Welding aluminium by laser (November 1996)

This paper is based on a similar article - 'Laser welding of aluminium' - presented at 'ICAWT '96', The International Conference on Advances in Welding Technology, Columbus, Ohio, USA, 6-8 November 1996

Introduction

Laser welding offers a rapid, flexible, low distortion, automated manufacturing route for many components. It is already used for steel and coated steel applications such as butt welding of sheets for tailored blanks, welding of pre-machined gear components and lap joints in sheets, as an alternative to resistance spot welding.

There are two main types of laser suitable for sheet metal welding:

• CO ₂ lasers;
• Nd:YAG lasers.

They are both available as sources giving either continuous or pulsed power output, and their main features are listed in Table 1.

Table 1 Features of the two laser types

Property	CO 2 lasers	Nd:YAG lasers
Lasing medium	CO 2 + N 2 + He gas	Single crystal rod neodymium doped yttrium aluminium garnet
Radiation wavelength	10.6 µm	1.06 µm
Excitation method	Electric discharge	Flash lamps
Consumables	CO 2, N 2, He, electricity	Flash lamps, electricity
Output powers	Up to 45 kW	Up to 4.5 kW
Beam transmission	Polished metal mirrors	Fibre optics or mirrors

Use of aluminium alloys as sheet, extrusions or castings in various industries is increasing with the continuing drive towards reduced weight, improved efficiency and recyclability. Accordingly, there is considerable interest in applying laser welding to aluminium alloys, especially for sheet metal processing where the advantages of low heat input, low distortion, high welding speeds and the potential for full automation and flexibility of manufacture can be fully utilised ^[1-4]. However, aluminium alloys have been notoriously difficult to join using lasers. Problems relate to their high surface reflectivity, high thermal conductivity and, for some alloys, low boiling point constituents. These and other material-related difficulties have led to weld and HAZ cracking, porosity, degradation in mechanical properties and inconsistent welding performance.

The work summarised here was undertaken to address some of these problems. Studies involving laser welding of mainly 5000 and 6000 series aluminium alloys in sheet form as butt and lap welds have been carried out, with and without filler material, to examine:

• crack sensitivity;
• mechanical properties (tensile, elongation, formability);
• porosity.

Crack Sensitivity

Two forms of weld-related cracking can be readily identified, HAZ liquation cracking and solidification cracking. In this work, occurrence of solidification cracking has been quantified in relation to material composition and welding conditions using a self-restraint cracking test. The test involves use of a tapered specimen, as a development of an earlier test used by Houldcroft.

Mechanical Properties

The performance of the laser welds has been evaluated by tensile and formability testing. Formability is particularly important to the potential use of aluminium for tailored blanks.

Porosity

Three forms of porosity have been identified; through thickness blow holes, large irregular pores and small spherical pores. Conventional radiography has been used to quantify the levels of porosity.

Following recent development in welding procedures, laser welding of aluminium alloy structures has wide potential in the aerospace, automotive, construction, domestic appliance, electronic, off-road vehicles and shipbuilding sectors. Table 2 summarises potential industrial uses for laser welded aluminium alloys. Both butt and lap joints in materials ranging from 0.5-10mm can be considered for laser welding applications.

Table 2 Summary of potential industrial applications for laser welding of aluminium alloys

Industry sector	Welding application areas	Status			Materials
		Production	Development	Future area	Typical alloys	Thickness
Electronic	Microwave packages	✔			6000/4000	0.5-2mm
	Radar systems		✔		6000/4000/2000	1-3mm
	Cable trunking			✔	6000	1-2mm
Packaging	Aluminium can materials			✔	1000	0.1-0.3mm
Domestic Appliance	Tube/sheet heat exchangers			✔	1000	0.4-0.8mm
	Double glazing frames	✔			6000	0.5mm
Automotive	Tailored blanks		✔		5000/6000	1-3mm
	Clinch joints		✔		5000/6000	1-2mm
	Space frame structures		✔		6000 cast	1-4mm
	Cast components		✔		Cast	2-4mm
Aerospace	Airframe components		✔		6000/8000/2000/7000	1-3mm
Sports Goods	Tubes for sports goods			✔	7000	0.5-1mm
Off-Road	Railway carriages			✔	6000	2-6mm
Ship	Lightweight ship structures			✔	5000/6000	3-12mm
Construction	Chemical/gas/pressure vessels			✔	6000	3-12mm
	Lightweight structures			✔	6000/5000	3-12mm

Experimental Details

Material

Aluminium alloy sheet of 1.6-2.0 mm thickness has been evaluated. A number of filler wires were also used to control weld compositions. The chemical composition and mechanical properties of the alloys are shown in Table 3.

Table 3. Material composition (wt%) and properties

Alloy	Form	Thickness, mm	Temper	Si	Fe	Cu	Mn	Mg	Cr	Zn	Ti	UTS, N/mm 2	Elongation, %
1200	Sheet	2.0	H4	0.11	0.33	-	-	-	-	0.01	0.01	122.5	11.7
5005	Sheet	2.0	-	0.05	0.37	0.04	0.05	0.77	-	-	-	165.0	-
5251	Sheet	2.0	H3	0.08	0.20	0.01	0.23	1.85	-	0.01	0.03	216.5	-
5754	Sheet	2.0	-	0.06	0.16	0.01	0.24	3.06	-	-	-	228.0	23.0
5083	Sheet	2.0	H2	0.15	0.34	0.02	0.63	4.82	0.05	0.03	0.02	309.5	20.7
6061	Sheet	1.6	T6	0.7	0.4	0.3	0.0	1.1	0.1	-	-	364.5	12.7
6082	Sheet	1.9	T6	0.92	0.42	0.05	0.50	0.90	0.02	0.05	0.02	341.0	16.0
4043	Wire	1.2	-	4.65	0.36	0.01	0.01	0.01	-	0.01	-	-	-
4047	Wire	1.2	-	12	0.8	0.3	0.15	0.1	-	0.2	-	-	-
5154	Wire	1.2	-	0.25	0.4	0.1	0.1	3.5	0.25	0.2	0.2	-	-
5556	Wire	1.2	-	0.09	0.34	-	0.57	5.60	0.07	-	0.07	-	-

Welding Procedure

Welding was carried out using two types of laser; a 5 kW CO ₂ laser, and 2 kW and 4.5kW Nd:YAG lasers (as yet only limited testing has been carried out on welds made with a 4.5kW experimental Nd:YAG laser facility located at Lumonics in Rugby, UK).

Butt welds were made using aluminium alloy sheets with or without filler wires to prepare the tensile and bulge test specimens. Helium was used for both top and underbead shielding during CO ₂ laser welding, whereas helium was used for top bead and argon for underbead shielding during Nd:YAG welding.

Assessment of Weldability

Hot crack test

A self-restraint crack test, a modified form of Houldcroft test was used ^[5-7]. The specimen tapered from 10 mm at one end to 40 mm at the other. The optimum dimensions were selected following a series of trials using sheet alloys of known crack susceptibilities. Full penetration melt-runs were made from the narrow end towards the wide end along the longitudinal centre-line of the specimens. No external clamping or restraint was applied during CO ₂ or Nd:YAG laser welding. At least three specimens were tested for each welding condition. Crack length was measured with the aid of dye penetrant. The ratio of crack length to weld seam length (which is equal to the specimen length) was used as the index of crack susceptibility.

Tensile test and biaxial bulge test

Weld properties in terms of failure strength and ductility/formability were assessed by transverse tensile tests and biaxial bulge tests.

Weld beads were not removed from the tensile specimens and the bead was positioned at the centre of the gauge length. Failure strength and elongation of the transverse tensile specimens were assessed.

In the biaxial bulge test, 200 x 200 mm butt welded sheets were deformed hydraulically until tearing occurred in the coupon. One or two tests were made for each welding condition, thus the accuracy of the formability data (i.e. the height to failure in mm) needs to be confirmed.

Radiography

To assess porosity, radiographic tests were carried out on the weld coupons. Pores larger than 0.5 mm diameter were counted from the radiographic film.

Results

Continuous Wave (CW) 5kW CO ₂ Laser Welds

Welding conditions and a summary of weld property results for laser welds made using a CO ₂ laser are shown in Table 4. A typical CO ₂ laser butt weld is shown in Fig. 1.

Table 4. Properties of aluminium sheet alloys CW CO 2 laser welded with and without filler wires (all are 2 mm thick except 1.6 mm thick 6061 alloy).

Material	Welding condition				Welding properties
Alloy	Filler wire	Laser power (kW)	Weld speed, m/min	Wire feed speed, m/min	Failure strength, N/mm 2	% of parent,%	Elongation,%	Failure area	Bulge test, mm	Porosity per 100mm
5251	-	5	6.0	-	170	78	2.2	Weld	6.53	6
	5154	5	5.5	2	173	79	3	Weld	7.47	8
	5556A	5	6.0	2	170	78	2.5	Weld	12.05	7
5754	-	5	5.0	-	211	93	11.0	Weld	13.9	10
	5556A	5	5.0	3	222	97	15.0	PM/HAZ	-	8
5083	-	5	5.5	-	277	89	9.3	Weld	19.5	2
	5154	5	5.5	2	252	81	6.5	Weld	15.2	4
	5556A	5	6.0	2	285	91	11	Weld	19.36	6
6061	-	4.5	7.5	-	189	58	1.2	Weld	9.85	13
	4043	5	5.5	2	234	72	2	Weld	-	1
	4043	4.5	5.0	5	189	58	0.7	Weld	-	3
	4047	5	5.5	2	235	72	1	Weld	-	1
	4047	4.5	5.0	5	200	61	1	Weld	-	8
	5556A	5	5.5	2	219	67	1.5	Weld	-	4
	5556A	4.5	5.0	5	219	67	1	Weld	-	7

Hot crack susceptibility

In melt runs, Al-Mg-Si alloys (6061, 6082) showed higher crack susceptibility than other alloys such as Al-6Cu (2219) and Al-Mg alloys (5251, 5083). ^[Fig.2]

Similarly, when laser welding lap joints, the Al-Mg-Si alloys (6061, 6013) also showed higher crack susceptibility than the Al-Mg alloys (5754, 5083) [Fig.3].

The hot crack susceptibility of lap welds using dissimilar alloy sheets (5754 and 6061) was in between those of each lap weld using the same alloy sheets, and depended on the stacking sequence of the two sheets ^[Fig.3].

In Al-Mg alloys, hot crack susceptibility increased with magnesium content, reaching the peak value at about 2wt% Mg and then decreased with further addition of Mg ^[Fig.4].

When use of filler wire was investigated, Al-12wt% Si filler (4047) was the most effective for reducing hot crack susceptibility of Al-Mg-Si alloy (6061) sheet [Fig.5]

Mechanical properties

The failure strength, elongation and formability of Al-Mg alloy laser welds were related to the composition of base metal and the fillers used. As Mg content increased in the base alloys and fillers, the failure strength increased. The elongation and formability (from the biaxial bulge test) of laser welds increased up to about 3wt% Mg (5754) and then decreased slightly with further increases in Mg ^{[Table 4, Fig.6, and Fig.7]}. The laser weld properties of Al-2wt% Mg (5251) were not improved substantially by filler additions, but those of Al-3wt% Mg (5754) and Al-4.5wt%Mg (5083) were improved by an Al-5wt% Mg (5556) filler wire. The failure of the tensile test in 5754 welded with 5556A wire was away from the weld in either the HAZ or parent material.

For Al-Mg-Si alloys, tensile failure strengths were 60 and 70% of base metal strength for the autogenous and wire feed welds of Al-Mg-Si alloy (6061). A slight increase on failure strength was obtained by addition of fillers, although ductility was not improved [Table 4].

Porosity

Levels of porosity ranged from 1 to 13 pores per 100 mm (pore sizes equivalent to 0.5 mm diameter average). There was no clear pattern in the results over different alloys or with use of filler wire.

2kW Nd:YAG Laser Welds

Welding conditions and weld property results for laser welds made using a 2kW Nd:YAG laser (CW with modulation for up to 5kW peak power) are shown in Table 5. A weld in 2 mm thick 5083 is shown in Fig.8.

Table 5. Properties of Nd:YAG laser welded aluminium alloy sheet (2 mm thick unless stated).

Materialmm	Welding conditions				Weld properties
	Ave. power, W	Peak power, W	Pulse freq., Hz	Speed, m/min	Failure strength, N/mm 2	% of parent	Elong., %	Failure area	Pores per 100mm	Bulge test mm	Crack test %
1200	2000	5000	100	0.8	84	69	4.5	HAZ	24	12.8	45
2219 (3.2)	4500	-	-	7.0	-	-	-	-	7	-	-
5005	2000	4000	100	0.8	132	80	3.5	HAZ	2	14.4	48
5251	2000	3400	100	0.9	200	92	16.8	Weld	2	25.4	65
5754	2000	3400	100	0.9	219	96	15.5	Weld	7	28	46
5083	2000	3400	100	0.8	278	90	10.2	Weld	1	23.9	1.5
5083	2000	3400	100	1.6	278	90	10.2	Weld	1	23.3	-
5083	4500	-	-	13.5	-	-	-	-	2	-	-
6061 (1.6)	2000	4500	100	1.4	230	63	2.0	Weld	2	15.6	98
6061 (1.6)	4500	-	-	14.0	-	-	-	-	0	-	-
6082	2000	4500	100	1.0	244	72	3.0	Weld	15	-	98
8090	4500	-	-	16.0	-	-	-	-	many	-	-

Hot crack susceptibility

The trend in hot crack susceptibility of pulsed Nd:YAG laser welds was similar to that for CW CO ₂ laser welds ^[Fig.9].

Peak crack susceptibility occurred at about 2wt% Mg for Nd:YAG laser welds in Al-Mg alloys. The 5083 alloy showed a far lower crack susceptibility among the alloys investigated.

Nd:YAG laser welds in Al-Mg-Si alloys (6061, 6082) showed high crack susceptibility, even when compared with the CW CO ₂ laser welds in these alloys.

Mechanical properties

Failure strengths of Al-Mg alloy welds were about 90% of base metal strength and weld strength increased as Mg content increased ^{[Table 5 and Fig.10]}.

Failure strengths of Al-Mg-Si alloy laser welds were about 60 to 70% of base metal strength [Table 5].
Elongation and formability of Nd:YAG laser welded Al-Mg alloys were lower than those of corresponding base metal ^[Fig.11], but higher than those measured from CO ₂ laser welds. Al-2wt% Mg (5251) and Al-3wt% Mg (5754) alloy welds using the Nd:YAG laser showed increased ductility compared with CW CO ₂ laser welds and the ductility value was higher than that of the other Al-Mg alloy welds ^[Fig.11].

The ductility and formability of Al-Mg-Si alloy (6061) welds were not as good as for the Al-Mg alloys (especially 5251 or 5754).

Porosity

Porosity levels varied from 1 to 24 pores per 100 mm. Most pores were present in welds in 1200 and 6082 alloys, with significantly less in the others.

Continuous Wave (CW) 4.5kW Nd:YAG Laser Welds

Welding conditions and test results available for welds made with a 4.5kW Nd:YAG laser are shown in Table 5. A typical section of a butt weld made using the 4.5kW Nd:YAG laser is shown in Fig. 12.

Welding speeds are in general much higher than with either the 5kW CO 2laser or the 2kW Nd:YAG laser. A 2.0mm thick 5083 alloy was welded at 13.5m/min, a 1.6mm thick 6061 alloy at 14.0m/min, a 1.6mm 8090 alloy at 16.0m/min and a 3.2mm thick 2219 alloy at 7.0m/min.

Except for the 8090 alloy, porosity in these welds is very low (Table 5).

Discussion

The feasibility of laser welding aluminium alloys has been proven, but it has not always been clear what effect the welding process has had on the mechanical properties of the joint in relation to laser type and alloy composition. This work has indicated clear trends in crack susceptibility, ductility, formability and tensile strength with changes in these variables.

The crack susceptibility test used was simple and easy to carry out and tended to differentiate well between the materials tested. In general it correlates well with other results and apparently with the 'closer-to-application' studies on welded joints. Further confirmation of these correlations is required.

The elongation measured from tensile tests followed a similar trend to the formability measured in the biaxial bulge test, hence confirming the results achieved. There were some tensile test failures away from the weld region indicating that the requirements of failure in the parent material for welding of steel sheet for the automotive industry are close to being met in laser welding of aluminium. Improvements in formability have been achieved by using filler material and different weld profiles and weld thermal cycles, and more improvements might be expected.

Porosity content has not shown any clear trends with alloy type or laser used, and in the ranges of porosity encountered (0 to 24 pores equivalent to 0.5 mm average diameter per 100 mm of weld), there does not appear to be any effect upon mechanical properties evaluated in this work.

Selection of 5000 Series Alloy

In general, 5000 series (Al-Mg) alloy welded sheets tended to give butt welds with low crack susceptibility and low levels of defects which might adversely affect properties of the joints. Peak formability and ductility appeared to be achieved at approximately 3wt%Mg, whereas peak weld strength and lowest crack susceptibility were achieved by increasing the magnesium content to at least 4wt%. Effects of other alloying elements were not studied in detail.

Selection of 6000 Series Alloy and Filler Wire

When 6000 (Al-Mg-Si) alloys were welded, the welds formed had lower strength (cf. base material), lower ductility and much higher crack sensitivity than the 5000 series alloys welded. This might be expected from these precipitation strengthened alloys which would regain some of their original properties on post-weld heat treatment if the weld-related cracking was low. Of all the as-welded properties, the crack susceptibility tended to be the easiest to control using filler wire. A 12wt%Si containing filler (4047) reduced cracking significantly in tests.

Selection of Laser Source

Comparison of performances of the CO ₂ (continuous wave) and the Nd:YAG laser shows differences in terms of the welding speed capabilities and the mechanical performance of the joints. The Nd:YAG laser can be delivered to the workpiece by flexible fibre optics, whereas a CO ₂ laser requires a more precisely mounted mirror based system. With continuous wave Nd:YAG lasers available in powers in excess of 4kW, processing speeds of around 10.0m/min when welding aluminium sheet (2 mm thick) are possible, compared with 5.0-7.5 m/min with a 5kW CO ₂ laser.

Increased speed of welding with an Nd:YAG laser might be explained by enhanced coupling at the aluminium surface, which normally occurs with use of 1.06µm light compared with 10.6µm of the CO ₂ laser. The W Nd:YAG laser would give full penetration welds at only 0.8-1.6m/min.

Possibly related to these differences in power and speed, and hence weld width differences, are the differences in mechanical performance, although the performance of welds made using the 4.5kW Nd:YAG laser has not been measured. In most materials tested, ductility and formability of the welds were higher using the Nd:YAG laser, while crack susceptibility and tensile strength were often similar. It may, therefore, be advantageous in many applications to use an Nd:YAG laser for welding, especially if complex manipulation of the beam is required.

Conclusions

Laser welding of 1.6 - 2.0 mm thick aluminium alloys using a 5 kW continuous wave CO ₂ laser, a 2 kW continuous wave (CW) modulated Nd:YAG laser and a 4.5 kW CW Nd:YAG laser followed by assessment of various mechanical properties has led to the following:

General

Laser welding can be applied to 1.6 - 2.0 mm aluminium sheet giving reliable butt and lap welds at processing speeds of around 6 m/min using a 5 kW CO ₂ laser, 1 m/min using a 2 kW Nd:YAG laser and 12-16m/min using a 4.5kW Nd:YAG laser.

Crack Sensitivity

Solidification crack sensitivity measured using a self-restraining crack test is highly dependent upon the composition of the alloy used but not so dependent on laser type. In 5000 series (Al-Mg) alloys, there is an increase in cracking up to ~2%Mg followed by reduced cracking with further increases in Mg. Much higher crack sensitivity was observed in 6000 series (Al-Mg-Si) alloys except when 4047 filler wire (Al-12wt% Si) was used.

Mechanical Properties

Formability (measured using a biaxial bulge test) and ductility (from a transverse tensile test) tend to be greatest for Al-Mg alloys containing approximately 3wt% Mg. Up to 15% elongation was shown by 5754 alloy in CO ₂ and Nd:YAG laser welding. Tensile properties tend to be high for these alloys, with over 95% of the parent strength in the weld especially when high Mg (~5wt%) filler wire is used. The 6000 series alloys in the as welded condition have lower tensile strength (58-72% of parent) and elongation (0.7-3.0%).

Porosity

Porosity ranged from 0 to 24 pores (equivalent to 0.5 mm diameter average) per 100 mm of weld bead, but showed no clear trends with alloy welded, laser used or mechanical properties after welding.

Selection for Application

Careful selection of alloy type and laser type is required for any given application. Special attention should be given to productivity, application complexity, weld strength, weld formability and cracking tendency in the requirement of any given design or manufacturing route.

Guidelines for laser welding aluminium alloys

Principles

• Need to overcome high reflectivity and thermal conductivity by exceeding threshold power density
  - CO ₂ laser : 4 x 10 ⁶ W/cm ²
  - Nd:YAG laser: 1.5 x 10 ⁶ W/cm ²
• In practice, achieved by, for example
  - CO ₂ laser: up to 5kW power, 150mm focal length lens, 20mm beam diameter.
  - Nd:YAG laser: 2kW average power with 5kW peak power, or approx. 5kW average power alone with a 0.5mm focus spot diameter.

Welding Conditions

• Tilt workpiece to avoid back reflections
• Gas shielding
  - coaxial or side jet: 30 l/min
  - helium
  - underbead shielding desirable

Typical conditions - Butt joints

Thickness mm	CO 2 laser		Nd:YAG laser
	Power, kW	Speed, m/min	Power, kW	Speed, m/min
2	5	6	2 (<5 peak) 5	1 12
6	5	1	possible but not developed at present
6	10	5

• Wire feed welding is possible

Welding Tolerances - 2mm thickness

Butt joints	CO 2 laser*	Nd:YAG laser**
Focus position	±1mm	±0.5mm
Gaps	<0.3mm	<0.5mm
Beam/joint misalignment	<0.3mm	<0.5mm
Surface mismatch	up to 50% sheet thickness	Up to 50% sheet thickness

Notes: CO ₂ laser* - 5kW power, 6m/min speed, 150mm focal length lens, 20mm beam diameter.

Nd:YAG laser** - 2kW power (up to 5kW peak), 1m/min speed, 100mm focal length lens, 1mm diameter fibre.

Welding of Different Alloy Types

Variations in optimum welding parameters and weld properties occur when different alloys are welded. For detailed parameters and expected results make reference to Tables 4 and 5.

References

1.	Jones I et al	'CO 2 laser welding of aluminium alloys', Proc. of LAMP '92, Nagaoka, Japan, June, 1992, pp.523-528.
2.	Cieslak M J et al	On the weldability, composition and hardness of Nd:YAG laser welds in aluminium alloys 6061, 5456 and 5086, Metal. Trans. B, 1988, Vol.19B, No.4, pp.319-329.
3.	Sepold G	'CO 2 laser welding of aluminium alloys for transport systems, IIW Doc. IV-539-90.
4.	Binroth C A et al	CO 2 laser welding with filler material of high strength aluminium alloys, Proc. of Power Beam Technology, Stratford-upon-Avon, U.K., 23-26 Sep. 1990, pp.119-127.
5.	Houldcroft P T	A simple cracking test for use with argon-arc welding, Brit. Welding J., 1955, Vol.2, No.10, pp.471-475.
6.	Matsuda F et al	Moving characteristics of weld edges during solidification in relation to solidification cracking in GTA weld of aluminium alloy thin sheet, Tran. of JWRI, 1980, Vol.9, No.2, pp.83-93.
7.	Matsuda F et al	A new test specimen for self-restraint solidification crack susceptibility test of electron beam welding bead, Tran. of JWRI, 1982, Vol.11, No.2, pp.87-94.