Sage Journals: Discover world-class research

Abstract

Conventional two-dimensional optical microscopy is the most common tool for capturing and quantifying keyhole collapse porosity and other defects formed during the laser welding of aluminium alloys. This approach is of limited utility since only a single longitudinal or transverse metallographic cross-section, comprising only a small fraction of the overall weld volume, can be extracted or analysed at a time. X-ray computed tomography tools, on the other hand, enable three-dimensional visualisation and analysis of porosity and defects to be made across the full weld volume. When comparing these two measurement techniques, two-dimensional microscopy tools significantly underestimated the fraction of defects present within the weld at levels between 50% and 80%. Three-dimensional x-ray computed tomography tools also provide a means for accurately characterising the location, size and morphology of irregularly shaped defects. By categorising the defect shapes using a Zingg's shape analysis, in which the distribution of defect morphologies could be quantified as disk, rod and spherical shapes, a transition in defect morphology moving from stationary beam to oscillating beam welds was detected and quantified.

Introduction

High-strength aluminium alloys are widely used in the automotive and aerospace industries,^1,2 but they are susceptible to the formation of process-related defects during laser welding.^3–14 At power density levels above 10⁵ W/cm²,¹⁵ a vapour cavity or keyhole forms,^16–18 driving the formation of large depth to width ratios characteristic of laser welds. However, the keyhole can become unstable when an imbalance forms between the vapour recoil pressure and surface tension forces at the tip of the keyhole, leading to the formation of internal porosity at the advancing solidification front.^19–26 Common high strength aluminium alloys are particularly susceptible to the formation of keyhole collapse porosity, thus limiting their more widespread use of laser welding in joining these alloys, particularly in critical applications.^27,28

Tools that oscillate the beam in linear, circular, sinusoidal and infinity patterns have shown promise for mitigating keyhole collapse porosity.²⁹ By integrating different beam oscillation patterns, porosity levels in Al laser welds have been reduced by nearly 91%,³⁰ with specific reductions in area fractions of defects decreasing from 34% to 2% in a 5083 Al alloy³¹ and from 27% to 0% in a 5A06 aluminium alloy.³² These reductions in porosity have been attributed to an increase in the solidification time in the molten pool, allowing the defects formed during the collapse of the keyhole to escape from the molten pool.³¹ In other cases, circular and infinity oscillation patterns were used. Reductions in keyhole collapse porosity observed in these welds were attributed to a reduction in the temperatures and corresponding vapour pressures along the rear wall of the keyhole. With these decreased temperatures, the keyhole wall becomes stabilised, thus minimising the possibility of the formation of a bulge at the keyhole wall that can lead to keyhole collapse.³²

Measurement of porosity in these laser welds, however, was limited to the use of traditional metallographic tools, in which longitudinal or transverse cross-sections were extracted from each weld. Even though this simple technique is widely used,^33–35 it provides limited information on the defect morphology, location and number density since each image captures only a small fraction of the overall weld volume. To enhance the accuracy of these metallography-based porosity measurements, additional techniques such as Dye Penetrant Testing and Magnetic Particle Inspection³⁶ have been used. However, these methods are confined to quantifying only surface and sub-surface defects. Serial sectioning techniques,^37,38 in which adjacent cross-section samples are extracted across a selected length of the weld and then reconstructed, provide a means for performing a three-dimensional analysis of the defect structure. This technique, however, is time-consuming and destructive, thus limiting further analysis, and has limitations to its accuracy since only a small fraction of the length of the weld line can be analysed at any one time.

X-ray computed tomography (CT) methods provide a much more detailed and comprehensive pathway for capturing the locations and morphologies of defects and other related structures.^39–41 For example, x-ray micro-CT analysis has been used to explore the impact of changes in processing parameters on the size, distribution and morphology of defects formed during the laser-welding of a 304L austenitic stainless steel.⁴² Using an interfacial shape distribution analysis on the x-ray CT data, the defects were found to be elliptical rather than spherical in shape. Since no direct comparisons were made with other two-dimensional characterisation techniques, the advantages of the three-dimensional x-ray CT scanning in quantifying defect volume, frequency and shape of these internal defects were not captured.

With the emergence of new tools for eliminating process related defects in laser welding, more accurate methodologies for quantifying the amount, morphology and distribution of defects are needed. By comparing conventional two-dimensional characterisation with three-dimensional tools, the improvements in performance of x-ray CT techniques were quantified in a series of Al laser welds fabricated both with and without beam oscillation. Three-dimensional x-ray CT tools significantly outperformed the standard two-dimensional ones in accurately capturing the overall amount of porosity, as well as the size, location and morphology of individual defects. Using image analysis and statistical tools to analyse these larger data sets, a more in-depth categorisation of the defect morphologies was made possible.

Experimental

Welding setup

A series of laser welds were performed on aluminium plates measuring approximately 25.4 mm in width, 50.8 mm in length and 3.2 mm in thickness, using a square-edge butt joint configuration. The joint edges were machined for proper fit-up and clamped securely using a weld fixture with clamps. To reduce the risk of solidification cracking and focus on keyhole collapse defects, dissimilar alloy plates were joined: AA6061 (Al-1 Mg-0.42 Si-0.25 Zn-0.19 Cu-0.27 Cr-0.21 Fe-0.15 Mn) on one side and AA4047 (Al-12.55 Si-0.05 Mg-0.05 Cu-0.43 Fe-0.18 Mn) on the other.

Welds with circular clockwise beam oscillation were produced using an IPG Photonics Model YLR-1000 multimode Yb-fibre laser (wavelength between 1070 and 1080 nm) with a FLW-D30 wobble head, producing a 55 µm beam diameter (d86.5) in sharp focus. Oscillation amplitude was fixed at 0.8 mm, and frequencies of 150 Hz and 475 Hz were applied. Laser power was set to 800 W and travel speed to 34 mm/s. Ultra-high purity argon (99.99%) was used for shielding at a flow rate of 0.0472 m³/s. An additional set of linear fixed beam welds (no oscillation) were also produced using a YLS-2000 multimode Yb-fibre laser with a custom processing head and a 150 mm final focusing lens, yielding a 300 µm beam diameter (d86.5). In order to achieve the same weld dimensions as those obtained with beam oscillation at the same linear travel speed, a higher laser power (1600 W) was used.

2D characterisation

Transverse cross-sections were extracted at three positions along the weld length (12 mm apart). Samples were mounted and prepared using standard metallographic procedures. They were ground with SiC papers (P320 to P4000), polished with 1 µm polycrystalline diamond slurry, and finally polished using 0.05 µm colloidal silica. Etching was done with Keller's Reagent (2.5 mL HNO₃, 1.5 mL HCl, 1 mL HF and 95 mL distilled water). Images were captured using a Zeiss SmartZoom 5 digital microscope. Defect area fractions, weld depth and width were measured using ImageJ software (see Figure 1 and Table 1).

Figure 1.

Overview of welding processing and two- and three-dimensional characterisation techniques for identifying defects in the Al laser welds.

Table 1.

Summary of weld dimension and defect measurements.

Weld depth (mm)	Weld width (mm)	2D measurements						3D measurements
Weld depth (mm)	Weld width (mm)	Orientation	Position	Total defect area (mm²)	Weld area (mm²)	Area fraction (%)		Total defect volume (mm³)	Weld volume (mm³)	Volume fraction (%)
Linear weld
1.51 ± 0.07	2.12 ± 0.12	Longitudinal	Left	0.09	19.05	0.48	0.96 ± 0.42	0.5	22.86	2.19
			Centre	0.21		1.12
			Right	0.05		0.27
		Transverse	#1	0.01	1.81	0.19	0.11 ± 0.18
			#2	0		0
			#3	0		0
			#4	0		0
			#5	0		0
			#6	0.008		0.44
Frequency = 150 Hz
1.46 ± 0.01	1.73 ± 0.03	Longitudinal	Left	0.16	18.54	0.84	0.86 ± 0.37	0.51	24.01	2.14
			Centre	0.23		1.23
			Right	0.09		0.5
		Transverse	#1	0.007	1.95	0.36	0.47 ± 0.41
			#2	0.01		0.5
			#3	0.02		1.23
			#4	0.005		0.25
			#5	0.001		0.007
			#6	0.008		0.46
Frequency = 475 Hz
0.99 ± 0.03	1.79 ± 0.03	Longitudinal	Left	0.32	12.57	2.57	1.03 ± 0.37	0.26	16.51	1.56
			Centre	0.04		0.35
			Right	0.02		0.18
		Transverse	#1	0	1.34	0	0.15 ± 0.21
			#2	0.005		0.38
			#3	0		0
			#4	0		0
			#5	0.006		0.46
			#6	0.001		0.07

3D CT scanning

A 14 mm section of the weld was preserved for non-destructive 3D imaging. Scans were performed using a GE v|tome|x L300 system at 210 kV accelerating voltage, 70 µA beam current and 900 ms exposure time, yielding a voxel size of 16 µm. A 0.5 mm copper filter was used to enhance contrast between the material and background. A total of 900 slices were acquired and reconstructed into a 3D volume using Avizo 9.7.0. All pores below the voxel size were excluded from analysis.

Image analysis methods

The defect morphologies in the two-dimensional metallographic cross-sections were quantified using standard parameters commonly used in particle analysis. Circularity, which is the measurement of the roundness or extent of deviation from a perfect circle of an individual defect, was determined using the relationship below:

Circularity = \frac{4 π A_{p}}{P^{2}}

(1)

where A_p is the surface area of a defect and P is the perimeter of the corresponding defect. A circularity value of one represents a perfect circle, and values below unity capture the extent of deviation from a perfect circle.

The convexity of the defects is a second metric that captures the presence or absence of inward curvatures or concavities along the perimeter of the defect and is defined using the equation below:

Convexity = \frac{P_{convex}}{P}

(2)

where P_convex refers to the convex hull perimeter,⁴³ representing the cumulative length of the boundary encompassing the smallest convex polygon that fully encloses the defect.

The aspect ratio characterises the elongated shape of the defects and can be evaluated using the following relationship:

Aspect ratio = \frac{Width}{Length}

(3)

When the defects are captured in three dimensions in the x-ray CT scanning, the metrics used to define defect morphology are modified. The overall morphology in three dimensions can be defined using a sphericity (y) metric, which is shown below:

ψ = {(\frac{36 π V_{p}^{2}}{A_{p}^{3}})}^{1 / 3}

(4)

where V_p is the volume and A_p is the surface area of the corresponding porosity. Here, y = 1 represents a perfect sphere.

Results and discussion

Metallographic techniques, in which multiple transverse or longitudinal cross-sections are extracted, are commonly used to capture the morphology, location and volume fraction of weld defects. This inexpensive and rapid approach is most useful in quickly capturing changes in weld dimensions and the appearance of defects with changes in processing parameters or materials. The utility of this approach is highlighted in Figure 1, in which the effect of beam oscillation on weld pool shape and size and the formation of processing defects is highlighted. Changes in the fusion zone shape and size are clearly captured and can be easily quantified in the transverse cross-sections with a rather high confidence level, since the weld dimensions do not significantly change with location along the weld length. Defects resulting from keyhole collapse are also visible in each transverse cross-section. However, the formation of weld defects typically occurs through the introduction of a process instability, such as keyhole collapse,⁴⁴ making their appearance less predictable and decreasing the likelihood that a statistically significant sampling of these defects can be captured in a single transverse cross-section.

Since an individual transverse cross-section captures only a single plane from the entire weld line, measurements across multiple cross-sections need to be obtained in order for any reliable analysis to be completed. Using this methodology, area fractions of the defects present in a baseline linear weld and two welds produced at a beam oscillation amplitude of 0.8 mm and frequencies of 150 and 475 Hz were measured and summarised in Table 1. In order to replicate metallographic measurements without destructive examination, three parallel longitudinal cross-sections were extracted from the x-ray CT scanning data at locations along the weld centreline and at parallel slices located 0.33 mm on either side. Six transverse cross-sections were also extracted at 0.5 mm intervals along the length of the weld in the same manner. Images capturing each of these longitudinal and transverse cross-sections are shown in Figure 2.

Figure 2.

Three-dimensional and two-dimensional characterisation of weld defects of (a) linear and beam oscillation welds produced with frequencies of (b) 150 Hz and (c) 475 Hz. The top row shows a three dimensional reconstruction of defect distribution within the welds. The middle three rows present longitudinal cross-sections along the weld direction, and the bottom two rows show transverse cross-section perpendicular to the welding direction.

Differences in the measurement of the defect levels across the transverse and longitudinal cross-sections are evident when comparing the average values of the area fractions of porosity in each weld. In the linear weld, transverse cross-sections shown in Figure 2 exhibited a range of defect levels, with several revealing no defects while others displayed defect area fractions between 0.44% and 0.49%. Transverse cross-sections extracted from the beam oscillation weld made at a frequency of 475 Hz displayed defect area fractions between 0.46% and 9%. Similar differences in the porosity measurements across these weld conditions were also observed in the longitudinal cross-sections. Even small changes in the location where the slices are extracted in the same weld have an impact on the area fraction of defects as well as their general size and morphology. For example, a shift of only 0.33 mm from the weld centreline can produce significant changes in the defects that can be captured and the resulting area fraction measurements. In the case of linear weld and the beam oscillation weld made at a frequency of 150 Hz, centreline longitudinal cross-sections exhibited porosity area fractions of 1.12% and 1.23%, respectively. These differences in defect area fraction measurements for the transverse and longitudinal cross-sections were accompanied by significant scatter or large standard deviations in the area fraction measurements. While present in both orientations, the scatter became even more pronounced in the transverse cross-sections, since the overall weld area is much smaller than that captured by the longitudinal cross-sections.

These differences in porosity measurements with both orientation and location along the weld highlight the shortcomings of solely relying on two dimensional measurements. Many of these shortcomings can be overcome by the application of three-dimensional x-ray CT scanning techniques. Three-dimensional representations of the linear and beam oscillation welds made at different frequencies are shown in Figure 2. By capturing the entire weld in three dimensions, the complex networks of defects that are present along the length of the weld can be identified. Using these three-dimensional images, measurements of the volume fractions of the porosity present in each weld can be made and are summarised in Table 1. Since these measurements are made over the entire weld volume, they capture all of the defects, resulting in consistently higher volume fractions than what was measured in the two-dimensional slices, regardless of orientation. For example, the three-dimensional characterisation revealed porosity levels of 2.19% for the linear weld and 2.14% for the beam oscillation weld made at a frequency of 150 Hz. Both values are much higher than those measured from the two-dimensional slices alone for both conditions. Similar results are also apparent for the beam oscillation weld made at a frequency of 475 Hz.

A comparison of the results of the three-dimensional scans with the two-dimensional measurements made along both longitudinal and transverse orientations is provided in Figure 2 and highlights a range of differences in the number, size and shape of the defects. Defect size measurements made in two dimensions along the longitudinal orientation along the weld centreline and in three dimensions are compared in Figure 3. By comparing the histograms breaking down the size distributions for individual defects using both techniques, it is clear that the measurements obtained in three dimensions capture not only more defects but also a wider range of defect sizes at both ends of the size distribution.

Figure 3.

Comparison between number of pores and pore volumes and diameters measured using three-dimensional techniques for (a and d) linear, (b and e) 150 Hz oscillation frequency and (c and f) 450 Hz oscillation frequency welds. The changes in sphericity as a function of the pore diameter are captured for the (f) linear, (g) 150 Hz oscillation frequency and (h) 450 oscillation frequency welds.

The locations of individual defects in the weld pool can also be obtained using both two-dimensional and three-dimensional measurements. These locations can be defined with respect to the weld depth and width, as shown in Figure 4, and a normalised depth can be determined using the following relationship:

Normalized depth = \frac{depth of porosity}{depth of weld}

(5)

Figure 4.

Measurement of defect locations using both two- and three-dimensional techniques for linear and beam oscillation welds with frequencies of 150 Hz and 475 Hz.

For the two-dimensional measurements of defect location along the weld depth, centreline longitudinal cross-sections were used when comparing the locations to the three dimensional results. The three-dimensional characterisation utilised coordinates obtained from x-ray CT scans, which were then converted to normalised depths. When measuring the weld width locations, multiple transverse cross-sections were taken along the weld length, ensuring at least one defect was present in each cross-sections for the two-dimensional technique.

The inherent limitation of two-dimensional analysis in capturing comprehensive data is evident in Table 2, with the two-dimensional characterisation methods consistently underestimating the number of defects by approximately 87% to 96% at locations along the weld width and depth. This significantly lower capture rate impacted the resulting position distribution across the depth and width of the welds. As shown in Figure 4, the defects are primarily distributed at normalised depths ranging from 0.4 to 1, meaning that the defects are primarily located at higher weld depths. This observation is further supported by the three-dimensional visualisation included in Figure 4, which clearly shows the clustering of defects near the root of the weld. The 3D representation provides an intuitive spatial understanding of defect distribution that complements the quantitative data and highlights the limitations of relying solely on two-dimensional sections for locating defects.

Table 2.

Summary of defect morphology measurements made using both two- and three-dimensional techniques.

	Number of pores			Circularity	sphericity	Convexity	Aspect ratio
	Two dimensional		Three-dimensional
	Transverse	Longitudinal	Three-dimensional
Linear weld	3	14	47	0.94 ± 0.05	0.99 ± 0.04	0.93 ± 0.05	0.69 ± 0.13
Frequency = 150 Hz	8	20	157	0.74 ± 0.13	0.91 ± 0.06	0.92 ± 0.04	0.59 ± 0.16
Frequency = 475 Hz	5	6	105	0.85 ± 0.5	0.94 ± 0.05	0.91 ± 0.06	0.68 ± 0.11

There is limited information that can be extracted concerning the shape or morphology of the individual defects from the two-dimensional analysis using the longitudinal or transverse cross-sections alone. Since a single plane is obtained along the weld line, only a two-dimensional representation of each three-dimensional defect is captured. The random two-dimensional longitudinal or transverse slices made along the weld length capture only a small portion of the defect, and their random orientation and location impacts the morphological analysis of the defect. As illustrated in Figure 5, the two-dimensional representation of a three-dimensional defect can be affected by the orientation of the defect and the angle at which the cross-section slices the particle.

Figure 5.

Schematic representation of how particle morphology measurements are impacted when two-dimensional techniques are used to characterise three-dimensional defects.

While some useful information on the circularity, convexity and aspect ratio of defects can still be obtained, three-dimensional measurements are still required to accurately capture the overall shape of a defect. Table 2 provides a summary of both two-dimensional measurements made on defects that resided along the centreline longitudinal cross-section of the welds and three-dimensional characterisation tools. Circularity values were typically between 10% and 20% lower than the corresponding sphericity values measured using three-dimensional techniques. For example, the average circularity value in the linear weld was 0.94, but the average sphericity obtained from the three-dimensional measurements was 0.99. Similarly, for beam oscillation welds with frequencies of 150 Hz and 475 Hz, circularity values were 0.8 and 0.7, respectively, while sphericities were 0.9 and 0.89, respectively. Measurement of the aspect ratios using these same three-dimensional techniques allowed a slight elongation of the shape to be determined.

Given the shortcomings of relying on the two-dimensional methods for characterisation of the internal defects, it is beneficial to focus on the results of the three-dimensional tools and identify a general metric for defining defect shape. Zingg's shape analysis⁴⁵ offers an approach to characterising the shape of defects in three-dimensional space. Unlike two-dimensional shape factors that assume defects to be circular, Zingg's shape analysis provides a detailed and accurate description of complex geometries and is applicable to three dimensions. This analysis⁴⁶ involves determining the specific shape characteristics of a three-dimensional defect based on the relative ratios of its three axes, defined by the longest dimension (l), intermediate dimension (i), and the shortest dimension (s). The elongation, which is defined by the ratio between the intermediate and longest dimensions and flatness, which is defined as the ratio between the shortest and intermediate dimensions, are used. By comparing the ratios between elongation and flatness, additional insight into the general shape of the defect can be obtained. As shown in Figure 6, these measurements can be used to define rod, blade, sphere and disk shapes. In the linear welds, the defects primarily fall within the spherical quadrant, corresponding with the high sphericity values (0.99) measured for the individual defects. With the addition of beam oscillation, lower sphericity values of approximately 0.9 were observed along with a shift in the defect morphology in the Zingg's analysis, which shifted to more disk- and rod-shaped defects. This level of fidelity in capturing differences in the pore shape is only possible using three-dimensional characterisation tools.

Figure 6.

(a) Definition of Zingg's shape classification for rod, sphere, blade and disk shape defects. Distributions of defect morphology measurements for the (b) linear, (c) 150 Hz oscillation frequency and (d) 450 Hz oscillation frequency welds.

Summary and conclusions

Accurately characterising the size, location and morphology of defects present in aluminium laser welds is a pre-requisite for producing defect-free and structurally sound welds. Two-dimensional microscopy-based techniques, which extract a relatively small fraction of the weld volume in the form of a single transverse or longitudinal cross-section, are the most common approach for microstructural characterisation but are limited in their accuracy. Three-dimensional x-ray CT, however, provides a full volume analysis in which all defects along the weld line are captured and precisely quantified. When comparing these two techniques, the superior capabilities of the three-dimensional characterisation tools in identifying the location, fraction, size and morphology of individual defects over the two-dimensional techniques was evident. By utilising these three-dimensional tools, more accurate measurements can be made and provide further insight into the mechanisms driving defect formation in laser welds made with and without beam oscillation. Additional findings are provided below:

The use of transverse and longitudinal cross-sections for measuring the amount of porosity in a weld consistently under-report the area fraction of defects when compared to three-dimensional x-ray CT techniques

Significant scatter is observed in area fraction measurements in both transverse and longitudinal cross-sections obtained at different locations along the length of the weld.

The location and distribution of pores across the weld length are more accurately captured using three-dimensional tools, far out-performing the longitudinal cross-sections which are more commonly utilised. In the linear and beam oscillation Al welds, defects were predominantly located near the root of the weld and within 0.2 to 0.3 mm of the fusion zone boundary.

Characterisation of keyhole pore morphology using two-dimensional characterisation techniques is affected by the locations where measurements are made and captures only a portion of the defect, producing inconsistent results when measuring circularity. The use of three-dimensional analysis allowed the entire defect to be captures and more accurate measurements of the sphericity to be obtained.

The implementation of Zingg's shape analysis of defects present in the linear and beam oscillation welds provided more in-depth information on the evolution of defect shapes. Primarily sphere-shape defects were observed in linear welds. With the addition of beam oscillation, the shape of the defects transitioned to more disk and rod shapes.

Footnotes

Acknowledgements

The authors would like to acknowledge the Manufacturing and Materials Joining Innovation Center (Ma2JIC) funded through the National Science Foundation Industry University Cooperative Research Center program (IUCRC) for its support in this research (Award Number 2052747). Additional thanks also go to Jack Herrmann for his extensive laser welding expertise and laboratory contributions to this work. Sandia National Laboratories is a multimission laboratory managed and operated by National Technology and Engineering Solutions of Sandia, LLC, a wholly owned subsidiary of Honeywell International, Inc., for the U.S. DOE's National Nuclear Security Administration under contract DE-NA-0003525. The views expressed in the article do not necessarily represent the views of the U.S. DOE or the United States Government.

ORCID iDs

Jeffrey Rodelas

Todd A. Palmer

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Directorate for Engineering, Department of Energy National Nuclear Security Administration (Grant Number 2052747, DE-NA-0003525).

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

References

Miller

Zhuang

Bottema

, et al. Recent development in aluminium alloys for the automotive industry. Mater Sci Eng: A 2000; 280: 37–49.

Heinz

Haszler

Keidel

, et al. Recent development in aluminium alloys for aerospace applications. Mater Sci Eng: A 2000; 280: 102–107.

Sheikhi

Malek

Assadi

. Prediction of solidification cracking in pulsed laser welding of 2024 aluminum alloy. Acta Mater 2015; 82: 491–502.

Zhao

Niu

, et al. Investigation on porosity suppression in deep-penetration laser welding by using computational fluid dynamics. J Laser Appl 2016; 28: 032011.

Shi

Jiang

, et al. Numerical study of keyhole-induced porosity suppression mechanism in laser welding with beam oscillation. Sci Technol Weld Joining 2021 Jul; 26: 349–355.

Chen

Jiang

, et al. Numerical investigation of asymmetric weld fusion geometry in laser welding of aluminium alloy with beam oscillation. Sci Technol Weld Joining 2022 Nov; 27: 595–605.

Tan

Zhang

Liu

, et al. Effect of beam oscillation frequency on porosity in laser welding of aluminum alloy lap-butt joints. Sci Technol Weld Joining 2024; 29: 124–133.

Oliveira

, et al. Modeling and numerical study of keyhole-induced porosity formation in laser beam oscillating welding of 5A06 aluminum alloy. Opt Laser Technol 2021 Jan 1; 133: 106540.

Wang

Liu

Yang

, et al. Study of porosity suppression in oscillating laser-MIG hybrid welding of AA6082 aluminum alloy. J Mater Process Technol 2021 Jun 1; 292: 117053.

10.

Gong

Xia

, et al. Influence of scan paths on flow dynamics and weld formations during oscillating laser welding of 5A06 aluminum alloy. J Mater Res Technol 2021 Mar 1; 11: 19–32.

11.

Soysal

Kou

. Effect of filler metals on solidification cracking susceptibility of al alloys 2024 and 6061. J Mater Process Tech 2019; 266: 421–428.

12.

Richardson

. Mechanism and possible solution for transverse solidification cracking in laser welding of high strength aluminium alloys. Mater Sci Eng: A 2006; 429: 287–294.

13.

Coniglio

Cross

. Mechanisms for solidification crack initiation and growth in aluminum welding. Metall Mater Trans A 2009; 40: 2718–2728.

14.

Kou

. A criterion for cracking during solidification. Acta Mater 2015; 88: 366–374.

15.

Tsukamoto

. Laser welding. Weld Int 2003; 17: 767–774.

16.

Lee

Farson

, et al. Mechanism of keyhole formation and stability in stationary laser welding. J Phys D: Appl Phys 2002; 35: 1570.

17.

Kaplan

. A model of deep penetration laser welding based on calculation of the keyhole profile. J Phys D: Appl Phys 1994; 27: 1805.

18.

Fabbro

. Melt pool and keyhole behaviour analysis for deep penetration laser welding. J. Phys. D: Appl Phys 2010; 43: 445501.

19.

Xiao

Zhang

. Problems and issues in laser beam welding of aluminum-lithium alloys. J Manuf Process 2014; 16: 166–175.

20.

Xiao

. Laser beam welding of aluminum alloys. In: Lasers in material processing and manufacturing III. SPIE, 2008 Feb 1, Vol. 6825, pp. 40–49.

21.

Katayama

Nagayama

Mizutani

, et al. Fibre laser welding of aluminium alloy. Weld Int 2009; 23: 744–752.

22.

Kawahito

Matsumoto

Abe

, et al. Relationship of laser absorption to keyhole behavior in high power fiber laser welding of stainless steel and aluminum alloy. J Mater Process Tech 2011; 211: 1563–1568.

23.

Matsunawa

. Problems and solutions in deep penetration laser welding. Sci Technol Weld Join 2001; 6: 351–354.

24.

Matsunawa

Kim

Seto

, et al. Dynamics of keyhole and molten pool in laser welding. J Laser Appl 1998; 10: 247–254.

25.

Pastor

Zhao

DebRoy

. Continuous wave-nd: yttrium–aluminum–garnet laser welding of AM60B magnesium alloy. J Laser Appl 2000; 12: 91–100.

26.

Huang

Hua

, et al. Numerical study of keyhole instability and porosity formation mechanism in laser welding of aluminum alloy and steel. J Mater Process Technol 2018; 252: 421–431.

27.

Huang

Hua

, et al. Role of welding speed on keyhole-induced porosity formation based on experimental and numerical study in fiber laser welding of al alloy. Int J Adv Manuf Tech 2019; 103: 913–925.

28.

Haboudou

Peyre

Vannes

, et al. Reduction of porosity content generated during nd:YAG laser welding of A356 and AA5083 aluminium alloys. Mater Sci Eng: A 2003; 363: 40–52.

29.

Rubben

Mohrbacher

Leirman

. Advantages of using an oscillating laser beam for the production of tailored blanks. In: Lasers in material processing. SPIE, 1997 Aug 18, Vol. 3097, pp. 228–241.

30.

Xiao

Zou

, et al. Weld formation mechanism during fiber laser welding of aluminum alloys with focus rotation and vertical oscillation. J Manuf Process 2018; 36: 149–154.

31.

Wang

. A study on laser beam oscillating welding characteristics for the 5083 aluminum alloy: morphology, microstructure and mechanical properties. J Manuf Process 2020; 53: 12–20.

32.

Oliveira

, et al. Modeling and numerical study of keyhole-induced porosity formation in laser beam oscillating welding of 5A06 aluminum alloy. Opt Laser Technol 2021; 133: 106540.

33.

Fetzer

Hagenlocher

Weber

, et al. Geometry and stability of the capillary during deep-penetration laser welding of AlMgSi at high feed rates. Opt Laser Technol 2021; 133: 106562.

34.

Rong

Huang

, et al. Keyhole-induced porosity formation during laser welding. J Mater Process Technol 2018; 252: 720–727.

35.

Kuo

Jeng

. Porosity reduction in nd–YAG laser welding of stainless steel and inconel alloy by using a pulsed wave. J Phys D: Appl Phys 2005; 38: 722.

36.

Ranganayakulu

Burra

Ravi

. Characterization of weldments defects through non destructive evaluation techniques. Indian J Sci Technol 2017; 10: 1–9.

37.

Madison

Susan

Kilgo

. 3D RoboMET Characterization. Sandia National Lab.(SNL-NM), Albuquerque, NM (United States); 2015.

38.

Desbois

Urai

Kukla

, et al. High-resolution 3D fabric and porosity model in a tight gas sandstone reservoir: a new approach to investigate microstructures from MM- to NM-scale combining argon beam cross-sectioning and SEM imaging. J Petrol Sci Eng 2011; 78: 243–257.

39.

Miyagi

Kawahito

Wang

, et al. X-ray phase contrast observation of solidification and hot crack propagation in laser spot welding of aluminum alloy. Opt Express 2018; 26: 22626–22636.

40.

Galos

Ghaffari

Hetrick

, et al. Novel non-destructive technique for detecting the weld fusion zone using a filler wire of high x-ray contrast. NDT & E Int 2021; 124: 102537.

41.

Roy

Ghosh

Bhattacharyya

, et al. Weld defect identification in friction stir welding through optimized wavelet transformation of signals and validation through X-ray micro-CT scan. Int J Adv Manuf Technol 2018; 99: 623–633.

42.

Madison

Aagesen

. Quantitative characterization of porosity in laser welds of stainless steel. Scr Mater 2012; 67: 783–786.

43.

Liu

Cashman

Rust

. Optimising shape analysis to quantify volcanic ash morphology. Geo Res J 2015; 8: 14–30.

44.

Matsunawa

Mizutani

Katayama

, et al. Porosity formation mechanism and its prevention in Laser welding. Weld Int 2003; 17: 431–437.

45.

Zingg

. Beitrag Zur Schotteranalyse. Zürich: A.-G. Gebr. Leemann & Co, 1935.

46.

Blott

Pye

. Particle shape: a review and new methods of characterization and classification. Sedimentology 2008; 55: 31–63.

Accurate measurement of porosity size and morphology in aluminium alloy laser welds

Abstract

Introduction

Experimental

Welding setup

2D characterisation

3D CT scanning

Image analysis methods

Results and discussion

Summary and conclusions

Footnotes

Acknowledgements

ORCID iDs

Funding

Declaration of conflicting interests

References