the total friction torque. Figure 5.20.: Measurements at 80◦C with grease MLi. (a) Specific film thickness λ. (b) Weighting factor φbl. Figure 5.21.: Specific film thickness and weighting factor at 80◦C. 69 5. Experimental results (a) Specific film thickness λ. (b) Weighting factor φbl. Figure 5.32.: Specific film thickness and weighting factor at 110◦C. Figure 5.33.: Stribeck curves for each grase at 110◦C. Such as with the measurements at 60 and 80◦Cthe Figures 5.14, 5.16, 5.18 and 5.20 show the total friction torque (a) and the the behavior of the different sources. As the temperature does not changes with the speed the viscosity does not changes either. It is also possible to see that with low speeds the Sliding torque (Msl) has 76 5.3. SKF model a higher value than the Rolling torque because there is contact between the metal surfaces due to the low specific film (see Figure 5.32). With the increasing of the speed the Sliding torque decreases and the Rolling torque (Mrr) increases because is necessary to spend energy to remove the grease in excess. In the Figures 5.25, 5.27, 5.29 and 5.32 are represented the specific film thickness (a) and the weighting factor. The difference between the values of the specific film thickness of the lithium and the polymeric greases is caused by the high difference between the viscosity values. The grease MLi is more viscous and consequently has the highest specific film thickness. The speeds with highest φbl have the lowest specific film thickness as predicted in the section 4.1.2. At 110◦C the viscosity is very low then any grease is in boundary film lubrication. The Stribeck curves (Figure 5.33) also confirms the values of the film thickness because the µsl decreases with the increasing of the Spparameter is is shown in the Figure 4.9 and for the same speed the friction coefficient decreases with the increasing of thickener content. 5.3.2. Friction coefficients The Table 5.2 shows the values of the friction coefficient for the different greases at the temperature of the performed tests obtained with the equations 4.7 and 4.8 and with the experimental results. Both values are temperature dependent and decrease with the increasing of the temperature. The µEHD as the lower values because the regime lubrication is closer to the boundary rather than the full film lubrication. Table 5.2.: Friction coefficients. 60◦C 80◦C 110◦C µbl µEHD µbl µEHD µbl µEHD M1 0,0982 0,0201 0,0893 0,0136 0,0817 0,0011 M2 0,0854 0,0236 0,0740 0,0131 0,0716 0,0023 M3 0,0698 0,0302 0,0577 0,0093 0,0485 0,0010 MLi 0,0837 0,0372 0,0689 0,0126 0,0561 0,0010 In the Figures 5.34, 5.35, 5.36 and 5.37, is shown the relation between µsl and the 77 5. Experimental results modified stribeck parameter. This curves are very similar to the Stribeck curves of the lubricants and is possible to relate with the results of the film thickness as is shown in the Figure 4.9. Figure 5.34.: Stribeck curves for M1. Figure 5.35.: Stribeck curves for M2. 78 5.3. SKF model Figure 5.36.: Stribeck curves for M3. Figure 5.37.: Stribeck curves for MLi. As is presented in the Figures 5.34, 5.35, 5.36 and 5.37 the friction coefficient reduces with the increasing of the Spparameter. This means that the µsl reduces with the increasing of the speed because in the measurements with controlled 79 5. Experimental results temperature only the rotation speed changes. For all greases the highest values of µsl are obtained at 60◦Cwhich correspond to the highest viscosity and consequently highest specific film thickness. The grease with highest sliding friction coefficient (at all speeds and temperatures) is M1 which correspond to the highest value of friction torque. The grease MLi has the lowest µsl for each temperature and speed despite the higher value of viscosity, because has the highest value of film thickness. 5.4. Free temperature measurements All the greases were tested at the same condition used with the measurements at controlled temperature but in this case the temperature of the bearing was not controlled to approximate to the real conditions. 5.4.1. Measured friction torque and temperature (a) Friction torque. (b) Stabilization temperatures. Figure 5.38.: Friction torque and stabilization temperature for all tested greases for low speeds. 80 5.4. Free temperature measurements (a) Friction torque. (b) Stabilization temperatures. Figure 5.39.: Friction torque and stabilization temperature for all tested greases for high speeds. In the Figures 5.38 and 5.39 are shown the measured friction torque at free temperature (a) and the stabilization temperature of the bearing at each speed. The results were plotted separately for low and high speeds due to the different stabilization temperature of the bearing (thermocouple III). The polymeric greases have almost the same stabilization temperature but the lithium greases has highest due to the higher friction torque caused by the highest viscosity. The friction torque is lower at high speed because the specific film thickness has higher values than with lower speeds as is shown in the Figures 5.40. In general the M3 has the lowest values of the friction torque of the three polymeric greases as was seen with controlled temperature. This means that the thickener content influences the grease performance because with the increasing of the thickener content the total friction torque reduces. 5.4.2. Calculated Film Thickness The values of the specific film thickness were calculated with the equation 4.9. To validate the results the values were compared with the curve of the φbl. As the temperatures changes with the speed the viscosity changes too as is shown in the 81 5. Experimental results Figure 5.42. Figure 5.40.: Specific film thickness for the tested greases at free temperature. Figure 5.41.: Weighting factor, φbl, for the tested greases at free temperature. 82 5.4. Free temperature measurements (a) Difference between the bearing stabilization and the room temperature. (b) Kinematic viscosity at stabilization temperature. Figure 5.42.: Grease viscosity at stabilization temperatures. In the Figures 5.40 and 5.41 are represented the specific film thickness (λ) and the curve of the φbl respectively. It is possible to observe that the grease with higher λis the MLi and the lubrication is in full film (see table 4.2) for almost all the speeds due to the higher viscosity of the lithium grease. The film thickness is calculated for each speed and is very influenced by the bearing and room temperature which justifies the changing in the evolution of the specific film thickness for the lithium grease between 1000 and 1750 rpm. As mentioned in the section 4.1.2 when the values of φbl tends to zero it means that the lubrication regime is in full film (λ > 3) which validate this results because almost all the speed for the bearing lubricated with MLi grease the φbl is approximately zero. The film thickness for the contacts lubricated with the polymeric greases does not vary for each grease but the stabilization temperature is lower (see Figure 5.42). 83 5. Experimental results 5.5. SKF model Figure 5.43.: Theoretical Friction torque for M1 at free temperature. Figure 5.44.: Theoretical Friction torque for M2 at free temperature. 84 5.5. SKF model Figure 5.45.: Theoretical Friction torque for M3 at free temperature. Figure 5.46.: Theoretical Friction torque for MLi at free temperature. The friction torque predictions are closer to the experimental ones (the maximum error is around 17 %) for all the measurements with free temperature (Figures 5.46, 5.43 5.44 and 5.45). For the measurements with free temperature there are more parameters changing. Along the measurements the speed was increasing such as in the controlled 85 6. Wear Figure 6.1.: Diagram of the particle size distribution in the direct reading ferrograph. [8] (bigger than 5 µm) precipitates first than the small ones. Therefore at the beginning of the tube, where are the biggest particles, is obtained the DLfactor which is influenced by the number of large em some small particles. In the end of the tube there is another sensor that gives us the DSfactor related with the particle smaller than 5µm (see Figure 6.1) [8]. With the index for large (DL) and small (DS) particles is possible to calculate the Wear Particle Concentration (CPUC), wear Severity (ISUC) and the Percentage of Large Particles (PLP) with the following equations. CPUC =DL+DS d(6.1) ISUC =DL2−Ds2 d2(6.2) PLP =100 ×(DL−Ds) DL+Ds (6.3) Where 92 6.1. Ferrography •d: Dilution factor Results The table 6.1 and 6.2 shows the index obtained with the Direct reading ferrography for low and high speeds respectively. Table 6.1.: Ferrography index for low speeds. Amostra M1 LS M2 LS M3 LS Mli LS Number of cicles d1 1 1 1 DL 1.6 5.2 11 4.2 DS 1.2 1.2 6.9 3.5 1.5 milion CPUC 2.8 6.4 17.9 7.7 ISUC 1.100 26.000 73.000 5.400 PLP (%) 14.29 62.50 22.91 9.09 Table 6.2.: Ferrography index for low speeds. Amostra M1 HS M2 HS M3 HS Mli HS Number of cicles d0.1 1 1 1 DL 48.1 49.9 7.9 12.2 DS 18.9 17.3 3 1.5 5.25 milion CPUC 670 67.2 10.9 13,7 ISUC 200000 2200 53 150 PLP (%) 43.58 48.51 44.95 78,10 To performed the testes was necessary to dilute the grease M1 (only the grease used in the measurements with high speeds) 10 times in order to get of the DL and DSbelow 100. For the tests with high speeds the greases M1 and M2 have the higher values of particles index. The extreme conditions of temperature and speed may have degraded the greases. The parameters of the others greases, including the lithium grease, are very similar. 93 6. Wear 6.1.2. Standard Ferrograph analyzer The Standard Ferrograph analyzer pumps the lubricant sample at a flow rate of the order of 0.25 ml per minute and the particles are deposited by size due to the high-gradient magnetic field under the treated transparent substrate. The particles are fixed to the slide and a washing process removes the residual lubricant. With a microscope is possible to analyze the distribution, shape and others parameters. Figure 6.2.: Lubricant being pumped across the transparent substrate to form a ferrogram. [9] Results In the observations with microscope were seen particles obtained due to the high temperatures since the color of the particles is different from the original color of 94 6.1. Ferrography the rolling bearing. This particles were observed for all the greases due to the high temperature and speed of the friction torque measurements (see Fotografia 3, 10, 12 and 15). It is visible that the high value of the particles obtained in the Direct Reading Ferrography was due to the thickener degradation (see Fotografia 5 and 9 ). This results are very influenced by the local where the sample is collected, than same error can be associated to this results. 95 7. Results discussion 7.1. Friction torque As is shown in the section 5.3 the values of the SKF R model with controlled temperature are more reliable because the viscosity does not changes with the different speeds (due to the steady temperature) and therefore less variables are changing in the system. It is also possible verify that for low speeds the friction torque has the highest values due to the small film thickness which can cause interaction between the surfaces and wear. For high speeds the film thickness increases and as the viscosity is constant, the friction torque decreases. Is also important to notice that the total friction torque sources change their values with the speed. For low speeds the sliding torque (Msl) is the main responsible for the total torque but for high speed becomes almost null and the rolling torque (Mrr) increases. This behavior is related to the film thickness formation which change the friction coefficients. When the specific film thickness is reduced, the φbl is closer to 1 what makes the Msl increasing too. For high specific film thickness, the φbl is almost zero and the Msl reduces because µEHL has low values too. In the free temperature measurements the stabilization temperatures are almost the same for the polymeric grease and are much lower than stabilization temperature of the lithium grease as can be seen in the Figure 5.42 which reduces the evacuated heat. This means that the thickener content does not influences the grease temperature. The stabilization temperatures of the grease MLi are higher because the friction torque is higher (for almost all the speeds at free temperature) which increases the heat generation. 101 Bibliography [1] G. W. Stachowiak and A. W. Batchelor. Engineering Tribology. Elsevier Science Publishers B.V., 1993. [2] J.A. Williams. Engineering Tribology. Oxford science publications. Oxford University Press, 1994. [3] B.J. Hamrock, S.R. Schmid, and B.O. Jacobson. Fundamentals of Fluid Film Lubrication. Mechanical engineering. Taylor & Francis, 2004. [4] Axel Cgristiernsson AB. New grease thickener system provides step change in grease life. Sealing Technology, page 16, 2008. [5] Thermo Nicolet. Introduction to fourier transform infrared spectrometry. Thermo Nicolet Corporation, 2001. [6] SKF. SKF General catalogue 60000EN. 2013. [7] G. M. Espejel. Using a friction model as an engeneering tool, volume 1. Evolution vol.1, 2006. [8] G. Pocock and S. J. Courtney. Some quantitative aspects of ferrography. Wear - Elsevier, 1980. [9] W. Seifert R. Bowen, D. Scott and V. C. Westcott. Ferrography. Tribology International, 1976. [10] Piet M. Lugt. A riview on grease lubrication in rolling bearings. Tribologu Transactions, 2009. [11] M. N. Webster P. M. Cann, J. P. Doner and V. Wikstrom. Grease degradation in rolling element bearings. Tribologu Transactions, 2001. 109 Bibliography [12] Jos´e Augusto de Sousa Pereira Brandao. Gear Micropitting prediction using the Dang Van high-cycle fatigue criteion. PhD thesis, Faculdade de Engenharia da Universidade do Porto, 2007. [13] Alexandre Sottomayor Jorge Seabra, Armando Campos. Lubrifica¸c˜ao elastohidrodinˆamica. 2002. [14] Michael Nosonovsky and Bharat Bhushan. Green Tribology. Springer, 2012. [15] Rui Bert˜ao Patricio Martins Marques. Friction torque in thrust ball bearings lubricated with polymeric greases. Master’s thesis, Faculdade de Engenharia da Universidade do Porto, 2013. [16] P. M. Cann. Grease liubrication of rolling element bearings - role of the grease thickener. Wiley InterScience, 2007. [17] Piet M. Lugt. Grease Lubrication in Rolling Bearings. John Wiley Sons, Ltd, 2013. [18] V. Mota and L.A. Ferreira. Influence of grease composition on rolling contact wear: Experimental study. Tribology International, 42(4):569 – 574, 2009. Tribology for engines and transmissions. [19] I. Couronn P. Vergne D. Mazuyer N. Truong-Dinh D. Girodin. Nature and properties of the lubricating phase in grease lubricated contact. Tribology Transactions, 2014. [20] I. Courine P. Vergne D. Mazuyer N. Truong Dinh. Effects of grease composition and structure on film thickness in rolling contact. Tribologu Transactions, 2002. [21] Meng Yonggang and Zheng Jie. A rheological model for lithium lubricating grease. Tribology International, 1998. [22] P.R. Griffiths and J.A. De Haseth. Fourier Transform Infrared Spectrometry. Chemical Analysis: A Series of Monographs on Analytical Chemistry and Its Applications. Wiley, 2007. 110 Bibliography [23] David Goncalves Beatriz Graca Armando V. Campos J. Seabra Johan Lecker Rene Westbroek. Formulation, rheology and thermal aging of polymer grease - part i: Influence of the thickener content. Elsevier, 2014. [24] Armando Campos Tiago Cousseau, Beatriz Graca and Jorge Seabra. Experimental measuring procedure for the friction torque in rolling bearings. Wiley InterScience, 2010. [25] Tiago Cousseau. Film Thickness and Friction in Grease Lubricated Contacts: Application to Rolling Bearing Torque Loss. PhD thesis, Faculdade de Engenharia da Universidade do Porto, 2013. [26] Armando V. Campos Tiago Cousseau, Beatriz M. Graca and Jorge H. O. Seabra. Influence of grease rheology on thrust ball bearings friction torque. Tribology International, 2011. [27] Puneet Katyal and Punit Kumar. Central film thickness formula for shear thinning lubricants in ehl point contacts under pure rolling. Tribology International, 2011. [28] D. Scott and V. C. Westcott. Predictive maintenance by ferrography. Wear - Elsevier, 1977. [29] Jack Poley. Oilbearings for monitoring hydraulic oil systems, a step stage approacg. Lubrication Engineering, 1989. 111 A. Measurements results A.1. Measurements results, standard deviation and bearing temperature 113 Speed [rpm] Mt [N.mm] σ Bearing temperature ºC Mt [N.mm] σ Bearing temperature Mt [N.mm] σ Bearing temperature Mt [N.mm] σ Bearing temperature 100 227,6638 19,0742 27,0200 210,7483 7,1529 26,7000 216,2000 17,1336 26,7710 245,3843 21,3661 26,1660 250 200,3643 7,8518 29,9470 196,9071 7,4085 29,4800 164,9100 17,6841 29,5280 233,1750 10,7961 33,0720 500 213,1700 11,2158 35,3740 173,3633 12,7834 34,4430 165,6860 11,3494 34,8070 242,3900 13,4290 40,1700 1000 192,2571 4,9723 44,0940 167,9986 6,6468 42,7900 186,5514 11,6091 43,2500 217,5186 27,4484 52,1400 1400 196,9100 17,3613 50,5890 173,5633 8,1923 49,0570 154,5186 12,4695 51,2520 202,4225 21,0277 60,3570 1000 145,4025 18,1663 42,9700 159,6571 15,1740 45,0760 156,5071 11,3829 45,1580 134,2475 21,7662 47,6740 1750 153,0571 20,0023 53,3160 129,8138 18,3463 58,2980 155,9971 14,6650 56,0410 156,8986 23,0799 60,1040 2500 157,3957 19,6802 62,4770 141,3043 17,7884 62,6210 120,0529 5,1197 62,8250 162,7343 28,2411 69,8550 3250 119,7257 16,4935 68,8380 129,3514 17,1078 67,7320 112,9743 16,9257 69,7450 160,5088 14,6057 78,1890 #N/A Speed [rpm] Mt [N.mm] σ Mt [N.mm] σ Mt [N.mm] σ Mt [N.mm] σ 100 304,7667 11,8036 323,1840 16,9684 209,8017 16,3931 210,5033 30,0340 250 207,7400 15,3125 175,0050 9,8073 161,2788 14,9676 181,3471 15,9844 500 195,5400 11,7546 170,2500 12,7078 140,2363 10,1105 182,2633 18,0975 1000 191,8829 6,7254 162,1100 8,6608 139,4386 10,1298 188,9043 15,0271 1400 179,5686 5,2734 144,2680 1,5876 137,1914 15,6061 202,4225 21,0277 1000 124,4371 15,3924 105,1371 13,1715 109,2286 8,0666 141,5620 13,9720 1750 128,1571 5,8263 119,5086 13,5427 108,4460 5,8441 157,7600 27,9370 2500 #N/A #N/A 3250 #N/A #N/A Speed [rpm] Mt [N.mm] σ Mt [N.mm] σ Mt [N.mm] σ Mt [N.mm] σ 100 253,0838 10,5877 233,8967 4,8738 206,4029 12,5247 160,9386 18,5999 250 202,4900 6,4247 179,5413 3,0287 118,4900 10,5505 174,5725 18,9193 500 186,9417 8,5986 163,3840 6,8676 206,4029 12,5247 134,6363 11,6292 1000 174,5014 8,6791 140,2214 8,2288 126,2171 7,7431 135,6100 18,5909 1400 160,2817 9,7114 135,6471 11,4014 98,7811 8,8642 160,3686 22,9127 1000 116,6883 15,0389 104,0533 11,3475 88,2013 6,0391 149,1583 8,5621 1750 138,8271 12,7241 104,6663 9,5171 85,0050 6,5934 79,7263 12,4356 2500 120,9886 15,9783 138,9667 17,7494 92,6475 8,0782 114,9833 29,7367 3250 91,2888 13,9544 94,6929 14,3678 112,6525 10,1517 130,7186 40,8748 60ºC 80ºC M3 MLi M1 M2 M3 MLi Free temperature M1 M2 M3 MLi M1 M2 Fast Transmitio n Slow Transmition Slow Transmition Fast Transmitio n Slow Transmition Fast Transmitio n 114 Speed [rpm] Mt [N.mm] σ Mt [N.mm] σ Mt [N.mm] σ Mt [N.mm] σ 100 254,2300 10,8797 201,4833 10,0856 211,3186 10,3487 245,3843 21,3661 250 224,6643 11,2581 208,2100 10,1449 179,1700 18,8680 233,1750 10,7961 500 234,8071 13,3075 196,6300 2,5943 142,5233 11,6578 242,3900 13,4290 1000 185,3843 16,3570 157,5214 4,4735 111,6243 10,8277 217,5186 27,4484 1400 157,8529 2,1258 156,1929 10,3995 97,7500 9,3224 202,4225 21,0277 1000 132,3038 23,6075 109,2800 11,8458 81,5557 9,1902 134,2475 21,7662 1750 104,2857 7,6191 94,5600 15,4374 71,5986 9,7041 156,8986 23,0799 2500 123,6900 14,8993 118,6700 13,6540 74,6133 3,4327 162,7343 28,2411 3250 129,2125 17,2269 110,3460 17,7334 74,3675 12,2076 160,5088 14,6057 110ºC MLiM3M2M1 Slow Transmition Fast Transmitio n 115 A. Measurements results A.2. Bleed Oil Viscosity Table A.1.: M2 Bleed oil viscosity at 60, 80 e 110◦C. Temperature [◦C] MLi Bleed oil Kinematic Viscosity [cSt] 60 22,6208 80 13,3404 110 7,4374 Figure A.1.: M2 Bleed oil viscosity Table A.2.: M3 Bleed oil viscosity at 60, 80 e 110◦C. Temperature [◦C] M3 Bleed oil Kinematic Viscosity [cSt] 60 23,6664 80 13,5889 110 7,6205 116 A.2. Bleed Oil Viscosity Figure A.2.: M3 Bleed oil viscosity Table A.3.: MLi Bleed oil viscosity at 60, 80 e 110◦C. Temperature [◦C] MLi Bleed oil Kinematic Viscosity [cSt] 20 426,8823 40 135,1716 60 56,6928 80 29,39555 110 14,0873 Figure A.3.: MLi Bleed oil viscosity 117