Impact of sulphuric acid on cylinder lubrication for large 2-stroke marinediesel engines: Contact angle, interfacial tension and chemical interaction

F.A. Sautermeister n, M. Priest, P.M. Lee, M.F. Fox

iETSI, School of Mechanical Engineering, University of Leeds, Leeds LS2 9JT, UK

a r t i c l e i n f o

Article history:Received 18 December 2011Accepted 4 June 2012

Keywords:Lubrication base oilLubrication emulsionSurface wettingLubricant degradation

a b s t r a c t

The effect of sulphuric acid on the chemical and physical behaviour of the piston ring lubricant in amarine engine cylinder was investigated. To reveal the basic influence of H2SO4 on the lubricant film,the saturated hydrocarbon Squalane (C30H62) was chosen as a simple model oil. The interfacial tensionbetween aqueous H2SO4 (0–98% w/w) and C30H62 was measured between !3 and 165 1C to understanddroplet formation in the lubricant. Interfacial tension decreases with increasing acid concentration andis temperature dependent.

The wettability of engine parts with corrosive sulphuric acid was characterised by the contact angle.The contact angle of H2SO4 (0–98% w/w) on a grey cast iron cylinder liner material (Wartsila, RT84) anda piston ring chrome-ceramic coating (Federal Mogul Goetze, CKS, ø960 mm) immersed in C30H62 wasmeasured over a temperature range from 20 to 165 1C. In general, larger contact angles were measuredunder higher temperature conditions and on chrome surfaces.

In addition to the physical measurements, chemical reaction between H2SO4 and C30H62 wasobserved which influenced the interfacial tension, visual appearance, phase separation and formation ofsolid matter. The reaction time was found to be faster than the neutralisation times of commerciallyformulated lubricants. The reaction products were analysed using FTIR spectroscopy and EDX to findoxidation and sulphonation.

& 2012 Elsevier Ltd. All rights reserved.

1. Introduction

At present, marine diesel engines burn refined but mostlyresidual fuels with a legislated maximum sulphur contentbetween 0.005% and 4.5% w/w sulphur. The intention is to reduceSOx exhaust gas emissions for environmental reasons. The fuelsulphur converts completely during combustion to form 100%sulphur dioxide, SO2. Further oxidation yields 0.3–7% sulphurtrioxide, SO3, which rapidly reacts with water to form sulphuricacid, H2SO4, at various concentrations [1]. Catalysis by vanadiumpentoxide, V2O5, at 450–600 1C increases the yield of SO3 [2].V2O5 arises in marine diesel engines from the combustion ofheavy fuel oil (HFO) with restricted maximum vanadium levels of600 mg/kg [3,4]. These reactions are specified below.

2SO2(g)þO2(g) !!V2O5450!600 1C

2SO3(g)þheat

SO3(g)þH2O(l)-H2SO4(aq)þheat

The resulting sulphuric acid condenses on the cylinder walland the higher the dew point, the more readily the condensationoccurs. The dew point equation of SO3 with H2O according Verhoff

[5] results in a large increase in dew point for small amounts ofSO3. The equation uses the partial pressures p in the gas phase.The resulting condensate is H2SO4.

Td ¼1000

2:276!0:0294$ lnðpH2OÞ!0:0858$ lnðpSO3Þþ0:0062$ lnðpH2O $ pSO3

Þ! "

ð1Þ

Applying typical values for an engine to Eq. (1), dew pointtemperatures Td are found to be substantially below the gastemperature in the cylinder; therefore no condensation occursin the gas volume. But the dew point temperatures are well abovecylinder wall temperatures; therefore condensation will takeplace at the wall. High concentrations of sulphuric acid can bepredicted from the vapour–liquid equilibrium diagram forsulphuric acid [6]. Therefore this study covers the range ofconcentrations from 0% to 98% w/w.

Depending upon concentration and temperature, H2SO4 dis-sociates in water to form HSO4

! , H3Oþ , SO4

2! and pure H2SO4

molecules as two steps [7]

H2SO4þH2O$HSO!4 þH3O

þ

HSO!4 þH2O$SO2!

4 þH3Oþ

The engine components can experience corrosive wear; how-ever, corrosion can only take place when droplets adhere to the

Contents lists available at SciVerse ScienceDirect

journal homepage: www.elsevier.com/locate/triboint

Tribology International

0301-679X/$ - see front matter & 2012 Elsevier Ltd. All rights reserved.http://dx.doi.org/10.1016/j.triboint.2012.06.002

n Corresponding author.E-mail address: mnfasa@leeds.ac.uk (F.A. Sautermeister).

Please cite this article as: Sautermeister FA, et al. Impact of sulphuric acid on cylinder lubrication for large 2-stroke marine dieselengines: Contact angle, interfacial.... Tribology International (2012), http://dx.doi.org/10.1016/j.triboint.2012.06.002

Tribology International ] (]]]]) ]]]–]]]

wall and are not removed by shear in the surrounding liquid.Adherence of a droplet to a wall under shear is quantified by thecapillary number, Ca¼m$U/g. When a critical value of Ca isreached, the droplets deform so extensively that they split orbecome completely detached from the wall. Above the criticalcapillary number, droplets with contact angles between 601 and901 will split leaving smaller droplets on the wall, while dropletsof 1201 are likely to detach from the wall completely [8]. Thecontact angles found in this study cover the whole range,depending on the surface material and acid concentration. Whilstthe dynamic viscosity, m, of the bulk liquid is known, the viscosityof the droplet has no influence on droplet displacement [9]. Thevelocity experienced by the droplet is given by the variation inpiston speed and droplet size.

Typical droplet diameters found in used engine lubricant arebetween 5 and 30 mm in stable emulsions [10]. Under ‘dry’ambient conditions water content can reach 0.5–0.65% w/w andunder tropical conditions a higher 0.75–1.3% w/w [10]. Water-in-oil emulsions of up to 45% w/w in marine cylinder lubricants arestable mainly due to the sulphonates contributed by the deter-gents which significantly lower the interfacial tension [11]. It isvery difficult to prove the influence of droplets on lubrication in alaboratory environment using conventional tribometers as thedroplets are unlikely to enter the lubricated interface [12].

The cylinder liner material for large 2-stroke marine dieselengines is a sand-cast grey cast iron and this study used a typicalcylinder liner sample from a Wartsila RT84 engine. A descriptionof corrosive attack of this material by sulphuric acid is given in[13]. The wear rates observed at Top Dead Centre for the latestlarge bore engines are around 0.01 mm/1000 h after 6000 h [14].

The coating materials used for the piston rings are more diversebut this study concentrates on the aluminium oxide filled galvanicchrome coating ‘‘CKS’’ from FM-Goetze [15], as predominantly usedin Wartsila 2-stroke engines. In the plating process, hot dilutesulphuric acid is used to etch porosities to give improved oilretention of the surface [16]. It therefore seems strange to findchrome-plated parts in an environment attacked by sulphuric acid.However, service experience with chrome plated piston rings andpiston ring grooves in marine diesel engines is good although wearrates are high compared to other heavy duty engines burning lowsulphur fuels. The top ring wear rates for the latest large boreengines are approximately 0.02 mm/1000 h after 6000 h [14].

The neutralisation of the sulphuric acid, formed by combus-tion, is taken care of by an alkaline reserve in the lubricants withtotal base number, TBN, levels of 40–70 mg KOH/g. The alkalinereserve is usually added into the lubricant by dispersing calciumcarbonate, CaCO3, particles as reversed micelles. An excess ofsurfactant disperses solid particles and acid droplets and‘‘form(s) a surfactant monolayer on the oil–acid interface’’ [17].‘Docking’ an additive-containing inverse micelle on to the mono-layer at the oil–acid interface forms a channel through which acidcan enter the micelle and react with the neutralising agent. Whena critical amount of surfactant is reached on the oil–acid interface,the inverse micelles now containing the neutralisation productscan leave the oil–acid interface. The adsorption of fresh unreactedmicelles was found to be rate limiting and increasing the micelleconcentration or TBN decreased the neutralisation time [17].Adding dispersants dramatically reduced the neutralisation effi-ciency of nitric acid due to blocking of the interface by thedispersing additives. For sulphuric acid, the dispersing additivescan transport the hydrated calcium sulphate crystals away fromthe interface into the lubricant. This is a reaction product whichblocks the interface and is removed by the dispersant, increasingthe neutralisation efficiency again to some extent [17]. Neutrali-sation times of sulphuric acid in fully formulated cylinder oils at200 1C are between 10 and 15 min [18].

1.1. Deposit formation

Other than corrosion, marine diesel engine cylinder systems alsosuffer from deposit formation arising from deposition of unusedadditives and lubricant degradation by fuel impingement [19].

Two stroke marine diesel engines are typically lubricated withAPI Group I base mineral oils [1], the least refined lubricantscontaining o90% w/w saturated hydrocarbons (alkanes) withthe remainder formed from unsaturated hydrocarbons such asalkenes and alkynes. The chain length of these oils lies betweenC15 and C50 [20] and for this study a C30 alkane was chosen astypical of this range.

The exothermic reaction of unsaturated hydrocarbons withSO3 and H2SO4, yields hydrocarbon sulphonates and water [21].Lubricant sampled from the engine, has been found to bedegraded by water droplets [10] and solid carbon [18]. The ratherinert alkanes can also undergo various reactions with sulphuricacid. When the methyl group, !CH3, moves in the molecularchain during isomerisation, a process used in refineries, a hydro-carbon with the same molecular weight but different properties isformed. Induction times for the reaction to take place of severalhours were observed for low temperatures but the time decreasedfor temperatures between 60 and 70 1C. The induction time wasexplained by the increasing amount of carbocations, which are anoxidation product of hydrocarbons, RH, by H2SO4. The inductiontime was eliminated by adding RH’s which oxidise easily in H2SO4

or by adding strong oxidising agents like chromium trioxide, CrO3.The surrounding atmosphere has no influence on the reactionproducts but the reaction products differ depending on acidconcentration and process temperature [22]. Possible isomerisa-tion products depend on the carbon chain length and the C30H62

used in this study has more than 4 billion possible isomers, whichmakes an analysis or prediction unfeasible.

In addition to isomerisation, alkane oxidation can also takeplace. Metal ions can act as catalysts of hydrocarbon oxidation;for example CrVI shows a higher oxidation potential than PtII–IV

which is more oxidising than pure H2SO4. With increasingsulphuric acid concentration the oxidation rate increases and ishighest for pure sulphuric acid [23]. In air, alkanes such as thesqualane used in this work are rather inert to oxidation. After 3 hat 150 1C a measurable mass loss due to evaporation of lightproducts occurs and a considerable increase in heavy products isobserved after 48 h at 180 1C [24].

Instead of an isomerisation reaction or oxidation, where C–Hbonds are broken, sulphonation can also take place where atertiary H-atom is replaced by a –SO3H group to form a sulphonicacid, which associate in aqueous solution to form micelles [25].

CnHnþ2þHOSO2OH-CnHnþ2SO3HþH2O

In industrial processes, concentrated and even fuming sulphuricacid are used in order to obtain a high yield of products. Theincreased formation of water slows the process down or even stopsit completely and it must be removed.

Oxidation in marine lubricants is seen as a plain thermal effect.Test methods introduce heating of the oil for extended time totemperatures well above 300 1C and evaluation of deposit forma-tion and viscosity increase [18,25]. This is surprising becausewhere low sulphur fuel is used for comparative engine tests asignificantly smaller deposit formation in the piston ring pack isobserved [37,26].

Surface active agents or ‘surfactants’ usually lower the inter-facial energy between immiscible substances by arranging them-selves at the interface. A surfactant needs a hydrophilic anda oleophilic side to affect a water/oil interface. Hydrocarbonsulphonates are a typical group of surfactants acting at theoil–water interface, consisting of a polar head group which

F.A. Sautermeister et al. / Tribology International ] (]]]]) ]]]–]]]2

Please cite this article as: Sautermeister FA, et al. Impact of sulphuric acid on cylinder lubrication for large 2-stroke marine dieselengines: Contact angle, interfacial.... Tribology International (2012), http://dx.doi.org/10.1016/j.triboint.2012.06.002

attaches to the water and a long hydrocarbon tail which remains inthe oil phase. When the surfactant molecules reach a certainconcentration at the interface, the interfacial tension is not loweredfurther, reaching a plateau, and thereafter micelles are formed; the‘critical micelle concentration’ or CMC. Micelles are droplets of oneliquid in another, stabilised by a surplus of surfactant. Dependingon their concentration, micelles can arrange in many formations, insome cases double layers of surfactants are formed on the interfaceand the interfacial tension drops for a second time, a point calledCMC II [27]. When this second drop in interfacial tension is found,the first CMC is named CMC I.

2. Materials

For interfacial tension measurements squalane (99% SigmaAldrich) and sulphuric acid (495% Fisher Scientific UK Ltd.analytical reagent grade) were used. Lower concentrations ofsulphuric acid were made with deionised water by weight.

Surface samples were prepared from a galvanic chrome-coatedpiston ring (Ø960 mm, CKS, FM-Goetze) and a grey cast ironcylinder liner segment (RT84, Wartsila Switzerland Ltd.). Thesamples were hand-ground and polished with a 15 mm BuehlerDiamond suspension and then cleaned in an ultrasonic bath witha Decon90 solution and then deionised water for 10 min each.After cleaning the samples were dried on a hot plate at 150 1C for5 min and oiled with squalane after a short cooling down periodto protect the surface.

3. Methods

3.1. Interfacial tension measurements

A standard Du-Nouy ring surface tension balance (WhiteTorsion Balance) [28] was modified to incorporate a temperaturecontrolled hot plate under the beaker containing the liquid. Thecorrect working of the arrangement was verified by surfacetension measurements on distilled water.

Measurements under three different conditions were taken.First the oil was added to the top surface of the (more dense) acidat room temperature and directly measured to see the timedependent development of interfacial tension. Secondly the samesamples were heated to 90 1C and then cooled down to measureat room temperature when all reactions between oil and acidwere finalised. Thirdly a new batch of samples was prepared byadding oil and acid at room temperature and stored for at least12 h. The samples were then cooled in a refrigerator and mea-sured until room temperature was reached. The temperature wasthen increased in steps of 15 1C and each step was held for 30 minto allow for stabilisation before measurements were taken. AK-type thermocouple (NiCr–Ni) was directly placed in the oilphase to avoid corrosion.

The platinum ring circumference was 40 mm. The innerdiameter of the sample beaker was 41 mm which defined thereaction area between squalane and sulphuric acid. The platinumring was cleaned after each measurement, following the order ofsolubility in deionised water, isopropanol and heptanes. Betweeneach step the ring was dried on laboratory tissue and finally it washeated until it glowed light red in a blue gas flame.

3.2. Contact angle measurements

The contact angle of various acid concentrations on surfacessubmerged in squalane was measured by the ‘pendant’ or ‘sessile’drop method [29]. To avoid corrosive attack the steel needle of a

glass syringe containing the aqueous acid was replaced by a 1 mmglass capillary with one tapered end to create a tight fit in thesmall hole of the syringe. A custom-built oil bath with parallelglass windows was placed on a temperature controlled hot platemounted on a 3-axis movable stage and the surface samples wereplaced in the oil bath. The evaluation was done by ImageJsoftware plug-in [29]. The correct operation and accuracy of thewhole arrangement was shown by measuring 1 mm diametersteel balls located in various smaller holes in a calibration plate.

3.3. Analysis of reaction products

Mixtures with acid concentrations of 0%, 0.5%, 10%, 25%, 40%,65%, 80% and 90% w/w were stored for several days between 80and 110 1C, 15 1C below their boiling point. The mixtures werefreshly prepared and out of contact with any metal that could actas a catalyst.

Solid reaction products were separated and neutralised inwater, washed with isopropanol and heptanes and dried at60 1C prior to analysis by Fourier transform infrared spectroscopy(PerkinElmer Spotlight 400-Spectrum100 FTIR microscope) andwith attenuated total reflectance (ATR) accessories.

Visual examination of the solids was achieved with a reflectedlight microscope (PolyvarMET) and an SEM (Philips XL30 withtungsten source operating at 20 kV, working distance of 10 mm).The SEM included EDX analysis capability for elemental identifi-cation in the near-surface layer (Oxford Instruments INCA Systemwith X-sight SiLi detector). To gain greater depth of field from thelight microscope, stacking of several pictures with incrementalfocal lengths was achieved using image analysis software [30]. Formeasuring droplet sizes in liquid samples a Zeiss Axiovert135TVtransmission microscope was used.

For FTIR analysis, acidic or aqueous liquid samples were placedin a CaF2 transmission cell, which were found to be resistant toconcentrated sulphuric acid and for the discoloured oil, KBr cellswith a 0.05 mm path length were used. The droplets observed inthe oil phase were small and stable and no chemical reaction withthe KBr was observed. The background scan was done with theempty cells in place. The FTIR ATR with diamond covered ZnSeoptics was used for acidic samples.

4. Results and discussion

4.1. Interfacial tension and chemical interaction

For the time variation measurements as described in Section3.1 it was found that the interfacial tension for all aqueous acidconcentrations against squalane at 20 1C decreases over time andstabilises after approximately 16 min. This behaviour was notobserved for pure water against squalane and the value found isin good agreement with Antonov’s rule [31]. The higher the acidconcentration, the sharper the decrease and the lower the result-ing values, Fig. 1.

The most plausible explanation for the decrease in interfacialtension with time is the reaction of squalane with the acid to formsurfactants which form a monolayer, micelles and consequentlymultiple layers. Applying the general rule that chemical reactionrates at least double for every 10 1C increment, the reactionbetween oil and acid at 200 1C would be complete after 0.22 s.A couple of minutes seem to be more realistic but it is stillfaster than the neutralisation time for commercial oils withconcentrated sulphuric acid at 200 1C with stirring, which is10–15 min [18].

The results obtained for equilibrium measurements asdescribed in Section 3.1 are shown in Fig. 2 (s H2SO4/C30H62).

F.A. Sautermeister et al. / Tribology International ] (]]]]) ]]]–]]] 3

Please cite this article as: Sautermeister FA, et al. Impact of sulphuric acid on cylinder lubrication for large 2-stroke marine dieselengines: Contact angle, interfacial.... Tribology International (2012), http://dx.doi.org/10.1016/j.triboint.2012.06.002

At this temperature and time-range, discolouration was observed forthe acid phase for highly concentrated sulphuric acid of '98% w/w.The discolouration commenced at approximately 70 1C, a valueconfirmed in the literature [32].

All mixtures below 40% concentration remained clear, the40% w/w mixture did not exhibit discolouration but insteadsome solid formation and slight clouding and all mixtures above40% w/w showed discolouration.

The results for the interfacial tension measurements in Fig. 2(s H2SO4/C30H62) show a sharp decrease in interfacial tension fora very low concentration (0.5% w/w) of sulphuric acid. From 0.5%w/w to 25% w/w a more shallow drop is observed, followed by aconstant level of interfacial tension between 25% and approxi-mately 65% w/w. A further decrease occurs until 90% w/w,where the lowest values are reached. The error bars include allpossible sources of error and do not reflect the repeatabilityof the measurements which showed a standard deviation of70.0002 N/m. The decrease in interfacial tension with increasingconcentration is a well established observation in interfacialscience. Usually the Langmuir–Szyskowski equation [33] is usedto describe interfacial tension on the oil–water interface for

known surfactants. At higher concentrations, more but unknownsurface active molecules can be formed, leading to the drop ininterfacial tension.

Also shown in Fig. 2 are the chemical species formed as acidconcentration is increased [7]. Point A in Fig. 2 marks the sharpdecrease in interfacial tension, which is explained by the largedifference in molecular concentration ratio between squalane, at1.92 mol/L, and sulphuric acid, 18.64 mol/L. This is illustratedfurther at point B where the HSO4

! ion reaches a concentration of1.98 mol/L which means that for every molecule of C30H62 at theinterface, there is a molecule of HSO4

! available. At point B, theinterface becomes saturated and the interfacial tension remainsconstant as the concentration of HSO4

! increases further. In colloidscience, this point is called ‘CMC I’. At point C the second dissocia-tion step reaches a maximum and from this point onwards as theconcentration of sulphuric acid increases, the second step ofdissociation into SO4

2! becomes increasingly suppressed. This effectincreases the rate of HSO4

! ion release and the interfacial tensionbegins to decrease slowly. At point D the number of acid species isgreater than the number of free H2O molecules and a second sharpdecrease in interfacial tension is observed. At point E the availableHSO4

! ions are in equilibrium with free H2SO4 molecules andinterfacial tension reaches its minimum. Finally, a slight increasein interfacial tension is observed when squalane is exposed to themost concentrated sulphuric acid.

Figs. 3 and 4 show the results for the temperature dependentinterfacial tension measurements as described in Section 3.1.

Discolouration and formation of solid particles was observedin solutions down to 40% w/w acid concentration when stored forseveral days at 90 1C. During the temperature dependent inter-facial tension measurements, discolouration of the samples con-taining higher acid concentrations was also observed. Themoment of discolouration was captured during the interfacialtension measurements for 80% w/w and 65% w/w acid concentra-tion, Fig. 3. During the measurements and before discolourationoccurred, the interfacial tension was directly correlated to thetemperature of the system and followed the upper values. When acertain high temperature was reached, or a lower temperaturewas held for several days, the sample started to change colour andthe interfacial tension decreased significantly. After the decrease,

Fig. 1. Time dependent development of interfacial tension between aqueousconcentrations (% w/w) of H2SO4 and C30H62.

Fig. 2. Results of interfacial tension measurements between squalane and sulphuric acid (s C30H62/H2SO4) at 21 1C in comparison to molecular concentration of species inthe acid phase, according to [7].

F.A. Sautermeister et al. / Tribology International ] (]]]]) ]]]–]]]4

Please cite this article as: Sautermeister FA, et al. Impact of sulphuric acid on cylinder lubrication for large 2-stroke marine dieselengines: Contact angle, interfacial.... Tribology International (2012), http://dx.doi.org/10.1016/j.triboint.2012.06.002

the correlation between interfacial tension and temperature waschanged irreversibly, as shown by the lower values in Fig. 3.

The upper values, without discolouration, were measured fornine acid concentrations and for each a closest fit function wasfound, which can be found in appendix A and are illustrated inFig. 4. For comparison and orientation the 21 1C values from thestirred and heated trial are added in the plot as single points.Measurements of the two different treated sample batches showgood agreement at most concentrations but are higher for 0.5%,65% and 80% in the case of the samples that were only stored forextended time at room temperature.

The interfacial tension between squalane and deionised waterdecreases linearly with increasing temperature but also is moreshallow than the linear surface tension of pure water and puresqualane for which the linearity is described by the Eotvos rule[28]. For all solutions containing sulphuric acid, the interfacialtension rises with temperature increase above approximately20 1C. For 0.5% w/w and 80% w/w solutions, the interfacial tensiondecreases disproportionately at temperatures below 20 1C. All

other solutions show a contrary disproportionate increase ininterfacial tension at temperatures below 20 1C.

The temperature–interfacial tension correlation for all concen-trations of sulphuric acid was non-linear. A similar fishhook orJ-shape behaviour was observed at the interface between waterand liquid paraffin [34] when POE (polyoxyethylene) stearyl etherwas dissolved in the water phase. This behaviour was alsoconcentration dependent. The authors explained the temperaturedependence in terms of the pulling forces on the polar headgroups of the surfactant. The hydrophilic head group and theoleophilic tail find a concentration dependent energetic idealposition at a certain temperature, where the interfacial tensiontherefore is at a minimum. When the temperature rises or falls,the head group changes its position to the oil or water side andthereby increases the interfacial tension.

No correlation was found between the temperature dependentextent of dissociation of H2SO4 in water [7] to the temperaturedependency of interfacial tension. This can be explained by achemical reaction which binds the HSO4

! ion strongly to the oilphase and excludes it from the dissociation process.

The change in colour and creation of solid particles at hightemperature or extended time of exposure and irreversible shiftin interfacial tension indicate a chemical reaction within thesystem which produces surface active substances. EDX analysis,Section 4.4, revealed both sulphur and oxygen in a dry reactionproduct. This indicates that either HSO4

! , SO42! or both reacted

with the squalane. The solid sample was formed at 80% w/w(14 mol/L) sulphuric acid concentration and 165 1C, where HSO4

!

is the predominant species in the acid and therefore the onlyavailable sulphur donator. The sulphur to oxygen ratio shouldtherefore be 1:4 based on the species but the EDX analysisshowed a ratio of 1:8. Although the EDX technique is moreprecise for heavier elements like the sulphur, with an error ofabout 72–5%, and less accurate for lighter elements like carbonor oxygen, with an error of about 710%, this possible errorcannot explain the surplus of oxygen. Therefore either 50% of thesulphur left the reaction in liquid or gaseous (H2S, SO2) form or50% of the oxygen originates from dissolved gaseous oxygen inthe liquid or was added from other liquid species, which areH3O

þ or molecular H2O which is available in excess of 5.07 mol/Lat 80% w/w, see Fig. 2. Experiments in the oven at lower

Fig. 3. Temperature dependent interfacial tension of 65% and 80% H2SO4 towards C30H62.

Fig. 4. Concentration and temperature dependency of the interfacial tensionbetween sulphuric acid, H2SO4, and squalane, C30H62. The values from Fig. 2 areadded for comparison.

F.A. Sautermeister et al. / Tribology International ] (]]]]) ]]]–]]] 5

Please cite this article as: Sautermeister FA, et al. Impact of sulphuric acid on cylinder lubrication for large 2-stroke marine dieselengines: Contact angle, interfacial.... Tribology International (2012), http://dx.doi.org/10.1016/j.triboint.2012.06.002

temperature over a longer time revealed that the formation ofsolids is concentration dependent and not due to dissolvedoxygen. Furthermore, no reaction between water and squalanewas observed. Formation of H2S would give a typical odour of fouleggs which was not observed but after storing the samples forseveral months a typical SO2 odour was evident. The oxidationpeak found in the FTIR spectrum reveals an extra sulphurindependent CQO bond, Sections 4.3.1 and 4.3.2, and thereforeit is likely that the squalane reacted with both HSO4

! and H3Oþ .

The surface tension of squalane to air was unaffected bychemical contact with sulphuric acid and was in good agreementwith values found in literature [35]. This indicates that anysurface active substances formed act only upon interfaces topolar fluids and not with the non-polar air.

4.2. Contact angle

Contact angle measurements were made for a piston ringchrome coating (CKS) and for a cylinder liner grey cast iron(RT84). Advancing angle, receding angle and static angle weremeasured. Advancing and receding angle were not averaged, as isnormal practice, because the parts in the engine do move and tounderstand transport phenomena it is important to know thedifference. The full results are given in appendix B.

In general it can be said that observed contact angles for the CKSpiston ring coating were larger than for the RT84 cylinder linermaterial. Receding angles yR are less than static angles yst, which inturn are less than advancing angles yA. In most cases, the contactangle is larger at higher temperatures, but not for all cases:

( yCKS_A: stays high between 0% and 80% w/w (H2SO4) with asharp drop at 98% w/w acid concentration.

( yCKS_R: unstable between 0.5% and 40% w/w but higher thanbetween 65% and 98% w/w where it is stable.

( yCKS_st: higher between 0% and 40% w/w than 65–98% w/w.( yRT84_A: high for 0% w/w, rising from low 0.5% and 80%, lower

than 0% w/w for 90% and 98% w/w.( yRT84_R: stable low between 0.5% and 98% w/w, lower than for

pure water.( yRT84_st: high for 0% w/w, stable low for 0.5–98% w/w.

For low interfacial tension one would expect a low contactangle but for the chrome surface the contact angles stayed stableat a high level for most of the acid concentrations.

With the critical capillary numbers Ca for droplet displacementgiven by [8], the measured contact angles y and interfacial tension g,it is possible to predict where droplets will remain stable on thecylinder wall. The velocities acting on the acid droplets are given bythe piston speed, oil film thickness and droplet diameter. Assumingplanar parallel plates, the velocity profile in the oil film is linear andtherefore the absolute shear stress is equal near both walls.

Latest measurements [38] and theoretical studies [15] showgood agreement in the range of minimum oil film thicknessbetween piston ring and liner. Minimum film thicknesses atTDC are between 2 and 5 mm and at higher piston speed reach10–12 mm. Typical droplet diameters found in engines arereported [10] to be between 5 and 30 mm. In this paper dropletsin the oil phase were found to be 1–8 mm diameter (Table 1).

The critical speed for droplet displacement from a wall [8]calculates to

Ucrit ¼Cacrit $ g

m ð2Þ

At 100 1C for 40% w/w H2SO4 concentration the followingvalues apply.

The maximum piston speed in the engine is 16m/s resulting inthe before mentioned 10 mm oil film. Assuming a ‘moving plate onstationary plate laminar couette flow’ for the piston ring-liner contact,the velocity distribution in the oil film is linear. For this case, thecritical speed for droplet displacement is reached at 1.4 mm from thechrome coated piston ring and at 2.7 mm from the grey cast ironcylinder liner wall. The volume of the droplets depends upon theirshape and the contact angle. The volume on the cylinder liner can bemuch greater than on the piston ring. At the piston reversal points,the droplet size can become bigger. This approximation illustratesthat small acid droplets can be picked up and transported by thepiston ring. In non-wetted, unsheared areas of the piston ring profile,the droplets may be bigger. For piston rings with asymmetric barrelshaped profiles, typically of axial piston ring heights of 20–24mm inmarine engines, one would expect to find acidic attack on the upperhalf of the ring profile.

Gas velocities in the piston ring gap are much higher than thepiston speed. However, the gas temperatures are well above theboiling point of sulphuric acid for the top ring and in additionthe influenced surface area is very small compared to the maincircumference of the piston ring. Therefore the piston ring gap isnot included in this consideration.

4.3. FTIR analysis

4.3.1. FTIR analysis of liquidsReacted, brown squalane which had been exposed to 80% w/w

H2SO4 at 165 1C for 15 min was analysed by transmission FTIR.Comparison with a fresh unreacted sample of squalane showedpeaks at 1723 cm!1 and 1739 cm!1 in the reacted sample, typicalindicators of oxidation to carbonyl CQO bonds.

The ATR has a working range down to 500 cm!1 which allowsanalysis of sulphur containing compounds. The squalane sampleswhich had been in contact with 65–98% w/w sulphuric acid showedan underlying H2SO4 spectra and a strong SO2 smell was revealedwhen opening the sample bottles after a storage time of two months.The acidity in the oil phase most probably originates from disperseddroplets as the acid phase of the samples revealed no change indissociation when compared with spectra from freshly prepared acid.This means that no water was produced during the reaction withsqualane or the levels were too small to be detected. Only the 65 and80% w/w acid samples that were heated to 165 1C, Fig. 3, showedtraces of squalane in the acid phase. This, however, may be due to thebroad IR response of acid and water which may mask the hydro-carbon peaks. Dark liquid, taken from the interface of the samesamples, revealed a strong carbonyl peak and strong acidity thatcorrespond well to the underlying bulk acid. The squalane peaks wereweak but present.

4.3.2. FTIR analysis of solidsA solid particle formed during the experiments with 80% w/w

H2SO4 at 165 1C for 15 min was analysed by an FTIR microscope.To reveal the peaks corresponding to minor components, the bulk

Table 1Critical speeds for droplet displacement, after [8].

Interfacial tension Cacrit CKS Cacrit RT84 Dynamicviscosity

gC30H62=40% H2SO4 at 100 1C For yCKS'901 For yRT84'151 mC30H62

0.0248 N/m 0.37 40.52'0.7 '0.004 Pa s(linear interpolated)

Result Ucrit 2.29 m/s 4.34 m/s

F.A. Sautermeister et al. / Tribology International ] (]]]]) ]]]–]]]6

Please cite this article as: Sautermeister FA, et al. Impact of sulphuric acid on cylinder lubrication for large 2-stroke marine dieselengines: Contact angle, interfacial.... Tribology International (2012), http://dx.doi.org/10.1016/j.triboint.2012.06.002

components C30H62 and H2SO4 were subtracted from the resultingspectrum. Four significant peaks were found at 3350 cm!1(O–H;carboxylic acids), 1750 cm!1(CQO bond; oxidation of alkane),1595 cm!1(C–C or CQC or CQO carboxylates) and 1340 cm!1

(C–H methyl; residual amount of squalane).The same particle was also analysed with the ATR to observe

lower wave numbers. The spectrum pattern for sulphuric acidwas revealed but the peaks shifted from the expected 80% to 30%w/w acid concentrations but with weakened water peaks.

4.4. SEM/EDX

At high temperatures, solids were formed on the interfacebetween squalane and sulphuric acid. The example shown inFig. 5 was formed within 3 min at 165 1C and a acid concentrationof 80% w/w. During rinsing with heptane, two edges of the samplerolled up, revealing the bottom side of the flake. Fig. 5 shows thedifference between the top and bottom of the sample. While the topside is smooth, the bottom side reveals an open or shrunk, foam likestructure with feature diameters between 10 and 50 mm. It seemslikely that the bottom surface pointed towards the acid side.

EDX elemental analysis revealed 70.317'7% carbon,26.347'2.6% oxygen and 3.37'0.07–0.17% sulphur. All resultsare given in at%.

4.5. Light microscopy

4.5.1. Light microscopy—liquidsThe used liquid samples were investigated under a transmis-

sion light microscope. For all acid containing samples, dropletswere found in the acid phase. When the acid phase changed

colour, more droplets of 5–70 mm diameter were found. When theoil phase stayed clear, no droplets were found in the oil phase.When the oil phase did change colour, tiny droplets of 1–8 mmdiameter were found in the oil phase. It was also observed thatdark particles, the structure of which was too small to berevealed, would stay on the outside of the droplets in the oilphase and were more likely to be found on the inside of thedroplet in the acid phase.

Fig. 6 shows an example of C30H62 on top of 80% w/w H2SO4

exposed to a temperature of 165 1C over 15 min; both phaseschanged colour. The samples were stored for several weeks atroom temperature before the phases were extracted for micro-scopic examination.

The difference in droplet size, with smaller droplets in the oilphase and larger droplets in the acid phase, suggests the forma-tion of surfactants with hydrophilic head groups and oleophilictails yielding micelles or inverse micelles. The density of headgroups per surface area should be the same for all droplets but inthe acid bulk phase, the oleophilic tails require space within themicelle droplet, therefore enlarging the volume of the droplets. Inthe oil phase, the tails point outwards from the inverse micelledroplets and thereby allow them to be smaller in volume.

4.5.2. Light microscopy—solidsFig. 7 shows a solid particle formed within 3 min at 165 1C and

an acid concentration of 80% w/w. Under the light microscope(200$ , stacked) a foam like structure was revealed with featuresizes comparable to the droplets found in the acid phase. It wasobserved that some bubbles shrank after rinsing with solvent,suggesting that the cells were filled with oil and the reaction

Fig. 5. SEM pictures of solid reaction product created between 80% w/w H2SO4 and C30H62 at 165 1C; (a) overview; (b) top side; (c) bottom side.

F.A. Sautermeister et al. / Tribology International ] (]]]]) ]]]–]]] 7

Please cite this article as: Sautermeister FA, et al. Impact of sulphuric acid on cylinder lubrication for large 2-stroke marine dieselengines: Contact angle, interfacial.... Tribology International (2012), http://dx.doi.org/10.1016/j.triboint.2012.06.002

leading to solidification is taking place at droplet interfaces,meaning on the head groups of surfactants to form micelles.

5. Conclusions

5.1. Thermal stability of the lubricant

The thermal stability of squalane was significantly reduced inthe presence of sulphuric acid. Oxidation and sulphonationoccurs, Sections 4.3 and 4.4, resulting in deposit, Sections4.4 and 4.5.2, and surfactant, Section 4.1, formation which canlead in the engine to ring sticking and can reduce the neutralisa-tion efficiency of lubricant additives [36].

The chemical reaction was time dependent, Section 4.1, butsignificantly faster than neutralisation in commercial lubricants[18] which means that the acid neutralising additives cannotprevent the lubricant to be degraded by the sulphuric acid.

The reactivity of API Group I base oils is expected to behigher due to the higher content of unsaturated and reactivehydrocarbons [1].

5.2. Emulsifying

The study has shown that squalane when brought into contactwith sulphuric acid will form stable emulsions due to reduction in

interfacial tension by forming surface active substances. This isnot significant for the engine, as fully formulated lubricantscontain dispersants which readily emulsify water droplets.

5.3. Corrosion

The contact angle measurements indicate that droplets adher-ing to the moving chrome coated piston ring will be more easilydisplaced and will be smaller than droplets on the stationary greycast iron cylinder liner.

Acknowledgements

Dr. Richard Walshaw; Leeds Electron Microscopy and Spectro-scopy Centre (LEMAS); Faculty of Engineering, for operating theSEM/EDX.

Mr. Michal Szymonik; Institute of Microwaves and Photonics,Faculty of Engineering, University of Leeds, for contact anglearrangement.

Mr. Japareng Lalung of D.G. Adams Laboratory, Institute ofIntegrative and Comparative Biology, Faculty of Biological Science,University of Leeds for induction and possibility to use thetransmission microscope.

Fig. 6. Transmission microscopic image of the oil and acid phase of the mixture 80% w/w H2SO4 and C30H62 after exposure to 165 1C.

Fig. 7. Solid reaction product after separation, left an overview of the approximately 2 mm$3 mm flake, right a 200$ microscopic image of the same flake.

F.A. Sautermeister et al. / Tribology International ] (]]]]) ]]]–]]]8

Please cite this article as: Sautermeister FA, et al. Impact of sulphuric acid on cylinder lubrication for large 2-stroke marine dieselengines: Contact angle, interfacial.... Tribology International (2012), http://dx.doi.org/10.1016/j.triboint.2012.06.002

Prof. Anne Neville; iETSI, School of Mechanical Engineering,University of Leeds and her group for use of laboratory area andequipment.

Dr. Nik Kapur; School of Mechanical Engineering, University ofLeeds for surface tension balance.

Peter Schmidt, Federal Mogul Goetze and Konrad Rass,WartsilaSwitzerland AG for piston ring and liner material samples.

CIMAC for access to their database.

Appendix A

See Table A1.

Appendix B. Results of contact angle measurements. ‘‘rec.range’’¼recommended range. ‘‘aver.’’¼averaged results.

See Table B1.

References

[1] CIMAC Working Group ‘‘Marine Lubricants’’. No. 26 guidelines for dieselengines lubrication impact of low sulphur fuel on lubrication of marineengines. Frankfurt/Main: CIMAC; 2007.

[2] & State of New South Wales through the Department of Education andTraining and Charles Sturt University. HSC online. Industrial chemistry: 3.Sulphuric acid. Online: /http://hsc.csu.edu.au/chemistry/options/industrial/2762/Ch953.htm#b3S.

[3] Schlager D. Wartsila. EP 1 752 560 B1 Europe; 2006.[4] CIMAC Working Group ‘‘Heavy Fuel’’. No. 21 recommendations regarding fuel

quality for diesel engines. Frankfurt/Main: CIMAC; 2003.[5] Huijbregts WMM, Leferink R. Latest advances in the understanding of acid

dewpoint corrosion: corrosion and stress corrosion cracking in combustiongas condensates. Anti-Corrosion Methods and Materials 2004;51(3):173–88.

[6] Land T. The theory of acid deposition and its application to the dew-pointmeter. Journal of the Institute of Fuel 1977:68–75.

[7] Young TF, Maranville LF, Smith HM. Raman spectral investigations of ionicequilibria in solutions of strong electrolytes. In: Hamer WJ, editor. Thestructure of electrolytic solutions. New York: John Wiley and Sons; 1959.p. 35–63 [book auth.].

Table A1Best fit equations as shown in Fig. 4.

Acid conc. in % w/w Closest fit of interfacial tension measurements in N/m, T in 1C Valid temperature range in 1C

0 ¼!1.2236E-5$Tþ0.03864 12–870.5 ¼!8.43983E-10$T4þ1.92715E-07$T3-1.64371E-5$T2þ6.87675E-4$Tþ0.02285 6–9010 ¼!2.86798E-11$T5þ8.95547E-09$T4!1.06049E-06$T3þ5.86861E-5$T2!1.3666E-3$Tþ0.03332 !3–10025 ¼5.79812E-10$T4!1.5079E-07$T3þ1.31115E-5$T2!2.50671E-4$Tþ0.01547 !5–10040 ¼3.50765E-10$T4!9.849E-08$T3þ9.68993E-06$T2!2.41447E-4$Tþ0.0156 !4–9965 ¼5.78778E-07$T2þ3.98494E-5$Tþ0.01738 17–13380 ¼1.46677E-08$T3!3.10118E-06$T2þ3.01794E-4$Tþ6.03057E-3 !5–12690 ¼2.34344E-10$T4!6.43761E-08$T3þ5.51694E-06$T2!1.41965E-4$Tþ2.78768E-3 2–12298 ¼1.94661E-10$T4!5.1293E-08$T3þ4.65421E-06$T2!1.0321E-4$Tþ1.58747E-3 !3–117

Table B1

Contact angle of aqueous H2SO4 on engineering material submerged in squalane (C30H62) in degrees

H2SO4 conc. in % w/w CKS (galvanic piston ring chrome coating) RT84 (grey cast iron cylinder liner) T in 1C

Advancingangle

Receding angle Static angle Corrosion Advancingangle

Receding angle Static angle Corrosion

Rec. range Aver. Rec. range Aver. Rec. range Aver. þ¼Yes Rec. range Aver. Rec. range Aver. Rec. range Aver. þ¼Yes

0 100–140 130 50–90 104 88 138 15–35 33 80–120 121 þ(slow) 20132 72 115 110 20 94 þ(slow) 50117 47 25 98 29 88 þ(slow) 100

0.5 127 88 90–110 106 30–60 56 10–25 14 25–45 29 þ 20105 66 97 41 22 39 þ 50115 59 96 47 16 31 þ 100

10 128 10–40 116 56 5–20 17 20–55 47 þ 20101 29 69 36 33 þ 50116 20 59 þ 36 21 þ 100

25 134 91 119 51 15 25–35 32 þ 20139 101 89 27 21 33 þ 50103 34 78 þ þ 100

40 110 35–65 41 80–100 88 30–70 67 15 30–50 39 þ 20119 50 95 40 þ 50122 57 85 þ 33 þ 100

65 116 10–25 19 46 þ 102 19 25–45 43 þ 20122 19 67 þ 115 þ 5061 18 37 49 þ 10024 19 þ 22 þ 145

80 132 5–20 75 100–120 119 20–60 55 þ 20119 12 50 107 þ 50

8 27 29 þ 100110 þ 109 25 þ 165

90 60–100 127 8 20–60 43 þ 50–90 68 5–20 29 þ 2079 10 52 75 3 57 þ 5044 12 27 þ 71 27 þ 10091 31 þ 59 17 þ 155

98 25–65 27 8 15–50 25 60–90 72 8 41 þ 2022 13 24 79 2 56 þ 5068 9 39 80 10 40 þ 100

62 0 þ 165

F.A. Sautermeister et al. / Tribology International ] (]]]]) ]]]–]]] 9

Please cite this article as: Sautermeister FA, et al. Impact of sulphuric acid on cylinder lubrication for large 2-stroke marine dieselengines: Contact angle, interfacial.... Tribology International (2012), http://dx.doi.org/10.1016/j.triboint.2012.06.002

[8] Schleizer AD, Bonnecaze RT. Displacement of a two-dimensional immiscibledroplet adhering to a wall in shear and pressure-driven flows. Journal of FluidMechanics 1999:29–54.

[9] Dussan VEB. On the ability of drops to stick to surfaces of solids. Part 3. Theinfluences of the motion of the surrounding fluid on dislodging drops. Journalof Fluid Mechanics 1987:381–97.

[10] Micali F, et al. Paper no. 74 suction air humidity influence on piston runningreliability in low-speed two stroke diesel engines. Bergen: CIMAC; 2010[CIMAC Congress].

[11] Lu Ya, Chen Xiaoxu. The effects of a demulsifier on the interface of marineengine oil and water. Lubrication Science 2011.

[12] Benner JJ, et al. Lubricating Properties of Water in Oil Emulsions. Transac-tions of ASME, Journal of Tribology 2006;128(2):296–311.

[13] Demmerle R, et al. New insights into the piston running behaviour of sulzerlarge bore diesel engines. Frankfurt: CIMAC Central Secretariat c/o VDMAe.V.; 2001 [International Council on Combustion Engines. p. 1–14].

[14] Spahni, Marc, Brunner H, de Jong R. Paper no. 86 the new wartsila 820 mm-bore engine series—advanced design and first running experience. Bergen:CIMAC; 2010 [International Council on Combustion Engines. p. 1–15].

[15] Cantow FG. Paper no. 209 piston rings for large bore engines. 2004 [Inter-national Council on Combustion Engines. p. 1–14].

[16] Burstein GT, Shreir LL, Jarman RA. 13.8 Chromium coatings. Corrosion, vol. 2.s.l.: Elsevier Butterworth Heinemann; 1994 [p. 99–110].

[17] Plucinski P, Nitsch W. Mechanism of mass transfer between aqueous phaseand water-in-oil microemulsion. Langmuir 1994:371–6.

[18] Garner T, et al. Paper no. 129 development of marine diesel cylinder oils.Kyoto: CIMAC; 2004 [International Council on Combustion Engines. p. 1–14].

[19] Class NK. Guidance for measures to cope with degraded marine heavy fuels,version II. Ohnodai, Midori-ku, Chiba: ClassNK, Nippon Kaiji Kyokai; 2008.

[20] The Petroleum HPV Testing Group. Test plan lubricating oil basestockscategory. API; 2003 [s.l.].

[21] Taylor, Peter. 1 Electrophilic addition reactions of alkenes. Alkenes andaromatics. Cambridge: Royal Society of Chemistry; 2002.

[22] Rudakov ES, Volkova LK, Tishchenko NA. Kinetics and mechanism of homo-geneous reactions of alkanes in sulphuric acid in the presence of carboca-tions. I. Oxidation, isomerization, fragmentation. Kinetics and catalysis1988:1168–73 [{nj, Russian translation].

[23] Rudakov ES, Zamashchinkov VV, Lutsyk AI. Kinetics of the reaction ofsaturated hydrocarbons with PdII, PTII–IV, HgII, CrVI, MnIII, MnIII–RuIII–IV,NO2þ , H2SO4, H2S2O8, and carbonium ions in sulphuric acid. Teoretcheskaya iEksperimental’naya Khimiya 1976(July–August):361–8 [Translation].

[24] Diaby M, et al. Kinetic study of the thermo-oxidative degradation ofsqualane (C30H62) modelling the base oil of engine lubricants. Journal ofEngineering for Gas Turbines and Power 2009.

[25] Kosswig K. Sulphonic acids, aliphatic. Ullmann’s encyclopedia of industrialchemistry. Weinheim: Wiley-VCH Verlag GmbH & Co. KGaA; 2000.

[26] Miyazaki M, et al. Evaluation of marine engine lubricating conditions using acylinder liner surface oil sampling apparatus. London: CIMAC; 1993 [CIMAC20th International Congress on Combustion Engines. p. 1–15].

[27] Shaw DJ. Introduction to colloid and surface chemistry. Oxford: Butterworth-Heinemann; 2003.

[28] Rode Sebastian. Oberflaechenspannung. Mainz: Johannes Gutenberg Univer-sitaet Mainz; 2010.

[29] Stadler AF, et al. A snake-based approach to accurate determination of bothcontact points and contact angles. Colloids And Surfaces A: Physicochemicaland Engineering Aspects 2006:92–103.

[30] Hadley A. CombineZ5.3; 17-February 2006.[31] Demond Avery H, Lindner Angela S. Estimation of interfacial tension between

organic liquids and water, vol. 27. Ann Arbor: American Chemical Society;1993.

[32] Stecher GP, et al. The Merck index of chemicals and drugs. Rahway: Merck &Co., Inc.; 1960.

[33] Joos P, Vollhardt D, Vermeulen M. Interfacial tension of sodium dodecylsulphate solutions at the hexane–water interface, vol. 6. Langmuir, AmericanChemical Society; 1990 [s.l.].

[34] Mitsui T, et al. Changes in the interfacial tension with temperature and theireffects on the particle size and stability of emulsions. Kolloid-Zeitschrift undZeitschrift fuer Polymere 1971:227–30.

[35] DECHEMA e.V. DETHERM thermophysical properties of pure substances &mixtures. /https://cdsdt.dl.ac.uk/detherm/detherm.phpS; 2005–2008[Online, Cited: 27-May 2010].

[36] Wu RC, Papadopoulos KD, Campbell CB. Acid-neutralising of marine cylinderlubricants: measurements and effects of dispersants. AIChE Journal2000:1471–7.

[37] Lancon D, Doyen V. Paper no. 98 the universal concept: the lubricationsolution to 2020 and beyond. Bergen: CIMAC; 2010 [International Council onCombustion Engines. p. 1–11].

[38] Saito Y, Yamada T, Moriyama K. Paper no. 243 the piston-running behaviourmonitoring of large bore low-speed marine diesel engine at sea by measure-ment of piston ring oil film thickness and iron content in cylinder drainoil. Bergen: CIMAC; 2010 [International Council on Combustion Engines.p. 1–14].

F.A. Sautermeister et al. / Tribology International ] (]]]]) ]]]–]]]10

Please cite this article as: Sautermeister FA, et al. Impact of sulphuric acid on cylinder lubrication for large 2-stroke marine dieselengines: Contact angle, interfacial.... Tribology International (2012), http://dx.doi.org/10.1016/j.triboint.2012.06.002