Environmentally Adapted Lubricants for Arctic thruster system

(1)

Arctic thruster system

Bharath Kumar Sundararajan

Mechanical Engineering, master's level (120 credits) 2017

Luleå University of Technology

Department of Engineering Sciences and Mathematics

(2)

1

(3)

2

Acknowledgements

This project has been carried out as a part of the Joint European Masters on Tribology of surfaces and interfaces. The work was done at the division of machine elements at Luleå University of Technology, Sweden. Though it is an individual project, it wouldn’t have been possible without the immense support and guidance from various people during the course of the project. I would like to take this opportunity to sincerely thank them.

 My supervisor and examiner, Professor Ichiro Minami, Tribo-chemistry, Division of machine elements, Luleå University of Technology for his constant support and encouragement. I have learnt a lot about tribo-testing and lubricant chemistry from him. He constantly motivated me to develop my technical reading, writing and presentation skills and strengthened my passion for research.

 Erik Nyberg, Doktorand, Division of machine elements, Luleå University of Technology for his guidance and training to use reciprocating and unidirectional tribometer.

 Marcus Björling, Researcher, Division of machine elements, Luleå University of Technology for his kind help in conducting the experiments with WAM ball on disc tribometer.

 Leonardo Pelcastre, Post doctor, Division of machine elements, Luleå University of Technology for his guidance and training to use WYKO optical interferometer and Scanning Electron Microscopy.

I would also like to thank all the staff members of Division of Machine elements for their continued support and help during the course of the project. Finally, I would like to thank my parents, my brother and my fellow colleagues and friends from TRIBOS master’s program who were always there by my side.

(4)

3

(5)

4

Abstract

Petroleum based lubricants used in ships often leak into the aquatic environment and affect the ecosystem. To avoid this, Vessel General Permit (VGP) 2013 has been enforced by the United States Environmental Protection Agency (USEPA), according to which lubrication systems in ships have to operate with environmentally acceptable lubricants (EAL).

According to VGP, EAL should contain at least 90% of readily biodegradable substances by mass. Several types of EAL based on synthetic ester fluids are commercially available.

Although the newly developed EALs can meet the standard tests, their performance under practical conditions are yet to be explored. Hence, it becomes essential to understand the pros and cons of using EAL as compared to conventional petroleum based gear lubricants.

Arctic Thruster Ecosystem (ArTEco) is a research consortium supported by ERA-NET MARTEC II project. The consortium comprises of 3 leading universities and 6 industries across Sweden, Germany and Finland. The role of Luleå University of technology is to conduct laboratory tribo-tests to compare conventional (petroleum-derived) lubricants and EAL. Steel ball-on-disc configuration having reciprocating, unidirectional and rolling/sliding motion under EHL to boundary conditions were deployed. The worn surface was studied using optical interferometer and its chemical composition was characterized using the Scanning Electron Microscope (SEM) and Energy Dispersive X-ray Spectroscopy (EDS).

EALs displayed analogous lubricating performance in comparison to petroleum based lubricant under boundary conditions in pure sliding motion. Friction fluctuations and higher wear were observed with EAL at temperatures of 75°C or higher, and resulted in irreversible changes in the oil contents by combination of heating and rubbing. However, pre-heating the EAL under static conditions lead to beneficial effects on the friction and wear performance. On the other hand, EALs displayed better friction and wear reduction than mineral based lubricants under EHL at different slide-to-roll ratios (0%-120%) and entrainment speeds. The results indicate superior rheological properties of synthetic ester fluids over petroleum based fluids. Analysis of worn surface indicated similar morphology with all lubricants, although a slightly higher roughness was observed with EALs. Presence of Phosphorous on worn surface was detected by EDS for all the lubricants. This suggests the formation of protective boundary film to reduce both friction and wear.

(11)

10

(12)

11

CHAPTER 1: INTRODUCTION

1.1. Global carbon footprint

Emission of greenhouses gases has been a hot topic of discussion in the recent years due to its impact on global climatic conditions. Approximately 65% of the total greenhouse gas emissions is composed of CO2 [1]. The carbon emissions and footprint have increased significantly since 1970. The CO2 emissions have increased by a whopping 90% [1]. The major source of CO2 emissions is from petroleum based fuels and products that are being used in various economic industries. Among the industries, transportation sector (road, rail, air and marine) contributes to around 14% of the total greenhouse emissions [2]. Among the various countries, China was found to be the primary emitter of CO2 contributing to approximately 30% of the total, followed by the United States of America with 15% and EU–28 with 9% [1].

The International Maritime Organization in its recent study have stressed on the need to reduce the emissions from ships. IMO has predicted that if the present scenario is left unattended, the CO2 contribution from ships alone would rise to 17%

of the total by the year 2050 [3]. The total carbon emissions are composed of both combustion and non-combustion emissions. Though combustion emissions contribute to the majority, it is important to control the non-combustion emissions as well. Non-combustion emissions include many sources, namely lubricant leakage, methane slip and venting, refrigerants, etc. [3]

1.2. Lubricant Leakage

There have been numerous studies on the lubricant demand in world market. It has been found that approximately 50 million tons of lubricant is being consumed every year worldwide [4]. The share of lubricant in the world market is displayed in Table 1 below. It is clearly seen that America, Asia and Europe are the major consumers of lubricants accounting for one-third of the total quantity respectively.

(13)

12 Asia /

Pacific North America West Europe Rest of

World Reference

37% 28% 13% 23% [4]

Asia America Europe Africa References

26% 32% 34% 5% [5]

28% 33.2% 34% 4.5% [6]

Table 1: Lubricant market share

Approximately 13% to 32% of the lubricant consumed is disposed back into the environment as “waste oil”. In the European Union alone, about 5.6 million metric tons of waste lubricating oils is being generated per year [7]. Among the countries in EU, Germany leads the list with 150 000 tons of lubricating oil being discharged into the environment every year [4]. The leakage of such oil, particularly from ships into water bodies such as freshwater lakes, seas and oceans significantly affect the aquatic organisms like plants and fish [8].

Lubricant leakage into Marine Environment

Majority of the ships operating on oceans and seas require large amounts of lubricant for the huge machineries (like stern tubes) on board as well under the deck. Oil leakage from these machineries have become a huge concern as they pollute the aquatic environment and hence, have become a topic of discussion of late. This kind of oil discharge into the aquatic environment is termed as “Oil Pollution” [9].

European Commission DG Joint Research Centre in the year 2001 [10]

commissioned a study on the Illicit Vessel Discharge in the Mediterranean Sea. It was found that amount of unauthorized operational discharge and leakage of lubricating oil was much more than accidental spills. Stern tube leakage was found to be the major source among all [10]. Another world-wide analysis in the year 2010

(14)

13

[11] estimated the amount of lubricating oil input into marine environment. It was found that on an average around 61 million litres of lubricating oil is discharged into the marine waters annually, of which 11% is contributed from leakage alone as shown in Table 2.

Operational

Discharge Lubricant Leakage Tank ship

spills Others References

50% 10% 19% 17% [10]

52% 11% 20% 20% [11]

Table 2: Lubricant discharge into Marine Environment

This calls for a change in the lubricant used for machineries on ships that can be environmentally acceptable and at the same time provide the same or higher level of performance as petroleum based lubricants. As discussed before, these lubricants must also comply with the VGP 2013 standards as prescribed by the US EPA. One class of lubricants that satisfy these criteria are Environmentally Adapted Lubricants (EAL) [9].

1.3. Environmentally Adapted Lubricants

The Encyclopedia of tribology defines Environmentally Adapted Lubricants or Bio- lubricants as those that are readily biodegradable in nature [12]. However, in order to obtain the VGP certification from US EPA, the lubricant has to satisfy the following three criteria.

1.3.1. Biodegradability

“Biodegradability is a measure of the breakdown of a chemical (or a chemical mixture) catalyzed by micro-organisms” [13]. It can be categorized based on the thermodynamic properties as follows.

(15)

14

“Primary biodegradation refers to the loss of one or more active groups in a chemical compound that renders the compound inactive with regard to a particular function” [14].

“Ultimate biodegradation, also referred to as mineralization, is the process whereby a chemical compound is converted to carbon dioxide, water, and mineral salts” [14].

On the other hand, based on the kinetics or rate of biodegradation, it can be classified into 3 types.

“Readily biodegradable means that some fraction of a compound is ultimately biodegradable within a specific time-frame, as specified by a test method” [14].

“A compound is considered inherently biodegradable so long as it shows evidence of bio-degradation in any test for biodegradability” [14].

Compounds that do not show any signs of biodegradability in any tests are considered to be non-biodegradable. There are several internationally recognized standard test methods for measuring the rate of biodegradability. The most commonly followed is the test method developed by the Organization for Economic Cooperation and Development (OECD): OECD 301-B, which is also termed as “CO2

evolution test” [15].

In this method, a measured volume of test substance (lubricant in this case) is aerated by CO2-free air in a dark environment or in diffused light. The CO2-free air is bubbled at the rate of 30-100 ml/min. The setup is then left for 28 days and the amount of CO2 produced is determined at regular intervals. The CO2 evolved during the process is trapped using BaOH or NaOH and its quantity is obtained by titration of remaining hydroxide. The rate of biodegradation is then directly related to the amount of CO2 produced. The sample is considered to be readily biodegradable if at least 60% of the material is completely converted to CO2 at the end of the 28 day period [15].

(16)

15 1.3.2. Aquatic Toxicity

The second criteria in the list is that EAL should be non-toxic and not affect the living organisms in the aquatic ecosystem. The test methods to check for aquatic toxicity include OECD 201-4 and 209-12. Among these tests, the most common ones are 72 hour algae growth test (OECD 201), 48 hour acute toxicity test for daphnia (OECD 202) and 96 hour toxicity test for fish (OECD 203) [9].

In the OECD 201 test standard [16], the aim is to determine the influence of a substance on the growth of freshwater micro-algae and cyanobacteria. These rapidly breeding test organisms are introduced into the test substance (lubricant in this case) for a period of 3 days (72 hours). Though the test duration is fairly small, the effect can be studied over several generations. The algal biomass (mg of algae/liter of test solution) is then used to quantify the growth and growth inhibition property of the test specimen using a mathematical equation [16].

1.3.3. Bioaccumulation

Bio-accumulation is the build-up of chemicals within the tissues of an organism over time. The longer the organism is exposed to a chemical and the longer the organism lives, the greater the accumulation of the chemical in the tissues [17]. This accumulation of chemicals also magnifies as we go up the food chain. The extent of bio-accumulation of any chemical compound can directly be related to its oleophilicity. Compounds that are lipophilic or hydrophobic tend to dissolve in the fatty tissues and cause bio-accumulation problems.

An EAL must comprise of chemicals that have a faster degradation rate so that their concentration does not build up in the tissues of living organisms. The 2 most common methods for testing the bio-accumulation tendency of EAL are OECD 107 and 117. In these tests, the substance is dissolved in a mixture of octanol and water and the dissolution in each phase is quantified using a numerical formula. As discussed above, substances that have a tendency to undergo bio-accumulation will have a tendency to dissolve more in octanol rather than water [18].

(17)

16

Having discussed about the important environmental aspects that a lubricant must possess, the next important step is to understand how these properties can be related to the base fluid of the lubricant.

1.4. Classification of Base Oil

Lubricants are formulated using one or more base fluids which accounts for 80 to 95% of the total weight, which varies depending upon the application. The percentage of additives can vary accordingly from 1% in some compressor oils and circulation oils to approximately around 20 or even 30% in some Metalworking fluids [19].

API classification of base oil is one of the most widely and commonly used one. All the oils are categorized into five groups as shown in Table 3 [20].

Table 3: American Petroleum Institute classification for base oils

1.4.1. API group I – III: Petroleum based oils

Mineral/Petroleum basestocks are composed of hydrocarbons that are refined from crude oil. They are the most widely class of lubricant for a period of over 100 years. They have a very long history of usage, troubleshooting and research. The processing of crude oil depends on desired quality of the base oil and also the chemical composition [21]. In general, mineral base oils are composed of 18 to 40

Base oil category Sulphur (%) Saturated

hydrocarbons (%) Viscosity Index (VI)

Group I (solvent refined) >0.03 <90 80 to 120 Group II (hydro treated) <0.03 >90 80 to 120 Group III (hydro cracking) <0.03 >90 >120

Group IV PAO Synthetic Lubricants

Group V All other base oils not included in above groups

(18)

17

carbon atoms in four basic hydrocarbon types depending on their structure as shown in Figure 1 [19].

Figure 1: Typical structure in Petroleum basestocks

Paraffin (Figure 1.a. and 1.d.) is the retired term for saturated hydrocarbons or alkanes. Naphthenes (Figure 1.b.) are cycloalkanes, especially cyclopentane, cyclohexane and their alkyl derivatives [22]. Base fluids composed of naphthenes are only used when low temperature properties are required and Viscosity Index is not of significant importance [19]. The main applications where such requirements are found are electric transformers, insulating oil, etc. Another important structure found in petroleum basestocks are aromatics (Figure 1.c.).

Apart from above mentioned hydrocarbon chains, they also contain other elements in trace amounts like Sulphur, nitrogen or oxygen. The final chemical composition of Mineral basestock depends upon a number factors such as the source of the oil, refining technique and its degree and the finishing process. Figure 2below indicates the typical composition of API Group I to III base oils [23] [24]. As expected, the level of saturated hydrocarbons is pretty high in API group III and has a very low percentage of unsaturation.

(a)

(d) (c)

(b)

(19)

18

The properties of the mineral base fluid is largely dependent on the composition of different HC structures discussed above. Table 4 below shows their influence on the viscosity index, oxidation stability and pour point of the mineral base fluids [23].

Table 4: Influence of hydrocarbon structure on mineral base fluid properties

Property Straight

Paraffin Branched

Paraffin Naphthenic Aromatic Viscosity

Index Better Better Poorer Poorer

Oxidation

stability Better Better Poorer Poorer if more

rings are present

Pour point Poorer Better Improves with longer paraffinic

chains

Improves with longer paraffinic

chains Additive

solubility Poorer Poorer Better Excellent

Viscosity Pressure relationship

Poorer Poorer Better Excellent

0%

25%

50%

75%

100%

API Group I API Group II API Group III

0%

25%

50%

75%

100%

API Group I API Group II API Group III

Figure 2: Mass fraction of different components in API Group I-III Source: Stipanovic, A.J. [23] Source: James.E.Anderson et.al [24]

(20)

19

The environmental aspects of mineral oil are also highly dependent on the composition of the fluid. The influence of the different hydrocarbon chains on toxicity and biodegradability is shown below in Table 5 [25].

Table 5: Influence of hydrocarbon structure on Environmental properties

Property Paraffin Naphthenes Aromatics

Toxicity None Lower Higher

Biodegradability Higher Moderately higher Lower

It is clearly seen that aromatics have the worst environment properties, as they are highly toxic and non-biodegradable as well. Thus, higher the aromatic content in the base oil, higher is its impact on the environment.

Very High Viscosity Index Fluids

One way to increase the performance of mineral oils is by the process of hydrocracking which would result in Very High Viscosity Index (VHVI) fluids [19].

These VHVI base fluids belong to API group III. The properties of the base fluid is improved by the above process due to the uniformity in molecular distribution. The color of this base fluid is very light, they have low volatility and a viscosity index of over 120. From environmental point of view, VHVI fluids are very poor as they are non-renewable and non-biodegradable. They are also toxic and pollute the aquatic environment. Hence, they are not suitable to be used as a basestock for EAL.

1.4.2. API Group IV and V

The basestocks from these 2 groups are mostly produced by specific chemical reactions (man-made) or they might be of biological origin. They can be derived from several sources such as petrochemicals, animal and vegetable oils and even coal derived feedstocks [19]. They were developed in the beginning of the 20^th century in countries where there was a lack in supply of mineral oil.

(21)

20

Despite the fact that mineral oils are relatively cheap and most widely used, there are various aspects where they still lack in terms of certain lubricating aspects. The most widely recognized basestocks in this category can be listed as follows

1. Synthetic Hydrocarbons (Poly-α olefins) 2. Halogenated Compounds

3. Silicone Polymers 4. Vegetable Oils 5. Synthetic Esters

6. Polyalkylene Glycols (PAG)

Of the above mentioned basestocks, Synthetic esters, vegetable oils, PAGs and low molecular weight PAOs have been found to be suitable candidates for EAL. The following sections will briefly describe the properties of the above mentioned base fluids.

Vegetable Oils

Before mineral oils came into existence, lubricants were largely derived from animal oils, lards and vegetable oils, which are mostly composed of esters of triglycerides [26]. The main sources of vegetable oils include sunflower oil, rapeseed oil, soybeans and palm oil. Nearly 100 million tons of vegetable oils are produced every year, out of which approximately one percent is used for lubrication purposes [21]. They are often used as chainsaw bar lubricants, straight MWFs, biodegradable grease, agricultural equipment oils, tractor oils, two stroke engine lubricants, etc. [26]

Vegetable Oils have often been considered to be a low cost, biodegradable replacement for mineral oils. They possess high biodegradability and are safe to the environment which makes them suitable candidates for EAL [27]. They have several performance based advantages over mineral oil including high viscosity index, high lubricity and high flash point [28]. They have also been found to perform better at extreme pressures [29]. However, they possess poor thermal oxidative stability and pose foaming problems [30]. Vegetable oils also exhibit poor corrosion protection

(22)

21

[31] and the presence of ester group makes it susceptible to hydrolytic breakdown [32]. Experiments conducted at low temperatures (close to -10°C), have shown vegetable oils to undergo precipitation resulting in flow obstruction, when exposed for a long period of time [33]. This shows that vegetable oils possess poor pour point properties which further limits their practical application.

There are very limited ways to improve these performance related issues of vegetable oils. Studies by Erhan et al, 2006 [30] shows some promising improvements in the above mentioned shortcomings. The thermal oxidative stability and pour point properties were found to be improved by using a proper mixture of anti-oxidants, pour point depressants, poly-α-olefins and high-oleic vegetable oils. The results were validated by comparing it against the performance of commercial vegetable oil lubricants.

Polyalkylene Glycols

PAGs are synthetic lubricants derived by the polymerization of ethylene or propylene oxide. They have alternating ether bonds (instead of hydrocarbon backbone) which makes the molecule highly polar. Due to their high polarity they exhibit good solubility in water, which is advantageous when seen from point of view of biodegradable fluids. However, this might pose problems in several applications as they tend to get contaminated with water easily [34].

It has been observed that PAGs have good viscosity-temperature performance and their performance does not get affected in the presence of water [35]. They also have a low pour point which means that they can operate at low temperatures without any issues [5]. All these significant observations have led to the development of thruster lubricants that are based on PAGs. Despite these advantages, there are quite a few major drawbacks when considering PAG based lubricants. They are not compatible with mineral oil which means that the changeover cost will be high, in fact the highest when talking about EAL [35]. They are also not compatible with varnishes, paints and seals. Though water soluble PAGs are biodegradable, as per the studies by Habereder et al., [36] , they have been

(23)

22

found to exhibit some increased levels of aquatic toxicity which is not a very desirable property for EAL.

Poly-α Olefins (PAO)

The lubricant basestocks of this category are usually hydrocarbons obtained by the process of polymerization of α-olefins followed by hydrogenation [25]. Linear alpha olefins are those with chemical empirical formula CnH2n. The biodegradable PAOs are the lower molecular dimers and trimers of the above mentioned alpha olefins.

Due to the low molecular weight, they have very high volatility which could be undesirable. However at lower temperatures this has proven to be beneficial and they exhibit excellent lubricating properties [34]. PAOs are non-polar in nature, which can pose problems with additive solubility, but it helps them to demonstrate excellent hydrolytic stability unlike synthetic esters.

They also exhibit good oxidation stability with suitable antioxidants and can perform on par with petroleum based oils, due to the absence of any double bonds or reactive functional groups [34]. Nowadays, PAOs are gaining popularity as biodegradable lubricants for hydraulics and engine oils, especially in cold climatic conditions and higher pressures in the order of 7000 psi. Unlike PAGs, they are compatible with mineral oils which minimizes the changeover costs.

Synthetic Esters

Synthetic esters have been in production since 1950s and were commonly used for jet engine lubrication. They are produced by the reaction of alcohol with organic or inorganic acids [21]. This reaction is termed as “Esterification”. As compared to the C-C bonds found in mineral oil, the ester bond linkages are much stronger, leading to excellent viscosity-temperature relationship and thermal oxidative stability [21].

This would generally result in a longer life of the lubricant. However, not all synthetic esters exhibit similar properties. Petterson et al., [37] in his study analyzed the film forming capability of six different synthetic ester base fluids and compared them to mineral oil. Out of the six fluids, 2 of them performed similar to mineral oil and formed a stable film and the rest showed less ability to form a stable

(24)

23

film. This proves that ester cannot be studied or treated as a group that can exhibit similar behavior [37].

Hence, the properties of synthetic esters need to be “tailored” to suit a specific application. But, this would require a deeper understanding on the connection between the molecular structure and base fluid properties. The Table 6 below shows the relationship between the chemical structure and composition of synthetic esters and base fluid properties.

Table 6: Relationship between chemical structure and base fluid properties of synthetic ester

Property Molecular

weight No. of ester

groups Branches Longest

linear chain Structural diversity

Viscosity + + + +

Viscosity Index + - +

Flash Point + + - + -

Pour point - + + - +

Biodegradability - + -

Note: ‘+’ symbol means positive correlation and ‘-’ symbol means negative correlation in the table.

Synthetic esters are a very able replacement to the traditional mineral oils and are most widely preferred as the base fluid for EAL. TMP (Trimethylolpropane) oleate is a commonly used synthetic ester base EAL. The reaction which is used for producing TMP oleate is shown in Figure 3 [38].

(25)

24

Figure 3: Synthesis of TMP oleate

Though synthetic esters are less readily biodegradable, more expensive and exhibit poor hydrolytic stability, they have proved to be much better in terms of thermal oxidative stability when compared to vegetable oils [36]. This means that they would have much longer service life. They are also compatible with mineral oil (unlike PAG) which would mean that the replacement costs would be minimal [29].

They also exhibit good solvency for additives and detergents unlike PAOs, which would enhance the lubricating properties of the base fluid.

All the above factors weigh in favor of them to be used as the base fluid for EAL.

Currently, majority of commercially available VGP certified EAL are formulated using synthetic esters with appropriate additive packages. However, these commercial lubricants have not been tested adequately, in marine gear contact conditions to guarantee effective and safe operation.

1.5. Lubricated contacts

Lubricants are a very vital component in the modern machineries so as to ensure smooth, efficient operation and longer service life. The lubricant usually forms a film between the contacting bodies that could be easily sheared, thus reducing friction and wear. Lubricated contacts between 2 bodies can be categorized into 2 types based on their geometry or shape: Conformal and Non-conformal.

(26)

25 1.5.1. Conformal and Non-conformal contacts

When two surfaces that are in contact fit well into each other geometrically, then it is called as conformal contacts. The apparent contact area is large in this case, which would mean that the applied load is well distributed over a larger contact area in comparison to the lubricant film thickness. Even when the load is increased, the load carrying area would remain fairly constant. Lubrication of such conformal contacts was well explained by Reynolds in his hydrodynamic lubrication theory in the year 1886 [39].

Figure 4: Conformal and Non-conformal contacts

On the other hand, non-conformal contact is found when the two surfaces does not fit well into each other. Hence, in this case the load is distributed over a relatively smaller contact area. In such non-conformal contacts, when there is an increase in load applied, there is a significant increase in the contact area. However, this would still be lower than the lubrication area between conformal surfaces [40].

Lubricated machine components can be categorized into 2 types, the first one where elements slide together and the second having elements that roll against each other. The contact between bodies in sliding motion are designed to have conformal contact. The common examples of such conformal contacts are journal bearings and slider bearings [40]. Machine elements having rolling-sliding motion are designed to have non-conformal contacts [41]. Gears, roller bearings, cams and followers are common examples of such non-conformal contacts.

Conformal

Non-Conformal

(27)

26 1.5.2. Lubrication regimes

Knowledge regarding the different lubricating regimes is very essential and would help in choosing the right type of lubricant for the given contact conditions and configuration. The most appropriate method of determining the lubrication regime is by calculating the lambda ratio, ‘λ’ [42].

Lambda ratio is defined as the ratio of minimum lubricant film thickness to the composite RMS surface roughness. Lambda ratio is often termed as specific film thickness [42]. It can be calculated using Equation 1.1 shown below. The minimum film thickness can be determined using the Hamrock-Dowson equation [40].

𝜆 = ℎ_𝑚𝑖𝑛

√𝑅_𝑞1² + 𝑅_𝑞2² (1.1)

Hydrodynamic lubrication regime

If the lambda ratio is very large (λ >3) [12], it would denote that the film is very thick and the surfaces are completely separated without any physical contact.

Hence, the frictional properties are governed by the rheological properties of the lubricant. This regime is termed as full film lubrication or hydrodynamic lubrication. The friction coefficient in this regime is dependent on the sliding speed, load applied per unit width and viscosity of the lubricant. The relationship is further explained in the Stribeck curve in the following section. This type of lubrication regime can be found in journal bearings.

Elasto-Hydrodynamic Lubrication (EHL) regime

The theory of EHL was first proposed by Ertel in the 1940s [43], where he explained two major benefits of high contact pressure. The first one being the local elastic deformation of the contacting surfaces and the second being increase in the viscosity of the lubricant. These effects were then coupled with the hydrodynamic theory to explain the lubrication mechanism under such high contact pressures.

(28)

27

EHL is usually found between non-conformal contacts, but could also occur between conformal contacts like heavily loaded journal and pad bearings [40]. The film thickness in this regime is very low, ranging anywhere between 0.1 and 1µm.

However, the elastic deformation of the surfaces and significant increase in lubricant viscosity helps in keeping the surfaces apart, thus reducing friction and wear.

Mixed lubrication regime

As the sliding speed decreases and load applied increases, the film thickness and surface roughness lie within comparable dimensions (1< λ<3) [12]. Full film lubrication is difficult to achieve under such conditions and the asperities from opposing surfaces begin to interact with each other. The contact load is now shared partially between the lubricant film and asperities in contact.

Boundary lubrication regime

Under very severe contact conditions, such as low speeds and high loads, hydrodynamic effects of the lubricant no longer play a role and there is no significant lubricant film (λ≤1) [12]. More and more asperities come into contact and the load is completely borne by them. The physical properties of the lubricant such as viscosity and density become insignificant and the chemical properties become more relevant. Lubricants containing additives are usually engineered to physically or chemically interact with the surface to produce boundary films or

“tribo-films”, which help in reducing friction and wear.

1.5.3. Stribeck curve

As seen from the above section, the tribological parameters that have significant impact on the lubrication environment are friction, load (P), lubricant viscosity (η) and relative speed (v). Stribeck in 1902 [44] represented the relation among the above mentioned parameters using the Stribeck curve which is shown below in Figure 5 [45].

(29)

28

Figure 5: Stribeck curve, effect of Hersey number on COF

He showed that the Coefficient of Friction (COF) is directly proportional to the viscosity of the lubricant and relative sliding speed and inversely proportional to the load applied per unit width. This could be mathematically represented by Hersey number [46].

𝐻𝑒𝑟𝑠𝑒𝑦 𝑛𝑢𝑚𝑏𝑒𝑟 =𝜂 ∗ 𝑣

𝑃 (1.2)

As seen from the graph, a lower Hersey number would mean that the film thickness is very low and vice versa. Hence, as the Hersey number increases, there is a sudden drop in the value of COF and a transition to mixed lubrication regime occurs. The rapid decrease could be attributed to the hydrodynamic action of the lubricant film that begins to carry majority of the load. The lowest value of COF usually marks the transition from mixed to hydrodynamic lubrication. There is a minor increase in COF as this transition occurs and is because of the viscous losses [47]. For machine elements operating in EHL conditions, the Stribeck curve might look a bit different (as shown in dotted lines). There is no increase in friction coefficient in this case due to modified thermal conditions and lower shear rates [47].

(30)

29

1.6. Arctic Thruster Ecosystem (ArTEco)

Arctic Thruster Ecosystem is a consortium funded by the European commission under the Maritime Technologies II Research Programme (MARTEC II) in the framework of European Commission ERA-Net scheme. They are focused on the research and development of ship propulsion system under arctic conditions [48].

The consortium consists of 6 leading industries and 3 universities, from Germany, Sweden and Finland, who are jointly taking efforts in developing thruster technology for the future.

From lubrication point of view (one of the prime area of concern for tribologists), the lubricant used in the ship propulsion system developed by ArTEco must be environmentally acceptable. To ensure this, the lubricant must comply with the standards set by The United States Environment Protection Agency (US EPA). These standards/criteria have been clearly defined in the Vessel General Permit [9]

document enforced in the year 2013. In order to achieve this goal, ArTEco partners are actively involved in testing various commercially available oil samples.

1.6.1. LTU collaborations

Being an integral part of this consortium, the role of Lulea University of Technology (LTU) in this project would be to carry out simplified laboratory ball-on-disc tribotesting of different oil samples. The results and findings are then shared with the other collaborating partners. Tampere University of Technology (TUT) are currently testing the same samples using twin disc setup. Technical University (TU) Dresden, another close collaborating member of ArTEco, use FZG gear tester to analyze the same EAL samples. Wartsilla test center at Finland, which has a large propulsion tester, carries out bench tests. Aged mineral oil sample used for the analysis in this work was provided by Wartsilla test center (after 1700 hours of use in the propulsion tester).

(31)

30

1.7. Objectives of the presented thesis

The thesis aims at experimentally investigating the friction and wear properties of different commercially available EAL using test parameters similar to the operating conditions and compare it to the performance of mineral oil. In order to achieve this goal, following steps were adopted.

 Study the friction and wear properties of oils using 3 different ball on disc tribometers.

 Analyze the influence of load and temperature on friction and wear performance of oils.

 Study the difference in tribological performance of oil samples in different lubrication regimes (boundary, mixed and EHL).

 Surface analysis (physical and chemical properties) of worn surface to understand the tribo-chemical reactions.

The following chapters explains the experimental methodology, results obtained and conclusions drawn from the experiments. Chapter 2 describes the oil samples analyzed, test rigs used and test parameters adopted to carry out the experiments.

Chapter 3 outlines the results that have been obtained so far by conducting the above tests. Chapter 4 draws some important conclusions that can be made from the results analyzed.

(32)

31

(33)

32

CHAPTER 2: METHOD

Petroleum based lubricants have a very long history of usage and troubleshooting.

On the other hand, EAL are gaining popularity only in the recent past. Hence, testing them under the conditions in which they would be used becomes a necessity and solid evidences are needed to show that EAL can perform equally or better than currently used mineral oils. This chapter discusses in detail the experimental method and design that have been adopted in order to achieve the same.

2.1. Lubricants

A total of four lubricant samples were analyzed for this thesis. The properties of all the oil samples are listed below in Table 7.

Table 7: Properties of lubricants

Sample code

Viscosity,

mm²s^-1 Flash

point (°C) Pour

point (°C) Description 40°C 100°C

EAL01 150 18.5 >200 <-30 EP gear oils with fully saturated esters and high degree of renewability, VGP approved

EAL02 145 18.6 256 -48 Readily biodegradable oil for gear boxes, roller bearings and slide bearings, VGP approved

PET01 150 15.0 198 -21 EP lubricant for highly loaded enclosed gears and bearings

PET01-A Aged PET01 after 1700h with

propulsion tester un Tuusula

(34)

33

All the lubricants are for gear applications with a viscosity of 150 mm²s^-1 at 40°C.

Out of the 4 lubricants, two of them are VGP certified EAL, formulated using synthetic esters as basestock with appropriate additive package. The remaining two are API group I petroleum based lubricants. Since they are commercial lubricants, their exact chemical composition and additive packages are unknown.

2.2. Tribotesting methods

A tribometer is an instrument or machine that is used for measuring the tribological quantities such as friction force, wear rate and coefficient of friction between two surfaces that are in relative motion with respect to each other. Sensors are deployed to measure the friction force, which is then used to calculate the friction coefficient by comparing it to the applied normal load. The test parameters and configuration could be precisely controlled and designed to simulate the actual operating conditions. The following sections would briefly describe the different tribometers used in this thesis work.

2.2.1. Optimol SRV tribometer

SRV stands for Schwingung Reibung Verschleiβ which when translated from German would mean reciprocating friction and wear. As the name suggests, in this machine a reciprocating sliding motion could be established between the two bodies in contact. A ball on disc configuration is used and the material specifications are shown below in Table 8 [49].

Table 8: Specification of ball and disc

Specimen Dimensions

(mm) C.L.A. Surface finish

(nm) Hardness

(HRC)

Ball Φ10 25±5 60±2

Disc Φ24*7.8 35<C.L.A<50 62±1

(35)

34

The load is applied on the ball which oscillates or reciprocates about its mean position with a constant frequency and amplitude. The disc on the other hand is fixed to a specimen block, which could be heated up to a temperature of 900°C to conduct high temperature tests. Approximately 0.3 cm³of lubricant is supplied in the contact region. The tests are conducted according to the ASTM D6425 standards [49]. The picture of the SRV tribometer at Tribolab facility in LTU and the schematic representation of the contact configuration is shown below in Figure 6.

Figure 6: Optimol SRV reciprocating tribometer at LTU

2.2.2. CETR UMT-2 micro-macro tribometer

In this machine, a unidirectional or rotating sliding motion is established between the two bodies in contact. A ball on disc configuration is used again as in SRV with similar specifications as shown previously in Table 8. The disc is mounted on the lower drive which can be set to rotate at a specific sliding speed. The ball is mounted on the upper specimen holder and remains stationary. The load is applied normally on the ball and the friction load sensor is used to measure the friction force and calculate the coefficient of friction. Approximately 0.5 cm³of lubricant is supplied in the contact region. Since a rotating motion is established in this machine, there is a certain centrifugal force acting that tries to push the oil outwards. Due to this reason, the sliding speeds are kept lower than that in the reciprocating machine.

However, the test duration is adjusted such that the sliding distance remains the

(36)

35

same in both the machines. The picture of the CETR UMT -2 micro macro tribometer at Tribolab facility in LTU and the schematic representation of the contact configuration is shown below in Figure 7.

2.2.3. Wedeven Associates Machine (WAM)

WAM ball on disc tribometer can be used to investigate the tribological properties of the different lubricants in different lubrication regimes, especially under EHL contact conditions. The picture of WAM machine in Tribolab of LTU is shown in Figure 8. A rotating ball is loaded on a disc which is also rotating and the result is a circular EHL contact. Since, both the ball and disc are driven by separate motors, they can rotate independently at different speeds. This helps in achieving a partly sliding-rotating motion. The spindle attached to ball could have a maximum speed of 25000 RPM and disc can rotate at a maximum of 12000 RPM.

The surface speed of the ball, “Ub”, can be calculated from the rotational speed of spindle that is connected to the ball, spindle angle and radius of the ball. Similarly the surface speed of the disc, “Ud”, can be calculated using the track diameter and

Figure 7: CETR UMT-2 micro macro tribometer

(37)

36

rotational speed of the disc spindle. The entrainment speed could then be calculated using the equation below.

𝑈_𝑒=𝑈_𝑏+ 𝑈_𝑑

2 (2.1)

As discussed before, by varying the speed of the disc and ball, a sliding-rolling contact can be established. The slide to roll ratio (SRR) can be defined using the equation below [50].

𝑆𝑅𝑅 =𝑈_𝑏− 𝑈_𝑑

𝑈_𝑒 (2.2)

From the equation, it can be established that SRR of 0 would indicate pure rolling motion and SRR of 2 would indicate pure sliding motion. Any value between 0 and 2 would indicate a partial sliding-rolling motion.

Figure 8: WAM ball on disc tribometer

Three kinds of tests were performed with this test rig. Stribeck sweep is made at SRR of 0.5 for all the lubricants to trace the friction behavior in different lubricant

(38)

37

regimes. Wear test is done to analyze the wear performance of the oil samples. A very high SRR of 1.5 is chosen so that sufficient wear occurs on the disc surface.

Finally, the behavior of friction coefficient with varying SRR is studied. With the WAM machine, it is possible to vary the SRR between -2.0 and 2.0 (positive SRR indicates ball rotates faster than disc and vice versa). When the surface roughness and thermal properties of the disc and ball are similar, then the friction behavior in negative SRR region would be identical to the one observed in positive SRR region as shown below in Figure 9 [47]. Hence, in the mu-slip analysis, friction behavior is traced at positive SRR values, between 0 and 1.2.

Figure 9: Friction coefficient versus positive and negative SRR

2.3. Microscopic techniques

After conducting the experiments in different ball on disc tribometers, the ball and disc were observed under 2 different microscopes to analyze the physical and chemical properties of the wear scar. Optical interferometer/profilometer is used to measure the wear scar diameter and wear rate, whereas Scanning Electron Microscopy (SEM) and Energy Dispersive X-ray Spectroscopy (EDS) are used to analyze the chemical composition of the worn surface.

(39)

38 2.3.1. Optical Profilometer/Interferometer

Veeco WYKO NT1100 optical profiler is used to measure the wear scar diameter and also study the worn surface morphology. There are two possible modes of operation with this profilometer: Vertical Shift Interference (VSI) and Phase Shift Interference (PSI). PSI is preferred for smooth surfaces since it can give sub nanometer level resolution (up to 0.1nm) and for smaller steps (160nm). For rougher surfaces and large steps (up to 1mm), VSI is preferred and a maximum resolution of 3nm can be achieved [51]. For this analysis, all the measurements are made in VSI mode. The picture of the profilometer at Tribolab in LTU is shown below in Figure 10.

Figure 10: Veeco Wyko NT1100 optical interferometer

The main advantage of such optical profilometer is that it is a non-contact technique and the surface that is being measured is not affected (nondestructive technique).

Three dimensional measurements are also obtained in this technique without moving the sample. One of the disadvantages of this techniques is that it might not detect the presence of thin films, as incident light may penetrate through it and the light would be reflected from the substrate.

(40)

39 2.3.2. Scanning Electron Microscopy (SEM)

Unlike Transmission Electron Microscope (TEM), SEM can be used to analyze bulk specimens. The picture of JOEL JSM-IT 300 SEM that is used for this analysis is shown below in Figure 11.

Figure 11: JOEL JSM-IT 300 Scanning Electron Microscope

Electrons are emitted from an electron gun (like tungsten filament) by heating it to a temperature of around 2800K. The electron beam emitted is then accelerated through a vacuum chamber between a cathode and anode (voltage difference varying between 0.1 keV and 50 keV). This electron beam is then focused on the specimen using a two or three stage lens system. The electron beam interacting with specimen surface is termed as primary electrons. This interaction results in interaction volume in the specimen, whose size depends on the diameter and energy of the primary electrons, angle of incidence of primary electrons and also the properties of the specimen being analyzed. The electrons and energy emitted from this interaction volume are detected with the help of detectors [52]. Electrons scattered after interacting with the specimen can be of 2 types: Elastic and inelastic scattering.

(41)

40

Inelastic scattering is a result of interaction of primary electrons with the electron cloud of the atoms on the specimen surface. As a result of this, the primary electrons loses their energy, which triggers the release of secondary electrons (SE), auger electrons, characteristic X-rays and cathodoluminescence [52]. Auger electrons and characteristic X-rays emitted are detected using suitable detectors and are used for qualitative and quantitative chemical analysis of the specimen. Secondary electrons can be used to give the topographical information of the specimen as they are emitted from surface.

When primary electrons interacts with the nucleus of the atom, they scatter elastically. In this case, the kinetic energy of the electrons remain unchanged.

However, the direction of such elastically scattered primary electrons change drastically and are termed as back scattered electrons (BSE). The different interactions that occur at the surface of the specimen are shown below in Figure 12.

Secondary electrons

Backscattered electrons

Characteristic X-rays Cathodoluminescence

Sample

Absorbed current Primary electrons

Figure 12: Electron interaction with specimen

(42)

41

2.4. Test conditions and parameters

2.4.1. Friction analysis

Friction analysis is done at varying load and temperature conditions to test the tribological properties of the lubricants. In the varying load friction analysis, the initial load applied is 100N and the load is increased in steps of 50N until a maximum of 300N. The oil is not heated in this test and is maintained at room temperature. In the varying temperature friction analysis, the load is fixed at 200N.

Oil temperature is initially set at 25°C which is then increased in steps of 25°C until 100°C. The oil temperature is then brought back to 25°C by the same way in which it was increased. For convenience, we term these tests as step load and step temperature tests. The test parameters used are summarized below in Table 9. Table 9: Test parameters for friction analysis

Type of Motion

Test standards ASTM D6425 ASTM G99

Load Applied, N 100 150 200 250 300

Max. contact pressure, GPa 2.1 2.4 2.6 2.8 3.0

Hertzian diameter, mm 0.30 0.35 0.38 0.41 0.44

Frequency, Hz 50 -

Stroke, mm 1 -

Revolutions, RPM - 80 100 120

Track diameter, mm - 12 10 8

Average sliding speed, m/s 0.1 0.05

Test duration, min 50 - 210 100

Sliding distance, m 300 - 1260 300

Oil temperature, °C 25-100 Room temperature

Reciprocating Load

Rotating Load

(43)

42

In case of varying load analysis, though the sliding speeds in reciprocating and unidirectional tribometers are different, the time duration is varied such that the sliding distance is the same, i.e., 300 metres. The varying temperature analysis couldn’t be carried out with the unidirectional machine as there is no heating element.

2.4.2. Wear analysis

For the wear analysis, the load is fixed at 200N. In the reciprocating tribometer, the temperature of the oil is maintained at either room temperature or 75°C. In case of unidirectional tribometer, wear test is conducted only at room temperature. A sliding distance of 540m is fixed for both the tribometers and the test duration is fixed accordingly. The test parameters for wear analysis are summarized below in Table 10.

Table 10: Test parameters for wear analysis

Type of Motion

Test standards ASTM D6425 ASTM G99

Load Applied, N 200

Max. contact pressure, GPa 2.6

Hertzian diameter, mm 0.38

Frequency, Hz 50 -

Stroke, mm 1 -

Revolutions, RPM - 80 100 120

Track diameter, mm - 12 10 8

Average sliding speed, m/s 0.1 0.05

Test duration, min 90 180

Sliding distance, m 540

Oil temperature, °C 25 / 75 Room temperature

Reciprocating Load

Rotating Load

(44)

43 2.4.3. WAM ball on disc tribometer

As discussed in the previous section, there are 3 types of tests that are conducted using the WAM ball on disc tribometer. The test parameters for each type of test is summarized below in the following sections.

Stribeck curve sweep

The Stribeck curve for all the lubricants are plotted by applying a constant load of 600N and varying the entrainment speed between 7.2 m/s and 0.007 m/s. By doing so, a transition from EHL to boundary through mixed lubrication is expected to be observed. The oil bath temperature is maintained close to 80°C. The material specifications of disc and ball and the test parameters used for Stribeck sweep are summarized below in Table 11.

Table 11: Test parameters for Stribeck sweep

Speci me n B all

Material AISI52100 steel

Diameter, mm 20.63

Rq, nm 25

D isc

Material AISI9310 steel

Diameter, mm 110

Rq, nm 600

T es t para me ter s

Load applied, N 600

Entrainment speed, m/s 7.2 - 0.007

Slide to roll ratio, % 50

Oil temperature, °C 80