Wear 376-377 (2017) 822–830
Contents lists available at ScienceDirect
Wear journal homepage: www.elsevier.com/locate/wear
Boric acid as a lubricating fuel additive – Simplified lab experiments to understand fuel consumption reduction in field test Elin Larsson n, Petra Olander, Staffan Jacobson Ångström Tribomaterials Group, Uppsala University, Box 534, 751 21 Uppsala, Sweden
art ic l e i nf o
a b s t r a c t
Article history: Received 16 September 2016 Received in revised form 23 January 2017 Accepted 24 January 2017
In field tests, a boric acid based fuel additive has led to reduced fuel consumption. The reduction was substantial, an average of 6 and 10% in passenger cars and diesel generators respectively. Aiming towards improved understanding of mechanisms behind the fuel saving, three methods to mimic the effect of the additive in the piston-ring/cylinder contact have been evaluated. A reciprocating cylinder/flat configuration with ball bearing steel against grey cast iron was used, and it was lubricated with base oil. The different methods were as following: A) repeated spraying of a small amount of the boric acid solution onto the surfaces, B) predeposition of a boric acid layer on the flat surface and C) a combination of method A) and B). The three methods all showed effects of the additive, spanning from about 20% to 50% reductions (in the latter case, from roughly 0.1 to 0.05 in coefficient of friction averaged over the stroke). The greatest potential of the additive was seen with local coefficient of frictions lower than 0.020 in tests at room temperature with Method C. This means a reduction of around 75% compared to the lowest levels measured for the reference tests run without the additive. The most stable friction test was Method A, where a small amount of boric acid solution was repeatedly sprayed onto the lubricated sliding surfaces. In this type of test, friction reductions of roughly 20% and 40% were found at 100 °C and room temperature respectively. The tribological and chemical mechanisms of boric acid in this test configuration are yet not fully understood and more studies are needed. However, the observed poor stability of the tribofilms containing boron and oxygen complicates such activities. & 2017 Elsevier B.V. All rights reserved.
Keywords: Boric acid Lubrication Friction Fuel consumption Fuel additive
1. Introduction Recent estimates give that one third of the fuel energy in an average passenger car is lost due to friction in the engine, transmission, tires and brakes, and only 21.5% of the energy is used to move the car [1]. For heavier vehicles, as trucks and buses, the same number is 33% and 34% respectively [2]. The transport sector accounts for approximately 20% of the total energy use in the world and 18% of the greenhouse gas emissions [1]. 73% of the total energy consumption in transports is consumed by road transport (light duty vehicles, trucks and buses), 10% by aviation, 10% by marine and 3% by rail [3]. To have a chance to fulfill the 2030 climate and energy framework of the European Commission and the Paris Climate Agreement, consumption of fossil fuels from transports must be reduced. The 2030 climate and energy framework targets greenhouse gas emission reductions of at least 40% (from 1990 levels), at n
Corresponding author. E-mail addresses:
[email protected] (E. Larsson),
[email protected] (P. Olander),
[email protected] (S. Jacobson). http://dx.doi.org/10.1016/j.wear.2017.01.105 0043-1648/& 2017 Elsevier B.V. All rights reserved.
least 27% share for renewable energy and at least 27% improvement in energy efficiency [4]. At the Paris Climate Conference 2015, 195 countries agreed to reduce their greenhouse gas emissions to limit the global warming well below 2 °C (above pre-industrial-levels) [5]. The climate and energy goal may feel difficult to accomplish when considering the expected increase in transportation demand. Even in an optimistic scenario, fossil fuels might remain the primary energy source for transports for the next two decades [3]. However, CO2 emissions can be reduced in several ways, e.g. by using new technologies to reduce friction. Since the friction losses in the piston assembly in an engine accounts for 45% of the total engine friction losses [1], a large amount of work has already been focused on their reduction. Suggested technologies include advanced surface texturing, coatings, lubricants and additives. In their scenario Holmberg et al. [1] estimate that such technologies could enable reductions of CO2 emissions with 290 million tonnes in the next 5–10 years and 960 million tonnes in the next 15–25 years. One candidate for this friction reduction is boric acid (B(OH)3), a solid lubricant that is abundant in nature and thereby cheap. Erdemir et al. [6] showed in 1990 that its layered structure provides lubricity. Low tangential forces are needed for sliding to
E. Larsson et al. / Wear 376-377 (2017) 822–830
occur within the material due to the weak interaction between the layers. Boric acid is also considered as relatively safe and has been used for numerous consumer products as well as in industrial applications. Detergents, insulation, metal working fluids and fertilizers constitute some examples. Some health concerns have been raised due to observed developmental and reproductive toxicity effects when laboratory animals are exposed to high levels of boric acid [7]. However, no such toxic effects have been observed in studies on highly exposed human populations, e.g. workers in mines. Boron compounds in tribology have been reviewed by Shah et al. as well as by Spikes [8,9]. Among many approaches, addition of various concentrations and sizes of boric acid particles to lubricating oils have been tested [10,11] and commercial products with emulsified boric acid particles are available. However, boric acid is not soluble in oil and dispersions of particles are often associated with agglomeration, which so far have rendered such products unsuccessful for use in engines. To prevent agglomeration, surfactants can be added to the lubricant to create a dispersion that is stable over a longer period of time. Such a lubricant was studied by Kim et al. [12] with resulting friction reduction. However, the long-time stability of the dispersion was not investigated. Sawyer et al. [13], instead studied the effect of boric acid powder delivery into stainless steel sliding contacts without oil lubrication. A coefficient of friction below 0.1 was achieved and the wear was reduced by 100 times or more. Another possibility is to use boric acid as a fuel additive, which is the focus of this study. To the best of our knowledge, no similar studies have been performed. The studied fuel additive is a commercial product consisting of boric acid dissolved in a solvent, mainly consisting of ethanol. It is mixed with the fuel to reduce the friction in the engine and thereby decrease the fuel consumption. In field tests, performed by an independent consultancy company, the fuel consumption in cars and stationary diesel generators was reduced with an average of 6 and 10%, respectively, when using the fuel additive [14,15]. The test details are summarized in Section 1.1. Aiming towards an improved understanding of tribological and chemical mechanisms behind the fuel saving, we examined the effect of the fuel additive in a simplified lab test mimicking the line contact of the reciprocating piston-ring/cylinder contact. Three methods to add the fuel additive into the sliding contact were tested. In all cases, a thin film of base oil was used as a simple representation of the engine oil. These simplified tests did not include mimicking of the combustion, gas pressure, fuel presence and oil replenishment. 1.1. Field test details Field tests of passenger cars and diesel generators with and without using the boric acid based fuel additive have been performed by an independent consultancy company [14,15].
823
1.1.1. Passenger cars Five different cars were used, chosen to represent an average of the current vehicle population. Both diesel and gasoline fueled engines, as well as both manual and automatic gearboxes were included. The 469 km test route was designed to represent general European driving conditions, with highways, country roads, mountain roads and urban driving. An experienced test leader controlled and managed the test rounds, vehicles, refueling and adding of the fuel additive. Fuel consumption, speed and weather conditions were measured during the test. Average speed (78.2 km/ h) and driving time was the same throughout the test rounds, as well as the road and weather conditions with temperatures between þ1 °C to þ6 °C. The cars were driven in a caravan, and changed places during the test. Every test round started in a heated garage to avoid cold starts and the coolers and brakes were cleaned before each test. Other factors that were controlled before and during the test include tire pressure, oil condition, etc. All cars drove four rounds without the additive, followed by a running in period and finally four rounds with the additive, which was mixed 1:1000 in the fuel, according to specification. All of the cars exhibited lower fuel consumption when the additive was used, with an average reduction of 6%. 1.1.2. Diesel generators Two Atlas Copco QAX 12–60 (50 Hz) diesel generator sets, placed indoor, were used to test the effect of the fuel additive. Both generators were run at 80% (9 kW), 50% (6 kW) and 30% (3 kW) of their maximum capacity. The power outputs were monitored during the tests. Each of the three operating capacities was first run without the fuel additive in five hours and then with the additive mixed in the diesel (1:1000). Using the fuel additive, the fuel consumption was reduced with an average of 10%.
2. Materials and methods 2.1. Test procedure A reciprocating cylinder on flat configuration was used to represent the piston-ring/cylinder contact. The tests were performed at room temperature and 100 °C. The high temperature was selected to represent a typical temperature at the cylinder wall, close to the top dead center and room temperature was chosen as reference. Three approaches to mimic the effect of fuel additive on the friction in the piston/cylinder contact were investigated, see Fig. 1. They differ with respect to how the boric acid additive is introduced into the tribosystem: A) Repeated spraying of a small amount of the boric acid solution onto the surfaces. B) Predeposition of a boric acid layer on the flat surface. C) Predeposition of a boric acid layer plus spraying, i.e. a combination of A) and B).
Fig. 1. Illustration of the three evaluated approaches. Method A) repeated fuel additive spraying, Method B) boric acid as a predeposited solid film on the flat surface and Method C) a boric acid film plus repeated fuel additive spraying. A thin oil film (PAO8) was predeposited onto the flat surface.
824
E. Larsson et al. / Wear 376-377 (2017) 822–830
In all cases, a thin film of base oil was added before the test. This film is intended to represent the engine oil film distributed by the piston rings over the cylinder surface in a real engine. Two references were used for comparison: D) Thin predeposited oil film. E) Thin predeposited oil film plus repeated spraying of ethanol (95%) onto the surfaces.
continuously measured using strain gauges. Both friction curves for the individual strokes and the average friction for each stroke are logged. In the 100 °C tests, the flat samples are heated through resistive heating of the sample holder. The temperature is measured on the side of the sample and controlled by a feedback loop.
Spraying of ethanol was chosen as a reference since ethanol is the main solvent in the product. The predeposited film of boric acid was prepared by pouring the fuel additive product onto the flat sample and letting the solvent (ethanol) evaporate. Note that the product was used as received for the film, and the spraying and not diluted 1:1000 with a fuel, as in the application. The sample was tilted to pour off excess fluid. The oil film was predeposited on the flat sample in a similar manner, using a mixture of 15 wt% PAO8 in hexane. The solvent quickly evaporates, leaving a thin oil film on the surface. The oil film thickness was estimated to be 5–6 mm from the mass increase of the sample. It should be mentioned that the test procedures are quite sensitive to small variations. Finding parameters that result in acceptable stability and repeatability of the tribological behavior has required extensive work. These procedures include the predepositioning of the oil film, the predepositioning of the boric acid film, the amount of fluid to spray and positioning of the spraying nozzle.
The normal load was 5 N, corresponding to an initial Hertzian maximum contact pressure of approximately 60 MPa. The frequency was 1 Hz (cycle/s), the stroke length 5 cm (hence a cycle length of 10 cm) and the test length 2000 cycles. The spraying fluid was pressurized by N2 at 3 bar and sprayed every 10th second (or every 10th cycle) starting after an initial period of 1000 cycles. All types of tests were performed several times.
2.4. Test parameters
2.5. Surface analysis After testing, samples were analyzed using Scanning Electron Microscopy (SEM), Energy Dispersive Spectrometry (EDS), X-ray Photoelectron Spectroscopy (XPS), and Vertical Scanning Interferometry (VSI). The cylinders and flat samples were cleaned in a bath of hexane for two times five minutes prior to analysis. An ultrasonic bath was used when cleaning the cylinders.
3. Results
2.2. Materials
3.1. Frictional behavior
The flat samples were alloyed grey cast iron cut from a cylinder liner of a marine two-stroke diesel engine. It was ground with SiCpaper (grit size 120 followed by 1000) in a cross hatch pattern (45° against test sliding direction). This procedure was performed to roughly represent the honing process of cylinder liners in combustion engines. The resulting surface had a Ra of 0.32 mm compared to 0.36 mm for an investigated engine cylinder liner reference surface, see Fig. 2. The cylinders had a diameter and length of 10 mm and were made of ball bearing steel (from SKF roller bearings).
The tribosystem is affected when boric acid is introduced and the friction reduction outcome is dependent on both test temperature and on how the boric acid is brought into the contact. The three methods including boric acid resulted in varying friction behaviors, see Fig. 3, spanning from about 20% at high temperature, to 50% reductions (from roughly 0.1 to 0.05) at room temperature. All tests have been repeated several times and the friction curves are not identical. However, to show the trends in a simple way, one representative curve per test is presented here.
2.3. Test setup The reciprocating motion is obtained by a motor running a crankshaft and a connecting rod to a linear bearing where the flat sample is mounted. The stationary cylinder sample is positioned on a lever, where the load is applied. The friction force is
3.1.1. Method A - Repeated spraying of a small amount of the boric acid solution onto the surfaces The most stable friction behavior, regardless of temperature, resulted from repeatedly spraying a small amount of the boric acid solution onto the sliding surfaces (i.e. according to Method A), see Fig. 3. Under these test conditions, the average friction was reduced by roughly 20% at 100 °C and 40% at room temperature. The friction
Fig. 2. Surface height profile (VSI) of a) a flat sample prepared for test (Ra ¼0.32 mm) and b) an engine cylinder liner (Ra ¼ 0.36 mm) as reference.
E. Larsson et al. / Wear 376-377 (2017) 822–830
825
Fig. 3. Friction curves for the three different test approaches A-C adding boric acid and two references. Each value in a curve is the average of a full cycle. a) Room temperature and b) 100 °C. Spraying starts at cycle 1000 for tests with spraying.
Table 1 Lowest coefficient of friction found during individual strokes in tests of Method A – Method E in both room temperature (RT) and 100 °C. Lowest coefficient of friction in test Method
RT
100 °C
A B C D E
0.020 0.035 0.018 0.080 0.075
0.110 0.100 0.080 0.134 0.127
falls already after the first spraying cycle, and continues to fall with further spraying until slowly leveling out. The friction is as low as 0.020 at room temperature and 0.110 at 100 °C at parts of the track, see Table 1. This corresponds to a friction reduction of about 75% in room temperature and 20% at 100 °C, when comparing with the two references (Method D and E). For the two references without boric acid addition (Method D and E in Fig. 3), the friction stays relatively stable over the whole test. The immediate friction reduction once spraying started in Method A (after 1000 cycles) confirms the effect of boric acid as a lubricating additive. This is also illustrated by the 3D-diagrams in Fig. 4, showing the development of friction curves over individual strokes. The friction reducing effect of adding the spray is clear over the whole stroke length, but strongest in the part closest to the spray nozzle. Also note that the friction is substantially higher at and close to the turning points, which of course influences the level of the average values shown in Fig. 3. 3.1.2. Method B and C - Tests with predeposited boric acid film At room temperature, the predeposited boric acid films clearly reduced the friction (Method B and C). The friction starts out low and soon stabilizes at around 0.07, see Fig. 3a. When adding the spraying of boric acid solution at cycle 1000 (Method C) the coefficient of friction decreased further to 0.05. This corresponds to a friction reduction of roughly 50% compared with the reference tests. Also here, friction varies along the track, being as low as 0.018 in some parts of the track, see Fig. 4b and Table 1. In the same type of tests performed at 100 °C, the friction shows a different behavior. It is initially high, around 0.30–0.40, which is higher than for the oil reference (Method D). After only a few cycles the friction falls to values around 0.12, i.e. lower than for the oil reference. The positive effect of boric acid then lasts a few hundred cycles until the friction increases and again becomes
Fig. 4. Coefficient of friction curves for each individual stroke during two room temperature tests. To simplify the graph, the friction in only one direction is presented. The friction is lower on one side of the stroke because the spray nozzle is positioned closer to that side. a) Method A, boric acid spray and b) Method C, boric acid film and spray. The coefficient of friction is as low as 0.020 and 0.018 in some parts of the track in Method A and Method C, respectively.
much higher than for the oil reference. The time period when the friction increased varied among the parallel tests, which is believed to depend on variations of the amount of additive predeposited on the sample. With a higher initial amount, the boric acid lasted longer within the sliding track resulting in longer period of low friction. The optimum amount of additive was, however, not investigated. When spraying starts after 1000 cycles in Method C at 100 °C, the friction becomes very unstable, varying between 0.10 and 0.2 during the first spray cycles and between 0.10 and 0.12 the last 70 spraying cycles.
826
E. Larsson et al. / Wear 376-377 (2017) 822–830
Fig. 5. Typical appearances of wear marks (SEM) on the cylinders after tests performed at a) room temperature and b) 100 °C. Large amounts of boron and oxygen are found within the darker areas in b). The arrow in a) indicates the sliding direction.
3.2. Composition and structure of the tribofilms 3.2.1. Cylinder samples All tests performed at the same temperature resulted in similar appearance of the cylinder wear mark. At the higher temperature,
the marks become darker and more pronounced. Boron and oxygen are found in these darker areas. Less boron was found on cylinders tested at room temperature compared to those at the higher temperature. This is evident from both EDS (performed on samples from all methods) and XPS analysis (performed on samples from Method A, see the following paragraphs for details). SEM-images of wear marks from tests with repeated spraying of boric acid (Method A) at both temperatures are demonstrated as an example in Fig. 5. The appearance of the flat samples on the other hand differs between different tests, depending on how the boric acid is introduced into the tribosystem as well as at what temperature the test was run, see Section 3.2.2. The chemical bonding of elements present at different depths within the wear scars on cylinders tested with Method A at both room temperature and 100 °C was studied with XPS, see Fig. 6. The temperature was found to have a significant impact on the film formation. Boron is present from the surface and down to more than 17 times deeper on the cylinder run at 100 °C compared to that run at room temperature (based on the more than 250 minutes more sputtering using the 200 V argon gun). On the 100 °C cylinder, the outermost surface (bottom of the diagrams) contains chemical species where boron has a lower binding energy than that found at larger depths. The shape of the peaks suggests that they consist of more than one peak, but the interpretation of this shift is not straightforward. The binding energy for boron in boric acid, B(OH)3, should according to references be higher than that of boron oxides (B2O3) [16,17]. Therefore, the gradual shift of the peak with increasing sputter depth suggests that the boric acid content increases with increasing depth. However, it is hard to get a definitive answer since the binding energies given for both boron oxide and boric acid in references, cover a quite wide region (NIST - Atomic spectra database). Also for oxygen, the binding energy is lower closest to the surface, and higher at larger depths. This suggests that further down, oxygen is bonded to boron. On the room temperature cylinder, the outermost layer is similar to that on the 100 °C cylinder, but there is no signal from boron-species at larger depths. Here, the chemical shift for oxygen instead suggests that it is in the form of iron oxide.
Fig. 6. XPS-spectra from different sputter depths of cylinders tested with Method A at room temperature (RT) and 100 °C. The spectra are offset for clarity, starting with no sputtering at the bottom and then 15 min sputtering at 200 V between each step upwards, as indicated by the arrow.
E. Larsson et al. / Wear 376-377 (2017) 822–830
827
Fig. 7. Appearance of the flat samples (SEM) tested using the three methods at room temperature and 100 °C. The darker patches/films contain boron and oxygen (as analyzed by EDS, example given in Fig. 8). The arrow in a) indicates the sliding direction. Method A: repeated spraying of a small amount of the boric acid solution onto the surfaces, Method B: predeposition of a boric acid layer on the flat surface, and Method C: a combination of Method A and Method B. These micrographs were obtained shortly after performing the tests. The appearance as well as the chemistry is changed during storing of the samples.
3.2.2. Flat samples Method A - Repeated spraying of a small amount of the boric acid solution onto the surfaces. When spraying the small amount of boric acid solution (Method A) at room temperature, a more or less fully covering film containing boron and oxygen was formed on the flat surface, see Fig. 7a. The same test procedure performed at 100 °C also results in a film (Fig. 7b), but then showing less boron and oxygen, and mostly located at the high plateaus that have been in direct sliding contact, see Fig. 8. From visual observation, it is evident that some of the sprayed fluid evaporates before it reaches the surface in the 100 °C case, explaining why these surfaces show less film. Interestingly, this temperature dependence is opposite to that on the cylinder side, where less boron and oxygen was found after the room temperature test (as described in Section 3.2.1). Method B and C – Tests with predeposited boric acid film. With the predeposition of boric acid as in Method B and C, a non-fully covering thick film is formed in the contact, see Fig. 7c and e. This film is retained, but flattened during the room
temperature tests. After spraying (Method C), the predeposited boric acid becomes coarser by adding more material to the preexisting network. At higher temperatures, however, the boric acid particles are pushed together into larger patches (Fig. 7d and f). 3.3. Long time stability and temperature dependence of the formed boron and oxygen containing films The stability of the predeposited films was found to depend on the test temperature. The deposited film gradually disappeared during heating as indicated by lower amount of boron and oxygen found outside the track after tests performed at 100 °C compared to those performed at room temperature. This is illustrated in Fig. 9. Further, the long-time stability of boron and oxygen containing films was shown to be poor. Considerably less boron and oxygen was found when the element analysis of the films was performed seven days after testing, see Fig. 10. These aspects complicate the analysis and interpretation of the analysis data. Due to their poor stability, all tribofilms were analyzed in connection with the tribological testing.
828
E. Larsson et al. / Wear 376-377 (2017) 822–830
Fig. 10. Example of EDS spectra of the boron and oxygen containing film on one of the plateaus in Fig. 8. One spectrum acquired directly after test and one seven days later.
4. Discussion
Fig. 8. SEM image with corresponding elemental EDS mappings of oxygen, boron, carbon and iron for the flat sample surface (Method A at 100 °C). Oxygen and boron is found on the sliding plateaus. The arrow indicates the sliding direction.
Fig. 9. Appearance of surfaces (SEM) outside the sliding track (Method B) for tests at a) room temperature and b) 100 °C. The small particles on the surface in a) is the predeposited boric acid. This non-covering film of boric acid is not stable at higher temperatures, explaining why less boron and oxygen are found in b) than in a).
The present results from simplified lab-scale tests support the positive findings from field tests and show the potential of boric acid as a fuel additive. They also provide new experience and some vital knowledge needed for further studies of the important mechanisms and the parameters affecting the friction performance. The tests show that a predeposited boric acid film needs several hundred sliding cycles to reach a low and stable friction (Method B and C in Fig. 3a). This suggests that the as-deposited boric acid particles need repeated tribological contact to become effective friction reducers. This transformation involves deformation and flattening of the initial relatively spherical particles, leading to local patches of tribofilm on sliding plateaus. Addition of boric acid to oil lubricated sliding surfaces resulted in average friction reductions of up to 50% and during some parts of the stroke by as much as 75% (Table 1). These largest reductions were consistently found in the tests run at room temperature. One possible explanation is that a larger fraction of the boric acid solution sprayed onto the sample actually reaches the surface than in the 100 °C tests, since ethanol evaporates at temperatures above its boiling point of 78.3 °C. This explanation is supported by the surface appearances after testing. A more or less fully covering film of boron and oxygen is found on the flat samples tested at room temperature, whereas the same elements are mostly found very locally on the sliding plateaus of the flat samples tested at 100 °C, see Fig. 7a and b. The differences found in tribofilm thickness and tribofilm chemistry between the room temperature and 100 °C cases are interesting. Especially, this includes the fact that less boron and oxygen was found on the cylinder in the room temperature tests, although there was more on the corresponding flat samples. However, at this stage we do not have enough data to fully interpret the implications of these differences or their connection to the resulting friction behavior. Our tests have shown that the mechanisms behind the tribological effects of the boric acid fuel additive are complex and call for further studies. Since this study is unique in treating the effect of boric acid as a fuel additive rather than as a lubricant additive, direct comparisons with published results are difficult. The more detailed analysis and interpretation of the tribofilms is further complicated by their lack of long-time stability. The difference between analyzing the surfaces immediately after the test, or several days later was substantial. Boron that was found present at high amounts when studied directly after the test, could just barely be detected when coming back for further analysis seven days later (see Fig. 10). Further, it is unclear how the films
E. Larsson et al. / Wear 376-377 (2017) 822–830
are affected by electron and X-ray beams as in SEM and XPS respectively. Furthermore, it was found that the stability of the additive itself was reduced at high temperatures, see additive outside the wear track in Fig. 9. This is probably related to results found in studies of boric acid evaporation and sublimation [18–20]. Gaillardet et al. [20] studied the loss of boron by evaporation from different solutions (H2O, HCl, HNO3, HF, acetone, ethanol and methanol) as well as the loss by sublimation when the dry boric acid residue is heated further (60 °C). It was found that losses of boron occurred in ethanol and methanol solutions and that the evaporation is enhanced by these solvents by the formation of a volatile organic complex. Trimethyl borate or triethyl borate were proposed as candidates (B(OCH)3 or B(OC2H6)3). Further, the boric acid is easily lost by sublimation, especially at higher temperatures. Boron losses by sublimation are also probable at 20– 25 °C but at a lower rate. To improve the understanding of boric acid as a fuel saving additive, a more detailed investigation of the stability of the tribofilms is needed. It is important to consider the differences between this simplified lab test and the piston-ring/cylinder-liner contact in a fired engine. Actual engine conditions are always difficult to mimic in a lab test. Here, a reciprocating sliding test was developed to represent some of the important conditions of the piston ring/cylinder contact in combustion engines. The low frequency of the reciprocating motion makes the test probably most relevant for the conditions close to the top and bottom dead center of a real cylinder, where the piston speed is lowest. The high temperature used (100 °C) was selected to represent a typical temperature at the cylinder wall, close to the top dead center. The most challenging condition to mimic may be the dual source lubricating situation in the real engine, involving both the feed of engine oil performed by the piston rings and the feed of the boric acid additive, sprayed into the cylinder as part of the fuel. The high temperatures during combustion and the presence of combustion products such as soot are also difficult to simulate. No attempts to simulate the presence of soot were included in this study. Three methods were designed to find a representative, repeatable and practical way of applying the combination of oil and fuel additive. Two corresponding methods, both including the lubricating oil but excluding the additive, provided reference baselines. In the first three methods, boric acid is added through predeposition and/or spraying of relatively large amounts of the fuel additive at moderate temperatures. In the engine, a small fraction of solution is mixed with the fuel and combusted under high pressure and temperatures. These differences may of course influence both the possibilities to form tribofilms, and their final chemistry and properties. Ideally, tribofilms formed in the test should be compared with those formed in a fired engine to confirm that the test is representative. Some attempts to do this have been initiated, but further efforts are needed. Our recent insights regarding long time stability of the tribofilms elucidate some of the problems of analyzing or even finding these types of films on components from field tested engines, which are typically only possible to analyze relatively long after being tested. Moreover, the tribofilm chemistry in the real application may be affected by additives in the fully formulated motor oil. To reduce complications and the risk for misinterpretations caused by competing additives, the conditions were simplified by using a simple base oil. Another difference is the lubrication regime. According to ‘conventional’ theory, piston-rings operate in the full film regime during the major parts of the stroke, whereas boundary and mixed lubrication operate at and close to the turning points. Furthermore, a major fraction of the energy losses are considered to be caused during the full film conditions, since these prevail over the major part of the stroke. However, the literature does not agree regarding
829
this distribution of energy losses between different lubrication regimes. For instance, in the estimation of Holmberg et al. [1], based on available literature and knowhow from motor specialists, 40% of the losses are attributed to hydrodynamic lubrication (HD) and 40% to elastohydrodynamic lubrication (EHDS), whereas 10% are attributed to mixed and 10% to boundary lubrication. Spencer [21] reviews this topic in his doctoral thesis, where he presents references showing a huge variation in this percentage breakdown and comments that this is natural considering the wide range of vehicles and type of drive cycles investigated, and the long period over which the numbers have been published. Based on the available literature numbers, it seems unlikely that the measured fuel consumption reduction of 6%, or more, could be achieved by an additive that is active only in the boundary and mixed lubrication regimes in the piston assembly. Either the published estimations for the distribution of energy losses are not valid for the conditions of the presented field tests, or the boric acid could have an effect also in the hydrodynamic regime. Alternatively there may be other mechanisms involved than friction reduction, such as enhanced combustion due to the additive, related to the study on the effect of boron nanoparticles by Wang [22]. A long list of friction reduction technologies were proposed by Holmberg et al. [1] as important tools to reduce the fuel consumption due to the friction losses in cars. Surface texturing, lowfriction coatings, etc. are often not suitable for cars already running. However, the fuel additive based on boric acid could potentially immediately be introduced in the whole installed fleet of cars, trucks, buses, stationary engines, etc. This would be a fast way to reduce fuel consumption and CO2 emissions until new technologies for reductions of energy consumption in the transport sector are used in full scale. This could result in a fuel consumption reduction of approximately 6%, based on the field tests of the fuel additive.
5. Conclusions Aiming towards a better understanding of the mechanisms behind the fuel saving effects of a boric acid based fuel additive, a number of simplified lab scale tests have been performed. Three methods were designed to find a representative way of applying the combination of oil and fuel additive. Two corresponding methods, both including the lubricating oil but excluding the fuel additive, provided reference baselines. All tests where the boric acid was added resulted in clearly lower friction than the reference tests, when tested at room temperature. At the 100 °C tests, the behavior was more complex, but all three methods showed lower friction than the references, at least during parts of each test. The effects were spanning from relatively small, up to 50% reductions (from roughly 0.1 to 0.05 average coefficient of friction values over the full cycles). The method based on repeated spraying of a small volume of the boric acid solution onto the sliding surfaces (Method A) resulted in the most stable friction. With this method, the friction reduction was 20% at 100 °C and 40% at room temperature. At room temperature, Method C demonstrated even greater potential of the additive, with local coefficients of friction below 0.020. This means a reduction of around 75% from the lowest levels measured for the reference tests run without the additive. The relation between the present results and the fuel saving effects of the additive found in field tests is not clear. Firstly, it is not known how well the friction values in our simplified test set up represents those in a fired engine. Secondly, the complex relations between the friction losses in the piston ring/cylinder contact and the fuel consumption require a deeper analysis.
830
E. Larsson et al. / Wear 376-377 (2017) 822–830
Microscopic and surface analytic post-inspection of the wear scars revealed boron and oxygen containing films. Their appearances were dependent on temperature as well as on the method of introducing the boric acid into the tribosystem. A more detailed analysis and interpretation of the tribofilms was complicated by their lack of long-time stability. Especially for tests performed at the higher temperature, it was found that the film gradually disappeared. The difference between analyzing the surfaces immediately after the test, or several days later was substantial. This insight may partly explain the problems of analyzing or even finding these types of films on field tested engines.
Acknowledgements This work has been supported by Vinnova VFS and Triboron International AB.
References [1] K. Holmberg, P. Andersson, A. Erdemir, Global energy consumption due to friction in passenger cars, Tribol. Int. 47 (2012) 221–234, http://dx.doi.org/ 10.1016/j.triboint.2011.11.022. [2] K. Holmberg, P. Andersson, N.-O. Nylund, K. Mäkelä, A. Erdemir, Global energy consumption due to friction in trucks and buses, Tribol. Int. 78 (2014) 94–114, http://dx.doi.org/10.1016/j.triboint.2014.05.004. [3] World Economic Forum, WEF Repowering transport Project white paper, Geneva, Switzerland, 2011. [4] 2030 climate & energy framework, Eur. Comm. Clim. Action, 〈http://ec.europa. eu/clima/policies/strategies/2030/index_en.htm〉 (accessed 02.08.16), 2016. [5] Paris Agreement, Eur. Comm. Clim. Action, 〈http://ec.europa.eu/clima/policies/ interntional/negotiations/paris/index_en.htm〉 (accessed 02.08.16), 2016. [6] A. Erdemir, G.R. Fenske, R.A. Erck, F.A. Nichols, D.E. Busch, Tribological properties of boric acid and boric acid-forming surfaces. Part II: mechanisms of formation and self-lubrication of boric acid films on boron- and boric oxidecontaining surfaces, Lubr. Eng. 47 (1991) 179–184. [7] CLH Report For Boric Acid, Bureau for Chemical Substances, Lodz, Poland, 2013.
[8] F.U. Shah, S. Glavatskih, O. Antzutkin, Boron in tribology: from borates to ionic liquids, Tribol. Lett. 51 (2013). [9] H. Spikes, Low- and zero-sulphated ash, phosphorus and sulphur anti-wear additives for engine oils, Lubr. Sci. 20 (2008) 103–136, http://dx.doi.org/ 10.1002/ls.57. [10] M.R. Lovell, M.A. Kabir, P.L. Menezes, C.F. Higgs, Influence of boric acid additive size on green lubricant performance, Philos. Trans. R. Soc. Math. Phys. Eng. Sci. 368 (2010) 4851–4868, http://dx.doi.org/10.1098/rsta.2010.0183. [11] H. Baş, Y.E. Karabacak, Investigation of the effects of boron additives on the performance of engine oil, Tribol. Trans. 57 (2014) 740–748, http://dx.doi.org/ 10.1080/10402004.2014.909549. [12] J.-H. Kim, K.K. Mistry, N. Matsumoto, V. Sista, O.L. Eryilmaz, A. Erdemir, Effect of surfactant on tribological performance and tribochemistry of boric acid based colloidal lubricants, Tribol. - Mater. Surf. Interfaces 6 (2012) 134–141, http://dx.doi.org/10.1179/1751584X12Y.0000000016. [13] W.G. Sawyer, J.C. Ziegert, T.L. Schmitz, T. Barton, In situ lubrication with boric acid: powder delivery of an environmentally benign solid lubricant, Tribol. Trans. 49 (2006) 284–290, http://dx.doi.org/10.1080/05698190600639939. [14] Fuel Consumption Test for Diesel Generators, Bilkonsult Cars & Tech Support AB, 2014. [15] Fuel consumption test, Bilkonsult Cars & Tech Support AB, 2014. [16] C. Higdon, B. Cook, J. Harringa, A. Russell, J. Goldsmith, J. Qu, P. Blau, Friction and wear mechanisms in AlMgB14-TiB2 nanocoatings, Wear. 271 (2011) 2111–2115, http://dx.doi.org/10.1016/j.wear.2010.11.044. [17] B.A. Cook, J.L. Harringa, J. Anderegg, A.M. Russell, J. Qu, P.J. Blau, C. Higdon, A. A. Elmoursi, Analysis of wear mechanisms in low-friction AlMgB14–TiB2 coatings, Surf. Coat. Technol. 205 (2010) 2296–2301, http://dx.doi.org/10.1016/ j.surfcoat.2010.09.007. [18] C. Feldman, Evaporation of boron from acid solutions and residue, Anal. Chem. 33 (1961) 1916–1920, http://dx.doi.org/10.1021/ac50154a040. [19] T. Ishikawa, E. Nakamura, Suppression of boron volatilization from a hydrofluoric acid solution using a boron-mannitol complex, Anal. Chem. 62 (1990) 2612–2616, http://dx.doi.org/10.1021/ac00222a017. [20] J. Gaillardet, D. Lemarchand, C. Göpel, G. Manhès, Evaporation and sublimation of boric acid: application for boron purification from organic rich solutions, geostand, Geoanal. Res. 25 (2001) 67–75, http://dx.doi.org/10.1111/ j.1751-908X.2001.tb00788.x. [21] A. Spencer, A Simulation Tool for Optimising Combustion Engine Cylinder Liner Surface Texture (Doctoral thesis), Luleå University of Technology, 2013 〈http://pure.ltu.se/portal/files/61378545/Andrew_Spencer.pdf〉 (accessed 01.09.16). [22] N. Wang, Boron Composite Nanoparticles for Enhancement of Bio-fuel Combustion (Master thesis), Louisiana State University and Agricultral and Mechanical College, 2012 〈http://etd.lsu.edu/docs/available/etd-08072012172506/unrestricted/wangthesis.pdf〉 (accessed 01.09.16).