Memoire Online - Study of tribological properties of titanium-based thin films applied to the rubbing parts of internal combustion engines piston

The evacuation process of a vacuum system is generally non-stationary. During evacuation, the pressure and flow of each point of the installation vary. The evacuation duration is the time it takes to obtain a certain pressure in the vacuum installation.

In the general case, the pumping time can be obtained using the nomogram shown in figure 3.4. To use it, first calculate the quotient V/?? by joining the points corresponding to V and Q on the respective straight lines. The obtained point on the line V/?? is joined to the point of the line P which corresponds to the pressure P which we wish to obtain. The point of intersection on the line t gives the necessary pumping time.

In their normal state, gases are electrical insulators. This is due to the fact that they don't contain free charged particles, but only neutral molecules. However, if they are given fairly strong electric fields, they become conductors. The phenomena that occur are called discharges in gases and are due to the appearance of electrons and free ions. The result of a discharge in a gas is the production of an ionized gas.

The degree of ionization value varies from very low values (of the order 10^-10) to one. When the degree of ionization is equal to unity, it is said that the gas is totally ionized or that it constitutes plasma.

The set of concepts, methods and results of the study of this state of matter constitutes the physics of plasmas. In our environment, the plasma phase is totally absent in the natural state, whereas on the cosmic scale, more than 99.9% of the visible matter occurs in this phase.

When heating a gas at a sufficiently high temperature (of the order of 10⁴K), the mean translated energy of its molecules can become of the same order as their ionization energy. Under these conditions, when two molecules collide, one of them can be ionized.

The probability of ionization as a function of energy is characterized by a threshold. This is the ionization energy. The ionization energies for all species are found in data banks.

- Lowly ionized gases: in which some ions and electrons move in the middle of a sea of neutral molecules. Ions and electrons interact only with neutral molecules;

- Highly ionized gases without interactions between particles: in this plasma the charged particles follow without any collision a trajectory determined by external electromagnetic fields; - Highly ionized gases with interactions between particles.

Coating with plasma involves of modifying the state the surface by one of the three following methods:

- Chemical reaction with the surface itself (oxidation, nitriding) or physicochemical transformation thereof (modification of adhesion, surface energy);

- Erosion of the surface either by chemical action (formation of pores between one or more atoms of the surface) or ion sputtering due to plasma ion bombardment of the surface atoms or chemically-assisted sputtering, which combines ion bombardment and chemical erosion.

The plasma is called «hot» when the temperature of the ions and electrons is high (greater than 10⁷k ). In this case, the medium is completely ionized.

The plasma is called «cold» when the electrons are hot (on the order of 10⁴k), and the ions are cold (from 300 to 10⁴k). The medium in this case is weakly ionized (ionization rate is close to 10^-6until 10^-2per neutral species).

Vapor deposition techniques allow us to have a deposit of thin layers (generally less than 10ìm) with very interesting physical and mechanical characteristics. Generally we distinguish two methods of deposition from the vapor phase: physical vapor deposition (PVD) and chemical vapor deposition (CVD).

In PVD processes, evaporation, sublimation or sputtering by ionic bombardment are used to transform the material of deposition in the vapor phase. These vapors are then condensed on the surfaces. The process can be summarized by the formula:

The absence of a chemical reaction gave the designation: physical deposition.

In CVD, one or more vapors (two in general) are used, which react with each other on a surface to form a defined compound and volatile products of the reaction:

Chemical vapor deposition(CVD)

Physical vapor deposition (PVD)

- Conventional chemical vapor deposition (CVD) - Metalorganic chemical vapor deposition (MOCVD)

- Plasma-enhanced chemical vapor deposition (PECVD)

- physical vapor deposition by evaporation: - PVD by direct evaporation

- PVD by Ionic deposition

- physical vapor deposition by sputtering

3.3.1. Chemical vapor deposition (CVD) techniques 3.3.1.1. Conventional chemical vapor deposition (CVD)

It is an isothermal process, in which the bar to be coated is placed in a closed chamber in contact with a mixture of a donor eventually alloyed (metal powder containing the element to be deposited), an activator (halogen compound) and an inert diluent which avoids self-sintering of the assembly. At 400 ° C, the activator reacts with the donor element to form the metal vapor. The subsequent heating at temperature between 800 ° C and 1100 ° C under a hydrogen atmosphere allows ensuring the production of the coating in a homogeneous manner.

For this process, the environment of the bar to be coated is continuously renewed by the circulation of a gaseous mixture of the material of the deposition formed outside the reaction chamber and carried by forced convection on the surface of the bar.

The need for high temperature in conventional chemical deposits to initiate the different chemical reactions leading to the metal transfer, several paths has been followed to do so, and two are the most developed for industry: the use of organic gaseous species (OMCVD) and the assistance of plasma (PACVD).

The idea is to reduce the temperature of the conventional CVD by substituting the halogenated compounds which have a higher decomposition temperature than the new organometallic compounds.

The advantage of such a process is the lowering of the processing temperature compared with conventional CVD processes, taking the example of chromium deposition, the temperature is

reduced from 1000 °C for the conventional CVD to 500 °C using by bis-benzene chrome Cr(C6H6)2, the rupture of the radical bonding results in the formation of chromium carbide.

On the other hand, the disadvantages of this process are: the high cost, the toxicity for a large number of them and the lack of stability over time.

Electronics is the most favorable field for the development of the OMCVD technique, with layers of some ten of nanometers.

A second way to reduce the temperature of the deposit to avoid any further heat treatment (see table 3.3), is to use the assistance of plasma and the use of chemically activated species such as ions and free radicals. These species are produced in the gaseous phase by electron-molecule collisions; the assistance of a plasma allows to obtain deposition speed varying from 1 to a few tens of ìm/h in a temperature range comprised between 150 and 500 ° C.

Deposition temperature (°C)
materiel	CVD	PACVD
Tungsten carbonate	1000	325-525
Poly silicon	650	200-400
Silicon nitride	900	300
Silicon dioxide	800-1100	300
Titanium Carbide	900-1100	500
Nitride of titanium	900-1100	500

The PVD technique consists of the formation of a coating under reduced pressure in three distinct stages: vaporization of the species to be deposited, transport of these species and finally the condensation and the growth of the deposit. There are two main processes of this technique, those using evaporation and those using sputtering.

Evaporated species (atoms, molecules) propagate in a straight line and condense on the cold zone (see figure 3.5). The transfer of the species takes place in molecular flow under a pressure of less than 10^-2 mbar, the rate of evaporation of the atoms can be found by the hertz-knudsen equation:

a: Stricking coefficient, it is the probability of having an adsorption; P: Gas pressure;

The disadvantage of this process is that the deposits have a weak adherence and are often powdery, because they are made in a vapor phase with a low energy. It is often necessary to heat the substrate towards 300 to 400 °C, or to use subsequent heat treatments to improve the adhesion.

Fig.3.5 Schematic representation of physical vapor deposition by evaporation 3.3.2.2 Physical vapor deposition by ionic deposition

The ionic deposition differs from the direct evaporation by the phase of transfer and condensation of the vapor. In this case the vapor must pass through plasma to arrive into a state more or less excited and more or less ionized. The plasma used to improve the regularity of the deposit on bars with complex shape, energy exchanges are particularly important in the presence of a reactive gas (see figure 3.6).

The vapor phase is created by electron beam at a pressure less than 5 · 10^-4 mbar, the working potential difference in the deposition chamber is between 1 and 5 KV.

By introducing a reactive gas in the vapor phase such as nitrogen or methane, it is possible to produce defined compounds of carbide or nitride. The main application of this technique is the realization of deposition of TiN, (Ti, Al)N and TiC.

In competition with the CVD technique, this technique is widely used for protection against corrosion and erosion, the thicknesses are limited by some ìm due to the presence of residual compressive stresses.

			Chapter 3: Vapor deposition and thin layer characterization techniques

Fig.3.6 Schematic representation of physical vapor deposition by ionic deposition 3.3.2.3. Physical vapor deposition by sputtering

Negative polarization in the order of 1 to 3 kV of an electrode (target) in the presence of a rarefied argon atmosphere at a pressure of about 1 to 10 Pa depending on the size of the reactor, leads to the creation of an electrical discharge between the target and the walls of the reactor. The walls of the reactor connected to the ground, act as an anode. The Ar + ions created in the discharge are accelerated in the cathodic cover and acquire the energy which they release when they impact on the surface of the target. This can cause the ejection of an atom by transfer of momentum, the implantation of the incident ion, the reflection of the incident ion neutralized by charge transfer or the emission of electrons that will serve to maintain the discharge (figure 3.7).

- The spraying rate or yield Y (Table 3.4): it's defined as the number of spraying atom per incident ion. It grows linearly with the ion energy, and inversely proportional to the energy of sublimation of the material;

- The secondary electronic emission coefficient: defined as the number of electrons emitted during impact with incident ion. Its value is close to 0.1 for most metals but reaches very high values in the case of many oxides or nitrides.

Fig.3.7 Main mechanisms resulting from the interaction of an energy ion and a surface

element	Y	element	Y	element	Y	element	Y	element	Y
C	0,12	Ti	0,51	Cu	2,3	Pd	2,08	W	0,57
Al	1,05	Cr	1,18	Zr	0,65	Ag	3,12	Pt	1,4
Si	0,5	Fe	1,1	Mo	0,8	Ta	0,57	Au	2,4
in blue, Metals with very high sputtering yield

Table 3.4 Sputtering yield of different elements by argon ions at 400 eV 3.3.2.4. Magnetron effect

Two main problems result from the diode process. On the one hand, the low ionization rate of the discharge leads to low deposition speeds (<0.1 ìm / h) and, on the other hand, the high heat of the sputtered atoms leads to the synthesis of porous coatings. In order to avoid these two disadvantages, the target is generally equipped with a magnetron device consisting of two concentric magnets of opposite polarities (Fig. 3.8). A pole piece closes the magnetic circuit on one side, while the non-magnetic target to allow the magnetron effect; it leaves the field lines closed within the gaseous phase, which affect the trapping of the secondary electrons and thus increasing their possibility of encountering an argon atom in the context of an ionizing interaction. Dense plasma is then generated at the gap of the magnets, which leads, despite

heterogeneous erosion of the target, to increase considerably the discharge current and, subsequently, a deposition speed in the order of 10 ìm / h.

The formation of thin layers by physical vapor deposition is the result of the condensation of the particles ejected from the target onto the substrate. It is carried out by a combination of nucleation and growth processes described in Figure 3.9.

At the moment of impact on the substrate, the incident atoms lose their kinetic energies and produce the limiting of their ability to diffuse into the substrate. This is true only if there is no external energy supplied to these particles by heating the substrate or ion bombardment. Since they are first adsorbed, they are known as adatoms. Those adatoms move on the surface until the thermal equilibrium with the substrate is reached.

During their displacement, the adatoms interact with one another; creating nucleus called clusters which continue their displacement by developing and colliding with each other. The clusters continue to grow in number and size until a nucleation density known as saturation density is reached.

The next step in the process of forming the thin layer is called coalescence. The clusters begin to agglomerate with each other by reducing the surface area of the uncoated substrate. The coalescence can be accelerated by increasing the mobility of the adsorbed species, for example by increasing the temperature of the substrate. During this step, new clusters may be formed on surfaces released by the approach of older clusters. The clusters continue to grow, leaving only holes or channels of small dimensions between them. Gradually, a continuous layer forms when the holes and channels are filled.

During three-dimensional growth, or Volmer-Weber growth, clusters form and their coalescence forms a film. This mode of growth is usually favored when the atoms forming the deposited layer are more strongly bonded to one another than to the substrate.

The Two-dimensional (2D) growth, or Frank-Van der Merwe growth, is favored when the binding energy between the deposited atoms is less than or equal to that between the thin layer and the substrate. So, the films form atomic layer by atomic layer.

The third mode of growth, called Stranski-Krastanov, is a combination of the two preceding modes: after two-dimensional growth begins, a change in growth mode is observed, while clusters formation becomes energetically favorable.

The growth mode of layers as well as the deposition conditions (substrate temperature, substrate nature, partial gas characteristics, etc.) influence on the crystallographic orientations and the topographical details of the clusters.

In fact, several models based on thin film growth approaches have been developed to study the influence of the deposition parameters on the microstructure of the layers. The first

description of the morphology of thin layers obtained by physical vapor deposition was presented by Movchan and Demchishin (fig.3.11) [MOV69]:

- First zone: the working pressure is greater than or equal to 1 mbar and the deposition temperatures are between 0.1 and 0.2 times the melting temperature of the deposited metal, this zone is favorable for applications or wettability of the surface is sought but not suitable for corrosion resistance or mechanical strength.

- Second zone: is called also the transition zone where the crystallization of the deposit is very fine with good properties in terms of resistance to corrosion and adhesion.

- Third zone: where the columnar crystallization is well identified, this zone is preferred for mechanical applications because of the good adhesion properties obtained.

Thornton [THO77] proposes a model that complements the previous one by considering the argon pressure in the sputtering. Its model shows a transition zone, called zone T, between zones 1 and 2 (fig.3.12). In this zone, the grains have a fibrous form without microporosities.

Later, Messier [MES84] showed that the fundamental process controlling the morphology of the T-zone was not only the gas pressure but also the effect of the gas pressure on the ionic bombardment of the growing film surface during the deposition by cathodic sputtering. He proposed a fifth zone M of morphology consisting of parallel columns with a domed surface (fig.3.13).

After the application of the layer on the sample, this layer must be characterized. First, we must study its morphology using chemical, crystallographic and microstructural techniques (Table 3.5), and then we characterize one or more properties of this deposit (mechanical, optical, electrical, chemical properties, etc.). These properties are generally studied with all

results of morphological studies, in order to draw a conclusion either to modify the deposition process or to apply it in industry.

The greatest difficulty in the characterization phase is that these techniques are indirect. So, the engineer's art consists in crossing the results obtained, to constitute a global and synthetic vision of the most realistic product possible, in order to be able to produce a list of functional characteristics.

Chemical characterization methods	Crystallographic characterization methods	Microstructural characterization methods
- Auger Electrons Spectroscopy
(AES)		- Atomic force microscope
- Energy Dispersion Microscopy		(AFM)
(EDS)		- Scanning Tunnel microscope
- Electron Energy Loss	- Selected area diffraction	(STM)
Spectroscopy (EELS)	(SAD)	- Optical Microscope (LM)
- - Electronic microprobe analysis	- X-ray diffraction (XRD)	- Scanning electron
(EMPA)		microscope (SEM)
- Raman Spectroscopy (RS)		- Transmission electron
- X-Ray Photoelectron spectroscopy		microscope (TEM)
(XPS)

Table 3.5 the classification of main morphological characterization techniques of a thin layer

3.4.1. Chemical characterization methods 3.4.1.1. Auger Electrons Spectroscopy (AES)

Fig.3.14 Distribution of the primary, Auger and backscattered electrons, as well as X-rays under bombardment of a focused primary electron beam of a diameter D

This technique consists in bombarding the layer with an energetic primary electrons (3 to 30 KeV), which penetrates into the matter at depths of the order of some micrometers and excites the atoms of the layer, the de-excitation can be carried out by emission of the secondary electrons, and of the Auger electrons after the relaxation.

These electrons are characteristic for each atom, and the study of their spectrum makes it possible to identify the chemical composition of the deposit, the atlas of the spectrums of the Auger rays is available in the literatures.

The AES technique is used to analyze the most superficial atomic layers for conductive and semiconductive samples.

This technique involves bombarding the layer by a beam of primary electrons; this beam can cause the emission of X-rays. The emitted spectrum breaks down into two radiations: a continuous radiation and a characteristic radiation which created after an inelastic impact between the atom and the electron. The characteristic radiation is specific for each element, and its detection allows for chemical analysis.

This technique involves exposing sample to a primary electron beam; some of these electrons will be subjected to inelastic impacts with the atoms of the sample, which causes the loss of their energy and a weak or random deflection of their trajectory. This loss of energy can be measured, and the spectrum can be used to determine the chemical characteristics of the sample.

This technique is based on the analysis of the X-rays emitted by the atoms of the sample after inelastic impacts with the electron beam. The X-radiation is sent to a monochromator which is a single crystal having well-known crystalline planes. This radiation is then diffracted, and the diffraction angle varies as a function of the wavelength.

The Raman effect results from the slight change in the frequency of the monochromatic light projected on the sample, this modification is caused by the vibrations and / or the rotations of the molecules. The spectrum analysis of the scattered light gives information on the molecular composition of the sample, because the spectrums of the vibrations are characteristic for each molecule.

Also called ESCA (Electron Spectroscopy for Chemical Analysis), this technique involves bombarding the sample with X-rays; the photons y of these rays enter the sample and ionize the atoms. The photons have kinetic energies E?? which correspond to the ionization energies E of the different orbits of the atom by the relation:

The energy of the electrons in each orbit is specific for each atom, so the knowledge of the energy of the projected photons h?? and the analysis of the kinetic energy of the photons makes it possible to identify the chemical composition of the sample. The average analysis depth by this technique is from 1 to 5 nm.

Crystallographic analysis techniques are based on the principle of X-ray diffraction (XRD) or electron-associated wave (SAD). These techniques make it possible to determine:

A primary beam of electrons passes through a thin sheet; the wave associated with the electrons is submitted to diffraction on the crystalline planes of the sample. The transmission electron microscope TEM allows obtaining the image and the spectrum of this diffraction. The SAD technique makes it possible to obtain the information for very small crystalline grains.

The X-rays are diffracted on a crystalline plane according to Bragg's law: A = 2d sin?? (3.19)

The diffracted X-Rays wavelengths vary between ë = 0.071 nm (source with molybdenum anode) and ë = 0.154 nm (source with copper anode).

The diffraction experiments are done differently for thick layers (more than 10 ìm) and for thin layers. The thick layers are tested with a diffractometer in which the radiation source and the detector move circumferentially around a sample, placed at its center, the radiation penetrates into the interior of the layer and the signal of the substrate is low. This technique should not be applied to thin films because the signal of the substrate would be dominant compared to that of the deposit. In this case, the surface diffraction technique (GRXD) in which the incident beam interacts with the surface of the layer at a small angle varying from 1 to 10 ° so that to have a penetration depth of less than 1 ìm, the detector moves to obtain the diffracted signal.

The microstructure analysis concerns the study of the layer surface (AFM, STM) as well as the study of grains inside the deposit (LM) or the internal grain morphology (TEM).

The STM is used to characterize the surface of conductive or semiconductive materials. This microscope has a tip whose end has the size of an atom.

The principle (figure 3.17), is to set a distance less than 1 nm (interatomic distance) between this tip and the surface of the sample, which allows the exchange of electrons and cause an electrical voltage of U = 2mV to 2V, the resulting current is called the tunnel electron current. Since this current is very sensitive to any changing in the distance between the tip and the surface, so this distance must be kept constant. This principle allows obtaining a map with a very good vertical resolution of the surface.

Unlike the STM, the AFM was developed to obtain surface profiles of non-conducting materials. The tip of the probe is equipped with a small diamond that touches the surface of the layer with a small force varying between F = 10^-6 and 10^-9 N to avoid any damage to it. The diamond is attached to a lever of which while the displacement occurs, the measurements are made by very sensitive sensors (fig.3.18).

Fig.3.18 Operating principle of the AFM 3.4.3.3. Optical or Light microscope (LM)

The use of the optical microscope requires special preparation of the layer surface. This microscope allows observing the porosity and the distribution of the pore size, the size of the unfused particles and their distributions, the thermal and mechanical deformation of the substrate at the interface with the deposit, the phase distribution in a composite deposit, etc. . .

The operating principle of the SEM is the projection of a primary electron beam which scans the sample surface and which accelerates by a voltage of U = 1 to 30 kV. The beam of electrons interact with the atoms of the sample either by inelastic impacts followed by their ionization and by the generation of secondary electrons with a kinetic energy of E?? < 50 eV coming from the superficial region of the deposit (some nanometers), or By elastic impacts with these

atoms and the generation of backscattered electrons with a kinetic energy of E?? > 50 eV coming from the deeper regions (hundreds of nanometers).

The secondary electrons are detected by a scintillation detector equipped with a grid carried at the positive potential to attract these electrons at low energy. The backscattered electrons are detected by a semiconductor detector.

The operating principle of this microscope is the emission of a beam of primary electrons passing through the sample. The image in this microscope is obtained by electrons that have not interacted with the sample. The electrons are accelerated by a voltage of several hundreds of kV to be able to cross slides, which must have a thickness less than 100 to 200 nm.

Three types of hardness are defined depending on the depth of indentation (Table 3.6). The loads used depend on the type of hardness. In nanohardness these charges range from some micronewtons to several hundredths of millinewtons, while microhardness requires loads ranging from a few tens of millinewtons to a few Newtons, in macrohardness the loads range from a few Newtons to several tens of Newtons.

The use of each type of hardness depends on the material being tested, so the nanohardness is used to characterize thin layers with thicknesses of less than one micrometer, whereas the microhardness used to study certain surface treatments such as shot blasting or Thermo-chemical treatments (carburizing, nitriding, etc.). The macro hardness is used to measure the hardness of the steels after a heat treatment for example.

With the appearance of nanohardness since 1990, it was possible to understand different mechanical characteristics of materials at the sub-micron scale. However, although this test is fairly simple to carry out in principle, its technological implementation remains delicate, and the interpretation of the results is not simple. Indeed, it is a test, although extremely powerful, which requires a particular attention during the testing, and the analysis of the information.

The nanohardness test consists in the simultaneous measurement of the force F applied to the indenter and of its penetration h into the material under investigation. The representation F = f (h) for the charge and the discharge constitutes the nanoindentation curve.

The inverse analysis by finite elements analysis of the indentation curves allows, to some extent, to have mechanical properties such as: modulus of elasticity, hardness tensile strength, etc.

The interpretation of the measured characteristics (and in particular the hardness) by the nanoindentation test requires simultaneous consideration of the ISE (Indentation Size Effect) and the exact shape of the indenter tip (Berkovich or Vickers), and without forgetting the effects of roughness.

We can observe hardness variations of the order of few percents due to the presence of internal stresses in the materials tested, this is particularly true for thin films.

It is assumed that the adhesion of protective film is an indispensable condition, because of that, the deposition of a coating film is always preceded by a mechanical and / or chemical treatment intended to clean and activate it in order to optimize the adhesion of the film to the substrate. It is also often the practice to deposit an intermediate film (bonding layer) between the substrate and the final coating, for example, a layer of silicon is deposited on steels before being coated with a DLC layer.

For the characterization there are two main families of techniques: non-destructive and destructive techniques. The first uses an optical or acoustic probe that explores the coated material and detects any defect or inhomogeneity (cracks, porosities, bubble etc.) at the film-substrate interface. The most used amongst these methods are the ultrasonic methods.

- Techniques used for ductile or weakly adherent coatings (paints, varnishes, polymers, etc.): like the peeling test, the swelling test...

- Techniques for metallic or ceramic coatings or when the adhesion is strong: as the scratch test.

This test consists in applying a force F, at an angle è, to a band of length b, deposited on a substrate (Figure. 3.19). This force increases gradually over time until the force F?? corresponds to the initiation of the flow of the film. The operation is then continued at constant speed until the complete flow. At the end of the test, it is necessary to check the condition of the uncoated strip and that of the substrate in order to check that there is no elongation or plastic deformation and that all the energy expended during the flow of the band interface was used to break the bonds at the substrate-band interface.

This test consists in making an opening in the substrate in such a way to retain the initial thickness of the film. A pressure P on the deforming film is then applied using an incompressible fluid. The film will begin to swell progressively up to a certain critical height ???? corresponding to a pressure P_?? and then begin to peel off from the substrate (Fig.3.20).

This test consists in applying, perpendicularly to the surface to be tested, a sharp Rockwell-type indenter (cone with an angle of 120 ° and a radius of curvature of 200 ìm) supporting a load of between 1 and 200 N which increases linearly over time with a loading rate of 10?? · m????^-1. When the indenter sinks into the surface of the material, it is made to slide with a velocity of 10 mm. m????^-1 (fig.3.21).

In the production of a high-temperature coating ( PVD or CVD), at the time of return to ambient temperature, due to the difference in the coefficient of thermal expansion between the coating and the substrate, internal stresses can appear in the film, and the deposition-substrate assembly can deform if the substrate is sufficiently thin (some tens to a few hundreds of micrometers thick), in this case the direction of convex or concave deformation indicates, whether they are tensile or compressive stresses .

The internal stresses have an important role because they can influence both the hardness of the coating, its adhesion to the substrate and also its resistance to wear and cracking.

The two main techniques used for the determination of internal stresses are X-ray diffraction and the Stoney method. The two methods consist in reducing the thickness of the substrate by machining to such a value in which the action of the residual stresses is manifested.

- The Stoney method (the determination of the internal stresses by measuring the radius of curvature):

From this technique, the mean internal stresses ó can be calculated using the expression:

h??, ????, ????: Are the thickness, the Young's modulus and the Poisson's ratio of the substrate;

This relation makes sense only when the ratio (h??/h??) is of the order 5 to10 % .

PVD and PACVD thin films are used to improve friction and wear resistance. Table 3.7 gives the main mechanical and tribological characteristics of the layers produced:

coating	TiN	(Ti,Al)N	TiCN	CrN	DLC
Color	Yellow gold	black	purple	silver	back
Hardness (HV)	2300 à 2500	2500 à 3200	3000 à 3400	1800 à 2200	3500 à 5000
oxidation résistance (°C)	400	800	300	600	400
Elaborating temperature (°C)	250 à 400	450	450	600	200 à 400
thickness (um)	2 à 5	2 à 5	2 à 6	3 à 8	1 à 4
Dry friction on 102 Cr6	0.55 à 0.65	0.50 à 0.60	0.45	0.40 à 0.55	0.05 à 0.07

- The coating of (Ti,Al)N is the most resistant to oxidation, so it will be used under severe thermal conditions.

- The DLC coating is the one which have the best tribological characteristics because of the presence in its microstructure of a large proportion of graphite bonds; it is therefore used for delicate lubrication conditions.

- The TiCN coating has a good hardness property, with good conductivity; it will therefore be used with strong mechanical stresses such as steels.

The piston used in our study as a substrate is HATZ E780. The materials used to make this piston are: Cast iron and aluminium-silicon eutectic alloy, the main production processes are: melding, forging [GER/https].

According to the XRD analysis it can be ensured that the material used for piston manufacture is the eutectic Al-Si alloy fig. 4.2.

The silicon content in aluminium for a eutectic Al-Si alloy is varied between 11.7 % and 14.5 % the average value is generally 12.5%. The most commonly used eutectic alloys for the manufacture of pistons are given in table 4.1 [EUR/http]:

Continent	4032 (AW-AlSi12,5MgCuNi)	4047A (AW-AlSi12 (A))	4045 (AW-AlSi10)
Si %	11-13.5	11.0-13.0	9.0-11.0
Fe %	1.0	0.6	0.8
Cu %	0.8-1.3	0.3	0.3
Mn %		0.15	0.06
Mg %	0.8-1.3	0.1	0.05
Zn %	0.25	0.2	0.1
Ti %		0.15	0.2
others %	0.2	0.2	0.2
Al %	Rest	Rest	Rest

The piston with cylindrical shape is cut in the workshop in order to obtain samples with dimension of (20×10×6) mm; the cutting operation is divided into two stages:

First stage: Consists of piston separation in two sections (Head and skirt) By means of a metal chainsaw

Second stage: Consists in cutting the upper part of the piston obtained by the preceding method with the aid of a hacksaw into samples of well-defined size.

In the process of surface preparation operations, polishing is essentially a finishing operation, the purpose of which is to:

- reduce roughness and surface topology by eliminating surface defects such as microcracks, porosities and inclusions.

Using a MECAPOL 230 polisher (fig.4.4), The 304 stainless steel substrates and the Al-Si substrate are polished, using abrasive papers with particles size varied from (200 to 1200) it is a kind of sandpaper, but with a larger particle size Fine and controlled.

Obtaining of the polished surface finish is done progressively (from the big size to the small one) in the presences of the water for the cooling and the evacuation of the debris.

- The polishing head makes it possible to process up to 6 samples Ø 50 mm flanged in a plate by central pressure.

The purpose of chemical cleaning is to remove contamination or impurities formed after polishing to ensure good adhesion of the deposit.

The samples are immersed in an acetone bath for about 5 minutes; then rinsed with distilled water and then dried with the aid of compressed air.

In our work, we will deposit a thin layer on three types of samples with different substrates in

- Five samples of a 304 stainless steel substrate on which an Al-Si layer is deposited;

We use the PVD sputtering technique to deposit thin layers of Ti-W-N on Al-Si, stainless steel 304 and stainless steel with intermediate layer of Al-Si, the evaporation technique was used to create the intermediate layer

1) Bottle of nitrogen gas - 2) Nitrogen passage line - 3) Primary pressure gauge - 4) secondary pressure gauge - 5) Primary and secondary pump ignition switches - 6) Secondary Pump Cooling Water Pipe - 7) deposition reactor - 8) Secondary diffusion pump - 9) Pallet primary pump - 10) Gas pressure control valve (nitrogen and argon) - 11) Throttle valve - 12) Power supplies

The deposition chamber is cylindrical in shape, with a high of 230 mm, diameter of 210 mm, the interelectrode distance is 45 mm, and the lower part of the chamber is connected to the pumping system.

The primary pump is an Alcatel-type pump that can achieves a vacuum of up to2 · 10^-2???????? (2.67 · 10^-2????????) And a flow rate of0.5 à 70 ????³ /?? .

The secondary diffusion pump used allows a vacuum to reach 10^-5???????? (1.33 · 10^-5????????) And a flow rate of 250 ????³ /?? .

The primary pressure gauge is Priani type, it measures pressures up to 10^-3Torr (1.33 · 10^-3mbar). The secondary pressure gauge is a Penning type gauge, the measured pressure can reach10^-5Torr (1.33 · 10^-5mbar).

The power generator supplies the system with a maximum electrical power of 20 Watt, a voltage of 2 kV and an electrical current of 10 mA.

- The application of argon alone for 5 minutes with a pressure of 9.5 · 10^-2Torr ; (0.127 mbar). Argon ions spray directly the target, allowing surface cleaning;

- The introduction of nitrogen into the plasma for one hour with the pressure of 0.1 Torr (0.133 mbar);

a)		b)
Fig.4.6: the deposits obtained on the PVD sputtering installation: - a) on Al-Si, - b) on stainless steel, - c) on stainless steel with intermediate layer of Al-Si	c)	b)

- in order To have a thin layer on stainless steel similar to aluminium-silicon eutectic alloy (of 12.5 % Of silicon content see chapter 2), il faut Put on the crucible 60 mg of powder contains 52.5 mg of the aluminium powder and 7.5 mg of the silicon powder.

1) Primary pressure gauge - 2) Secondary pressure gauge - 3) Primary and secondary pump ignition switches - 4) Secondary Pump Cooling Water Pipe - 5) Deposition reactor - 6) Secondary diffusion pump - 7) Pallet primary pump - 8) Throttle valve - 9) To power supply

Fig.4.8: Inside the reactor chamber: a) Al-Si powder on the crucible, b) The samples of stainless
steel on the substrate port

An indispensable complement to elemental chemical analysis for the identification of compounds, X-ray diffraction allows the fine characterization of massive crystallized materials

Fig.4.9: Diffractometer D8 ADVANCE of CDTA 4.4.1. Characteristics of the device

X-ray diffraction gives access to many information contained in the arrangement of atoms within a crystallized material and The 3D geometrical type of arrangement (network), the distances between atoms (mesh size, typically a few Å) which constitute schematically a "unique" identity card for each compound.

The BRUKER D8 Advance diffractometer is equipped with a BRAGG-BENTANO geometry goniometer. In this type of diffractometer, an incident monochromatic X-ray beam is diffracted by the sample at certain specific angles, According to Bragg's law: _2dhklsinè = A, in which dhkl Refers to the interreticular distance of family plans (hkl), è Is the angle of incidence taken from the surface of the planes (hkl) ,ë The wavelength of the scattered photons.

As Lambda does not vary during a measurement, it is enough to vary the angle Theta to locate all diffraction angles. By means of a converter - a scintillation counter, the intensity of each point of the measurement is thus observed.

When the X-ray beam is diffracted, its representation is a peak. A Theta scan produces an X-ray diffraction pattern.

The recording of the signal by a suitable detector makes it possible to visualize the angles and Intensities of the diffraction peaks obtained. The indexing of these peaks is carried out using specific databases that have more than 350 000 records. Allowing the identification of the compound (s).

During the interaction of a light beam with the material, several phenomena can occur. A part of the light beam is reflected, a part is diffused and a part can be transmitted through the sample (figure 4.11). During propagation in a dense medium, different phenomena appear: refraction, absorption, diffusion, and possibly other non-linear effects. The absorption may then induce photoluminescence or non-radiative de-excitation processes.

The scattering of light is manifested by the deflection of a part of the light beam in multiple directions. The majority of the scattered light is of the same energy as the incident light. This phenomenon of elastic diffusion is called Rayleigh scattering. However, a small portion of the scattered light (about one photon out of 106) present gain or loss of energy relative to the incident light. This is the phenomenon of Raman scattering. In a classical approach, this phenomenon of inelastic diffusion is explained by the creation of an induced dipole which oscillates at a frequency different from that of the incident light. Indeed, under the action of a monochromatic electromagnetic wave of frequency ù whose electric field oscillates according to:

The Stokes diffusion and the anti-Stokes diffusion would be of the same intensity. However, experimentally the intensity of the Stokes diffusion is always (outside resonance) higher than that of the anti-Stokes diffusion.

The Raman spectrum of a compound indicates both the type of binding of a compound and its crystalline structure.

The scanning electron microscope provides surface images of virtually all solid materials at scales ranging from the magnifying glass (x10) to that of the transmission electron microscope (x 500,000 or more).

Conventional SEM operates in an ordinary vacuum (10-5 to 10-6 mbar); The samples may be massive, ranging in size from a few ìm (particles) to about tens of cm in diameter, or even more (industrial samples). They must endure the vacuum. The preparation is generally simple.

The SEM at low pressure allows observation in a vacuum of up to 30 mbar, making it possible

to examine moist or oily samples, insulators without prior metallization (ceramics, corroded metals).

? The SEM is coupled to the EDS for the elementary microanalysis.
? Maximum voltage 30 kV.

The indenter used on our characterization equipment CSM_NHT (Nano-Hardness Testers is a Berkovich indenter (pyramidal with triangular base geometry). We can apply a normal force between 0.3mN and 500mN.

?? = 2 ??

During an indentation test an acquisition system records the applied force as a function of the penetration depth of the tip. These two parameters are measured during a charge phase and a discharge phase. The result is a load-displacement curve.

The two main properties measured by the nanoindentation are the modulus of elasticity (E) and the hardness (H). The model used for the calculation of these two properties is that developed by Oliver and Pharr. According to this model the hardness is given either geometrically (fig.4.16) or analytically by:

A_c : The contact area between the indenter and the sample which is given in the case of a Berkovich indenter by:

In our nanoindentation tests, we used the Oliver-Pharr model with the following parameters:

This device is equipped with an optical microscope to select the area to be indented. An X-Y

motorized table with a repositioning accuracy of 1 ìm allows the programming of complex

On the CSM tribometer (fig.4.18), a ball, a tip or a plane is placed in contact with the surface of the sample under a predetermined load. The device is mounted on a lever arm, and associated with a displacement sensor.

The coefficient of friction is determined during the test by measuring the deflection of this elastic arm. The wear rates for the ball and the pawn are calculated by determining the loss of volume during the test. The formula for determining the wear rate WS is given by:

Tribometer analysis facilitates the study of friction mechanisms for a wide variety of material couple with or without lubricating agent. In addition, the control of test parameters such as speed, contact pressure, frequency, test duration and environmental parameters make it possible to reproduce the actual stresses of use of these materials.

Instruments can provide rotational or alternative movement of the sample. A particular point of these instruments is based on the possibility of interrupting the test as soon as the coefficient of friction reaches a predefined value or, when a number of cycles is realized. In addition, the tribometer is equipped with a containment enclosure in order to use the instrument under controlled atmospheric temperature conditions.

	Nano	Micro
Normal Strength Range	50 ìN - 1 N	until 60 N
Maximum tangential force	10 ìN - 1 N	10 N
Maximum temperature	-	1000 oC
Rotation speed	1 - 100 rpm	0.3 - 500 rpm
Rotating test radius	30 ìm - 10 mm	30 mm
Linear travel speed	10 - 500 ìm	60 mm
Length of linear travel	Until 10 mm/s	until 100 mm/s
Frequency	0.1 - 10 Hz	1.6 Hz
Penetration depth measurement	20 nm - 50 ìm	until 1.2 nm

The polarization measurements were carried out in a three-electrode glass cell: a working electrode, a platinum counter electrode and a saturated calomel reference electrode (SCE). This cell, shown in figure 4.18, is designed to maintain a fixed distance between the three electrodes. The passage of the current in the cell is carried out through the counter-electrode.

Electrochemical measurements were conducted using a computer-controlled, potentiostat, model parstat 4000 (Fig.4.19).

The purpose of the EVANS diagram is to rapidly obtain, by a laboratory measurement, the corrosion rate and the corrosion current density denoted _icorr from the polarization curves

The potential measurement is carried out in a given aggressive medium at a given temperature by means of a potentiostat to which are connected three electrodes.

?
· The reference electrode E.R which serves as a reference for the dissolution potential, ?
· The counter electrode C.E, which is a platinum electrode used to drain the major part

of the current flowing through the circuit when the working electrode is subjected to an

To measure the corrosion current density, the representation E = f (i) is modified, we switch
coordinates to E = f (log / i) or EVANS diagram in order to linearizes the polarization curve and

makes it possible to determine Experimentally the value of corrosion current density _icorr. THE CURRENT is obtained by extrapolating the linear parts of the anodic and cathodic polarization curves. This point of assistance, in cases where it is quite clear, gives access to the characteristics of corrosion (corrosion potential and the corrosion current density), these two linear parts, respectively called the anodic and cathodic Tafel branches.

We see clearly on this diagram the Tafel domain in which the potential varies linearly with the logarithm of the intensity according to the equations of Tafel

EF_e,i0(Fe),EH, i0(H2/Fe) are respectively The equilibrium potentials and the exchange current of the couples Fe2+/Fe et H+/H2 ; b_a et b_c Are the Tafel constants of oxidation and reduction, (ba>0)

This graphic construction therefore provides access to these two important corrosion data. The calculation of Tafel's equations is based on equality:

The phases of the thin film were evaluated by the XRD analysis, and the results are shown in the figure 5.1.

We can see that the deposited layer is an amorphous layer which does not have a crystallographic alignment with a random orientation of the grains, so it is inaccessible by X-ray diffraction.

Fig.5.1: XRD spectra of the Stainless steel bars with and without a layer of Ti-W-N

According to fig. 5.1, a bump can be observed on the angular value 2è = 23.5°, which corresponds to tungsten silicon (Si-W) in its amorphous form, which is requiring a subsequent heat treatment to obtain the crystal structure of layer.

Fig.5.2: XRD spectra of the Stainless steel bars with and without double layers of Al-Si and Ti-W-N

In the XRD spectra of fig.5.2, we can see again the bump correspond to the silicon of tungsten (Si-W) at 2è = 23.5 °, we notices also iron peaks belonging to the substrate which are iron á in 46 ° and iron y at 44 °, which means that the covering of the surface is not perfect and the formation of an amorphous layer.

Fig.5.3: XRD spectra of the Stainless steel bars with double layers of Al-Si and Ti-W-N and with a layer of Ti-W-N

Figure 5.3 shows that the intermediate layer deposited by evaporation between the substrate and the Ti-W-N layer influences in the layer diffraction properties in a manner which permits the appearance of peaks corresponding to the substrate.

Fig.5.4: XRD spectra of Aluminum-Silicon bars with and without a layer of Ti-W-N

The spectra obtained by the XRD analysis of the Ti-W-N layer on the aluminum substrate shows that a layer with peaks similar to that of the substrate with an amorphous structure was formed.

The results obtained by the RAMAN on the Ti-W-N layer for different substrates are given by the following curves:

Fig.5.7: RAMAN specter of stainless steel bar with double layers of Al-Si and Ti-W-N

It is good to notice that the spectrum obtained for different bars are almost the same, which means that the same elements have been observed in the deposited layer for different substrates.

- Peaks with the wavenumber 125 ????^-1 and 675 ????^-1 can be corresponded to titanium dioxide ????02 peaks [IOP/http];

- Peak with a wavenumber close to 62.5 ????^-1 can be corresponded to a titanium nitride ?????? peak [CHE94];

- Peak with a wavenumber close to 807 ?`??^-1 can be corresponded to a tungsten oxide W2⁰3 peak [COU64];

- Peak with a wavenumber close to 172 172 ????^-1 observed in the spectra of stainless steel with double layers of Al-Si and Ti-W-N can be corresponded to an aluminum oxide ????203 peak [AND02].

From the RAMAN spectrum results, it can be said that our layer contains titanium Ti, tungsten W and nitrogen N, in the form of tungsten oxide, titanium nitride and titanium dioxide. All of these peaks (except the ????02 peak in stainless steel substrate) are observed in different bars.

In the contrary of the XRD results in which amorphous results are obtained, the RAMAN results are identified all the elements of the thin layer, which can be explained by the operating principle of each instrument. The XRD gives results for a crystallographic analysis, so it allows us essentially to determine the crystalline composition of the layer, which is practically difficult without a recrystallization of the deposition by an appropriate heat treatment.

RAMAN, on the other hand, is used for chemical analysis. It allows us directly to observe the chemical composition by the diffusion of the light at the frequency of the radiation caused by the vibrations of the molecules.

Figure 5.8 shows pictures from the Scanning Electron Microscopy of a stainless steel bar with the double thin layer of Al-Si and Ti-W-N.

In the picture b) we observe the formation of a layer made up with a small nanoparticle of the order of 50 nm in diameter, and particles of large diameter which can reach 500 nm incorporated in the surface, the interaction of these particles creates a form of deposition called the clusters, those clusters collide one another in order to increase their number and size until the coalescence density where the agglomeration commences, and the uncoated substrate surface reduces.

Fig.5.8: SEM pictures, magnification of the double layers Al-Si and Ti-W-N on stainless steel with apparition of

The morphology and the topography of the layer were studied by the SEM on a cross section of a silicon bar with the thin layer (fig.5.9). The cross section of the bar shows a columnar structure terminating with domes at the surface, with grain diameter between 50 and 400 nm.

The mean thickness measured is about 1.1 rim, which lasted one hour of deposition, what means that the growth rate is equal to 0.30 nm/ s. That rate is a very large, which means that the deposition is not an atom by atom layer. It is reasonable to assume that non-uniform nucleation at the substrate surface at the beginning of deposition is the main cause of the nonuniform growth of the layer, which cause to the random Nano-crystallization, and this can explain the XRD results.

The charge-discharge curves of different bars for different depths and effort are given in the

annex, the mechanical properties obtained for each test are given in the following curves, besides the tables of the results are given in the Annex II.

Fig.5.10: Scatter plot of modulus of elasticity as a function of the max depth (nm)

Fig.5.12: Scatter plot of modulus of elasticity as a function of the max depth (nm)

Fig.5.14 scatter plot of modulus of elasticity as a function of the max depth (nm)

There are in fact four factors to optimize in order to increase the hardness of a layer: the nature of the material of deposition; its microstructure and its architect which are more or less interdependent, and also there is the ionic bombardment during the growth of the film.

If we make a comparison between The hardness values obtained and the hardness values of layers with titanium, we find that our values are of low hardness, this can be explained by the low content of interstitial element (nitrogen) which is 5%, what means a large percentage of

substitutional elements (the transition elements of titanium and tungsten). Indeed, when the content of transition elements increase, the average size of the crystallites decreases [TAK08/A], the decrease in the size of the crystallites is the origin of the decrease in the hardness. This effect is found in the Hall and Petch law, which reflects the fact that the tensile stress of material is inversely proportional to the square root of the mean crystallite size [TAK08/A].

Fig.5.16: Scatter plot of modulus of elasticity as a function of the max depth (nm)

We can see in this latter bar, that its hardness is very high compared to that measured in the preceding bars and reach a value of 17172 MPa. Indeed, large harnesses are measured in the

case of multilayers, and the cause of this phenomenon is generally attributed to a combined effect including the blocking of the dislocations at the interfaces between the layers and the intrinsic structural differences in each layer [CIE99].

5.5. Interpretation of tribometer results 5.5.1. The coefficient of friction

The coefficient of friction is not an intrinsic property for a single material, but it is a value which depends on a couple of materials. In our study, the antagonist material of our bars is the aluminaAl2O3. The conditions of the study are given in the previous chapter.

Figure 5.18 shows the coefficient of friction variation for the Al-Si bare in terms of distance traveled, and Table 5.5 shows the statistical results of the coefficient of friction obtained.

Fig.5.18: Curve of the coefficient of friction variation for the Al-Si bare as a function of distance traveled

According to the curve, it can be seen that, from 0 m to 7.4 m, the coefficient of friction value is low (between 0.06 and 0.26), after that it increases abruptly to reach higher values (between 0.62 and 0.85), this is due to the existence of an aluminum oxide (alumina) layer on the substrate surface with very low thicknesses, and which has a low coefficient of friction. If the values corresponding to the alumina are eliminated, the following results are obtained:

Fig.5.19: Curve of the coefficient of friction variation for the Al-Si bar as a function of distance traveled
without the alumina values

The curve in Figure 5.20 shows the friction coefficient variation for the Al-Si bar with the Ti-W-N layer, and Table 5.7 gives the statistical results obtained.

Fig.5.20: Curve of the coefficient of friction variation for the Al-Si bar with Ti-W-N layer as a function of distance

Comparing the results obtained for a piston without and a piston with a thin layer of Ti-W-N, we noticed that the coefficient of friction decreases from 0.5253 to 0.0857, that is to say the use of this thin layer Allows us to reduce piston friction (if we consider that the antagonist material is alumina) by about 613%.

By way of comparison of the coefficient of friction of the layer for different substrates, we can give the results obtained from this layer deposited on the stainless steel 304 and on a layer of Al-Si deposited by evaporation on the stainless steel 304.

Fig.5.21: Curve of the coefficient of friction variation for the stainless steel bar with Ti-W-N layer as a function of

Fig.5.22: Curve of the coefficient of friction variation for the stainless steel bar with double layers of Al-Si and Ti-W-N as a function of distance traveled

From these results it can be estimated that the mean value of the coefficient of friction for the torque Ti-W-N / Al2O3 is 0.08, which represents in fact a low value.

If we compare the results obtained between the bar with and a without the thin layer of Ti-W-N in terms of dissipated energy in the contact during the work, using the Coulomb's law:

u represent the sliding friction coefficient which has a value of 0.52 for the Al-Si bar without Ti-W-N layer and 0.08 for the Al-Si bar with Ti-W-N layer;

The dissipated energy in the torque Al-Si/Al2O3 is equal to 5.2 Joules for every 10 meters of piston movement i.e. 0.52 J/m, but in the torque Ti-W-N/Al2O3 it's equal to 0.8 joules for each 10 meters that is to say 0.08 J/m.

It should be noted that the most of this dissipated energy in the torque (more than 95%) [COU64] is a heat energy, which requires for the piston without layer more expense in lubrication and cooling. The reduction of these losses results a gain in the output and therefore in the size and mass of the cooling systems and so of the engine.

We can add that the propagation of this heat inside the piston under the effect of the thermal conduction causes the initiation of the cracking and the creation of thermal stress inside the piston.

The reduction of the friction coefficient by about 613% can be explained by the studies done by BUCKLEY [BUC67] and COURTEL [COU64] which point that the friction is minimal on the crystalline planes with the maximum atomic density, and on these planes in the directions of maximum atomic density.

If we make a comparison of the friction factors in different atomic structures we can notice that the crystalline materials with a hexagonal structure have generally lower values of the coefficient of fiction than those in cubic structures.

The origin of this phenomenon is that the atoms of the hexagonal structures are strongly bound with all their neighbors and have just some freedom to establish interfacial bonds with the atoms of the antagonistic materials. On the other hand, the coefficient of friction depends on the force required to break the interfacial binding between the torque surfaces [GRA08].

The structure of aluminum is a face-centered cubic structure, but Titan and nitrogen have a hexagonal structure, and tungsten has a centered cubic structure, that structure has sliding planes less than those of face-centered cubic structure.

The method allows us to obtain the wear rate was given in chapter 4. This method based on the measurement of the wear track using the optical microscope. The representation of the wear track for the different bars is given in the following figures, and the values of volume of tearing mater and that of wear rates are given in Table 5.6.

Fig.5.26 The wear track on the stainless steel bar with double layers of Al-Si and Ti-W-N

Al-Si without layer		Al-Si with Ii-W-N
volume of tearing mater (mm³)	Wear rate (mm2/N)	volume of tearing mater (mm³)	Wear rate (mm²/N)
2325.52	0.2326	1309.18	0.1309
Stainless steel with Ii-W-N		Stainless steel with Al-Si and Ii-W-N
volume of tearing mater (mm³)	Wear rate (mm2/N)	volume of tearing mater (mm³)	Wear rate (mm²/N)
798.79	0.08	3614.29	0.3614

We can see from these results that the wear rate for a piston without a layer has an acceptable value of 0.2326 mm2 / N, but it can be reduced by almost the half when we add the Ti-W-N layer.

For the Ti-W-N layer, the wear rate value is varied between 0.08 and 0.13, which represents very good results for a torque of metal / metal which has an average of values between 1 to 0.1 mm²/N [HOL58].

In this test, we search to resolve the question of whether the thin layer of Ti-W-N is able to protect the metal substrate against corrosion in a solution of 5% NaCl at 25° C.

The results presented in this paragraph concern the thin layer of Ti-W-N produced by the cathode sputtering process which is shown in the EVANS diagrams in order to obtain the corrosion rate of the layer.

The polarizability curve log(i) = f(i) allow us to determine the corrosion potential E(I = 0) using the Tafel lines and also the corrosion current. This quantity has a huge importance to evaluate the corrosion rate by the Faraday law.

We can notice that the two curves are parallel, which indicate that the two bars have an identical corrosion mechanism.

Nevertheless those results are totally in contradiction with the fact that the different compounds formed in the layer (TiN, TiO2, W2O3...) have good corrosion resistance characteristics in aggressive medium. We can explicate this contradiction by saying that this layer is deposited in the form of column with significant porosity between the grains formed, and this will cause to concentrate the phenomenon of corrosion on small zones, that type of corrosion called pitting (see chapter 2), which acceleration the degradation of the layer .

? Variation of the layer deposition parameters, the increasing of the deposition energy lead to more compact and denser structure;

Fig.5.31: Curve of polarizability of the bars of the stainless steel with and without the layer on a logarithmic scale

6.1. Mechanical analysis of the internal combustion engine piston using SOLIDWORKS

The piston is one of the most important components of the internal combustion engine. Piston fails mainly due to mechanical stresses and thermal stresses. In this paper is determined by using the finite element method, stress and, deformation and displacement distribution.

The study considers a flat sample of aluminium - silicon alloy with dimension of (20X20 mm), subjected to 50 KN of normal force which is equivalent to 100 bar.

The material used for the piston is aluminium -silicon eutectic alloy. The properties used are in table 1.

6.1.2. Finite element analysis of the sample 6.1.2.1. Sample with thin coating

Fig.6.3. shear stress due to mechanical load for the sample with the thin film coating

Fig.6.4. VON MISES stress due to mechanical load for the sample with the thin film coating

Fig.6.5. Normal stress due to mechanical load for the sample with the thin film coating

	Sample without coating	Sample with coating
Shear stress	Sc = 2	Sc =4.9
VON MISES stress	Sc = 2.2	Sc = 5.4
Normal stress	Sc = 1.6	Sc = 3.9

Table.6.3. table summarizing the minimum values of safety coefficient 6.1.3. Analysis of the result

Table 6.3 shows clearly the impact of the thin films to the resistance characteristic of the sample, just by using a 1um of hard thin films the minimum value of safety coefficient jumped with 2.5 times to a much safer zone

Analysing the figures 6.4 and 6.8 we noticed a small drop on the VON MISES stress in the sample with the thin coating, even if this result is not what we hoped for, but it proof that a thin film can provide much greater characteristic just by using a small amounts of materials.

Study of tribological properties of titanium-based thin films applied to the rubbing parts of internal combustion engines piston

3.1.4. Duration of evacuation process