Laser Welding of Aluminium Alloy Liverpool L69 3GH, United Kingdom

21st International Congress on Applications of Lasers and Electro-Optics, Scottsdale, October 14-17, 2002 (ICALEO 2002) ISBN 0-912035-72-2 Laser Weld...
Author: Cameron Carroll
2 downloads 1 Views 523KB Size
21st International Congress on Applications of Lasers and Electro-Optics, Scottsdale, October 14-17, 2002 (ICALEO 2002) ISBN 0-912035-72-2

Laser Welding of Aluminium Alloy 5083 *

P.Okon**, G.Dearden*, K.Watkins*, M.Sharp+, P.French+ Laser Group, Department of Engineering, University of Liverpool, Brownlow Street, Liverpool L69 3GH, United Kingdom + Lairdside Laser Engineering Centre, Birkenhead, Wirral, United Kingdom

Abstract There are two laser welding mechanisms, keyhole mode and conduction mode. Keyhole welding is widely used because it produces welds with high aspect ratios and narrow heat affected zones. However keyhole welding can be unstable, as the keyhole oscillates and closes intermittently. This intermittent closure causes porosity due to gas entrapment. Conduction welding, on the other hand, is more stable since vaporisation is minimal and hence there is no further absorption below the surface of the material. Conduction welds are usually produced using low-power focused laser beams. This results in shallow welds with a low aspect ratio. In this work, high-power CO2 and YAG lasers have been used to produce laser conduction welds on 2mm and 3mm gauge AA5083 respectively by means of defocused beams. Full penetration butt-welds of 2mm and 3mm gauge AA5083 using this process have been produced. It has been observed that in this regime the penetration depth increases initially up to a maximum and then decreases with increasing spot size (corresponding to increase in distance of focus above the workpiece). Results of comparison of tensile strength tests for keyhole and conduction welds are shown. This process offers an alternative method of welding aluminium alloys, which have a high thermal conductivity. 1. 1.1


Background Laser welding is at the frontier of welding technology and the use of keyhole welding has been adopted increasingly by various sectors of the manufacturing industry. The possibility of welding materials of varying thickness quickly, efficiently, and with resultant small heat affected zones continues to attract more and more industrial interest in laser welding. However, keyhole welding (> 106W/cm2) is not without its problems such as instability, keyhole oscillation, and intermittent closure of the keyhole that often leads to porosity. In some alloys, the high weld speed (and hence high rates of cooling) can lead to embrittlement in the weld or heat affected zone. Laser conduction welding (LCW) (< 106W/cm2) is comparatively stable and may offer an alternative means of welding traditionally difficult materials such as aluminium alloys. Advances in laser keyhole welding and investigations into its accompanying difficulties are well documented in the literature1,2. Analytical, numerical and empirical studies have been undertaken in order to achieve better understanding of the process. Some pioneering investigations have been made in the area of LCW, but the development of LCW as a high quality CO2 laser welding process that can be practically implemented, remains to be undertaken. Tsai et al3 modelled the response of laser conduction-mode weld pool width dimensions to a step change in power input, using a 2D heat flow analysis. Pitscheneder et al4 carried out a multi-parametric numerical and experimental investigation into the role of sulphur content and welding parameters on the fluid flow, heat transfer and resulting geometry of conduction mode laser spot weld pools. . Pitscheneder et al5

investigated the development of a quantitative model for behaviour of solid precipitates in the weld pool. Esposito et al6 analysed experimental results of laser conduction seam welding of low carbon and stainless steel samples. A method was formulated to calculate values of all the related parameters including the laser focal spot radius required for the best utilization of energy. It was found that the penetration depth increased with reduction in power density although the investigation was limited to distances of –10mm to + 10 mm out of focus while the thickness of the material used was 1.2mm with the laser power equal to 300W. Bos et al7 performed a heat conduction analysis of the weld pool and weld pool convection studies. The prediction of pool aspect ratios and consequently weld dimensions based on given aspect ratios and welding parameters was achieved. Williams et al8 investigated the suitability of direct diode lasers for conduction welding of AA2024 and titanium alloys. The results were interpreted by thermal modelling. Nakamura et al9 proposed a technique to detect the transition between laser keyhole welding and laser heat conduction welding by monitoring optical and acoustic emissions from a laser-irradiated point. Russo et al10 carried out numerical and experimental investigations into weld-pool surface phenomena in conduction mode laser welding from for various materials including aluminium. Paul et al11 simulated temperature profiles and fluid flow fields through numerical solutions of Navier-Stokes equation and the equation of conservation of energy for low power laser welding in conduction mode. Computed results obtained for weld pool temperature, secondary dendrite arm spacings, and weld pool surface topography compared favourably with experimental results obtained using high manganese stainless steel, USS Tenelon. Zhao et al12 carried out an experimental and theoretical study to achieve a quantitative understanding of the effect of various welding variables on vaporisation and composition change during conduction mode welding of aluminium alloy 5182. The emphasis of most of the work done so far in the area of laser conduction welding (LCW) has been the investigation of the process as a means of understanding the complex nature of laser welding as a whole. Low powers at the focus have been mainly utilised and most of the work has been in the area of spot welding. Investigations into the effects of high-power defocused laser beams have so far been restricted to situations where the distances from the focus have been in the range –10mm to +10mm for both conduction and keyhole welding6,13. In this work the distances of the work piece below the focus are in the range of +5mm to + 100mm which truly takes the welding into the conduction welding regime. Semi-quantitative analysis It has been observed that penetration depth increases with spot radius during low-speed laser welding in the conduction mode regime. The laser beam is defocused. Below is a 1D semi-quantitative analysis of the phenomenon. The temperature at a depth z below the surface of a semi-infinite solid at a time t after heat flow starts is given by14 1 2H z T ( z, t ) = (αt ) 2 ierfc( ) (1.2.1) 1 K 2 2(αt ) where H = I(1 –Rs) I is beam intensity Rs is surface reflectance α is thermal diffusivity K is thermal conductivity t is time The function ierfc is given by 1.2

ierfc( x) =

1 π

[exp(− x 2 ) − x(1 − erf ( x))]

where erf ( x) =

2 π



−ξ 2


From (1.2.1) 1 2 H αt 2 T (0, t ) = ( ) (1.2.2) K π Let zm be depth at which temperature reaches melting point, Tm when the surface temperature has just reached the boiling temperature, Tb. From equations (1.2.1) and (1.2.2) 1 T ( z m , t ) Tm z = = π 2 ierfc( m 1 ) (1.2.3) T (0, t ) Tb 2(αt ) 2 From (1.2.2) 1 2

1 2

Tb Kπ 2H Putting this in (1.2.3), we obtain Hz m T ierfc( ) = 1m 1 (αt ) =


KTbπ 2 π 2 Tb Equation 1.2.4) shows that the product Hzm is constant as long as the surface temperature is at boiling temperature. 1 Thus zm α . This means that the welding depth will increase with decreasing intensity for H as long as the surface temperature remains at boiling. This result is comparable to that obtained by Duley(1976) and Prokhorov(1990) which is as follows15,16: 1.2 KTm Tb zm = (1.2.5) ( − 1) H Tm Equation (1.2.5) shows that zm is inversely proportional to H. As the intensity decreases, a point is reached where the boiling temperature can no longer be sustained. Let the temperature at this point be Tp. Then Tm < Tp < Tb Equation (1.2.4) becomes Hzm T ierfc( ) = 1m (1.2.6) 1 KTpπ 2

π 2Tp


But from (1.2.2), Tp = ∴ (

Hzm KTpπ

1 2

) ={

2 H αt 2 ( ) K π zm

2(αt )

1 2




Therefore, ierfc(

KTpπ and

Tm π Tp

) = ierfc{

2(αt )

zm 2(αt )


= ierfc{

1 2

1 2

1 2

1 2




2R and R>> zm v 1 For small values of x, ierfc(x) ≈ − x ≈ 0.56 – x17 π Combining (1.2.7), (1.2.8) and (1.2.9) and (1.2.10) we have

Now, for large spots, t =

1 2

zm = 2(αt ) [0.56 −

1 2

3 2

πTm v R K 3 2

1 2

] = 1.12. 2.α

1 2

2 α P Equation (11) may be written as

R v

1 2

1 2

πTm KR 2 P

1 2

zm = AR − BR 2 for constant P ,K, Tm, v and α

(1.2.9) (1.2.10)




where A = 1.12 2

α2 v

and B =

1 2

πTm K P

Semi-quantitative Variation of Penetration Depth with Spot Radius 3.5 3.25 3

Penetration Depth (mm)

2.75 2.5 2.25 2 1.75 1.5 1.25 1 0.75 0.5 0.25 0 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2

Spot Radius (mm)

Figure 1 – Result of semi-quantitative simulation of laser conduction welding process Figure 1 shows the result of a simulation using values, P=500W, v = 10mm/s and K, Tm and α remaining constant. What is important here is the penetration depth/spot radius variation. It is also important to observe that the latent heat of vaporisation is over 20 times greater than the latent heat of fusion (Lf ~ 397KJ/kg ; Lv ~ 9492kJ/kg). This means that a significantly greater amount of energy is required to vaporise the metal liquid at boiling point than is required to melt the solid metal at melting point. Therefore most of the additional heat absorbed at the boiling point will be utilised in boiling the liquid metal rather than cause vaporisation in the conduction mode regime. This phenomenon is only observable at relatively low welding speeds. Thus the choice of welding speed is critical depending on the type of laser and the laser power involved. At

greater speeds the reduced interaction time results in a relatively lower heat absorption by the workpiece. 2.

Experimental methodology

2.1 Laser welding trials Experiments were carried out to establish the dimensions of the fusion zone in laser welding as the laser beam diameter incident on the sample was increased. The increase in the laser beam diameter was achieved by increasing the standoff distance between the lens and the work piece. The spot radius for a real laser beam is given by18


zλM 2 2 rc = r0 1 + ( ) 2 πr0


2 M 2 Fλ r0 = πD


rc = calculated spot radius at distance z D = diameter of unfocused laser beam at lens λ = laser beam wavelength F = focal length of lens All spot radii values used were calculated. 2.2 Shielding gas Coaxial shield gas delivery was adopted for all samples. Argon gas was used in order to minimise oxidation due to atmospheric exposure during welding. 2.3 Sample preparation The samples were sandblasted prior to welding in order to increase absorptivity. After welding, the samples were sectioned with a band saw and mounted with a Buehler Pneumet II mounting press machine. The hot mounting was performed for 15 minutes at a temperature of 1000C. Each mount was stone ground using a Struers Prepamatic polishing and grinding machine. This was performed gradually until the desired surface finish was achieved. The samples were then macroscopically etched with sodium hydroxide solution. 2.4 Lasers Two types of lasers were used for experiments. The first was a Multi Wave-AutoTM Lumonics AM356 continuous wave YAG laser. The maximum power that could be delivered was 4.5kW with a optic fibre of diameter 600 microns. The second was a PRC/OPL CO2 laser with a maximum power delivery of 1.8kW and a TEM01* mode laser beam. 3. Results and discussion 3.1 Penetration depth/spot radius relationship Conduction welds were produced using high-power defocused laser beams on 2mm and 3mm gauge AA5083. Figure 2(i - vii) shows micrographs of conduction welds made from 3mm gauge AA5083. The numbering is done in order of increasing spot radius. The expected hemispherical shape characteristic of conduction welds was obtained. The variation of the

penetration depth with increasing spot radius or distance from the focus becomes apparent on close examination. In all cases it was found that the aspect ratios (penetration depth divided by weld width) of the welds were less than one. Penetration depth/spot radius curves plotted showed an increase (up to a maximum) and then a decrease of penetration depth with increase in spot radius in the conduction weld regime. This can be attributed to the interplay between decreasing power density and increasing interaction time, as the spot radius becomes larger19. It is suggested that the laser beam distribution and pre-heating due to the interaction of the outer fringes of the laser beam with the work piece also contribute to this phenomenon. Figure 3 shows the variation of penetration depth with spot radius for 2mm and 3mm gauge AA5083 welded speeds of 600mm/min and 720mm/min.

(v) (i)

(iii) (vi)




Figure 2 – Conduction weld for 3mm gauge AA5083 using YAG laser ; Average power = 2000W ; Weldspeed = 600mm/min


3.2 3


2.8 2.6



Penetration Depth (mm)

Penetration Depth (mm)

1.6 1.4 1.2 1 0.8 0.6

2 1.8 1.6 1.4 1.2 1 0.8 0.6

2mm AA5083, v - 600mm/min (Sandblasted)

0.4 0.2 0.3


3mm AA5083, v = 10mm/s (Sandblasted) 3mmAA5083, v = 12mm/s (Sandblasted)

0.4 0.2







Spot Radius (mm)





0 0.5 0.75


1.25 1.5 1.75


2.25 2.5 2.75


3.25 3.5 3.75


Spot Radius (mm)

(i) (ii) Figure 3 – (i) Variation of penetration depth with spot radius for 2mm AA5083 at 600mm/min welding speed using CO2 laser and (ii) Variation of penetration depth with spot radius for 3mm AA5083 at 600 mm/min and 720 mm/min welding speeds using YAG laser. Figure 4 shows micrographs of laser keyhole and conduction welds. Figure 4i is a micrograph of a butt weld obtained using the YAG laser at a welding speed of 360mm/min and the work

piece situated at distance of 20mm below the focus. This position produced the maximum penetration depth for the laser setup. It can be seen that the weld was a fully penetrating one. Figure 4ii is a laser keyhole weld produced by a CO2 laser at maximum power of 1750W with welding speed of 1600mm/min on 2mm AA5083. The groove on the weld surface constitutes a an area of weakness. Although the surface was wire-brushed and cleaned, a pore was clearly revealed. Pores are seen in the 2mm AA5083 laser keyhole weld in Figure 3iii(a) whereas the conduction welds made on the same material in Figures 3iii(b)(bead-on-plate) and 3iii(c)(butt-weld) are relatively free from pores. The welding parameters are as given below in the figure caption. Again the conduction welds are fully penetrative.

a b









Figure 4 – (i) 3mm AA5083 LCW(YAG,Butt Weld), v = 360mm/min,(ii) 2mm AA5083 KHW, v =1600mm/min(CO2) (iii) 2mm AA5083 (a) KHW, v= 1100mm/min, (b) LCW, v = 500mm/min, (c) LCW (Butt Weld), v = 450mm/min 3.2 Tensile strength tests Samples used for tensile strength tests had the shape and dimensions as shown in figures 5 and 6. The dimensions chosen are meet International Standard Organization (ISO)

Figure 5 – Shape for tensile strength testing

requirements and the samples were tested on an Instron machine using a strain rate of 0.001mm/s and a cross head speed of 4.5mm/s.

W1 (mm)

W (mm)

t (mm)

l (mm)

L (mm)






Figure 6 – Dimensions for samples used in tensile strength tests

The results obtained were as follows: Fracture Load for LKWs = 4.775kN Fracture Load for LCWs = 5.733kN Theoretical Fracture Load for base metal = 6.96kN; U.T.S. = 290MPa Tensile strength tests indicate that LCWs are stronger than LKHs. One reason for this is the alloy constituent depletion (in this case magnesium) during laser keyhole welding because of the high power densities involved. Secondly a small groove resembling a notch is formed in the weld during laser keyhole welding and this invariably becomes an area of structural weakness. Conclusions Semi-quantitative and experimental investigations into laser conduction welding (LCW) of AA5083 using high-power defocused CO2 and YAG laser beams reveal that penetration depth increases up to a maximum and then decreases with increasing spot radius. More specifically, Semi-quantitative analysis of the LCW process shows that the penetration depth increases with increase in spot radius when the surface temperature remains at boiling temperature. This trend is reversed when the boiling temperature is no longer sustained due to a further increase in the spot radius. This confirms that the penetration depth/spot radius variation occurs well within the conduction-welding regime, as ideally there should be no vaporisation during LCW. It is suggested that this phenomenon can be attributed to (i) the interplay between decreasing power density and increasing interaction time, (ii) the laser beam distribution and (iii) pre-heating due to the interaction of the outer fringes of the laser beam with the work piece. Tensile strength tests indicate that LCWs are stronger than LKHs. One reason for this is the alloy constituent depletion during laser keyhole welding because of the high power densities involved. Secondly the small groove formed in the weld during laser keyhole welding constitutes an area of structural weakness. This process may offer an alternative method of welding resulting in slower rates of cooling which may enhance weld bead properties of traditionally difficult materials such as aluminium alloys used in the aerospace industry. Full penetration laser conduction welds produced on such materials by means of high-power defocused laser beams could reduce problems encountered with keyhole welds thereby greatly reducing if not eliminating keyhole defects.

References 1. Ready,J.F,ed. LIA Handbook of Laser Materials Processing, Magnolia Publishing, Inc., 2001 2. Duley,W.W., Laser Welding, New York, John Wiley and Sons, Inc., 1998: Chapter 4 3. Tsai,F. et al “Modelling of Conduction Mode Laser welding Process for Feedback control”, Transactions of ASME , Vol. 122, 2000, pp. 420-428 4. Pitscheneder,W. et al., “Numerical and Experimental Investigation of ConductionMode Laser Weld Pools”, Mathematical Modelling of Weld Phenomena 4, ed Cerjak,H 1998 5. Pitscheneder,W. et al., “Experimental and Numerical Investigation of Transport Phenomena in Conduction Mode Weld Pools”, Mathematical Modelling of Weld Phenomena 3, ed Cerjak,H ,1997 6. Esposito,C., Duarelio,G., Cingolani,A., “On the Conduction Welding Process of Steels with CO2 Lasers” , Optics and Lasers in Engineering 3,1982, pp. 139-151 7. Bos,J.A., Chen,M.A, “On the Prediction of Weld Pool Size and Heat affected Zone in Shallow-Pool Welding”, Transport Phenomena in Materials Processing and Manufacturing ASME, HTD Vol. 336 ,1996 8. Williams,S.W et al “Direct Diode Laser welding of Aerospace Alloys”, LaserOpto, Vol. 33, No.4, 2001 9. Nakamura, S. et al “Detection Technique for Transition between Deep Penetration Mode and Shallow Penetration Mode in CO2 Laser Welding of Metals ”, J.Phys.D:Appl. Phys. 33, pp.2941-2948, 2000 10. Russo,A.J. et al “Two-Dimensional Modelling of Conduction-Mode Laser Welding”, L.I.A. Vol 44 (ICALEO) , pp. 8-16 ,1984 11. Paul,A. and DebRoy,T., “Free Surface Flow and Heat Transfer in Conduction Mode Laser Welding”, Metallurgical Transactions B, Vol. 19B, pp. 851-858, 1988 12. Zhao,H. and DebRoy,T., “Weld Metal Composition Change during Conduction Mode Laser Welding of Aluminium Alloy 5182”, Metallurgical Transactions B, Vol. 32B, pp. 163-172, 2001 13. Matsumura,H.,Orihashi,T., Nakayama,S et al “CO2 Laser Welding Characteristics of Various Aluminium Alloys” , Proceedings of LAMP’92, Nagaoka, Japan ,June 1992 14. Wilson,J., Hawkes,J.F.B., 1987 “Lasers”, (Prentice Hall) 15. Duley,W.W., CO2 Lasers, Effect and Applications, New York: Academic Press, 1976 16. Prokhorov,A.M. et al, Laser Heating of Metals, IOP Publishing, 1990 17. Verbal Communication with Prof. W.M. Steen 18. Sun,H., “Thin Lens Equation for a Real Laser Beam with Weak Lens Aperture Truncation”, Opt. Eng.37,No. 11 ,1998 19. Magee,J., Okon,P. and Dowden,J., “The relationship between spot size and penetration depth in laser welding”, IIW Doc. IV-764-2000 presented at the International Institute of Welding conference in Florence, Italy in July, 2000.

Suggest Documents