PARAMETRIC STUDY OF LASER CLADDING OF AS 21 MAGNESIUM ALLOY WITH ALUMINIUM SILICON/ TUNGSTEN CARBIDE POWDER

Magnesium Technology 2003 Edited by Howard I. Kaplan TMS (The Minerals, Metals & Materials Society), 2003

PARAMETRIC STUDY OF LASER CLADDING OF AS 21 MAGNESIUM ALLOY WITH ALUMINIUM SILICON/ TUNGSTEN CARBIDE POWDER

M Mandagie 1,2 M Brandt 1,2 Y Durandet1,2 M Jahedi 1,3 1

CRC CAST Metals Manufacturing, 2 IRIS Swinburne University of Technology 3

CSIRO Manufacturing and Infrastructure Technology Abstract magnesium alloy with a mixture of aluminium silicon (40 % wt) and tungsten carbide (60 % wt) powder to improve its wear properties. The effects of laser power, scan speed, powder feed rate, and laser spot size on the clad layer thickness were examined. The results indicate that crack and porosity free clad layers up to 2 mm thick can be produced under optimum laser and powder mass flow rate conditions

Magnesium alloys are of growing interest for a number of applications in a range of industries because of their low densities and high specific strength. An impediment to their increased application is their low resistance to both wear and corrosion compared to that of steel or aluminium alloys. This is due to their relatively low surface hardness and high chemical affinity for numerous elements respectively. This paper describes the results of experiments investigating Nd:YAG laser cladding of AS21 .

1. Introduction attractive range of bulk material properties magnesium alloys have relatively low resistance to wear and corrosion which is a serious impediment to their wider application.

The light-weight properties and high strengthto-weight ratio of magnesium alloys compared to those of steel or aluminium have made them the material of interest for a number of applications in the automotive and aerospace industries (Zeumer, 1998). In the automotive industry, the drive to lower fuel consumption and emission of harmful pollutants has seen the range of automotive components made from magnesium alloys expand. The automotive components now made from magnesium alloys include the steering wheel, transmission housing, gear box housings, wheel rim and door inner (Zeumer, 1998). However, despite their

Wear and corrosion are essentially surface related phenomena that can be reduced or minimised by suitable tailoring of the surface microstructure and or composition without affecting the bulk. Laser surfacing is one technique that can be used to improve the surface properties of materials (Steen, 1998). Galun et al (1999) performed laser surface alloying of commercially pure magnesium, cast WE 54 T6 and extruded Al 80 and AZ 61 alloys 81

WC particles in Al/Si matrix for improving the wear resistance of magnesium alloys. The main objective of the study was to determine the operating condition for producing crack and porosity free clad layer. The study has investigated the effects of laser power, scan speed, powder feed rate and laser spot diameter on the formation of the coatings.

using powders of aluminium, copper, nickel, or silicon with particle sizes of 10 to 45Pm. Alloyed layers were produced with thicknesses of 500 to 1200 Pm and were generally free of cracks and porosity. The highest wear resistance for laser alloying with single element was obtained with layers containing 30 to 33 % at of copper. With Al and Ni alloying elements, the wear depth produced using the multiple scratch test method was reduced by 90 % compared to untreated pure magnesium.

2. Experimental Details Experiments were performed using a mixture of Al/Si (40 % wt) and WC (60 % wt) powder with AS21 magnesium alloy substrate. The Al/Si powder consisted of 88 % Al and 12 % Si with an approximate powder particle size of 45 Pm. The composition of the substrate is given in Table 1. The substrate was cut from cast ingots into 10 mm thick slices and its surface machined. Before cladding, the substrate surface was cleaned with alcohol.

Hiraga et al (2000) used a CO2 laser to clad AZ91E alloy substrates with powders of TiC, SiC and Al-30% Si aluminium alloy at varying powder feed rates. Their results show that the wear resistance of the coating, as measured by the pin-on-disk method, was significantly improved compared to the bulk and that further improvement is possible. This paper reports on the results of laser cladding experiments investigating the use of

Table 1. Composition of AS21 ingot cast substrate Composition AS 21

Al

Zn

Mn

1.9 –2.5

0.15 0.20 0.35

The experiments were carried out using a 2.5 kW Nd:YAG laser equipped with a 10 m length of 0.6 mm diameter step-index glass fiber, terminated with a 200 mm focal length processing head. Figure 1 shows the Nd:YAG laser cladding set up. Argon was used as the shrouding gas at a flow rate of 2.75 l/min. To minimise the reaction of the substrate with air, a steel shielding box continuously purged with argon was also used in the cladding process. Two series of experiments were conducted, one

Si

Cu

Ni

0.70 – 1.2

0.04

0.002

involving single tracks of length 32 mm and the other overlapped tracks in order to produced a continuous clad layer. For the single track series, the main process parameters were laser power, spot size, powder feed rate and scan speed, as summarized in Table 2. The scan speed was increased in increments of 100 mm/min. For the overlapped track series, the main process parameters were laser power and scan speed as shown in Table 3.

82

Nd:YAG Focusing Optics

Molten pool

Alloy powder injecting nozzle

Clad layer

Figure 1. Nd:YAG laser cladding set up Table 2. Processing parameters used to produce single track clad layer. Clad sample A B C D E F

Laser Spot diameter (mm) 3.75 3.75 3.75 4.6 4.6 4.6

Laser Power (kW) 1.4 1.4 1.4 1.8 1.8 1.6

Powder Feed Rate (g/min) 5.13 7.50 10.13 5.13 7.50 5.13

Scan Speed (mm/min) 100 - 300 100 - 300 100 - 400 100 - 400 100 - 500 100 - 300

Table 3. Processing conditions used to produce clad layers with overlapped tracks. Clad sample A1, B1, C1

Laser Power (kW) 1.6

Powder Feed Rate Scan Speed (g/min) (mm/min) 10.13 300 - 500

D1

1.4

10.13

300

After cladding, the samples were cut, cleaned with alcohol, cold mounted and polished to 1 Pm for measurement and analysis. 83

Track increment pattern (mm) 25 mm long x 1.5 increment x 14

temperature of the substrate increases due to greater coupling of the beam. It is also observed that the formation of a clad layer is a function of scan speed, starting later at higher scan speeds. This again reflects on the coupling of laser light into the powder and substrate.

3. Results and Discussion 3.1. Single Track Clad Layer Figure 2 shows the appearance of some single clad tracks produced using processing parameters given in Table 2. The surface of each clad track is relatively rough and its width, when using the laser power of 1.8 kW, increases from the start to finish as the

C

B

A

32 mm start

finish

E

D

Figure 2. Surface appearance of single clad tracks produced under different processing parameters. For each sample group the scan speed increases from right to left. It is evident that for each sample group, the faster the scan speed the less of the powder alloy is deposited onto the substrate. For example, for tracks in group A, produced using the laser beam spot size of 3.75 mm, laser power of 1.4 kW, and powder feed rate of 5.13 g/min, sufficient laser power and powder are available to form a clad layer at 100 mm/min, while at 300 mm/min the laser power and powder per unit length are insufficient to form a clad layer. The trend of decreasing layer thickness with increasing scan speed holds true for different laser power, powder feed rate, and laser spot size.

melting of the substrate occurs. The amount of melting is a function of laser power, scan speed and powder mass flow rate, increasing with increasing laser power, decreasing scan speed and mass flow rate. These parameters also affect the distribution of WC particles in the clad layer. At lower scan speed and/or higher mass flow rate (Figure 3a) more WC particles are found in the clad layer. Also, these particles are located at the bottom of the clad layer. At higher scan speed and/or lower mass flow rate (Figure 3b) less WC particles are deposited per unit length and these particles are distributed more uniformly through out the clad layer.

A typical cross section of a clad layer is shown in Figure 3 where it can be seen that significant

The distribution of WC particles is governed by melt pool convection currents and its 84

Mg the WC particles will sink to the bottom of a melt pool at large dwell times as shown in Figure 3a. Due to the rapid stirring in the melt pool, at short dwell times the particles become trapped throughout the solidified clad layer depth.

solidification time. Solidification time can be estimated from the dwell time, ratio of spot diameter to scan speed. Dwell time in Figure 3a is 27 ms, while in Figure 3b is 2.5 ms. As the density of WC particles at 15.7 g/cm3 is about seven times greater that that of Al and a)

b)

3.41 mm

1.97 mm

8.42 mm

5.60 mm

Figure 3. Effect of scan speed on clad layer formation for a) 100 mm/min and b) 300 mm/min for laser power of 1.8 kW, spot size of 4.6 mm and 5.13 g/min powder feed rate Illustrated in Figure 4 is the dependence of clad thickness and width on scan speed for different laser powers. It can be observed that the layer thickness decreases from about 3.4 mm at 100 mm/min and 1.8 kW to 2 mm at 400 mm and 1.8 kW. The width (measured at half length) decreases from 6.1 mm at 100 mm/min and 1.8 kW to 4.7 mm at 400 mm/min and 1.8 kW. The 4

7

3.5

6

3

Clad Width (mm)

Clad Thickness (mm)

trend of decreasing thickness with increasing scan speed is also observed for laser powers of 1.6 kW and 1.4 kW.

2.5 2 1.5 1

Laser pow er : 1.8 kW

5

Laser pow er : 1.6 kW

4 3

Laser pow er : 1.4 kW

2 1

0.5 0

0 100

200

300

400

100

Scan speed (mm/min)

a)

200

300

400

Scan speed (m m /m in)

b)

Figure 4. Effect of scan speed on a) layerl thickness and b)layer width for laser spot size of 3.75 mm and powder feed rate 5.13 g/min. The increase in clad layer thickness with increasing laser power is due to the increase in

the amount of substrate that is melted at higher laser power. The decrease in layer thickness 85

with increasing scan speed is associated with decreasing line energy and powder per unit length available for fusion with the substrate. In addition, experiments indicate that the width of clad layer is also dependent on the laser spot size and hence power density. A clad track produced with 1.4 kW laser power, 5.13 g/min powder feed rate, 100 mm/min scan speed and laser spot diameter of 3.75 mm has a width of

5.29 mm. A track produced under the same laser processing parameter except for the laser spot size of 4.6 mm, has a width of 6.36 mm. 3.2. Multiple Track Cladding Figure 5 shows the appearance of overlapped clad layers under different operating conditions using a laser spot diameter of 3.75 mm and increment between tracks of 1.5 mm.

A1 (400 mm/min)

B1 (500 mm/min)

C1 (300 mm/min)

D1 (300 mm/min)

Figure 5. Surface appearance of overlapped clad layers. Each clad layer is formed from right to left. For the same laser power and powder feed rate, faster scan speed produced a smoother surface (Figure 5 B1) than the slower one (Figure 5 C1). However, at the faster scan speed more time is needed for the clad layer to be formed similar to single tracks. It is not until the end of the first track that the coating started to be formed. The effect of laser power on clad layer structure for the constant operating condition is shown in Figure 6. It can be observed that laser power not only affects the clad layer thickness but also the distribution of WC particles within it.

In the case of higher power (Figure 6a), significant melting of the surface occurs and the WC particles are predominantly present at the bottom of the clad layer. Whereas, at the lower laser power, a thicker layer was produced with different distribution of the WC particles. In this case the layer is formed on the surface of the substrate and the WC particles are distributed more uniformly throughout the layer.

86

Start of clad

a)

1.2 mm

b)

Start of clad 3 mm

Figure 6. Effect of different laser power on clad layer formation for a) 1.6 kW and b) 1.4 kW for mass flow rate of 7.50 g/min and scan speed of 200 mm/min. Hiraga, H., Inuoe, T., Kojima, Y., Kamado, S., Watanabe, S., “Surface Modification by Dispersion of Hard Particles on Magnesium Alloy with Laser”, ‘Magnesium Alloy 2000’: 2000, pp: 253-260.

4. Conclusion A series of experiments was conducted to investigate laser cladding of AS21 magnesium alloy substrate with a mixture of WC/AlSi powder. The results indicate that it is possible to deposit a crack and pore free clad layer. The result also indicates that there are relationships between laser processing parameters and the formation of the clad layer. The slower the scan speed, the thicker the total clad layer produced due principally to more energy deposited into the substrate. The distribution of tungsten carbide particles in the clad layer was shown to be effected by the duration the melt pool stays molten. Further work is in progress to produce samples for wear testing.

Schuman, S., Friedrich, F., “ The Use of Magnesium in Cars-Today and in Future”, Magnesium Alloy and Their Application Conference Proceeding, Wolsburg, Germany, 1998. Subramanian, R., Sircar, S, Mazumder, J, “Laser Cladding of Zr on Mg for Improved Corrosion Properties”. Environmental Degradation of Ion and Laser Beam Treated Surfaces, The Minerals, Metals & Material Society, 1989.

5. Acknowledgment The CRC for CAST Metals Manufacturing (CAST) was established under the Australian Government’s Cooperative Research Centres Scheme.

Wang A.H and Yue, T.M, “ YAG Laser cladding of Al-Si Alloy on Mg/ SiC composite for the Improvement of Corrosion Resistance”. Composite Science and Technology, 2001, vol 61, no 11 pp 1549 - 1554.

6. References Galun, R., Weisheit, A., Mordike, B., “Surface Treatment of Magnesium Alloys with Laser Melting and Laser Alloying”, 5th European Conference on Laser Treatment of Material, Bremen-Vegesack, 1994.

Zeumer, N., Honsel A.G, Arudy, “Magnesium Alloy in the New Aeronautic Equipment”, Magnesium Alloys and Their Application Conference, Germany, 1998.

87