**5. Laser surface cladding of aluminium alloys**

Aluminium-based metal matrix composites (Al-MMCs) have high strength, hardness and wear resistance, and find application in various industrial sectors, such as automotive and aerospace industries (Anandkumar et al,. 2007). The major drawbacks of these materials are their high coat and complex production methods compared to conventional alloys, but for many applications, like rapid tooling, the bulk stress levels are compatible with the use of high-strength Al alloy, the required wear resistance being achieved by coating the component with a high wear resistance materials such as a ceramic-reinforced Al-matrix composite (Anandkumar et al,. 2007). Aluminium alloys have been cladded with ceramics such as SiC, B4C, TiC due to their high hardness and thermal stability and various other metallic materials such as Ti, B, Ni etc. to enhance their surface properties (Anandkumar et al,. 2007). These ceramic reinforcement particles have a low reflectivity; therefore they absorb a considerable amount of laser energy (Anandkumar et al., 2009) and may reach very high temperatures, which will lead to intense reactions between the reinforcement and the liquid metal or to particle dissolution in the melt pool. The tendency of reactivity of reinforcement particles with depends on their temperature, which depends on the interaction time between the particles and the laser beam (Anandkumar et al., 2009).

In this case, the velocity of injected powder is an important factor that affects on the interaction time and particles temperature. The temperature variation of injected powder particles is calculated by several researchers using mathematical modeling. Huang et al. (Huang et al., 2005) calculated the beam attenuation and particle temperature variation due to the interaction of an off-axis powder stream with a laser beam on the basis of Lambert-Beer law and Mie's theory. They found that the temperature of injected powder particles increases with decreasing the angle between the powder jet and the laser beam from 45 to 0°, because the particles trajectory through the laser beam is longer.

Also, a mathematical model for calculation of particles temperature under laser beam irradiation is established by jouvard and co-workers (Jouvard et al., 1997). Figure 28 shows an off-axis blown powder laser cladding process diagram used for jouvard model.

Fig. 28. Diagram of laser beam-powder stream interaction (Anandkumar et al., 2009)

Laser Surface Treatments of Aluminum Alloys 139

*s*

*T T I C ddz*

where, the integral limits zs and zo are the z coordinates of the pointwhere the particle enters

In case of modification using many ceramics, especially carbides, it has been found that they also chemically react with Al and form compounds which decrease the strength of Aluminum alloy. For example, ceramic particles of SiC, tend to react and dissolve in molten Al alloy, and leading to the formation of Al4C3 and ternary Al-Si-C carbides during solidification (Viala et al., 1990; Hu et al., 1996) according to the reaction at a temperature

 4Al(l) + 3SiC(s) → Al4C3 (s) + 3Si (11) At higher temperatures (above 1670 k), the reaction product is the ternary carbide Al4SiC4,

 4Al(l) + 4SiC(s) → Al4SiC4 (s) + 3Si (12) The presense of this phase in microstructure of Al alloy is shown in figure 29a (Anandkumar et al., 2009). The hardness of Al4C3 is very lower then Al4SiC3 (300 and 1200 Hv, respectively) but, unlike Al4SiC4, it is brittle and tends to react with water, forming

Accordingly, the presence of Al4C3 in the surface microstructure results in poor mechanical properties and low long-term stability (Anandkumar et al., 2007) and its formation must be

According to the equation 10, Anandkumar et al. (Anandkumar et al., 2009) calculated the SiC particle temperature at two different jet velocities for laser surface cladding of Al-Si alloy. They found that when the particle injection velocity is 5 m/s, the particles are exposed to the laser radiation for a shorter time and they absorb correspondingly less energy. As a result, the temperature of the particles reaching the melt pool is much lower and no significant reactions occur between SiC and molten aluminum, leaving the composition of the melt essentially unchanged. During cooling this liquid solidifies as primary α-Al

Fig. 29. Microstructure of clad tracks prepared at particle injection velocities of (C) 1 m/s

0 , *o*

*A*

*vmc* η

the beam and impinges the substrate, respectively (Anandkumar et al., 2009).

range between 940 and 1620 k (Anandkumar et al., 2007):

formed by the reaction (Anandkumar et al., 2007):

aluminium hydroxide (Anandkumar et al., 2007).

dendrites and α-Al+Si eutectic (Figure 29b).

and (d) 5 m/s (Anandkumar et al., 2009)

avoided.

*p pp z*

( ) ( )

*<sup>z</sup> <sup>n</sup> <sup>p</sup> x y*

−

( 1)

= + <sup>−</sup> ∫ (10)

They reported that, the temperature of a powder particle (T) interacting with the laser beam can be calculated by (Jouvard et al., 1997):

$$T = T\_0 + \frac{I\_{\left(x, y\right)} \eta A\_p t}{m\_p c\_p} \tag{7}$$

Where

T0 initial temperature of the particles (25°C)

I(x,y) laser radiation intensity

η absorptivity of the particle material

Ap cross-sectional area of the particle

mp mass of the particle

cp specific heat capacity of the particle material

t laser beam-particle interaction time

In this equation following simpler assumptions is considered:


As the interaction time (t) is given by d/vp, where d is the distance traveled by the particle through the laser beam and vp its projected velocity component, Eq. (7) can be written as:

$$T = T\_0 + \frac{I\_{\left(x, y\right)} \eta A\_p d}{m\_p c\_p v\_p}.\tag{8}$$

The trajectories of the particles are represented by a series of lines diverging from O (Figure 28) and the energy absorbed by particle is calculated as a line trajectory integral through laser beam, because the intensity of the beam depends on x and y and finally also on z. To establish a function describing the laser beam attenuation in the z direction, the interaction region is divided into n layers of thickness Δz and the fraction of radiation intensity (C) absorbed by the particles in each layer is calculated using the following equation:

$$\mathbf{C} = I\_{\{x,y\}} \eta A\_p \Delta z \left(\frac{N}{V}\right)\_{\prime} \tag{9}$$

where (N/V) is the density of the powder stream, which depends on the powder feed rate and the injection velocity (Gingu et al., 1999). The particles in the n-th layer absorb part of the incoming radiation intensity and the remaining intensity is regarded as the input intensity for the n+1 layer and so on. The final temperature of the particles is computed by solving Eq. (10) using Wolfram Research Mathematica®6 software,

They reported that, the temperature of a powder particle (T) interacting with the laser beam

( ) ,

*x y p p p*

= + (7)

= + . (8)

⎝ ⎠ , (9)

*I At*

*m c* η

0

2. the powder particles are spherical and uniformly distributed in the powder stream,

4. the effect of gravity and the drag exerted by the surrounding gas on particle movement

6. the fraction of the laser beam energy absorbed by a particle is given by the absorptivity

As the interaction time (t) is given by d/vp, where d is the distance traveled by the particle through the laser beam and vp its projected velocity component, Eq. (7) can be written as:

( ) ,

The trajectories of the particles are represented by a series of lines diverging from O (Figure 28) and the energy absorbed by particle is calculated as a line trajectory integral through laser beam, because the intensity of the beam depends on x and y and finally also on z. To establish a function describing the laser beam attenuation in the z direction, the interaction region is divided into n layers of thickness Δz and the fraction of radiation intensity (C)

*x y p pp p*

*mcv* η

*I Ad*

0

*T T*

absorbed by the particles in each layer is calculated using the following equation:

solving Eq. (10) using Wolfram Research Mathematica®6 software,

(,) *xy p <sup>N</sup> CI Az*

η

⎛ ⎞ = Δ ⎜ ⎟

where (N/V) is the density of the powder stream, which depends on the powder feed rate and the injection velocity (Gingu et al., 1999). The particles in the n-th layer absorb part of the incoming radiation intensity and the remaining intensity is regarded as the input intensity for the n+1 layer and so on. The final temperature of the particles is computed by

*V*

*T T*

can be calculated by (Jouvard et al., 1997):

T0 initial temperature of the particles (25°C)

cp specific heat capacity of the particle material

In this equation following simpler assumptions is considered:

are negligible and all particles have the same velocity, 5. the shadow effect of the particles on each other is accounted for,

7. the temperature distribution in each particle is uniform,

1. the laser beam is parallel and has a Gaussian energy distribution.

of the particle material (η) for the laser radiation wavelength,

8. latent heat effects due to melting are neglected (Anandkumar et al., 2009).

3. energy loss by convection and radiation is negligible (Fu et al., 2002),

η absorptivity of the particle material Ap cross-sectional area of the particle

t laser beam-particle interaction time

I(x,y) laser radiation intensity

mp mass of the particle

Where

$$T = T\_0 + \frac{\eta A\_p}{v\_p m\_p c\_p} \prod\_{z\_\*}^{z\_\*} \left( I\_{\{x, y\}} - \mathbb{C} \right)^{(n - 1)} d d z \tag{10}$$

where, the integral limits zs and zo are the z coordinates of the pointwhere the particle enters the beam and impinges the substrate, respectively (Anandkumar et al., 2009).

In case of modification using many ceramics, especially carbides, it has been found that they also chemically react with Al and form compounds which decrease the strength of Aluminum alloy. For example, ceramic particles of SiC, tend to react and dissolve in molten Al alloy, and leading to the formation of Al4C3 and ternary Al-Si-C carbides during solidification (Viala et al., 1990; Hu et al., 1996) according to the reaction at a temperature range between 940 and 1620 k (Anandkumar et al., 2007):

$$\text{Al}\_{\text{(l)}} + \text{3SiC}\_{\text{(s)}} \rightarrow \text{Al}\_{\text{C}}\text{C}\_{\text{(s)}} + \text{3Si} \tag{11}$$

At higher temperatures (above 1670 k), the reaction product is the ternary carbide Al4SiC4, formed by the reaction (Anandkumar et al., 2007):

$$\text{4Al}\_{\text{(l)}} + \text{4SiC}\_{\text{(s)}} \rightarrow \text{Al}\_{\text{4SiC}\_{\text{(s)}}} + \text{3Si} \tag{12}$$

The presense of this phase in microstructure of Al alloy is shown in figure 29a (Anandkumar et al., 2009). The hardness of Al4C3 is very lower then Al4SiC3 (300 and 1200 Hv, respectively) but, unlike Al4SiC4, it is brittle and tends to react with water, forming aluminium hydroxide (Anandkumar et al., 2007).

Accordingly, the presence of Al4C3 in the surface microstructure results in poor mechanical properties and low long-term stability (Anandkumar et al., 2007) and its formation must be avoided.

According to the equation 10, Anandkumar et al. (Anandkumar et al., 2009) calculated the SiC particle temperature at two different jet velocities for laser surface cladding of Al-Si alloy. They found that when the particle injection velocity is 5 m/s, the particles are exposed to the laser radiation for a shorter time and they absorb correspondingly less energy. As a result, the temperature of the particles reaching the melt pool is much lower and no significant reactions occur between SiC and molten aluminum, leaving the composition of the melt essentially unchanged. During cooling this liquid solidifies as primary α-Al dendrites and α-Al+Si eutectic (Figure 29b).

Fig. 29. Microstructure of clad tracks prepared at particle injection velocities of (C) 1 m/s and (d) 5 m/s (Anandkumar et al., 2009)

Laser Surface Treatments of Aluminum Alloys 141

Fig. 31. Temperature distribution of SiC particles injected at 1 m/s along the X axis (y=0)

Fig. 32. Temperature distribution of SiC particles injected at 1 m/s along the Y axis (x=0)

Speeds, or both, i. e. with a higher interaction time τ and/or a higher absorbed energy, melting of the injected particles occur, as mentioned above. When the power fraction absorbed by the powder is higher than that by the substrate, only limited melting of the substrate occurs and therefore cladding is formed, with a low dilution rate of the incoming

(Anandkumar et al., 2007)

(Anandkumar et al., 2007)

powder.

Due to the lower absorption of laser beam energy by aluminum alloy, the temperature of metallic particles is always much lower than that of ceramic particles. The maximum temperature attained by the particles as a function of their injection velocity is shown in figure 30 (Anandkumar et al., 2009). Also, the temperature distribution of SiC particles injected at 1 m/s along the X axis (y=0) is shown in figure 31. Particles arriving at the X axis traveled the same time through the laser beam: particles reaching the surface near the laser beam axis are subjected to higher radiation intensity and reach higher temperatures, while the temperature decreases towards the periphery of the powder stream as the beam intensity decreases. By contrast, the temperature of the particles increases linearly along the Y axis (Figure 32). Two factors explain this evolution. On one hand, the length of the particle's path through the laser beam varies along the line: it is zero for particles reaching the leading edge of the melt pool and increases with Y up to the trailing edge of the melt pool, where it reaches its maximum value. On the other hand, attenuation of the laser beam by the particles, which decreases from the leading to the trailing edge, further enhances the particle's temperature increase in this direction. The present results show that particle injection velocity is a key parameter in control of the microstructure and properties of metal matrix composite coatings produced from metal–ceramic powder mixtures by laser cladding and laser particle injection (Anandkumar et al., 2009). The particles injection velocity must be kept higher than a certain threshold to avoid excessive heating of the ceramic particles reaching the melt pool and potential reactions between the reinforcement material and the liquid metal.

Other laser parameters such as the power of laser and scanning rate have an important effect on the properties and features of clad layers. Sallamand and Pelletier (Sallamand & Pelletier, 1993), (during laser cladding of aluminium-base alloy with Al-Si and Ni-Al powders), found that at low laser powers or high scanning speeds (or both), some of the injected particle are unmelted and some porosity is sometimes detected as shown in figure 33. Also, with higher power or lower scanning.

Fig. 30. Maximum temperature attained by the particles as a function of their injection velocity (Anandkumar et al., 2007)

Due to the lower absorption of laser beam energy by aluminum alloy, the temperature of metallic particles is always much lower than that of ceramic particles. The maximum temperature attained by the particles as a function of their injection velocity is shown in figure 30 (Anandkumar et al., 2009). Also, the temperature distribution of SiC particles injected at 1 m/s along the X axis (y=0) is shown in figure 31. Particles arriving at the X axis traveled the same time through the laser beam: particles reaching the surface near the laser beam axis are subjected to higher radiation intensity and reach higher temperatures, while the temperature decreases towards the periphery of the powder stream as the beam intensity decreases. By contrast, the temperature of the particles increases linearly along the Y axis (Figure 32). Two factors explain this evolution. On one hand, the length of the particle's path through the laser beam varies along the line: it is zero for particles reaching the leading edge of the melt pool and increases with Y up to the trailing edge of the melt pool, where it reaches its maximum value. On the other hand, attenuation of the laser beam by the particles, which decreases from the leading to the trailing edge, further enhances the particle's temperature increase in this direction. The present results show that particle injection velocity is a key parameter in control of the microstructure and properties of metal matrix composite coatings produced from metal–ceramic powder mixtures by laser cladding and laser particle injection (Anandkumar et al., 2009). The particles injection velocity must be kept higher than a certain threshold to avoid excessive heating of the ceramic particles reaching the melt pool and potential reactions between the reinforcement material and the

Other laser parameters such as the power of laser and scanning rate have an important effect on the properties and features of clad layers. Sallamand and Pelletier (Sallamand & Pelletier, 1993), (during laser cladding of aluminium-base alloy with Al-Si and Ni-Al powders), found that at low laser powers or high scanning speeds (or both), some of the injected particle are unmelted and some porosity is sometimes detected as shown in figure

Fig. 30. Maximum temperature attained by the particles as a function of their injection

liquid metal.

33. Also, with higher power or lower scanning.

velocity (Anandkumar et al., 2007)

Fig. 31. Temperature distribution of SiC particles injected at 1 m/s along the X axis (y=0) (Anandkumar et al., 2007)

Fig. 32. Temperature distribution of SiC particles injected at 1 m/s along the Y axis (x=0) (Anandkumar et al., 2007)

Speeds, or both, i. e. with a higher interaction time τ and/or a higher absorbed energy, melting of the injected particles occur, as mentioned above. When the power fraction absorbed by the powder is higher than that by the substrate, only limited melting of the substrate occurs and therefore cladding is formed, with a low dilution rate of the incoming powder.

Laser Surface Treatments of Aluminum Alloys 143

Fig. 34. Microstructure of typical cladding (AI-7at.%Si + injection of (AI, Si, Ni) powders; laser powcr P = 2800 W; scanning speed t, = 1.5 cm s-1; diameter d of the laser beam on the

Fig. 35. Evolution of the size of the dendrites from the bottom to the top of the cladding (AI-7at.%Si+injection of (AI,Si,Ni) powders; laser power P= 2800 W; scanning speed v = 2.0 cm s-1; diameter d of the laser beam on the sample, 1.25 mm) (Sallamand & Pelletier, 1993)

sample, 1.25 mm; magnification G = 80) (Sallamand & Pelletier, 1993)

Fig. 33. Micrograph of cladding before optimization of the processing conditions showing that pores are detected. (AI-7at.% Si+injection of (AI, Si,Ni) powders; laser power P= 1900 W; scanning speed v = 1.5 cm s-1; diameter d of the laser beam on the sample 1.25 mm; magnification G = 100) (Sallamand & Pelletier, 1993)

About the Ni powder, when adding the Ni powder into the melted aluminum alloy zone, it is the need for good homogenization of the nickel. The diffusion-controlled process can be enhanced by increasing the temperature, but then vaporization and plasma formation above the sample have to be avoided in order to obtain regular treated zones. It can also be enhanced by increasing the interaction time; however, an increase in the lifetime of the melted pool yields an increase in the melted depth and, consequently, a higher dilution rate of the nickel (Sallamand & Pelletier, 1993). The microstructure of a typical cladding is shown in Figure 34. It appears to be mainly dendritic. The orientations of the dendrites are not very regular; two explanations can be proposed:


Fig. 33. Micrograph of cladding before optimization of the processing conditions showing that pores are detected. (AI-7at.% Si+injection of (AI, Si,Ni) powders; laser power P= 1900 W; scanning speed v = 1.5 cm s-1; diameter d of the laser beam on the sample 1.25 mm;

About the Ni powder, when adding the Ni powder into the melted aluminum alloy zone, it is the need for good homogenization of the nickel. The diffusion-controlled process can be enhanced by increasing the temperature, but then vaporization and plasma formation above the sample have to be avoided in order to obtain regular treated zones. It can also be enhanced by increasing the interaction time; however, an increase in the lifetime of the melted pool yields an increase in the melted depth and, consequently, a higher dilution rate of the nickel (Sallamand & Pelletier, 1993). The microstructure of a typical cladding is shown in Figure 34. It appears to be mainly dendritic. The orientations of the dendrites are not very

1. A cross-section effect of a three-dimensional network occurs, where dendrites are perpendicular to the solidification front which progresses from the bottom to the top of

2. Convection movements in the melted pool can modify the regularity of the growth direction, since they induce perturbations, both in the thermal gradients and in the chemical composition (Sallamand & Pelletier, 1993). Nevertheless, the main result is the existence of a fine and dendritic microstructre, without cracks, pores or undissolved nickel, aluminium or silicon particles. Therefore the duration of the melted pool was long enough to achieve first complete melting of the injected particles and then good interdiffusion of the different elements. It may be observed in Figure 35 that the geometrical features of the dendrites are progressively modified from the interfacial zone to the surface of the sample; a progressive refinement occurs. This phenomenon is due to the evolution of the solidification rate during the process itself: as shown by many workers this rate starts from zero at the interface and increases to a maximum

value at the end of the phenomenon, on the surface of the specimen.

magnification G = 100) (Sallamand & Pelletier, 1993)

regular; two explanations can be proposed:

the sample.

Fig. 34. Microstructure of typical cladding (AI-7at.%Si + injection of (AI, Si, Ni) powders; laser powcr P = 2800 W; scanning speed t, = 1.5 cm s-1; diameter d of the laser beam on the sample, 1.25 mm; magnification G = 80) (Sallamand & Pelletier, 1993)

Fig. 35. Evolution of the size of the dendrites from the bottom to the top of the cladding (AI-7at.%Si+injection of (AI,Si,Ni) powders; laser power P= 2800 W; scanning speed v = 2.0 cm s-1; diameter d of the laser beam on the sample, 1.25 mm) (Sallamand & Pelletier, 1993)

Laser Surface Treatments of Aluminum Alloys 145

The transmission of an incident laser pulse throughout a water layer is expected to be controlled significantly by its pulse duration and / or to its rise time. Indeed, the faster energy deposition may generate the better laser-target coupling in plasma confined regime

Payer et al. (Peyre et al., 2005) studies the influence of laser intensity, wavelength, and pulse duration on the pressures generated in plasma. Results are presented in figures 38, 39.

Fig. 37. Peak plasma pressures obtained in WCM as a function of laser power density at

Fig. 38. Influence of laser intensity and pulse durations on the pressures generated in plasma confined with water regime (λ=1.06 μm)-compaison with the analytical model of confinement

1.064mm, 0.532mm and 0.355mm laser wavelength (Ding & Ye, 2006)

with water (Peyre et al., 2005).

(25 ns) (Peyre et al., 2005)
