Influence of matrix and alloying on the high cycle fatigue properties of compacted graphite iron for cylinder heads

(1)

IN

DEGREE PROJECT MECHANICAL ENGINEERING, SECOND CYCLE, 30 CREDITS

STOCKHOLM SWEDEN 2020,

Influence of matrix and alloying on the high cycle fatigue properties of compacted graphite iron for

cylinder heads

ERIK FROM

(2)

Abstract

Continued improvements in combustion processes, introduction of biofuels, decreasing fuel consumption and increasing specific power will inevitably lead to an increase of the combustion pressure, which can be accompanied with a temperature increase. This will increase the loading on many components and drives the development of materials more resistant to the harsh conditions close to the engine. The cylinder head is one of the components that will be affected by the increased loads. Therefore, thermomechancial loads (TMF) originating from the cyclic temperature variations from the heating up and cooling down cycles of the engine and the high cycle fatigue load (HCF) emanating from the repeated combustion pressure pulses may cause failure of the component.

Today, most cylinder heads for heavy trucks are cast in either lamellar graphite iron (LGI) or compacted graphite iron (CGI). In both cases, the graphite particles are embedded in a pearlitic matrix. A possible improvement of the endurance life of the component is to add silicon to the composition. Silicon promotes the formation of ferrite and the hypotheis is that the more ductile ferritic matrix enclosing the graphite particles will improve the resistance to macro crack initiation and propagation and thus the life of the component. The purpose of this master thesis is to investigate the effect of a ferritic matrix as well as other added alloying elements such as molybdenum and nickel on the high cycle fatigue properties of CGI. The HCF properties of three ferritic alloys with different amounts of molybdenum and four pearlitic alloys with additions of molybden or nickel were compared at a fatigue life of 2 miljon cycles and a stress ratio of R=0.1.

The results show that the silicon alloyed ferritic base material, without molybdenum, had a fatigue strength that on average was about 10 % higher than the corresponding pearlitic base alloy without alloying elements. Adding 0.25 wt % molybdenum improved the fatigue strength for both the pearlitic and the ferritic alloy with about 5-10 %. Further, for the pearlitic materials, molybdenum also improved the static mechanical properties while no such effect was seen for the alloys with a ferritic matrix. It was also found that nickel did not alter the HCF or the static properties significantly on the materials with pearlitic matrix.

(3)

Sammanfattning

Förbättringar i förbränningsprocesser, introduktion av biobränslen, minskad förbrukning och ökad specifik effekt kommer oundvikligen leda till ökat förbränningstryck, tillsammans med en temperaturökning. Detta bidrar till ökad last i komponenter, vilket driver utvecklingen av material för att klara av de förhållanden som uppstår i motorn. Cylinderhuvudet är en komponent som blir direkt påverkad av dessa ökade laster. Termomekaniska laster som uppstår på grund av temperaturvariationer från motorns start-stopp cycler och högcykellaster från upprepade förbränningspulser kan leda till brott av komponenten.

Idag är cylinderhuvud för tunga lastbilar gjutna i lamellärt gråjärn eller kompaktgrafitjärn. I båda fallen är grafitpartiklarna omslutna av en perlitisk matris. En möjlig förbättring av komponentens livslängd är att tillsätta kisel för att få en ferritisk matris. Hypotesen är att den mer duktila ferritiska matrisen som omsluter grafitpartiklarna kommer förbättra resistensen mot initiering och propagering av makrosprickor, vilket i sin tur påverkar komponentens liv.

Syftet med detta arbete är att undersöka effekterna av en ferritisk och perlitisk matris och andra legeringsämnen såsom molybden och nickel på högcykelutmattningsegenskaperna (HCF) hos kompaktgrafitjärn. HCF-egenskaperna hos tre ferritiska legeringar med olika mängder av molybden och fyra perlitiska legeringar med tillsatser av molybden eller nickel jämfördes vid två miljoner cykler och ett spänningsförhållande R=0.1.

Resultaten visar att det ferritiska referensmaterialet med legeringsämnet kisel hade en utmattningshållfasthet som var cirka 10 % högre än det perlitiska referensmaterialet. Tillsats av 0.25 wt % molybden förbättrade utmattningshållfastheten för både de perlitiska och ferritiska legeringarna med cirka 5-10 %. Tillsatsen av molybden för det perlitiska materialet ökade även statiska mekaniska egenskaper, men ingen liknande effekt kunde ses för legeringar med en ferritisk matris. För material med en perlitisk matris hade nickel ingen påverkan på HCF-egenskaper eller statiska egenskaper.

(4)

Preface

This report is a master thesis in material science, performed at Scania in Södertäljde, Sweden. I would like to thank every one that has helped and supported me throughout the project. I had a challenging, yet intriguing task. A special thank to:

Peter Skoglund, Scania, for his invaluable advice, effort and enhance.

Bo Alfredsson, KTH, for taking me on as a thesis student.

(5)

1 Introduction

As the world is an ever changing place, it needs a environmentally conscious and stable transport chain that can handle current and future needs. Trucks play an important role in this chain, and are a necessary part that delivers goods to create a sustainable, as well as, developing society. However, as demands for low weight, high torque, low emission and lower fuel consumption creates new challenges for manufacturers. Increased temperature and pressure of diesel engines may lower their life expectancy, causing them to fail prematurely [1]. Thus, the need for stronger materials is evident.

A relatively new material, compacted graphite iron, or CGI, is found in a range of applications, especially as an important material for cylinder heads. It consists of soft, randomly oriented, graphite particles, as well as ductile ferrite and high strength pearlite iron [2, 3]. The graphite particles, usually in a matrix of pearlite, morphology is very important when analysing the properties of cast iron. By changing the shape of graphite by adding magnesium as an inoculant in the casting process, it is possible to obtain compacted graphite iron, which has shown favourable characteristics compared to other cast iron qualities [4].

As engines are designed for a finite life expectancy, and since the loads within the engine are of high pressure and of cyclic repetitions, one way to measure this expectancy is with high cyclic fatigue, or HCF experiments. It is therefore of interest to investigate the load at which 2·10⁶cycles are obtained. Therefore, some sort of iterative method must be employed.

Further material characteristics are evaluated to obtain an optimised cast iron for cylinder heads for Scania´s new generation diesel engines.

1.1 Background

Scania is an international company that produces trucks, buses and engines for a wide range of applications. The company is producing a new generation of engines in conjunction with MAN that will hold for the current, and if possible, future emission targets. Further, this thesis is performed within the framework of a project together with with Linköpings university and Volvo to research new materials for high cycle fatigue.

(7)

1.2 Thesis objective

The thesis intends to investigate the effect of a ferritic and pearlitic matrix on CGI with respect to high cycle fatigue properties. Special care will be taken with respect to the composition of the material, since it is known that addition of silicon, which is the main alloying element used to obtain ferrite, results in improved mechanical properties compared to pearlitic CGI. [4]. Therefore, inclusion of alloying elements such as silicon, molybdenum and nickel, and their effect on the microstructure and high cycle fatigue properties, will also be studied.

The specimens will be characterized with respect to composition, hardness, matrix, fatigue performance and static mechanical properties. The results will be compared with a pearlitic and a ferritic reference alloy. An important part of the work will be to investigate and characterize the different microstructural constituents of the alloys and relate them to the fatigue performance. This thesis work is part of a project where Volvo and Linköpings university also participate.

1.3 Delimitations

The number of fatigue analyses depends on the amount of test specimens made by the foundry, which are provided by Scania. The number of tensile, fatigue and hardness tests needed are therefore taken into account. The number of tensile specimens are also determined in consultation with Scania.

It is evident that for higher number of test specimens, the accuracy of the resulting fatigue data increases. However, it takes time to perform a high amount of HCF tests, and since the thesis work is restricted within a specified time frame, it is not possible to perform a high amount of HCF tests. Further, there is also an economical, and an environmental aspect to consider, which gives further reason to limit the amount of test specimens manufactured and tested.

The influence of temperature on the mechanical properties are proven to be of great importance. [5, 6]. To add, the loads applied to an engine and its components are complex with different temperatures, load amplitudes and frequencies. The tests performed in this study are uni-axial tensile fatigue tests of a dog bone shaped test specimen. Because of this, the other loads mentioned have been neglected, and instead, predetermined load scenarios have been used for the experiments. Thus, the results generate a statistic analysis of the influence of different alloying elements on ferritic and pearlitic cast irons.

(8)

2 Materials and microstructures

This section discusses the production of the materials, their composition and structure.

2.1 Casting

During casting, the solidification and cooling process are of great importance, especially since the cooling time needs to be relevant for cylinder heads. Different rates of these variables creates different microstructures and material characteristics. To control the mentioned variables, different cast thicknesses were used. Further, the geometries regarding the feeders to the dies were also of different sizes. The term medium cooling and medium solidification (MM) represents the conditions found in the combustion plane of the cylinder head during casting. Simulations were performed by Scania to produce a cast that achieves cooling and solidification rates that are relevant to the cylinder head with the use of different thicknesses of the plates, and different diameters for the feeders, see Figure 1.

Figure 1: The cast with red arrows that represent the orientation of how the specimens are cut.

The solidification and cooling rate is highly important as it affects the microstructure and in turn the mechanical properties. Higher mechanical properties such as higher strength and hardness are obtained with higher cooling rates. Although, there are a few drawbacks. If the cooling rate is too high, the result may be white cast iron which has high hardness, but is very brittle [3], and is therefore not applicable to cylinder heads. The thickness of the plate

(9)

thus becomes 16.3 mm and the feeders obtain a diameter of 128.5 mm. The result is a cast from which standard dog bone specimens are cut from.

The plates and the feeders, shown in beige, have different thicknesses. The plate with the arrows is the one used for this thesis. As previously mentioned, a specific cooling and solidification rate was chosen. The cooling rate is mainly controlled by the dimensions of the feeders (pontoons) on the side of the plates, which are the thick part of the cast, and the solidification rate is determined by the thickness of the thinner part of the plate, where then specimens are cut and fabricated from.

2.2 Materials

Cast iron is excellent for complex geometries, has a low price and high wear resistance [7].

Further desirable properties are that during solidification, cast iron has a low shrinkage compared to steel, good wear resistance and vibration damping [3]. It is therefore suitable for diesel engines and cylinder heads. However, since the material is highly dependent on the cooling rate during casting, the cast´s thickness is of great importance, as can be seen in Figure 2.

Figure 2: The effect of the casting thickness [3].

From Figure 2 it can be stated that since a diesel engine has a complex geometry, higher

(10)

evaluating the tensile strength in a particular area of an engine.

The microstructure of iron is highly dependent on the carbon content. With the inclusion of carbon, graphite forms during the solidification process which directly affects the hardness of the material, as well as the castability and ferrite promotion. However, if the carbon content is too high, the result is degradation of the ductility, coarser structure and decrease of fracture toughness. Compared to steel, graphite is weaker, which allows for less elastic motion. When sharp graphite particles are present, the initiation of cracks is easier compared to steel. The crack initiation phase of cast iron takes place as a separation between graphite particles and the iron matrix and often occurs at low loads. [8].

There are three main structures of graphite. Flake, compacted and spheriodal, which are obtained with increasing amounts of magnesium (Mg). Mg, added as inoculants in the liquid, affects the nucleation and growth of graphite. During the solidification, the nuclei of graphite are surrounded by austenite. The growth of graphite diffuses carbon in austenite. Therefore, the crystal grows in a spherical, rather than a flake shape. [8]. The Magnesium´s effect on the graphite shape and as a consequence on the tensile stress can be seen in Figure 3 [8].

Figure 3: Tensile stress as a function of strain for different cast irons [3].

From Figure 3 it is shown that compacted graphite iron is an intermediate regarding tensile

(11)

strength. In the following chapter the different classes of cast iron are explained and discussed.

2.2.1 Flake graphite iron (FGI)

Flake graphite iron, lamellar or Grey iron is mainly constituted of iron, carbon and silicon.

It´s microstructure is composed of needle-like shaped graphite in an iron matrix. The main characteristic is the high level of connectivity of lamellar graphite in three-dimensions that leads to a better thermal conductivity, compared to spherical graphite. The negative side effect of this connectivity is less attractive mechanical properties, especially regarding fatigue, and as previously stated, tensile strength [3]. A visual representation of FGI can be depicted in Figure 4.

Figure 4: Flake graphite [9].

(12)

2.2.2 Spheriodal graphite iron (SGI)

High amounts of magnesium is needed to obtain spheriodal graphite (ductile iron). Like the name suggests, the graphite is in the shape of a sphere, and as a result of the shape, it is stronger and can elongate more than gray iron. It is resistant to fatigue but not as machinable as FGI. However, ductile iron has worse thermal transfer characteristics and a higher tendency to retain porosity in its microstructure compared to gray iron [3]. A visual representation of SGI can be depicted in Figure 5.

Figure 5: Spheriodal graphite [10].

2.2.3 Compacted graphite iron (CGI)

To obtain CGI, magnesium is added in order to remove flake graphite. This amount is lower compared to the amount used to acquire spherical graphite iron. The result shows individual worm shaped graphite, which are thicker and not as long as flake graphite, connected to their nearest neighbours within the eutectic cell. The interconnected graphite provides better thermal conductivity and damping capacity than spheroidal graphite iron (SGI). It should be added that not all of the graphite is compacted. Less than 20 % is spheriodal graphite, where the rest is compacted graphite [4].

CGI have strength properties close to those of spheriodal graphite, higher fatigue strength compared to FGI and intermediate conductivity. It is therefore possible to state that CGI is intermediate to FGI and SGI in its microstructure and properties. It should be noted that the mechanical and physical properties of CGI are further determined by the pearlite/ferrite ratio.

To control the amount of the respective value, alloying elements need to be added where tin

(13)

and copper are added to promote pearlite. A visual representation of CGI can be depicted in Figure 6.

Figure 6: Compacted graphite [11]

2.2.4 Ferrite and pearlite

Pearlite is proven to promote desirable mechanical properties for iron-carbon alloys [3].

However, with the addition of silicon, ferrite may be promoted, where properties such as increased ductility and impact failure energy, are attained [5]. Addition of molybdenum to a ferritic material, and molybdenum and nickel to a pearlitic material can be performed to obtain desirable properties. However, the change in high cycle fatigue properties has not been well documented, and it is therefore of interest to further analyse the influence of these alloying materials.

2.2.5 Alloying elements Molybdenum

This material is a carbide stabilizer, retards the transformation of pearlite, the pearlite phase field in the continuous cooling transformation diagram (CCT) is shifted towards longer times, and affects the ferrite formation. For a pearlitic microstructure, addition of molybdenum decreases the interlamellar spacing, which increases strength and improves toughness [12].

(14)

Nickel

Nickel promotes solid solution strengthening mechanism, is good for low-temperature toughness and improves high temperature strength. Further, it is resistant to oxidation and corrosion. During machining, the material undergoes a hardening effect because of the pressures of the cutting operation. This property is often not favourable for applications where the material used needs to possess well defined material properties. Therefore, too much nickel might not be applicable to products who require high amounts of machining [13].

Large amounts of nickel does, as mentioned, supply the required hardenability, but combining it with molybdenum is more effective because of the synergetic effects of these two elements.

Silicon

Silicon has a minor effect on the spacing of pearlite. Increasing the Si content, the material resilience decreases, causing brittleness, but also inhibits carbide precipitation during austempering. This helps stabilizing austenite. When silicon is added, it segregates around graphite nodules and is low at the joints of eutectic cells. Thus segregation of the material leads to gradients in silicon distribution and is the reason for initiation and propagation of cracks [14].

2.2.6 Compositions of compacted graphite iron

Seven different materials were examined. Their reference name and desired composition can be seen in Table 1.

Table 1: Materials and their nominal composition in w %.

Material FBas FMo0.15 FMo0.25 PBas PMo0.25 PNi0.5 PNi1

C 3.22 3.22 3.22 3.65 3.65 3.65 3.55

Si 4 4 4 2.3 2.3 2.3 2.35

Mn 0.2 0.2 0.2 0.2 0.2 0.2 0.2

P 0.03 0.03 0.03 0.03 0.03 0.03 0.03

Mo 0.05 0.15 0.25 0.05 0.25 0.05 0.05

Ni 0.05 0.05 0.05 0.05 0.05 0.05 1

Cr 0.01 0.01 0.01 0.05 0.05 0.05 0.05

Sn 0.01 0.01 0.01 + + + +

Cu 0.02 0.02 0.02 + + + +

c_eqv 4.24 4.24 4.24 4.24 4.24 4.24 4.24

Sn and Cu are adjusted to levels necessary for perlitic grades (+).

(15)

c_eqv is calculated from c_eqv = c + Si/4 + P/2. The effect of molybdenum on both pearlite and ferrite was investigated. However, the alloying effect of nickel was only investigated for pearlite. Tin was added to promote pearlite. The composition for the materials attained from the foundry are depicted in Table 2. FBas represents Ferritic reference material, FMo0.15 contains 0.15 % molybdenum and, logically, FMo0.25 contains 0.25 molybdenum. Using the same principle, P stands for pearlitic, which means that the remaining materials are pearlitic, which contain 0.25 % molybdenum, 0.5 % nickel and 1 % nickel respectively.

Table 2: Materials and their composition in w % measured by the foundry.

C 3.17 3.16 3.16 3.59 3.58 3.59 3.55

Si 4.14 4.09 4.06 2.33 2.36 2.34 2.35

Mn 0.14 0.14 0.139 0.318 0.319 0.318 0.316

P 0.027 0.03 0.029 x x x x

Mo 0.001 0.146 0.248 0.01 0.23 0.0 0.00

Ni 0.0123 0.0121 0.01222 0.02 0.02 0.47 0.91

Cr 0.015 0.0153 0.0149 0.022 0.022 0.022 0.022

Sn 0.005 0.005 0.005 0.082 0.084 0.082 0.081

Cu 0.0172 0.0161 0.0154 0.99 0.996 0.998 1.01

ceqv 4.22 4.24 4.19 4.18 4.19 4.19 4.16

CE 4.54 4.54 4.5 4.52 4.53 4.53 4.47

MGM 23.45 27 33.18 28.25 31 28.34 24.39

MGI 34.78 46.34 59.03 49.74 44.37 41.46 49.43

CE, MGM and MGI are casting parameters. There exist rest products during the foundry process from previous moldings and scrap metal. However, the amount of these elements are comparably small and have therefore no, or little, effect on the final properties of the material. For comparison with Table 2, one test specimen at each alloy was sent for chemical analysis at Degefors laboratories, see table 3. The materials entire composition measured by the foundry and Degefors laboratories can be viewed in Table A.1 and A.2, Appendix A.

(16)

Table 3: Materials and their composition in w % measured by Degefors laboratories.

C 3.28 3.25 3.14 3.76 3.70 3.65 3.77

Si 3.88 4.11 3,92 2.44 2.43 2.44 2.43

Mn 0.12 0.13 0.13 0.298 0.3 0.29 0.29

P 0.029 0.033 0.029 0.025 0.027 0,026 0.027

Mo <0.02 0.16 0.23 <0.02 0.22 <0,02 0.02 Ni <0.02 <0.02 <0,02 0.03 0.04 0.47 0.91 Cr <0.01 0.02 <0,01 0.02 0.02 <0,01 <0.01 Sn <0.005 <0.005 <0.005 0.081 0.075 0.075 0.074

Cu 0.007 0.011 0.008 0.91 0.9 0.9 0.88

It is important to mention that comparing Table 2 and 3, some substances are lower than a certain value in 3. That is not because they are not present, but rather because that they are below the level of which the methods used could measure. Further, there are some noticeable differences, more specifically in the carbon, tin and copper levels.

(17)

2.3 Failure mechanism

Since cast iron contains graphite, the failure of the material is initiated by the separation of the steel matrix and soft graphite. Although, this procedure occurs simultaneously at several initiation sites in the matrix. Crack propagation follows separation, where the crack connects to other cracks initiated at other graphite particles. The crack then reaches a critical size where final fracture occurs, which happens rapidly [8, 16]. When failure occurs, it is important to evaluate the fracture surface, if one is obtained. A cross section of a tested specimen can be seen in Figure 7.

Figure 7: The cross section of a cracked specimen.

The cross section shows a very rough surface which is common for cast iron with high carbon content. A problem with this occurrence is the fact that it is difficult to see the crack initiation, making problematic to further evaluate the cracked face without the use of specific tools.

(18)

3 Testing and test rig

Testing is an important part in obtaining the mechanical properties and desired knowledge of materials. The dimensions of the test specimens are given in Figure 8.

Figure 8: The dimensions of the tensile specimen.

There is an area in the middle of the specimen that has a decreased radius in order to obtain localized stress concentrations, guaranteeing that failure of the specimen will appear in this section. The thicker parts are clamped by the testing rig.

(19)

The testing rig is a Rumul electromagnetic one axial machine which can reach a maximum load of 150 kN. By being driven by electricity and magnets instead of hydraulics, it can reach higher frequencies. The load is measured by the machine. The machine evaluates the specimen’s eigenfrequency by ramping up the frequency until the eigenfrequency is reached.

This procedure is done by the trained personnel at Scania, since the ramping time and speed needs to be specified so that the overshoot is not too large. In essence, it is a simplified PID controller. The test rig can be viewed in Figure 9a, and the clamped specimen in Figure 9b.

(a) The test rig. (b) The clamped specimen

Figure 9: The testing rig and the clamping of the specimen.

In order to determine when the specimen has reached failure, there needs to be a definition of when this happens. One way to define this is a reduction in the eigenfrequency as a result of the geometry change and stiffness of the specimen during crack initiation and growth.

The limiting frequency change is set to 3 % of the standard unchanged eigenfrequency. It should be noted that the cycle-limit is set to 2·10⁶ cycles.

(20)

The application of the load is also of importance. A way to define this is R = S_min

S_max, (1)

where S_min is the minimum applied load and S_maxis the maximum applied load [15]. In this fatigue test, R is 0.1. This means that there will always be tension in the test specimen. A visual representation of how the load is applied is shown in Figure 10.

Figure 10: Load application [15].

By deciding what the maximum amplitude is, it is possible to express the maximum and minimum stress as

S_min = S_mean− S_amp (2)

and

S_max = S_mean+ S_amp, (3)

where Samp is the amplitude of the applied stress. As stated before, this is chosen, which together with R, makes it possible to express the mean stress as

S_mean = S_amp(1 + R)

1 − R . (4)

(21)

Inserting (4) into (2) and (3) gives

S_min = S_amp(1 + R)

1 − R − S_amp (5)

and

Smax = Samp

(1 + R)

1 − R + Samp. (6)

Earlier tests have generated reference values regarding the stress that gives 2·10⁶ cycles. To transform stress into a load, the relationship

σ = S

A (7)

where A = πr², is used, with R defined in (1), r being the specimen radius and S_amp as the amplitude stress [16]. The results obtained from the tests were inserted into Matlab to calculate the mean value and standard deviation.

3.1 Microscopy

To classify the microstructures, an optical microscope is used. It makes it able to get a clear representation of the microstructure, where the graphite was classified according to iso 945- 1:2008 [17]. There are three criteria which the graphite microstructure is assessed. Form, distribution and size. The microscope is connected to a computer with a program which performs this evaluation automatically.

3.2 Hardness testing

In order to obtain the hardness of the materials. The Brinell hardness test was employed. It consists of applying a constant load, 187.5 kg, for a time period of 13 s, using a 2.5 mm in diameter hardened steel ball, on a flat part of the examined material. By using the following expression

HB = 2P

πD(D²−pD²− D²_i), (8)

where D is the diameter of the indenter, D_i the diameter of the indentation and P the applied force, the hardness number HB is obtained [18].

It should be noted that the greatest source of error is the measurement of indentation. The edge is often poorly defined which makes it difficult to measure the correct diameter. The procedure is performed in conjunction with a machine to reduce the impact of human

(22)

3.3 Static mechanical properties

Initially, when a stress is applied to a tensile specimen, the strain will be linear, making it possible to evaluate Young´s modulus. However, if the stress is increased beyond a certain limit, plastic strain may occur, and the point where plastic strain occurs gives the yield strength. The classic strain-stress plot is shown in Figure 11.

Figure 11: Tensile test plot with stress as a function of strain.

From Figure 11, E is Young´s modulus, defined as E = ∆σ

∆ε (9)

An important definition is R_p0.2, which is defined as the amount of stress that will result in a remaining plastic strain of 0.2 %. The reason being that the transition from linearity to non- linearity is for cast irons often continous, meaning that there is no singular point to identify it as in the apparent yield observed in for example low-carbon steels. Therefore, the distinction between the elastic and the plastic regime is defined by the Rp0.2criteria, called proof stress.

Ultimate strength is defined as Rm, however, because of plasticity, the strain can go beyond this region until failure is reached, described as A_t. [19]

(23)

3.4 High cycle fatigue

Fatigue failure can be defined in several ways. The formation of a small, detectable crack, a percentage decrease in the tensile load caused by the crack nucleation and growth, a decrease in the ratio of unloading to loading moduli due to the presence of a crack, or final fracture [16].

To get a visual representation of the limit load for different stress levels, an S-N curve can be used. In an S-N curve, the number of cycles to failure is recorded for several different loads.

However, in this report, the fatigue strength at 2·10⁶ cycles has been determined. The test is performed according to ISO 12106 and 12107:2012 standards [16, 19].

In order to reach a high enough statistical confidence for the load exerted on the specimen at 2·10⁶ cycles, some sort of interpolation procedure must be used. Therefore, the staircase method is applied [16].

With high cycle fatigue, the load levels will never exceed the elastic region. However, if plastic strain is reached during the load cycle, the stress-strain will not follow a linear relationship and instead hysteresis curves shown in Figure 12 will be formed [20].

(24)

3.4.1 Staircase method

To acquire an acceptable value of the load limit for 2·10⁶ cycles, an iterative procedure must be employed. One of these is the staircase method. At least six specimens are required for exploratory tests, but at least 15 specimens are needed for reliable data [16]. The stress level is either increased or decreased with a fixed step, depending if the test resulted in a non-failure or failure, respectively.

The load step is set to 5 MPa, which has been employed earlier for similar tests at Scania. An example of how the staircase is implemented can be viewed in Figure 13.

Figure 13: Example of the staircase procedure for the pearlitic CGI with 1.0 % nickel.

The circles represent non-failure, which if obtained, means that the applied stress is increased by 5 MPa. When failure is observed, the load is reduced. The initial results with a cross will not be included in the statistical analysis [16].

3.4.2 Statistical analysis

The results are evaluated by counting the frequencies of failure and non-failure for the different stress levels. The stress levels are arranged in ascending order as S₀<S₁<...<S_l, where l is the number of stress levels for the least obtainable result. This means that the stress levels are those that gave either failure or non-failure. Since the initial stress levels are

(25)

either too high or too low, they are only evaluated one test before the stress levels have reached the first local, or global, peak. The stress step d is 5 MPa. The mean value of the fatigue limit is

µ_x = S₀+ d A C ± 1

2

(10) where ±1/2 is −1/2 when the test resulted in failure, and +1/2 when the test resulted in non-failure for the stress level S₀. The standard deviation is

σ_x = 1.26d (D + 0.029) . (11)

A, B, C and D are defined as

A =

l

X

i=1

if_i, (12)

B =

l

X

i=1

i²f_i, (13)

C =

l

X

i=1

f_i (14)

and

D = BC − A²

C² . (15)

fi is the number of obtainable results for the given stress level i. However, (15) is only valid when D > 0.3[19].

(26)

4 Results

4.1 Microstructure

The microstructures for the the ferritic and pearlitic cast iron are shown in Figures 14, 15 and 16.

(a) Reference ferritic CGI (b) Ferritic CGI with 0.15 % molybdenum Figure 14: Ferritic CGI

(a) Ferritic CGI with 0.25 % molybdenum. (b) Reference pearlitic CGI Figure 15: Ferritic and pearlitic CGI.

(27)

(a) Pearlitic CGI with 0.5 % nickel. (b) Pearlitic CGI with 1 % nickel.

Figure 16: Pearlitic CGI.

Figure 17: Pearlitic CGI with 0.25 % molybdenum.

In Figure 15b, it is possible to observe corrosion, brown spots, which is the result of iron reacting with the air. The size and shape of the microstructures are evaluated according to

(28)

Table 4: Graphite classification with the values being in %.

Shape index FBas FMo0.15 FMo0.25 PBas PMo0.25 PNi0.5 PNi1

I 3.08 2.58 1.7 3.2 3.6 4.5 8.32

I_Dev ± 1.2 ± 0.8 ± 1.2 ± 4.5 ± 1.2 ± 2.5 ± 5.8

II 19.3 22.3 19.7 22.7 17.2 23.2 25.7

IIDev ± 3.7 ± 5 ± 2 ± 9.3 ± 4.4 ± 3.3 ± 3.7

III 36.2 32.1 33.6 39.8 30 32.5 33.1

III_Dev ± 3.5 ± 5.1 ± 2.4 ± 10.5 ± 3.1 ± 5 ± 4.5

IV 23.7 23.3 25 24.6 26.8 22.5 20.4

IVDev ± 2.5 ± 4.9 ± 3.1 ± 4.7 ± 5.4 ± 1.8 ± 3.2

V 9.76 13 10.7 7.76 13.3 8 7.62

V_Dev ± 2.4 ± 2.3 ± 2.9 ± 2.3 ± 1.7 ± 1.5 ± 3.2

VI 7.94 6.72 9.3 1.92 9.2 9.2 4.82

VIDev ± 1.5 ± 2.4 ± 1.6 ± 0.7 ± 2 ± 2.9 ± 2.8

Size index FBas FMo0.15 FMo0.25 PBas PMo0.25 PNi0.5 PNi1

3 0 3.58 0 5.64 3.5 2.2 3.12

3_Dev ± 0 ± 2.6 ± 0 ± 6.1 ± 5.2 ± 1.9 ± 2.7

4 9.86 13.3 13.7 26.4 16,1 13.7 16.9

4_Dev ± 4.6 ± 3.5 ± 4.3 ± 6.8 ± 2.9 ± 4.1 ± 3.7

5 35 37.3 34.4 34.5 40.8 42.2 35.6

5Dev ± 3.8 ± 3.7 ± 1.5 ± 7.1 ± 4.9 ± 5 ± 4.3

6 36.1 30.3 37.4 21.3 27.6 28.2 28.4

6_Dev ± 1.4 ± 4.1 ± 2.9 ± 3.4 ± 3.6 ± 2.7 ± 3.7

7 15.8 11.6 11.8 8.36 8.6 11.3 12.2

7Dev ± 0.5 ± 1.5 ± 1.3 ± 0.8 ± 1.2 ± 0.7 ± 3

8 3.26 3.96 2.8 3.76 3.4 2.5 3.96

8_Dev ± 0.4 ± 0.4 ± 0.7 ± 0.6 ± 0.4 ± 0.4 ± 0.6

Graphite area 13.4 14 13 14.7 11.4 11.8 12.8

Graphite areaDev ± 0.3 ± 0.4 ± 0.2 ± 3.1 ± 0.8 ± 1.2 ± 0.8

Nodularity 13.8 12.9 18 5.48 19.2 15.8 8.38

Nodularity_Dev ± 1.1 ± 3.4 ± 2.3 ± 2.4 ± 3 ± 2.6 ± 4

The results from the graphite classification show greater nodularity for pearlitic and ferritic cast iron with 0.25 % molybdenum. Size index 6 and 7 are also higher for the ferritic- compared to the pearlitic materials. It should be noted that the pearlitic cast iron with 0.25 % molybdenum was observed with islands of graphite. A similar aggregation of graphite could be seen for the ferritic cast iron with 0.25 % molybdenum. In Table 5, the average value and standard deviation for graphite-, ferrite- and pearlite percentages are presented [21].

(29)

Table 5: Graphite, ferrite and pearlite proportion [21]

Graphite [%] 10.9 10.76 10.93 10.75 10.77 10.49 11.09

Graphite_Dev [%] ± 0.04 ± 0.08 ± 0.26 ± 0.18 ± 0.23 ± 0.15 ± 0.12

Ferrite [%] 89 89 89 10.7 10 11.3 9.3

FerriteDev[%] - - - ± 0.94 ± 0.82 ± 0.94 ± 0.47

Pearlite [%] - - - 78.6 79.3 78.2 79.6

The ferritic alloys contain no pearlite and are thus not presented in Table 5.

4.2 Static mechanical properties

The yield strength, Young´s modulus E, R_p0,2, R_m, A_tand Brinell hardness, for the different materials are shown in Table 6. The data is based on averages of three measurements on each material.

Table 6: Material characteristics.

E [GPa] 141 140 143 112 130 120 125

EDev[GPa] ± 0.8 ± 2 ± 0.7 ± 0.8 ± 1.1 ± 1.9 ± 1.4

R_p0,2[MPa] 393 390 396 331 360 347 362

R_p0,2Dev [MPa] ± 1.5 ± 1.2 ± 0.7 ± 0.6 ± 1.2 ± 1.9 ± 1.2

R_m[MPa] 441 435 443 442 494 447 450

R_mDev[MPa] ± 1.1 ± 1.1 ± 1.3 ± 4.1 ± 2.7 ± 1.5 ± 7.8

A_t[%] 5.2 5.6 5.1 2.5 2.5 2.1 1.7

A_tDev [%] ± 0.4 ± 0.4 ± 0.2 ± 0.1 ± 0.2 ± 0.2 ± 0.2

Brinell hardness 186 181 185 229 233 229 239

Brinell hardness_Dev ± 0.5 ± 1.4 ± 0.9 ± 3.9 ± 4.2 ± 1.9 ± 1.2 The results in Table 6 show that by comparing the different levels of added molybdenum in ferritic cast iron, and the different levels of added nickel in pearlitic cast iron, it can be concluded that in order to obtain the higher, or similar material characteristics such as the reference basic materials, the alloying elements need to be above a certain amount, since below this amount, no effect is seen.

(30)

4.3 Fatigue

The results for high cycle fatigue tests can be observed in Figure 18. It should be noted that the materials with the name EN-GJV-400 and 450 are materials in order to have a frame of reference. The results for these materials are obtained from earlier tests performed at Scania.

Further, the number of tests performed for the seven materials were six or higher in order to reach the requirement of exploratory tests. In Figures B.1, B.2, B.3, B.4, B.5, B.6 and B.7, Appendix B, the number of tests for each specimen, with the results used for the statistical analysis, are shown [16].

Figure 18: Results from the high cycle fatigue tests.

From these results, it can be stated that addition of nickel does not improve the fatigue behaviour of pearlitic cast irons. Both the basic pearlitic and ferritic obtained similar, or higher average stress values than the reference EN-GJV-400 and 450 materials. Further, comparing the results for 0.15 and 0.25 % molybdenum for ferritic cast iron shows that molybdenum needs to be added above a certain amount in order to have a positive effect on high cycle properties. The specific values of these tests are summarized in Table 7 and 8.

(31)

Table 7: Results from the high cycle fatigue tests for the ferritic and pearlitic CGI.

HCF [MPa] 118 112 125 106 116 110 105

HCF_Dev [MPa] ± 5.6 ± 9.5 ± 2.3 ± 2 ± 2 ± 2.3 ± 2.3

Table 8: Results from the high cycle fatigue tests for the reference materials.

Material EN-GJV-400 EN-GJV-450

HCF [MPa] 96 109

HCFDev[MPa] ± 2 ± 5

4.4 Comparison between ferritic and pearlitic CGI

Reference ferritic cast iron proves to attain higher stress levels for high cycle fatigue compared to pearlitic cast iron. Further, Young´s modulus, proof stress, and strain at failure At are higher for all ferritic materials, while the pearlitic materials have a higher hardness.

The reference pearlitic CGI attained similar stress levels for high cycle fatigue compared to EN-GJV-450, whereas the reference ferritic CGI and ferritic CGI with 0.25 % molybdenum reached higher stress levels compared to EN-GJV-450.

5 Discussion

A possible explanation for the higher fatigue results for ferritic, and pearlitic, cast iron with 0.25 % molybdenum could be the fact that the nodularity was increased, or that molybdenum acts as a carbide stabilizer. It has been concluded that spherical graphite iron is resistant to fatigue [3]. Comparing E, R_p02 R_m and the fatigue results for pearlitic cast iron with 0.25 % molybdenum and the reference pearlitic cast iron, gives increased confidence in the statement that spherical iron is resistant to fatigue. From theory, it has also been concluded that spherical iron have improved static mechanical properties. However, it should be noted that for ferritic CGI, a similar increase in the for static mechanical results was not observed.

(32)

By plotting the fatigue strength as a function of tensile strength in Figure 19, it is possible to see further connections.

Figure 19: Fatigue strength as a function of tensile strength.

From this, it is possible to see that both pearlitic and ferritic CGI with molybdenum gives improved tensile and fatigue strength. The ferritic material with 0.15 % molybdenum is lower than the reference material, which might be because of the large standard deviation.

The weighted mean aspect ratio of the graphite is a factor of the width and length. If this value goes towards 1, it can be viewed as nodular. This ratio can be plotted with the tensile strength on the y-axis, which gives Figure 20 [21].

Figure 20: Tensile strength versus weighted mean aspect ratio [21].

(33)

Table 9 converts the used material names in this report into the names in Figure 20.

Table 9: Material convert table.

Name F20 F26 F36 P01 P04 P10 P16

Figure 20 shows that pearlitic cast iron with 0.25 % molybdenum has increased nodularity, and the same can be said for the ferritic counterpart, although not as prominent. Thus, there is further reason to believe the thesis that these alloys have a higher nodularity, which in turn improves the fatigue behaviour and static material properties. This could be due to molybdenum, or an effect of different inoculation observed in Table 3.

The results for the ferritic materials showed favourable fatigue behaviour, which may go against results regarding ultimate tensile strength and other material characteristics [3]. It is therefore important to note that the materials used in this thesis are hardened with silicon, giving different material behaviour compared to ordinary ferritic cast iron.

In order to avoid local concentrations, five different pictures at different areas of the cross section for each material were used to evaluate the graphite classification, which in turn resulted in an average value. The area for these five pictures was approximately 4 mm². It should be noted that corrosion was observed, which does affect the automatic process that performed the graphite classification. However, because of the usage of several measurements, and ethanol and cotton in order to remove the majority of oxidation, the effect of corrosion could be reduced.

Ferritic cast iron with 0.15 % molybdenum had a large standard deviation compared to the other ferritic materials. This can be credited to the fact that a local minimum was observed early in the test, see Figure B.2, Appendix B. However, as the tests continued on, the stress reached a higher stress equilibrium. This subsequently resulted in a large standard deviation, which could be the result of the cast irons inherent behaviour to yield variation in material properties.

Because of the extraordinary situation with Covid-19, Scania took measures to reduce the risk of spreading the virus by reducing the amount of working days to two days a week.

Since this thesis was dependent on testing on site, in order to be completed in a relatively acceptable time span. The number of tests for each material was reduced to six instead of 15. As a result, the accuracy of the results are not as high compared to the tests that were done with 15 specimens (ferritic with molybdenum and pearlitic base material) before the decision to do 6 tests, which is the minimum needed for extrapolary tests.

(34)

on mechanical properties. As heat effects the fatigue and corrosion behaviour and could decrease the life expectancy of the material, further tests with the inclusion of heat would not only represent real world applications, for cylinder heads at least, but also give a higher confidence in the results. Further, with increased nodularity, there is a reduction in thermal conductivity, which is not an advantageous property for applications such as cylinder heads in diesel engines. It is therefore not possible to conclude that ferritic CGI with 0.25 % molybdenum should be used for diesel engine applications without further investigation.

6 Conclusion

Based on the results, it is concluded that:

• Ferritic high carbon cast iron is a suitable material for cylinder heads in diesel engines with regards to high cycle behaviour and material characteristics.

• Ferritic cast iron has improved A_tand R_p0,2compared to pearlitic cast iron.

• The addition of molybdenum to ferritic CGI needs to be above a certain level in order to have similar or better high cycle fatigue results and material characteristics with respect to the reference ferritic material.

• The amount of molybdenum needed to give better material behaviour needs to be further investigated.

• The reason why molybdenum improves material characteristics needs to be further investigated, but could be because of the fact that the nodularity increases.

• Nickel in the examined compositions did not affect the high cycle fatigue properties.

7 Acknowledgement

The thesis is done within the framework of a project partly funded by Vinnova, and Fordonsstrategisk Forskning och Innovation (FFI). Vinnova Dnr: 2017-05491.

(35)

References

[1] Y. Liu, Y. Li, J. Xing, S. Wang, B. Zheng, Dong, Wei Li. Effect of graphite morphology on the tensile strength and thermal conductivity of cast iron. Materials characterizationn; 2018. 144; 155-165 [cited date 28 February 2020]. Available from https://www.sciencedirect.com/science/article/pii/S1044580318305540?via%3Dihub [2] Dr. S. Dawson, T. Schroeder Compacted Graphite Iron: A Viable Alternative.

Engineered Casting Solutions AFS; 2000. [cited date 28 February 2020]. Available from https://sintercast.com/library/technical-papers-english/

[3] J.R. Davis Cast irons - ASM speciality handbook. 1st edition. The materials information society. United states of America: ASM International; 1996

[4] R. Ghasemi, L. Elmquist, H. Svensson, M. König, A. E. W. Jarfors Mechanical Properties of Solid Solution-Strengthened CGI. International Journal of Cast Metals Research; 2016. vol 29; nos 1 and 2 [cited date 2 March 2020]. Available from https://www.tandfonline.com/doi/abs/10.1080/13640461.2015.1106781

[5] A. Alhussein, M. Risbet, A. Bastien, J.P. Chobaut, D. Balloy, J. Favergeon Influence of silicon and addition elements on the mechanical behavior of ferritic ductile cast iron.

Materials Science and Engineering A; 2014. 605; 222-228 [cited date 28 February 2020].

Available from https://doi.org/10.1016/j.msea.2014.03.057

[6] T. Matsushita, A. Gómez Saro, L. Elmquist and A. E. W. Jarfors On the thermal conductivity of CGI and SGI cast irons. International Journal of Cast Metals research; 2018. vol 31; NO. 3 [cited date 2 March 2020]. Available from https://doi.org/10.1080/13640461.2017.1379263

[7] C.L. Zou, J.C. Pang, M.X. Zhang, Y. Qiu, S.X. Li, L.J. Chen, J.P. Li, Z. Yang, Z. F. Zhang The high cycle fatigue, deformation and fracture of compacted graphite iron: Influence of temperature. Materials Science and Engineering A; 2014. 724; 606-615 [cited date 28 February 2020]. Available from https://doi.org/10.1016/j.msea.2018.01.025

[8] N. Collin Thermo-mechanical fatigue of cast iron for engine applications.

[Master thesis on the internet]. Stockholm, Sweden: Royal institute of technology; 2014 [cited date 4 March 2020]. Available from https://kth.diva- portal.org/smash/get/diva2:768318/FULLTEXT01.pdf

[9] G. Vander Voort. Metallographic Procedures for Cast Irons. VAC aero international;

2015. [cited date 25 March 2020]. Available from https://vacaero.com/information- resources/metallography-with-george-vander-voort/1444-metallographic-procedures- for-cast-irons.html

(36)

sugar industries applications. Wear; 2015. Volumes 390–391. 113-124. [cited date 25 March 2020]. Available from https://doi.org/10.1016/j.wear.2017.07.007

[11] W. Guesser, T. Schroeder, S. Dawson. Production Experience

With Compacted Graphite Iron Automotive Components. AFS

Transactions; 2001. 01-071. [cited date 25 March 2020]. Available from http://www.tupy.com.br/downloads/guesser/productexpgrafitaauto.pdf

[12] Chen, X., Zhai, Q., Dong, H. et al. Molybdenum alloying in cast iron and steel. Adv. Manuf. 8; 2020. 3–14 [cited date 17 March 2020]. Available from https://doi.org/10.1007/s40436-019-00282-1

[13] J. H. Weber, M. K. Banerjee. Nickel and Nickel Alloys: An Overview. Reference Module in Materials Science and Materials Engineering; 2015. [cited date 25 March 2020].

Available from https://doi.org/10.1016/B978-0-12-803581-8.02572-8

[14] P. Beckley, G.K. Sujan. Silicon Alloys. Reference Module in Materials Science and Materials Engineering; 2016. [cited date 25 March 2020]. Available from https://doi.org/10.1016/B978-0-12-803581-8.02572-8

[15] Elements of Metallurgy and Engineering Alloys. ASM International;

2008. 05224G. [cited date 27 March 2020]. Available from https://www.asminternational.org/documents/10192/1849770/05224GChapter14.pdf [16] Metallic materials – Fatigue testing – Statistical planning and analysis of data.

International Organization for Standardization; 2012. 12107:2012 [cited date 28 February 2020]. Available from https://www.iso.org/standard/41501.html

[17] Microstructure of cast irons – Part 1: Graphite classification by visual analysis.

International Organization for Standardization; 2009. 945-1:2008 [cited date 28 February 2020]. Available from https://www.iso.org/standard/41501.html

[18] Metallic materials – Brinell hardness test – Part 2: Verification and calibration of testing machines. International Organization for Standardization; 2017. 6506-2; STD- 922630. [cited date 12 March 2020]

[19] Metallic materials – Tensile testing – Part 1: Method of test at room temperature.

International Organization for Standardization; 2009. 6892-1:2009 [cited date 28 February 2020]. Available from https://www.iso.org/standard/51081.html

[20] T. Marohniüc, R. Basan, M. Franulovic Evaluation of the possibility of estimating cyclic stress-strain parameters and curves from monotonic properties of steels.

Procedia Engineering 101; 2015. 277-284 [cited date 12 March 2020]. Available from 10.1016/j.proeng.2015.02.029

[21] J. Ahlsén, H. Titus, J. Wrangstål En jämförelse av mikrostruktur och mekaniska egenskaper för kompaktgrafitjärn. Linköpings university 101; 2015. 277-284 [cited date 11 June 2020]. Available from https://liu.se/

(37)

A Appendix - Material composition

Material composition measured by Scanias foundry, Table A.1, and measured by Degefors laboratories, Table A.2.

Table A.1: Materials and their composition measured by the foundry.

Material FBas FMo0,15 FMo0,25 PBas PMo0,25 PNi0,5 PNi1

C 3.17 3.16 3.16 3.59 3.58 3.59 3.55

Si 4.14 4.09 4.06 2.33 2.36 2.34 2.35

Mn 0.14 0.14 0.139 0.318 0.319 0.318 0.316

P 0.027 0.03 0.029 x x x x

Mo 0.001 0.146 0.248 0.01 0.23 0.0 0.00

Ni 0.0123 0.0121 0.01222 0.02 0.02 0.47 0.91

Cr 0.015 0.0153 0.0149 0.022 0.022 0.022 0.022

Sn 0.005 0.005 0.005 0.082 0.084 0.082 0.081

Cu 0.0172 0.0161 0.0154 0.99 0.996 0.998 1.01

Al 0.0043 0.0043 0.0041 x x x x

S 0.0106 0.0114 0.0112 0.0106 0.0106 0.0106 0.0081

Co 0.00206 0.0028 0.0029 x x x x

Nb 0.0049 0.0049 0.0049 x x x x

Ti 0.0034 0.0035 0.0037 x x x x

V 0.0023 0.0023 0.0022 x x x x

Pb 0.0023 0.0024 0.0026 x x x x

Sn 0.005 0.005 0.005 0.082 0.084 0.082 0.081

Mg 0.0059 0.008 0.0064 x x x x

As 0.0028 0.0031 0.0032 x x x x

Zr 0.0012 0.0007 0.0009 x x x x

Ce 0.0157 0.0163 0.018 x x x x

N 0.0042 0.0035 0.0044 x x x x

Zn 0.0014 0.0015 0.0013 x x x x

B 0.0008 0.0005 0.0003 x x x x

La 0.0072 0.008 0.0081 x x x x

Bi 0.0016 0.0018 0.0019 x x x x

Sb 0.0031 0.0035 0.0036 x x x x

W 0.0051 0.0055 0.0056 x x x x

Te 0.0038 0.0042 0.0043 x x x x

Se 0.0026 0.002 0.0025 0.019 0.02 0.017 0.019

Fe 92.4 92.3 92.2 92.5 92.2 92 91.6

c_eqv 4.22 4.24 4.19 4.18 4.19 4.19 4.16

(38)

Table A.2: Materials and their composition measured by Degefors laboratories.

Material FBas FMo0,15 FMo0,25 PBas PMo0,25 PNi0,5 PNi1

C 3.28 3.25 3.14 3.76 3.70 3.65 3.77

Si 3.88 4.11 3,92 2.44 2.43 2.44 2.43

Mn 0.12 0.13 0.13 0.298 0.3 0.29 0.29

P 0.029 0.033 0.029 0.025 0.027 0,026 0.027

Mo <0.02 0.16 0.23 <0.02 0.22 <0,02 0.02

Ni <0.02 <0.02 <0,02 0.03 0.04 0.47 0.91

Cr <0.01 0.02 <0,01 0.02 0.02 <0,01 <0.01 Sn <0.005 <0.005 <0.005 0.081 0.075 0.075 0.074

Cu 0.007 0.011 0.008 0.91 0.9 0.9 0.88

Al <0.002 <0.002 0.002 <0.002 0.004 0,002 <0.002

S 0.01 0.008 0.009 0.012 0.01 0.008 0.009

Co <0.02 <0.02 <0,02 <0.02 <0.02 <0.02 <0.02

Nb x x x x x x x

Ti <0.01 <0.01 <0,01 <0.01 <0.01 <0,01 <0.01 V <0.01 0.01 <0,01 <0.01 <0.01 <0,01 <0.01

Pb x x x x x x x

Sn <0.005 <0.005 <0.005 0.081 0.075 0.075 0.074

Mg x x x x x x x

As 0.001 0.001 0.001 0.001 0.001 0,001 0.001

Zr x x x x x x x

Ce x x x x x x x

N x x x x x x x

Zn x x x x x x x

B <0.001 <0.001 <0.001 <0.001 <0.001 <0,001 <0.001

La x x x x x x x

Bi x x x x x x x

Sb x 0.0035 x x x x x

W <0.01 <0.01 <0,01 <0.01 <0.01 <0,01 <0.01

Te x x 0x x x x x

Se x x x x x x x

Fe 92.56 92.2 92.44 92.36 92.22 92,06 91.53

(39)

B Appendix - Staircase results

Results from the staircase analysis for the specified materials are shown in Figures B.1, B.2, B.3, B.4, B.5, B.6 and B.7. Circles represent no-failure and crosses represent failure. Further, in accordance with the standards for staircase analysis, the results before the first local peak is observed, are not accounted for.

Figure B.1: High cycle fatigue results for the ferritic reference material.

(40)

Figure B.2: High cycle fatigue results for ferritic CGI with 0.15 % molybdenum.

(41)

Figure B.3: High cycle fatigue results for ferritic CGI with 0.25 % molybdenum.

(42)

Figure B.4: High cycle fatigue results for the pearlitic reference material.

(43)

Figure B.5: High cycle fatigue results for pearlitic CGI with 0.25 % molybdenum.

(44)

Figure B.6: High cycle fatigue results for pearlitic CGI with 0.5 % nickel.

(45)

Figure B.7: High cycle fatigue results for pearlitic CGI with 1 % nickel.

(46)

TRITA - SCI - GRU 2020:235

Influence of matrix and alloying on the high cycle fatigue properties of compacted graphite iron for cylinder heads

Influence of matrix and alloying on the high cycle fatigue properties of compacted graphite iron for

cylinder heads

ERIK FROM

Abstract

Sammanfattning

Preface

Contents

1 Introduction

1.1 Background

1.2 Thesis objective

1.3 Delimitations

2 Materials and microstructures

2.1 Casting

2.2 Materials

2.3 Failure mechanism

3 Testing and test rig

3.1 Microscopy

3.2 Hardness testing

3.3 Static mechanical properties

3.4 High cycle fatigue

4 Results

4.1 Microstructure

4.2 Static mechanical properties

4.3 Fatigue

4.4 Comparison between ferritic and pearlitic CGI

5 Discussion

6 Conclusion

7 Acknowledgement

References

A Appendix - Material composition

B Appendix - Staircase results