Life assessment of rubber articles in fuels

(1)

Examensarbete 30 hp Juni 2013

Life assessment of rubber articles in fuels

Emmy Selldén

(2)

(3)

Teknisk- naturvetenskaplig fakultet UTH-enheten

Besöksadress:

Ångströmlaboratoriet Lägerhyddsvägen 1 Hus 4, Plan 0

Postadress:

Box 536 751 21 Uppsala

Telefon:

018 – 471 30 03

Telefax:

018 – 471 30 00

Hemsida:

http://www.teknat.uu.se/student

Abstract

Life assessment of rubber articles in fuels

Emmy Selldén

The choice of rubber material for use in sealings and hoses in the fuel system is of great importance. If a wrong type of rubber is used, premature failure during service may occur. This impacts the environmental performance, the safety during driving, uptime and economy of the transport. In this diploma work, rubbers for use in sealing and hoses in the fuel system have been evaluated to assess which materials have the potential to be used under long-term use in contact with commercial fuels.

Three commercial fuel hoses, nitrile rubber (NBR), hydrogenated nitrile rubber (HNBR), ethylene-acrylic rubber (AEM) and fluorocarbon rubber (FKM) of varying types and compositions have been evaluated in diesel with 7% RME (rapeseed methyl ester), 100% biodiesel of RME and ethanol fuel. Tests were performed by immersing the materials in fuel and measure the compression set and changes in properties like volume, hardness, tensile strength and elongation at break.

The results showed that one NBR material, one AEM and all FKM are potential materials for long term use in diesel with 7% RME. All types of NBR and two types of FKM (terpolymers, peroxide cured) may be used in ethanol fuel. NBR and HNBR were the only rubbers evaluated in biodiesel. NBR and HNBR with an ACN content of ~30% might be used in 100% RME at lower temperatures for shorter periods. The aging resistance in air was good for HNBR, AEM and FKM but poor for NBR.

Sponsor: Scania CV AB

ISSN: 1650-8297, UPTEC K 13012 Examinator: Karin Larsson Ämnesgranskare: Jöns Hilborn Handledare: Maria Conde

(4)

I

Svensk sammanfattning

(Swedish summary)

I lastbilens och bussens bränslesystem finns det packningar, tätningar och slangar av gummimaterial. Gummi är ett material som består av så kallade polymerer, d.v.s. långa kedjor mer repeterande enheter, som är sammanbundna till varandra i vissa

gemensamma punkter. Typ av polymer som används, men också andra tillsatser i gummit, påverkar gummits egenskaper. Vissa typer lämpar sig bättre när det är kallt, andra när det är mycket varmt. En del klarar av att användas i kontakt med kemikalier medans andra bryts ned.

I bränslesystemet kommer gummit i kontakt med bränsle vilket gör att det är viktigt att det gummit man använder klarar av att användas i bränslet. Ett felaktigt materialval kan innebära att komponenten drabbas av förtidigt mekaniskt brott vilket medför en

säkerhetsrisk och innebär att fordonet inte kan köras lika länge som tänkt, vilket också påverkar ekonomin.

I takt med att koldioxidutsläppen ökar och tillgången på olja minskar, utvecklas nya bränslealternativ till diesel. Två sådana alternativ är biodiesel och etanol. Biodiesel utvinns från växtolja och fett medans etanol kan utvinnas ur socker, stärkelse och

cellulosa från växter. Att byta från ett bränsle till ett annat innebär dock problem när det kommer till materialval. Skillnader i kemisk sammansättning hos de olika bränslena gör att gummit påverkas olika beroende på bränsle.

I det här examensarbetet har tester av ett antal gummi utförts i olika sorters bränsle för att göra en bedömning av vilka sorters gummi lämpar sig för användning under lång tid i tunga fordon. Det har gjorts genom att sänka ner prover i olika bränslen vid förhöjda temperaturer. En förhöjd temperatur gör att kemiska reaktioner, såsom åldring och absorption av media, går snabbare, vilket gör att man på några veckor vid en hög temperatur kan uppskatta bränslets påverkan under en lång tid vid lägre temperatur.

Eftersom att gummikomponenterna även utsätts för luft i verkligheten, har även exponeringar i luft vid förhöjd temperatur utförts. De exponerade materialen har utvärderats med avseende på volymsvällning, ändring i hårdhet och mekaniska egenskaper samt sättning. Det sistnämnda är en mycket relevant egenskap för att bedöma risken för läckage.

Resultaten visar att typ av gummi påverkar och att vissa gummityper lämpar sig bättre än andra i olika bränslen. Fluorgummi visade sig till exempel fungera bra i både diesel med 7 % biodiesel och etanolbränsle. Även luft hade smärre inverkan på dessa material.

Nitrilgummimaterialen uppvisade stora skillnader i diesel med 7 % biodiesel, beroende på sammansättning. Alla sorters nitrilgummi klarade sig däremot bra i etanolbränsle, men dåligt i luft. En specialvariant av nitrilgummi kan också komma att användas i diesel med 7 % biodiesel. En typ av etenakrylgummi svällde mycket i etanolbränsle, men klarar i övrigt av både varm luft och diesel med inblandning av 7 % biodiesel.

(5)

II

Acknowledgments

I would like to thank all the people at Scania that have helped me during my diploma work, especially my supervisor Maria Conde for great support during the whole process.

I would also like to thank Martin Bellander for good advice and Christian Sjöstedt for help in the lab and for good music during lab sessions. Thank you Jenny Johansson and Karin Agrenius, at SP Technical Research Institute of Sweden, for help during the project and for letting me visit you in Borås. Thank you Erica Forslund at Trelleborg Ersmark AB for providing the samples and for technical support.

Great thanks to the whole team at UTMC, Materials Technology at Scania, you have all been very kind and helpful. Finally I would like to thank my friends and family for supporting me.

Emmy Selldén

(6)

III

Acronyms and glossary of rubbers

Nomenclature of rubbers according to Swedish standard SS-ISO 1629 ACN Acrylonitrile

AEM Copolymer of ethyl acrylate (or other acrylates) and ethylene. Ethylene-acrylic rubber AR Aramid Reinforcement

ASTM American Society for Testing of Materials ATR Attenuated Total Reflectance

B100 100% biodiesel B7 Diesel with 7% RME

CO Polychloromethyloxirane. Epichlorohydrin rubber.

CPE Chlorinated Polyethylene Rubber CR Chloroprene Rubber

CS Compression set

DLO Diffusion Limited Oxidation

ECO Copolymer of ethylene oxide and chloromethyloxirane. Known as epichlorohydrin copolymer or rubber

ED95 Ethanol fuel of 95% ethanol and 5% additives FAME Fatty Acid Methyl Esters

FKM Fluoro rubber having substituent fluoro, perfluoroalkyl or perfluoroalkoxy groups on the polymer chain

FPM Same as FKM

FTIR Fourier Transform Infrared Spectroscopy

GECO Terpolymer of epichlorohydrin-ethylene oxide-allyl glycidyl ether HNBR Hydrogenated acrylonitrile-butadiene rubber

IRHD International Rubber Hardness Degree

NBR Acrylonitrile-butadiene rubber, known as nitrile rubber PVC Poly Vinyl Chloride

RME Rapeseed Methyl Ester SIS Swedish standards institute SS Swedish Standard

(9)

1

1 Introduction

This diploma work has been performed at Scania CV AB, a manufacturer of heavy trucks, buses and industrial and marine engines. Material selection is an important part when developing new components. In this part, a background to the use of rubber components in contact with fuel will be given, followed by aim and goals for the project.

1.1 Background

The main applications of rubber materials in heavy vehicles are in components for sealing, vibration damping, fluid transportation and in interior details. In particular, tightness of seals and hoses for fluid transport are of great importance for the

environmental performance, the driving safety, uptime and economy of the transport.

Fuel hoses are used in several parts of the fuel system in low and medium pressure applications. They are used in the tank area in the fuel filler system and in the engine as feed hoses to direct the fuel towards the fuel injector. Rubber hoses are also used to connect the tank and engine compartment and consist of feed hoses that transport fuel to the engine and as return hoses that transport unreacted fuel back to the tank. Fuel hoses consist typically of several layers of materials. The inner rubber layer is in direct contact with the fuel. Behind the inner layer, reinforcement is used. Depending on

temperature it can, for example, be polyester, cotton or aramid. Outermost is an external rubber layer that faces the outer environment and it should be able to resist weather, fuel, high external temperature, ozone, coolant, vibrations etc. Depending on the materials used for the inner and outer layer, an intermediate layer might be used. The intermediate layer gives less permeation and gives better adhesion between the layers and the reinforcement [1].

Sealings are used to prevent leakage and exclude contaminants. Several materials can be used like metal and rubber. Rubber has low hardness which allows for lower sealing pressures and it has elasticity making it possible to maintain the pressure [2]. Rubber is therefore common in sealings and gaskets. Many types of sealings and gaskets are available on the market. Commonly used are radial, axial and O-ring sealings. Radial sealings consist of several components, where one is a gasket cuff made of rubber. It is used to prevent the transportation of fluid between two parts where, for example, one part is stationary and the other is rotating. Axial sealings are commonly used to exclude external contamination. O-rings are circular sealings used for both radial and axial sealing. The O-ring is placed in a groove between two parts. When subjected to a load, it deforms and seals [2]. One example of an important gasket is the cylinder head gasket between the engine block and the cylinder heads in the engine. This gasket comes in contact with oil, unreacted fuel and degraded fuel.

The temperature in the fuel system varies and there are areas that are colder and warmer. For example, the fuel transport for injection in the engine transports cold fuel

(10)

2

and the temperature is rarely over 60⁰C. For transportation of unreacted fuel, the temperature is higher.

To reduce repair cost and to ensure safety it is wished that rubber articles have a lifetime close to the truck life.

To estimate how good rubber components will perform in fuel, accelerated tests are commonly performed according to Swedish standard SS-ISO 1817 or American standard ASTM D471, these tests are normally short, for example 70h [3]. There is a need to perform these tests at longer times, about 1000h, in real fuels to better estimate how good different rubber components will perform during long time in service. In this diploma work, three fuel hoses and thirteen rubbers will be examined in air and some in commercial fuels. The rubber components are three types of nitrile rubber (NBR), two types of hydrogenated nitrile rubber (HNBR), four types of ethylene-acrylic rubber (AEM) and four types of fluorocarbon rubber (FKM). NBR and HNBR are used in lower temperature applications while AEM and FKM are used at higher temperatures.

1.2 Aim and goals

This study aims to obtain relevant data to predict the long-term properties of rubber components used in applications in commercial fuels such as diesel with 7% RME (rapeseed methyl ester), biodiesel and ethanol fuel. This information will help to give safer recommendations on the life assessment of different rubber materials in fuels.

The questions to be answered are:

 Which of the chosen rubber materials have the potential to be used in fuel applications, for long-term use, in commercial fuels like diesel with 7% RME, biodiesel and ethanol fuel?

 How does polymer type and differences in composition of the different rubber components impact on aging and fuel resistance?

(11)

3

2 Theory

In this section, information from literature relevant for the understanding of the project is presented. It begins with a description of typical rubbers used in fuel and is followed by information on different fuels. A brief overview of the interaction of rubber and fluids is given and accelerated tests are discussed. Finally, results from previous research are presented.

2.1 Rubber materials

Rubber materials consist of special polymers showing high elastic properties. The main part of rubber consists of elastomers which are long, chainlike molecules that can be stretched at a great extent and then has the ability to recover to its original shape [4].

The elasticity originates from the movable and sparsely cross-linked molecules of these materials [5]. In order to improve the physical properties, rubbers are vulcanized, which is a chemical process where cross-links are formed between the polymeric chains [6].

Rubber materials consist of several components in addition to elastomers like cure system, fillers, softeners, aging protective agents and other additives. A brief description of some elastomer types are given below, followed by an overview on different

additives.

2.1.1 Description of some elastomers used in rubber Nitrile rubber

Nitrile rubber (NBR) is a copolymer of butadiene and acrylonitrile (ACN), see Figure 1.

The amount of ACN affects several properties like petroleum oil and fuel resistance, tensile strength, hardness and low temperature properties. A higher amount of ACN gives better petroleum oil and fuel resistance, improved tensile strength and increased hardness, but at the cost of low temperature properties [7]. In general NBR has good resistance to oil, aliphatic and aromatic hydrocarbons and vegetable oils, but poor resistance to polar solvents like esters and ketones, where it swells, because NBR is a polar rubber [8] [7]. Because of the double bond present in the backbone, NBR is vulnerable to oxygen, ozone and UV light.

Figure 1. Repeating units of NBR. To left: butadiene unit, to right: acrylonitrile unit.

(12)

4 Hydrogenated nitrile rubber

In hydrogenated nitrile rubber (HNBR), some or all double bonds in NBR has been removed by hydrogenation, see Figure 2. This makes HNBR more resistant to oxidation than NBR and gives it improved temperature and chemical resistance [9].

Figure 2. Repeating units of HNBR. To left: fully hydrogenated butadiene unit, to right:

acrylonitrile unit.

Ethylene-acrylic rubber

Ethylene-acrylic rubber (AEM) is a copolymer of methyl acrylate (or other acrylates) and ethylene, see Figure 3 . Its trade name is DuPont™ Vamac® and is available in several grades, some used in this diploma work. Vamac® G is the base grade, Vamac®

GLS has greater swelling resistance in oil and diesel fuel compared to Vamac® G and has a higher amount of methyl acrylate. Vamac® HVG is similar to Vamac® G but has higher viscosity [10]. Generally, AEM shows poor chemical resistance towards aliphatic,

aromatic and chlorinated hydrocarbons. Better resistance is shown towards mineral oils, natural fats and some salts [7].

Figure 3. Repeating units of AEM. To left: ethylene unit, to right: methyl acrylate unit.

Fluoro rubber

Fluoro rubber, (FKM, sometimes abbreviated FPM in some standards), has a fluorinated carbon-carbon backbone. There are many types of monomers available and ASTM D1418 has divided FKM into five types:

 Type 1: Dipolymer of hexafluoropropylene and vinyldiene fluoride, see Figure 4.

The fluorine content is usually ~66% and it is used for general purposes [11].

 Type 2: Terpolymer of tetrafluorethylene, vinylidene fluoride and

hexafluoropropylene, see Figure 5. The fluorine content is 66-70% which gives it improved resistance towards oil, solvents and fuels [11].

 Type 3: Terpolymer of tetrafluoroethylene, a fluorinated vinyl ether and vinylidene fluoride, see Figure 6. The fluorine content is usually 64-67%. It has improved low temperature properties but worse chemical resistance [11].

(13)

5

 Type 4: Terpolymer of tetrafluoroethylene, propylene and vinylidene fluoride.

 Type 5: Pentapolymer of tetrafluoroethylene, hexafluoropropylene, vinylidene fluoride, ethylene and fluorinated vinyl ether.

Figure 4. Repeating units of FKM type 1. To left: hexafluoropropylene, to right: vinylidene fluoride.

Figure 5. Repeating units of FKM type 2. To left: hexafluoropropylene, middle: vinylidene fluoride, to right: tetrafluorethylene.

Figure 6. Repeating units of FKM type 3. To left: vinylidene fluoride, middle:

tetrafluoroethylene, to right: perfluormethylvinylether (example of a fluorinated vinyl ether).

Due to the high bonding energy of C-F bonds and shielding of polymer backbone by fluorine, FKM has good high temperature resistance and resistance to oxidation, ozone, fuel and petroleum oils [12]. FKM swells in polar solvents such as low molecular esters and ketones [8]. Higher fluorine content increases the temperature and chemical

resistance [7], but the low temperature performance and compression set is worse [13].

(14)

6

FKM is commonly cured with peroxide or bisphenol. Peroxide cured FKM generally give weaker cross-links and results in worse aging resistance compared to bisphenol cured FKM [12]. The resistance to acids, steam and hot water is better compared with

bisphenol [14].

One of the trade names of FKM is DuPont™ Viton®. Two types that are studied in this diploma work are Viton® GBL-S and Viton® GFLT. GBL-S is of type 2 and GFLT of type 3 where the fluorinated vinyl ether is perfluormethylvinylether, both has an additional cure site monomer [15].

Epichlorohydrin rubber

Epichlorohydrin rubber is a group of three types of rubbers of halogenated polyethers, all with the epichlorohydrin monomer. An elastomer of epichlorohydrin homopolymer is designated CO. If epichlorohydrin is copolymerized with ethylene oxide one obtains an elastomer designated ECO, see Figure 7 [12]. GECO is a terpolymer of epichlorohydrin, ethylene oxide and allylglycidylether [16]. Epichlorohydrin rubber is polar, where CO is most polar. It has good resistance to petroleum fuels, alcohols, oxygen, ozone and light [7].

Figure 7. Repeating units of CO and ECO. To left: epichlorohydrin unit, to right: ethylene oxide unit.

Chloroprene rubber

Chloroprene rubber (CR) consists of chloroprene units, see Figure 8. Due to the chlorine atoms present, CR is a polar rubber. The chlorine atoms give the rubber better

resistance to weather and ozone. The swelling resistance in vegetable oils and animal fat is better compared to non-polar diene rubbers, but less compared to NBR [12].

Figure 8. Repeating unit of CR. Chloroprene unit.

(15)

7 Chlorinated polyethylene rubber

Chlorinated polyethylene (CPE) is produced by chlorinating polyethylene, see Figure 9.

Polyethylene is a crystalline polymer and not a rubber. By introducing chlorine,

crystallization is prevented and an elastomer is obtained. How rubbery the polymer will be depends on the degree of chlorination. Chlorinated polyethylene rubber is polar and hence oil resistant and due to the saturation it is less sensitive towards oxygen, ozone and light [7].

Figure 9. Repeating units of CPE. Chlorinated polyethylene.

2.1.2 Additives Cure system

Sulfur is the most used cross-linking agent and is used when the elastomers are

unsaturated. Sulfur reacts chemically with double bonds to form cross links between the chains. To speed up the vulcanization, zinc oxide, stearic acid and accelerators are

added. Zinc oxide reacts with stearic acid [8] to work as an activator [17]. Accelerators are usually organic chemicals. Some accelerators give a slow cross linking, some gives fast cross-linking and some are used to delay the cross-linking [8].

When no double bonds are available, peroxide can be used. Peroxide does not need zinc oxide, stearic acid and accelerators. Sometimes so called co-agents are used to improve the vulcanization. The peroxide acts by removing a hydrogen atom from the polymer chain and creating a radical. A radical on one site can react with a radical on another site and hence create a cross-link. Peroxide cured systems gives a better compression set than sulfur but has reduced tensile strength. Other cross-linking agents than sulfur and peroxide are metal oxides, which are used on halogen containing elastomers, and some amines alternatively bisphenols which can be used on fluoroelastomers and polyacrylate [8].

Fillers

Fillers give reinforcement to the rubber, thereby increasing the mechanical strength and stiffness. The size, shape and surface chemistry of the filler determine whether the reinforcement will be high or low. Carbon black is commonly used as filler and is the reason why rubbers often are black. Due to its surface activity the mobility of the rubber is reduced as it adsorbs at the surface of the carbon black [7]. Other fillers used are silica, clays and chalk [8].

(16)

8 Softeners

Softeners, also known as plasticizers, are used to increase the deformability (elongation) of a polymeric material as described in ASTM D 1566. Softeners can be used, for

example, to reduce hardness, reduce viscosity of uncured material and improve low temperature properties. One of the major sources of softeners is petroleum oils. In order to use oil, the elastomer has to have low or no oil resistance. If the elastomer is oil

resistant, polar liquids, like ester, can be used [8].

Aging protective agents

Antioxidants, also called stabilizers, are added to neutralize free radicals to protect the rubber from aging, which is caused by oxygen and accelerated at elevated temperatures.

To protect the material from ozone, antiozonants are added [8].

2.2 Fuels

To reduce CO2 emissions and decrease the dependence of oil, the use of other fuels alternative to diesel has increased in recent years, in particular biodiesel (consisting of fatty acid methyl esters, FAME) and bioethanol [18]. Ethanol fuel is used in buses and trucks, biodiesel is used in several types of vehicles. Biodiesel can be used in its pure form (designated B100, meaning 100% biodiesel) or in blends with diesel.

There are European emission standards, designated Euro I, II, III and so on, that regulates how much emissions heavy vehicles may emit. To be certified according to a Euro standard, reference fuels are used to assure that the standard is fulfilled. Euro VI is the latest standard and comes into force during 2013. Earlier (Euro V), diesel with 5%

FAME was approved as reference fuel. With Euro VI a reference fuel of diesel with 7%

FAME is introduced [19]. In this section a description of diesel, biodiesel and ethanol fuel will be given.

Diesel

Diesel is produced by distillation of petroleum crude oil. Petroleum crude oil consists of hydrocarbons like paraffins, naphtenes and aromatics. Paraffins¹ has the general formula CnH2n+2 and is divided in to normal paraffins, which are long, straight chains of hydrocarbons, and isoparaffins, which are long chains with branches, see Figure 10 [20].

Naphtenes² are saturated hydrocarbons with some carbons in a ring. In diesel, the rings in naphtene, have five or six carbon atoms [20]. Aromatics are unsaturated

hydrocarbons arranged in rings of six carbons. The carbons in the ring are joined by aromatic bonds [20]. After some refinery processes a group of hydrocarbons called olefins might be present. Olefins are hydrocarbons having one or more double bonds [20]. Other compounds than hydrocarbons, containing sulfur, nitrogen and oxygen, are also present in petroleum crude oil [20].

1 In petrochemical chemistry the term paraffins is used for acyclic alkenes (saturated carbons) [58].

2 Other names are cycloalkanes and cycloparaffins. [20]

(17)

9

Diesel, with a boiling point of 150-380°C [21], is a mixture of hydrocarbons of 10 to 22 carbon atoms [21] and consists of 50-70% paraffins, 30-45% naphtenes and 3-5%

aromatics [22]. The ratio and length of the different hydrocarbons gives the diesel different properties like boiling point, freezing point, density, heating value³, viscosity and cetane number. The cetane number is a measure of the ignition quality, i.e. how readily a fuel starts to burn once it has been injected into the cylinder. Aromatics tend to swell elastomers so the amount of aromatics present in the diesel is of importance for elastomeric behavior [21]. Additives are added to diesel to improve fuel handling, system performance, thermal stability and to control contamination [20].

Figure 10. Different hydrocarbons present in petroleum crude oil.

Biodiesel

Biodiesel consists of mono alkyl esters produced from feedstock of vegetables and animals. Common vegetable plants used as feedstock are soybean (most common in USA), rapeseed (most common in Europe) and palm oil [23]. Vegetable oils and animal fats consist of triglycerides which are hydrocarbons bonded to a glycerol molecule [20].

The triglycerides can be converted to FAME through a process called transesterification.

The transesterification is carried out by reacting the triglycerides with an alcohol,

commonly methanol, under the presence of a base [20] [23], see Figure 11. The resulting product, FAME, has alkyl chain lengths of 12 to 22 carbons depending on feedstock and is used as biodiesel [20].

Different feedstock gives different types and amounts of fatty acids, which influences the oxidation resistance of fuel [24]. Biodiesel does have some residual byproducts from the transesterification like glycerol, acylglycerols and methanol [25]. The amount of these byproducts and free fatty acids are regulated by different standards like ASTM D6751 and European standard EN 14214.

3 The heating value is the amount of heat released for a certain amount during combustion.

(18)

10

Because of the structure of unsaturated fatty esters in the biodiesel, there are some oxidative stability problems [25]. Oxidation of biodiesel can convert esters into carboxylic acids which gives enhanced corrosion and degradation of fuel properties, peroxides are also formed [26]. Compared to diesel, biodiesel takes up more water, which might promote microbial growth. It has increased polarity and solvency, which can cause the degradation of some elastomers [26].

Figure 11. Reaction scheme for the transesterification of vegetable oils.

Ethanol fuel

Ethanol is produced from sugar, starch or cellulosic biomass. Depending on plant it is produced in different ways. If sugar canes are used the sugar can be fermented directly.

If the source is starch, which is the case when using e.g. maize, it has to be converted into glucose before fermentation [27].

A commercial ethanol fuel is ED95, consisting of 95% ethanol and the remaining 5% are ignition improver, lubricants and other additives [28], see Table 1.

Table 1. Example of content of ED95. Information given from safety sheet by SEKAB.

Substance Weight%

Ethanol 90-92

Glycerol etohxylate 4-7 Methyl-t-butyl ether < 3 Isobutanol < 1 Lubricant < 2

2.3 Degradation of rubber and interaction with fluids

Degradation of rubber and other polymeric materials means irreversible deterioration of the physical and chemical properties [29]. There are many types of factors that can cause the degradation. These are temperature, light, ionizing radiation, humidity, fluids, bio-organisms, mechanical stress and electrical stress [30]. The degradation can be due to bond scissions in the polymer chain, breaking of cross-links and formation of new cross-links. It can also be due to extraction of and chemical attack on additives in the rubber [29]. Chain scission is seen as a deterioration of mechanical properties while the formation of additional cross-links is seen as an increase in hardness and modulus [31].

(19)

11

Thermo-oxidative degradation is a common reaction mechanism. High temperatures give rise to the formation of radicals by cleavage of carbon-carbon and carbon-hydrogen bonds. The radicals react with oxygen, forming peroxide radicals which can continue to react with the rubber [32]. A commonly used reaction scheme can be seen in Figure 12.

Figure 12. Common reaction scheme for oxidation of rubber. RH is the rubber polymer.

Reaction scheme as suggested in [32].

Fluids can cause chemical degradation, swelling, cracking and extraction of additives of rubber. Swelling is caused by absorption of fluid in to the polymer network. A general rule is that polar substances dissolves better in polar liquids and non-polar substances dissolves better in non-polar liquids [33]. Biodiesel is chemically different than diesel and contains more polar esters. Therefore swelling of polar elastomers is greater in biodiesel than in diesel. Swelling is observed when more liquid is absorbed than soluble components are being extracted from the rubber. If, in contrary, the volume decreases it might be due to soluble components being replaced by less dense solvent molecules or that the extraction of additives is greater than the absorption of solvent [33]. Volume change is increased at higher temperatures [31].

The volume increase is often accompanied with a decrease in hardness due to plasticization when fuel is absorbed. If larger changes in hardness are seen, it might indicate chain scission or formation of additional cross-links [34]. Additional cross-links might give an increase in hardness. This can also be observed if there has been loss of softeners. Chain scission can result in a decreased hardness.

If the volume change is purely physical, change in tensile strength and elongation at break are slightly reduced. If a large deterioration is observed, it is probable that chemical reactions has occurred (this applies for aged samples too). Tensile strength is usually decreased if cross-links have been attacked, while chain scission might be

observed as a reduction in both tensile strength and elongation at break [34]. Additional cross-links can result in an increased tensile strength.

(20)

12 2.4 Accelerated tests

Since the life time of rubber components is typically some years, it is not practical to perform tests for such a long time. To estimate how well a material will perform during its life time, accelerated tests are used, commonly by increasing the temperature. There are standards used in the industry on how to perform these tests. SS-ISO 188 is used for accelerated aging⁴ and SS-ISO 1817 is used to determine the effect of fluids by so called fluid resistance tests. In this section the Arrhenius equation will be introduced followed by a short description of fluid resistance tests.

2.4.1 Arrhenius equation

The Arrhenius equation gives the relation between the reaction rate and the temperature for a chemical reaction:

(1)

The equation can also be expressed as:

(2) Where

k = the rate constant for the reaction [time unit^-1] A= a pre-exponential factor

R = the gas constant (8.314472 JK^-1mol^-1) T = the temperature [K]

Ea = the activation energy [Jmol^-1]

With the Arrhenius equation, tests performed at higher temperatures can be used to predict the performance of a material at longer times at lower temperatures and can therefore be used for life time predictions [8]. If the activation energy is known for the reaction that dominates, or if the average activation energy for several reactions is known, the time needed for accelerated testing, t1, at a certain temperature, T1, may be calculated. The exposure time for the accelerated aging test would correspond roughly to the wished lifetime, t2, and operation temperature, T2. Using the Arrhenius equation:

(3)

4 Aging refers to degradation caused by oxygen.

(21)

13

If the activation energy is unknown, one way to decide the time for accelerated tests, is to assume an Ea. Typically, Ea in the order of 80-150 kJmol^-1, is found [35] [36] [37].

Assuming an Ea is of course a rough estimation, but to obtain data for Arrhenius extrapolations in order to determine Ea is very time consuming.

Drawbacks of using the Arrhenius equation

Several assumptions are made when using the Arrhenius equation. It is assumed that the same reactions occur under service conditions as under testing conditions. The

Arrhenius equation describes the temperature dependence for one chemical reaction, in reality there can be several reactions occurring and the reactions can be complex and not as easy as the equation suggests. It is also assumed that the activation energy is independent of temperature [29].

There are several studies proving non-Arrhenius behavior, which is seen as a non-linear behavior when plotting data. Some studies are reviewed by M. Celina et al [38]. Kenneth T. Gillen et al [39] has reviewed the limitations of using the Arrhenius equation. It is described that oxidation, which is often given by a simple equation described by the Arrhenius equation, in fact is a set of chemical reactions. By steady state analysis it is predicted that the Ea can be non-constant.

It is also discussed that diffusion-limited oxidation (DLO) can give non-Arrhenius behavior. DLO means that oxygen is consumed within the material faster than oxygen can be resupplied from the surroundings. The surface is not affected by this, but DLO can be seen deeper within the material where less oxidation occurs [39]. Another

mechanism that can give non-Arrhenius behavior is when two pathways give rise to degradation in a material. If one reaction has a lower Ea compared to the other, this reaction will not be evident until lower temperatures.

2.4.2 Fluid resistance tests

Normally, fluid resistance tests are done by immersing test pieces in liquids, such as fuel.

The effect of liquid on the rubber is evaluated by measuring certain properties like mass change, volume change, hardness and tensile-stress properties [40] [41] [42] [43].

Drawbacks of immersion tests

The disadvantage of immersion tests and other standard laboratory tests is that the experimental conditions differ from real service conditions and hence material selection can be incorrect [44]. Gordon Micallef et al [44] have compared standard laboratory testing (according to standard ASTM D2240 hardness, ASTM D412 stress-strain and ASTM D471 fluid immersion) with testing under service conditions for different fluoro rubbers in different fuels. Under service conditions, water contamination is common and, especially in biodiesel since water is more soluble in biodiesel than in diesel [44].

The water contaminated fuel gave a large deterioration of some of the elastomers compared to standard laboratory tests and it is suggested that the water causes

hydrolysis of esters in biodiesel which open up for other chemical reactions than in fuel

(22)

14

without water. This is just one example of how real service conditions can give differences in properties of great importance and this is important to bear in mind.

2.5 Previous research

There are general recommendations on what type of rubbers that has good or poor resistance to certain fluids. “The Los Angeles Rubber Group” has put together a chemical resistance guide, which can be found in DuPonts Chemical resistance guide [45]. Some of the results are summarized in Table 2. Available are also ratings for the suitability of use for some elastomers in different chemicals at room temperature, the result for ethanol and diesel oil is presented in Table 3. Ratings are based on data from several suppliers and manufactures and the criteria used for rating was volume swell resistance,

compression set resistance and aging resistance when applicable. How the data was obtained is not stated.

Table 2. General chemical resistance for some elastomers, information taken from [45].

Elastomer Generally resistant to Generally attacked by

NBR Many hydrocarbons, fats, oils,

greases, hydraulic fluids, chemicals

Ozone, ketones, esters, aldehydes, chlorinated and nitro hydrocarbons

HNBR Similar to NBR but with

improved chemical resistance and higher service

temperatures

Ozone, ketones, esters, aldehydes, chlorinated and nitro hydrocarbons

ECO Similar to NBR with ozone

resistance Ketones, esters, aldehydes, chlorinated and nitro hydrocarbons

AEM Weather, ozone, hydrocarbon

lubricants/greases, hydraulic fluids

Aromatic hydrocarbons, esters, gasoline, ketones FKM

Dipolymer, 66% fluorine All aliphatic, aromatic and halogenated hydrocarbons, acids, animal and vegetable oils

Ketones, low molecular weight esters and alcohols and nitro containing compounds

Table 3. Ratings for the suitability of some elastomers in ethanol and diesel oil at room

temperature, ratings taken from [45]. 1 = little to minor effect, 0-5% volume swell, 2 = minor to moderate effect, 5-10% volume swell, 3=moderate to severe effect, 10-20% volume swell, 4=not recommended. Time and way of testing is not stated.

Elastomer Rating in ethanol Rating in diesel oil

NBR 1 1

HNBR 1 1

ECO 2 1

CR 1 3

AEM 4 1

FKM, dipolymer 2 1

FKM, terpolymer 1 1

(23)

15

A.S.M.A Haseeb et al [41] have performed immersion tests in different concentrations of palm biodiesel in diesel at 25°C and 50°C for 500h. Tests were performed on NBR, CR and FKM (Viton A), the contents of these rubbers are not specified which is unfortunate since the amount of ACN in NBR is of great importance [7]. NBR and CR showed

deterioration in properties while negligible changes were shown for FKM.

F.N. Linhares et al [43] conducted immersion tests at 70°C for 70h in Brazilian biodiesel (ethylic biodiesel from coconut oil and castor bean oil). Three different samples of NBR with different ACN content were tested (28, 33 and 45% ACN). It was concluded that Brazilian biodiesel can degrade NBR but that increasing ACN content can prevent the degradation. NBR with 45% ACN appeared to be resistant to the biodiesel used in this study.

Gordon Micallef and Axel Weimann [44] performed immersion tests in diesel and in diesel blended with 30% RME of several types of FKM of varying type and fluorine content. After immersion at 150°C for 336h the volume change was under 10% for all types and the changes in mechanical properties was not of the degree that it would affect the actual performance of the rubbers.

E. Frame and R.L McComeric [46] have published a technical report of the compatibility of some elastomers in diesel blended with 20% biodiesel (from soybean) and diesel blended with 15% ethanol. The rubbers tested were NBR, NBR with high ACN content, peroxide-cured NBR, FKM filled with carbon black and FKM without carbon black.

Immersion was performed at 40°C for 500h. All samples, except from NBR with high ACN content, showed decreased break load after immersion in ethanol blended diesel as compared to diesel. This was not seen in the biodiesel blend. Volume swell was larger in ethanol blended diesel compared to biodiesel blend and diesel. The overall conclusion is that all the tested rubbers seem to be compatible in diesel blended with 20% biodiesel but less compatible in diesel with 15% ethanol.

Wimonrat Trakarnpruk et al [42] studied the impact of 10% biodiesel (from palm oil) in diesel on six types of rubbers in 100°C for 23, 670 and 1008h. The rubbers were NBR, HNBR, NBR/poly vinyl chloride (PVC), acrylic rubber, FKM – dipolymer and FKM- terpolymer. Mass change, volume change, hardness change, tensile strength and elongation were measured. None of the materials showed a significant change in properties.

Concluding remarks on previous research

No research has been found on the effect of ethanol fuel on rubber components. There are recommendations for pure ethanol, as seen in Table 3, but since there are additives in ethanol fuel that can affect the properties, further studies would give valuable

information.

Studies on RME are sparse and even though there are studies of biodiesel derived from other feedstock, a study on testing in RME would be interesting since different feedstock

(24)

16

gives different types and amounts of fatty acids [24] and thereby also might affect the impact on rubbers.

In many studies the acceleration factor is too high, i.e. materials are tested at too high temperatures for too short time. There is a risk of accelerating tests too fast. At

temperatures higher than the rubber component is usually subjected to, new chemical reactions can occur. It can for example be melting of material and migration of additives.

Short term tests, like one week, is therefore not good for predicting long-term properties since the temperature has to be raised significantly to correspond to the operation time.

In this case, extrapolation to lower temperatures may result in wrong assessment of the expected life time. Slower accelerations for longer times at lower temperatures,

decreases the risk of unwanted chemical reactions, hence better for predicting the long- term performance. Since studies on rubber components for several weeks are rare, tests at 1000 hours or longer would give the data needed to better predict the life time.

3 Methods

Fluid resistance tests by immersion in the different fuels are chosen to achieve the aim and goals of this diploma work. This is because they are used frequently in other

investigations, as previously described, and they are relatively easy to perform. Aging in air is carried out in addition to fluid resistance tests since warm air will be available in real service conditions. This will indicate if there is a risk of embrittlement of the materials during service.

SS-ISO 1817 forms the basis of the experimental set up for fuel exposures and involves immersion in fuels and evaluation of change in properties before and after exposure.

Some of the methods described in SS-ISO 1817 are selected. These are: change in hardness, volume and tensile stress-strain properties. The performance of rubber materials in sealings is evaluated by so called compression set according to SS-ISO 815-1.

Material characterization, to provide information on molecular changes in rubber before and after immersion, is provided by Fourier Transformed Infrared Spectroscopy (FTIR).

This chapter gives a description of the rubber components analyzed, fuels used and the analyses performed.

3.1 Rubber components analyzed

Thirteen rubber materials were obtained from Trelleborg Ersmark AB in form of compression molded sheets. Three fuel hoses were obtained from external companies.

The materials in the rubber sheets were: three types of NBR, two types of HNBR, four types of AEM and four types of FKM. These materials are the same used in a diploma work by Sara Wengström, Scania CV AB, 2012 [47], that aimed to study their low

temperature properties. The fuel hoses consisted of several layers and were: one hose of FKM/ECO/AR/ECO where AR is aramid reinforcement, one hose of HNBR/CPE with reinforcement and finally a hose of NBR/CR with reinforcement. Information on the different rubber sheets and hoses is given in Table 4.

(25)

17 Table 4. Information on the rubber components analyzed.

5 TR10 is the temperature where the rubber retracts 10% from an original stretch in frozen condition [56].

6 The type of phosphate is unknown. Common types are tributhoxyethyl phosphate and tricresyl phosphate [65].

Assigned

name Type Rubber Trade name Color TR105

(⁰C) Cross linking agent Filler Softeners Other information

NBR_1 sheet NBR Black -32 Sulfur Carbon black Ether, 5% ACN 30.5%. Standard

blend.

NBR_2 sheet NBR Black -50 Sulfur Carbon black Ether, 10% ACN 19%. Low ACN blend.

NBR_3 sheet NBR Black -40 Sulfur Carbon black Phosphate,⁶

10% ACN 29.5%. Low

temperature blend.

HNBR_4 sheet HNBR Black -15 Peroxide Carbon black None ACN 34%, fully saturated.

HNBR_5 sheet HNBR Black -32 Peroxide Carbon black Dioctyl

sebacate, 7-9% ACN 21%, partly saturated.

AEM_6 sheet AEM Vamac HVG Black -32 Diamine Carbon black Adipate,

12 parts Standard composition.

AEM_7 sheet AEM Vamac GLS Black -31 Diamine Carbon black Adipate,

18 parts Low swell.

AEM_8 sheet AEM Vamac HVG Black -39 Diamine Carbon black Adipate,

22 parts Different fillers than AEM_6.

AEM_9 sheet AEM Vamac

HVG/G Black -34 Diamine Carbon black Adipate,

14 parts

FKM_10 sheet FKM Green -15 Bisphenol Barium sulphate None F 66%, type 1. Standard

copolymer.

FKM_11 sheet FKM Viton GFLT Black -25 Peroxide Carbon black None F 67%, type 3. Low temperature

FKM_12 sheet FKM Black -13 Bisphenol Carbon black None F 68%, type 2. Standard

terpolymer.

FKM_13 sheet FKM Viton GBL-S Black -17 Peroxide Carbon black None F 67.5%, type 2. Increased fluid resistance

hose_14 hose FKM/ECO/AR/ECO Black

hose_15 hose HNBR/CPE Black

hose_16 hose NBR/CR Black

(26)

18 3.2 Fuels used

Three fuels were used for exposure of samples: diesel with 7% RME (will be given the abbreviation B7), biodiesel (B100) and ethanol fuel. B7 is a reference fuel used for certification of Euro VI engines, it was provided from Preem. The biodiesel consists of 100% RME and was provided from Preem. The ethanol fuel was ED95 provided from SEKAB.

3.3 Choice of time and temperature for exposures

The time of exposure was roughly estimated by using equation 3. The wished life time for the rubber articles was set to 40000h. The continuous temperature in colder parts of the fuel system was set to ~65⁰C and warmer parts to ~85⁰C. For about a total of 10% of the wished rubber article life time, corresponding to 4000h, the temperature was

assumed to be elevated, with a temperature of ~80⁰C in the colder areas and ~120⁰C in the warmer areas.

Activation energy of 96.5 kJ mol^-1(1eV) was assumed and the temperature of exposure was chosen so that the acceleration factor would not be too high. The calculated time for exposure, using equation 3, was rounded off to correspond to whole weeks. The result of the Arrhenius calculation is presented in Table 5. The short term exposures (168h) at higher temperatures were performed to see how the materials are affected at elevated temperatures. To see the change in properties with time, an additional time was added at about 500h for exposures at the lower temperature.

The exposure times and temperatures shown in Table 5 was used for aging in air.

Table 5. Assumed time and temperature in service with corresponding time and temperature for exposure.

Time in service, t2 (h)

Temperature in service, T2 (⁰C)

Time of

exposure, t1 (h)

Exposure

temperature, T1 (⁰C)

Used for rubber and hoses with:

40 000 ~65 1008 105 NBR, HNBR

40 000 ~85 1008 135 AEM, FKM, HNBR

4000 ~80 168 115 NBR, HNBR

4000 ~120 168 165 AEM, FKM

For exposures in B7and B100, the temperature for 168h and 165⁰C was lowered to 150⁰C, due to experimental set up limitations related to safety during testing in highly flammable fuels.

For ED95, exposures were conducted for 504h and 1008h at 70⁰C, close to the boiling point of ED95 (boiling point is ca 78⁰C), with exception for compression set that was conducted at 115 and 150⁰C for 168h.

NBR and HNBR are usually used in colder areas while AEM and FKM are used in warmer areas. HNBR can however withstand higher temperatures for shorter times and was

(27)

19

therefore tested at some of the higher temperatures. An overview of the time and temperatures for exposures in fuel and air is presented in Table 6.

Table 6. Time and temperatures used for exposure in different fuels and in air of the rubber materials and hoses.

504h 1008h 70⁰C

504h 1008h 105⁰C

168h 115⁰C

504h 1008h 135⁰C

168h 150⁰C

168h 165⁰C

B7 NBR, HNBR NBR, HNBR AEM, FKM AEM, FKM,

HNBR

B100 NBR, HNBR NBR, HNBR AEM, FKM AEM, FKM,

HNBR ED 95 NBR, HNBR,

AEM, FKM

Compression set NBR, HNBR

Compression set AEM, FKM, HNBR

Aging

in air NBR, HNBR NBR, HNBR AEM, FKM,

HNBR AEM,

FKM

3.4 Sample preparation

For all tests, with exception for compression set, dumbbells were used. Dumbbells from rubber sheets and hoses were punched out, with size according to SS-ISO 37 type 2. The parallel length and width of the narrow portion was 25mm and 4mm respectively. For measurements of volume change, dumbbells cut in half were used.

Test pieces for compression set were punched out using a circular die, of size according to SS-ISO 815-1. The diameter was 13 mm.

3.5 Aging in air and exposure in fuels ED95 and B7

Due to the low flash point of B7 (ca 68⁰C) and ED95 (ca 10⁰C), exposures were

performed by SP Technical Research Institute of Sweden in Borås, in autoclaves. Not all of the rubber components could be tested at SP so ten materials was selected: NBR_1, NBR_2, NBR_3, HNBR_5, AEM_7, FKM_10, FKM_11, FKM_13 and hose_14 and hose_16. SP performed the compression set for these samples and measured the volume change for ED95 exposed samples. The rest of the samples were sent back so the other tests could be performed.

B100

Only the exposure for 168h at 115⁰C in B100 could be conducted. When starting exposure at 150⁰C, severe smoke generation was observed. It was concluded that the flash point of B100 was significantly lower than measured by the supplier. Therefore the other exposures in B100 could not be carried out. Exposures in B100 were performed by hanging the samples on steel wire in flasks of 250 ml and fill the flasks with B100, see Figure 13a).

(28)

20 Aging in air

Aging in air was performed in cell type ovens where only rubbers of same polymer type were aged in the same cell. Samples were hung by hooks lined with

polytetrafluoroethylene on a stand and then placed in the cell, see Figure 13 b) and c).

a) b) c)

Figure 13. Exposure in B100 and air. a) exposure in biodiesel, b) samples for aging hanging on stand, c) cell type oven used for aging.

(29)

21 3.6 Analysis of samples

Following is a description of the theory and execution of analysis.

3.6.1 Volume change

The volume of a test piece can be determined by fluid displacement methods.

Archimedes principle states that when a test piece is immersed in a fluid, an upward force will act on the sample. The magnitude of the force equals the weight of fluid being supplanted and the volume of supplanted fluid equals the volume of the test piece [48].

By weighing the sample in air and liquid before and after exposure in fuel, the percentage change in volume, ΔV100, can be calculated by

(5) [49]

Where

ρ = the density of the liquid used for displacement mi = the mass after exposure in fuel

mi,liq = the mass in liquid after exposure in fuel (including the mass of a sinker if it is used)

ms, liq = the mass an eventual sinker m0 = the initial mass

m0,w = the initial mass of the sample when weighed in water (including the mass of a sinker if used)

ms,w = the mass of the eventual sinker

Water can be used as the liquid for displacement if the fuel is immiscible with water.

Equation 5 can then be expressed as

(6) Where

mi,w = the mass in water after exposure in fuel (including the mass of a sinker if it is used).

Measurements of volume change after exposures in B7, B100 and ED95 were carried out by fluid displacement in water and equation 6 was used for calculating the volume change, see Figure 14. For test pieces exposed to B7, weighing was performed after about 24h, for test pieces exposed to B100 and ED95, weighing was performed 30min after terminating the exposure. Three test pieces were used for each exposure.

It was discussed whether it is correct to weigh test pieces exposed in ED95 in water or not. Ethanol is soluble in water and another fluid for displacement might have been more appropriate. However, SP informed that the balance was stable during the

weighing and that no visible volume change could be seen when immersing in water. It

(30)

22

was concluded that weighing in water instead of other media should be of minor importance.

Figure 14. Measurement for volume change. To left: weighing in air, to right: weighing in water.

3.6.2 Change in hardness

Hardness is the resistance to indentation. For hardness measurements of rubber, the international rubber hardness degree (IRHD) scale can be used. It ranges from 0 to 100 where 0 is the hardness of a material having an elastic modulus of zero and 100 is the hardness of a material having infinite elastic modulus [50]. The hardness measurement is performed by using a spherical indentor. The hardness is given by measuring the difference in penetration depth of the indentor between a small contact force and a large force applied on the sample [50]. The penetration is then converted to IRHD. Tables for this can be found in SS-ISO 48.

The change in hardness, ΔH, before and after aging or fuel exposure is calculated by

(4)

Where

ΔH0 = the initial hardness

ΔHi = the hardness after aging or immersion

Hardness was measured with a Bareiss digitest hardness tester (IRHD micro), according to method M (microtest) in SS-ISO 48, measuring the hardness over 30 seconds. For measurements on unexposed material, five readings were conducted. For fuel exposed and aged materials, three readings on three different test pieces were performed.

Measurements on hoses were conducted both on the inside and outside of the hose, see Figure 15. The measurements took place about 24 h after terminating the exposure.

(31)

23

a) b) c)

Figure 15. Hardness measurements. a) Measuring the inside of hose, b) outside of hose c) the hardness tester

3.6.3 Tensile testing

In tensile testing, a dumbbell is clamped in a tensile testing machine and stretched at a uniform speed until it breaks. The force needed to stretch the sample and the extension of the sample is recorded. By dividing the force with the initial cross section area, the tensile stress, σ, is obtained:

(6)

Where

F = the force [N]

A = the cross section area [mm²] of the narrow part of the dumbbell.

The maximum tensile stress during measurement to rupture, is called tensile strength, see Figure 16a).

The extension per unit length is called elongation or strain, ε, and is calculated by

(7)

Where

L = the measured extension

L0 = the original length of the narrow part of the dumbbell.

The elongation at rupture is called elongation at break, εB, see Figure 16a).

(32)

24

a) b)

Figure 16. Tensile testing. a) example of stress-strain curve for a rubber indicating tensile strength and elongation at break, b) Tensile tester used for testing.

Tensile testing was carried out with an Alwetron TCT 50 tensile tester with a load cell of 500N, see Figure 16b). A pretension of 0.5N and a test speed of 500mm/min were used.

For unexposed and aged samples, five dumbbells were used. For fuel exposed samples, 3-4 dumbbells were used. Measurements were performed about 24h after terminating the exposure.

When conducting tensile testing of hose_16 after 168h at 115°C in B7, the temperature in the lab was elevated and ~4°C higher than normal.

3.6.4 Compression set

Compression set is used to measure the ability of a rubber to recover from an applied compression. Test pieces and spacers are placed between two steel plates that are tightened. The thickness of spacers determine the compression the test pieces will be subjected to, for rubber with hardness 10-95 IRHD a compression of 25% is normally used [51]. After exposure, the test pieces are released and the thickness of the samples is measured after recovering at room temperature for a given time, see Figure 17. The compression set, CS, is calculated by

(8)

Where

h0 = the initial thickness

h1 = the thickness after recovery hs = the thickness of the spacers used.

(33)

25

As can be seen in equation 8, test pieces that recover fully have a CS of zero, while test pieces that do not recover have a CS of 100%.

Figure 17. Compression set. A test piece (black) is placed between two steel plates that are screwed together, a spacer is used (light grey). After disassembling the equipment, test pieces are left for recovering before measuring the thickness.

For compression set, equipment like the ones in Figure 18 a-b) was used. Three circular discs were piled up, forming one sample, with a total height of approximately 6mm, see Figure 18 c). Spacers were chosen so that the compression of each sample was 25 2%.

For aging, CRC Silicone lubricant was sprayed onto the steel plates as release agent. For exposure in fuel, the equipment was immersed in fuel. For aging, the equipment was placed in the bottom of the cells in the oven. At the end, the equipment was left to cool down for 75 15 min (method B in SS-ISO 815) before it was disassembled. Test pieces were then left to recover for 30 minutes before the thickness was measured.

For exposures performed by SP, the cooling time was ~120 min. PTFE spray was used as release agent.

Five test pieces were used for aging and three test pieces were used for exposure in fuel.

Compression set was performed on samples from rubber sheets and not on hoses.

Figure 18. a) and b) equipment for compression set, c) three piled up discs forming one sample.

3.6.5 FTIR

Fourier transform infrared spectroscopy (FTIR) is used to measure how much a sample absorbs infrared radiation, it gives information on the molecular bonds present in the sample. A bond between atoms in a molecule can be assumed to be a spring that can be bent and stretched, this is referred to as vibrations. If incoming infrared radiation has the same frequency as the vibration of the molecule, it can be absorbed. Only molecules

(34)

26

having an electric dipole moment that changes during the vibration can show infrared absorption [52]. Infrared spectra are complex due to the many vibrations coupling over the entire molecule. There is a region in the infrared spectrum, below around 1500cm^-1, that gives information about the molecule as whole and is useful when identifying a material, this is called the fingerprint region.

For some samples, like rubber, reflection techniques are used along with FTIR. In attenuated total reflectance (ATR) spectroscopy, the sample is placed on a crystal. The beam is reflected in the crystal and penetrates a small portion of the sample where some radiation is absorbed, the remaining signal is detected [52]. ATR-FTIR is an easy-to use and fast method where no advanced sample preparation is needed.

FTIR analysis was performed with a PerkinElmer Spectrum 100 with uATR (universal ATR) between 4000 and 650cm^-1 with a resolution of 4cm^-1 and 4 scans, see Figure 19.

Unexposed rubber, aged and fuel exposed rubber for 1008h and all fuels were analyzed.

For black rubber, a Ge-crystal was used due to the high absorption of carbon black. Ge has a high refractive index which allows deeper penetration of radiation into the material. For fuels and green rubber, a diamond/ZnSe crystal was used. For rubber samples from sheets, FTIR was conducted on a new cross section. For hoses, the inner and outer layer was pulled apart. Analysis was conducted on the inside, outside, inside towards reinforcement and outside towards reinforcement, see Figure 19.

Figure 19. a) Cross section of fuel hose. FTIR was performed on the inside, outside, inside towards reinforcement and outside towards reinforcement. b) uATR-FTIR equipment.

a) b)

Life assessment of rubber articles in fuels

Examensarbete 30 hp Juni 2013

Life assessment of rubber articles in fuels

Emmy Selldén

Abstract

Life assessment of rubber articles in fuels

Svensk sammanfattning

Acknowledgments

Contents

Acronyms and glossary of rubbers

1 Introduction

2 Theory

3 Methods