Circumventing Solidification Cracking Susceptibility in Al–Cu Alloys Prepared by Laser Powder Bed Fusion

Abstract

Laser powder bed fusion (LPBF) of Al–Cu alloys shows high susceptibility to cracking due to a wide solidification temperature range. In this work, 2024 alloys were manufactured by LPBF at different laser processing parameters. The effect of processing parameters on the densification behavior and mechanical properties of the LPBF-processed 2024 alloys was investigated. The results show that the porosity increases significantly with increasing laser power, while the number of cracks and lack-of-fusion defects increase distinctly with increasing scan speed. The solidification cracking susceptibility of the LPBF-processed 2024 alloys prepared at different processing parameters was analyzed based on a finite element model, which was accurately predicted by theoretical calculations. Dense and crack-free 2024 samples with a high densification of over 98.1% were manufactured at a low laser power of 200 W combined with a low laser scan speed of 100 mm/s. The LPBF-processed 2024 alloys show a high hardness of 110 ± 4 HV_0.2, an ultimate tensile strength of 300 ± 15 MPa, and an elongation of ∼3%. This work can serve as reference for obtaining crack-free and high-performance Al–Cu alloys by LPBF.

Introduction

The laser powder bed fusion (LPBF) technique uses a high-energy laser to locally melt metallic powder based on a layer-by-layer fabrication principle, allowing to manufacture metallic components with high geometric freedom.¹ With increasing demands for complex components, LPBF has attracted growing attention and has been successfully applied to manufacture Al-,^2–4 Fe-,^5–7 Ti-,^8–10 and Ni-^11,12 based alloys. Among these alloys, aluminum alloys are commonly used as structural materials due to their low density and high specific strength. It is, therefore, of great importance to obtain high-performance aluminum alloy components prepared by LPBF to extend the range of structural applications.¹³

During the LPBF process, the interaction time between the laser and metallic powder is extremely short, complicating the metallurgical processes due to the large temperature gradients and cooling rates of the molten pool.¹⁴ The as-fabricated aluminum alloys tend to form metallurgical defects, such as balling, porosity, and cracks, because of high laser reflectivity and thermal conductivity and affinity to oxygen.¹⁵ This puts forward strict requirements for laser processing optimization to obtain high-performance aluminum alloys with favorable manufacturing quality. Previous studies have mainly focused on high-castability Al–Si alloys (i.e., AlSi10Mg, Al–12Si).^16,17 The Si content in these alloys is close to the eutectic composition of the Al–Si phase diagram, exhibiting good fluidity and laser processability.¹⁸ However, the enhancement of mechanical properties is very limited for structural applications. Al–Cu alloys are utilized in large quantities as structural materials because of their high specific strength, good fatigue properties, and heat treatable strengthening.^19,20

Considering their steep solidus line, it needs to pass a large semisolid zone upon solidification.^21,22 In such a case, hot cracks tend to form in the final stage of solidification in the LPBF process, thus essentially limiting the manufacturing quality of LPBF-processed Al–Cu alloys.

To suppress the cracks in these Al–Cu alloys, attempts have been made to optimize the processing parameters of LPBF. Zhu colleagues^22,23 have reported that crack-free Al–Cu alloys were obtained at a laser volumetric energy density over 340 J/mm³. It was found that cracks can be eliminated with decreasing scanning speed at laser powers of 200 and 180 W, owing to increased track width and remelted depth. However, the authors did not further investigate the case of laser powers over 200 W due to equipment limitations.

Although Tan et al²⁴ have fabricated Al–Cu–Mg alloys in a wide process window with laser powers of 200–375 W and scan speeds of 600–1650 mm/s, cracks were still observed in all the specimens. It is worth noting that crack-free Al–Cu–Mg–Si alloys were manufactured at a low scan speed of 165 mm/s.²⁵ The heat-treated specimens showed a tensile strength of 455 MPa and an elongation of 6.2%, which are as high as those of wrought 2024-T6 alloys. Even though substrate preheating,²⁶ design of support structures,²⁷ and optimization of scanning strategy²⁷ have been further proposed, cracks cannot be thoroughly eliminated from the as-fabricated Al–Cu alloys.

It is well established that cracking of LPBF-processed Al–Cu alloys can be effectively suppressed by careful control of the processing parameters, while the corresponding mechanism and the correlation of the crack susceptibility and processing parameters still remain unclear.^28–30

In this work, 2024 Al–Cu alloys were prepared by LPBF with a laser power variation of 200–400 W and a scan speed variation of 100–800 mm/s. The influence of the laser processing parameters on the densification behavior and the mechanical properties of the 2024 alloys manufactured by LPBF was investigated. The formation mechanisms of metallurgical defects such as porosity and cracks in the LPBF-processed 2024 alloy specimens were revealed. The cracking susceptibility of 2024 alloy to laser processing parameters was evaluated by combining finite element simulation and theoretical calculations.

Experimental Procedures

This work is not involved animal trials or human subjects. The IRB approval or waiver statement is not necessary.

Powder material

Spherical gas-atomized 2024 alloy powder supplied by TLS Technik with some satellites on the powder particles was used as starting material in this work (Fig. 1a). According to X-ray diffraction, the 2024 alloy powder was mainly composed of α-Al phase (Fig. 1b). The alloy powder had an average particle size of ∼36.4 μm, as analyzed by the ImageJ software (Fig. 1c), falling in the standard range of powder for LPBF process (30–50 μm). The chemical compositions of the powder commercially provided by the supplier and standard AA2024 are given in Figure 1d. It is found that the main element composition of the as-used Al-Cu powder falls in the composition range of standard AA2024 alloy. Both Mg and Mn elements of the used powder are also in the range of standard AA2024. Notably, the chemical composition of LPBFed samples characterized by the energy dispersive spectrometer (EDS) analysis shows some deviation from the used powder and standard AA2024 alloy. This is closely related to the error of Mg composition in this sample.

FIG. 1.

(a) SEM image revealing the powder morphology, (b) XRD pattern, and (c) particle size distribution for the 2024 alloy powder. (d) Comparison of chemical composition for the powder material and AA2024. The inset SEM image in (a) shows the alloy powder at high magnification. LPBF, laser powder bed fusion; SEM, scanning electron microscope; XRD, X-ray diffraction.

The content of Mg is difficult to be detected by EDS analysis, which is probably attributed to the combination of the influence of close and strong Al peak, Mg evaporation, and EDS detection limit.

LPBF process

The 2024 alloy specimens were fabricated using an LPBF-150 device. The LPBF-150 was self-developed by Nanjing University of Aeronautics and Astronautics and equipped with a YLR-500 fiber laser (with a maximum laser power of 500 W, a spot size of 70 μm, and a wavelength of 1.064 μm), a high-precision powder-layering apparatus, an inert gas protection system, and a computer control system. To prevent oxidation during the LPBF process, the alloy specimens were manufactured in a high-purity argon atmosphere to ensure that the O₂ content was below 50 ppm. Cuboid specimens with a length × width × height of 10 mm × 10 mm × 5 mm were built for forming quality analysis and microstructure characterization, and specimens with a length × width × height of 70 mm × 10 mm × 5 mm were produced for tensile tests. The specimens for the tensile tests were built perpendicular to the tensile direction and then machined according to the ASTM-E8 standard.

The laser processing parameters were as follows: laser power p = 200–400 W, scan speed v = 100–800 mm/s, hatch spacing h = 60 μm, and layer thickness t = 30 μm. An island scanning strategy with 37° rotation between layers was used to reduce the residual stresses of the specimens. The detailed processing parameters are listed in Table 1.

Table 1.

Sample Designations and Laser Processing Parameters Applied to Prepare the Different 2024 Alloy Specimens in This Work

Samples	Laser power (W)	Scan speed (mm/s)	Samples	Laser power (W)	Scan speed (mm/s)
A	200	100	G	300	400
B	200	200	H	300	800
C	200	400	I	400	100
D	200	800	J	400	200
E	300	100	K	400	400
F	300	200	L	400	800

Microstructure and property characterization

The relative density of bulk as-fabricated specimens was measured using the Archimedes method. The cross-sections parallel to the building direction (BD, parallel to Z direction) of the LPBF-processed 2024 alloy specimens were ground and polished according to standard metallographic procedures. The defects of the specimens were examined using an optical microscope (OM, XJP-300). The porosity defined by the ratio of pore area to total area is calculated by ImageJ software from at least three OM images at different areas for each specimen. The crack length is obtained by manually tracing the cracks using lines and counting the total line lengths using ImageJ software. The crack density is calculated by the crack length per unit area. Then the specimens were etched with Keller's reagent (1.0 mL hydrofluoric acid, 1.5 mL hydrochloric acid, 2.5 mL nitric acid and 95.0 mL distilled water) for 15 s, and characterized using a scanning electron microscope (TESCAN LYRA3). The grain size and orientation were characterized by electron backscattered diffraction (EDAX Velocity Super), and the kernel average misorientation was calculated to analyze the stress distribution.

The microhardness was measured on the polished specimens using a Vickers indenter (HXS-100AY) with a load of 200 g and a loading time of 10 s. The mean hardness value for each specimen was determined from 25 points with a space of 0.1 mm between each measurement. The tensile properties of the specimens were measured using a universal tensile tester (MTS E45.105) with a fixed strain rate of 2 mm/min.

Finite element simulation

Due to the very short existence time of the molten pool with complex thermodynamics, it is difficult to measure its temperature distribution accurately. To analyze the thermal behavior of the molten pool for the 2024 alloys, a finite element model (FEM) of the LPBF process was developed using the COMSOL software. The FEM was divided into half along the YZ plane (plane built from Y axis and Z axis), which contains the centerline of the molten pool (Fig. 2). This takes advantage of the symmetry of the molten pool to improve the simulation efficiency. The formed part and the powder bed were tetrahedrally meshed, while the molten pool and the surrounding area were hexahedrally meshed to improve the simulation accuracy. The thermophysical parameters of 2024 alloys at different operating temperatures, such as thermal conductivity, thermal expansion coefficient, and heat capacity, were calculated.

FIG. 2.

The finite element model used for simulating the molten pool in the LPBF process. BD, building direction.

Results and Discussion

Effect of laser processing parameters on the metallurgical defects

Figure 3 displays OM images of the cross-sections parallel to the BD of the samples fabricated under different LPBF processing parameters. Two typical metallurgical defects are observed in these specimens: pores and cracks. Based on their morphologies, the pores are divided into small spherical pores, near-spherical pores, and irregularly shaped voids. The diameter of the small spherical pore ranges from 5 to 20 μm in all specimens. These pores stem from the adsorbed gases existing in starting powder caught in the molten pool during the LPBF process. There are several near-spherical pores with a mean diameter of 73 ± 20 μm in the 2024 alloy samples fabricated at a high laser power of 400 W and a low scan speed of 100–200 mm/s (Figs. 3i, j, and 4a). These spherical pores result from the transition of the heat transfer from conduction mode to keyhole mode at a relatively high volumetric energy density.^31,32 In this case, multiple reflections of the laser radiation occur on the wall of the molten pool, significantly increasing the melting depth. Meanwhile, a large amount of vapor is generated, causing a strong recoil pressure toward the bottom of the molten pool.

FIG. 3.

(a–l) OM images of the LPBF-processed 2024 alloy samples fabricated at different laser powers and scan speeds. OM, optical microscope.

Once this recoil pressure is larger than the surface tension and the hydrostatic pressure of the melt, the melt on the upper wall of the molten pool collapses inward. Hence, the liquid solidifies at the bottom of the molten pool, forming large keyholes.³³ On the other hand, high-vapor-pressure Al and Mg in the 2024 alloy tend to evaporate at high temperatures. In such a view, the vapor cannot escape from the molten pool in the rapid cooling process, thus contributing additionally to the formation of near-spherical pores.³⁴ Irregularly shaped voids with a width of over 500 μm are observed in the samples C, D, H, and L (Fig. 3c, d, h, l), and unmolten alloy powders appear in these voids (Fig. 4b). This indicates that at a low laser power and a high scan speed, the laser energy input is insufficient to completely melt the alloy powder, resulting in the formation of lack-of-fusion defects in the samples.

FIG. 4.

Metallurgical defects revealing (a) a large near-spherical pore and (b) an irregularly shaped void with unmolten powder in the LPBF-processed 2024 alloys.

Figure 5 shows the variations of the relative density of the LPBF-processed 2024 alloy with the scan speed at different laser powers. As the scan speed increases, the relative density decreases, whereas it increases with increasing laser power. As seen from Figure 3, the number of irregularly shaped voids formed in the LPBF-processed 2024 alloys increases significantly with increasing scan speed, leading to a low densification level at a high scan speed (Fig. 5). With increasing the laser power from 200 to 400 W, the porosity of the 2024 alloys prepared at high scan speeds of 400 and 800 mm/s decreases from ∼3.5% to ∼1.5% and from ∼6.4% to ∼2.4%, respectively (Figs. 3a and 6a), exactly corresponding to an increased relative density. In contrast, the porosity of the 2024 alloys fabricated at low scan speeds of 100 mm/s and 200 mm/s increases from ∼0.2% to ∼3.1% and from ∼0.2% to ∼2.4%, respectively, as the laser power increases (Figs. 3a and 6a).

FIG. 5.

Variations of the relative density of the 2024 alloys processed at different laser powers and scan speeds.

FIG. 6.

(a) Porosity and (b) crack density of the 2024 alloys fabricated at different laser powers and scan speeds, (c) processing parameter optimization map of metallurgical defects.

Instead, their relative densities increase. This is probably because a high laser energy density promotes evaporation of Al and Mg in the as-fabricated specimens that have low melting point and low density,²³ leading to an increase in the theoretical density of the 2024 alloys. Nevertheless, the effect of element vaporization on the theoretical density is ignored in the calculation of the relative density. It is reasonable to expect that the calculated relative density of the specimens in the latter case is much higher than the estimated values.

Cracks are observed in the LPBF-processed 2024 alloys, except for those specimens prepared at a low scan speed (Fig. 3). Figure 6b shows the crack density of as-fabricated specimens at different laser processing parameters. The crack length of the specimens increases with increasing scan speed. At a high laser power of 400 W, the crack density increases from ∼17 to ∼2550 μm/mm², when the scan speed increases from 100 to 800 mm/s (Fig. 6a). This is attributed to a shortened time the liquid exists in the molten pool at high scan speed, limiting the time the liquid has to sufficiently fill gaps between dendrites, thus promoting cracking.

Figure 6c illustrates the different types of defects observed for the as-fabricated specimens for varying laser processing parameters. The specimens fabricated at low laser powers of 200 and 300 W and low scan speeds of 100 and 200 mm/s show favorable processability with few cracks and pores. Large near-spherical pores appear in the specimens fabricated at a high laser power of 400 W and a low scan speed of 100–200 mm/s, whereas long cracks and lack-of-fusion defects are observed in the specimens fabricated at a high scan speed of ≥400 mm/s. These findings show that the laser processing parameters are the key factors responsible for the formation of metallurgical defects in the LPBF-processed 2024 alloys. It is well established that good laser processability and high manufacturing quality for 2024 alloys can be achieved at low laser powers of 200–300 W and low scan speeds of 100–200 mm/s.

Effect of laser processing parameters on microstructure development

As the scan speed is a dominant factor influencing the formation of cracks,³⁵ sample A and sample C fabricated at the same laser power that show a large difference in crack density were selected to reveal the mechanism of cracking. Many columnar grains are observed in both types of samples (Fig. 7a, b). As a large temperature gradient perpendicular to the bottom of the molten pool develops in the LPBF process,²² columnar grains tend to grow in this direction upon solidification. These columnar grains prefer to epitaxially grow along the grain orientation of the previously fabricated layer, resulting in the formation of coarse columnar grains across molten pools. It is worth noting that some equiaxed grains with random orientation are observed in sample A (Fig. 7c). This can be ascribed to high-intensity convection of the melt in the keyhole mode at a relatively high energy density, resulting in fragmentation of coarse columnar dendrites, disruption of their growth, and thereby producing equiaxed grains between them.³⁶ In contrast, the columnar grains of sample C show a strong <001> texture parallel to the BD (Fig. 7d).

FIG. 7.

(a, b) EBSD orientation maps, (c, d) inverse pole figures, (e, f) KAM maps, (g–i) misorientation angle distribution for the LPBF-processed 2024 alloy samples fabricated at different scan speeds and a laser power of 200 W: (a, c, e, g) sample A at a scan speed of 100 mm/s, (b, d, f, h) sample C at a scan speed of 400 mm/s. EBSD, electron backscattered diffraction; HAGBs, high-angle grain boundaries; KAM, Kernel average misorientation; LAGBs, low-angle grain boundaries.

A careful examination shows that the cracks in the sample C are formed at the boundaries of columnar grains (Fig. 7b). Large orientation differences are observed near these cracks distributed parallel to the BD and consistent with the crack direction (Fig. 7f), indicating that high plastic deformation at the columnar grains is correlated to the generation of cracks. On the contrary, the orientation differences in the equiaxed grains of the sample A are relatively small, deviating from the BD (Fig. 7e). Lower stresses through free movement of dendritic fragments induced in the keyhole mode effectively suppress the formation of cracks in this case.^20,37 These results indicate that coarse columnar grains initiate cracks in the specimens fabricated at a high scan speed of ≥400 mm/s, whereas a low scan speed facilitates crack suppression by crushing the coarse columnar grains.

Based on the misorientation angles of grains (Fig. 7g, h), two different types of boundaries were identified: high-angle grain boundaries (HAGBs) are defined as a misorientation angle over 15° (black) and low-angle grain boundaries are defined as a misorientation angle of 2°–15° (red). Figure 7i depicts the misorientation angle distribution for grain boundaries in LPBF-processed 2024 alloy samples fabricated at different scan speeds and a laser power of 200 W. With increasing scan speed from 100 to 400 mm/s, the proportion of HAGBs shows an increased trend from 53.7% to 58.0%. As the HAGBs are more sensitive to hot cracking,³⁸ cracking tends to preferentially occur at HAGBs. This well corresponds to the observation of cracking present in the samples fabricated at 400 mm/s.

Evaluation of solidification cracking susceptibility

Due to a wide solidification temperature range of 775–913 K (Fig. 8a) of 2024 alloys and high temperature gradients during the LPBF process, coarse columnar crystallites are preferentially formed in as-fabricated 2024 alloys.³⁹ A solid-liquid zone (namely mushy zone) with high viscosity and high susceptibility to cracks often appears in the final stage of solidification. This is the zone where long and narrow gaps between dendrites tend to form. The mushy zone can restrict the liquid to fill these gaps between dendrites thoroughly, and solidification shrinkage further expands these gaps, thereby forming solidification cracks in the specimens.⁴⁰ As proposed by the Feurer model,²¹ solidification cracking is prone to emerge in the samples if the solidification shrinkage rate is larger than the liquid backfill rate in the mushy zone (Fig. 8b). Therefore, the solidification cracking susceptibility (SCS) is expressed as follows⁴¹:

FIG. 8.

(a) Solidification curve of the 2024 alloys, (b) schematic diagram of solidification cracks formed during the LPBF processing of 2024 alloys.

S C S = \frac{V o l u m e t r i c s h r i n k a g e}{V o l u m e t r i c f l o w r a t e},

(1)

where $V o l u m e t r i c s h r i n k a g e$ is the solidification shrinkage rate, and the $V o l u m e t r i c f l o w r a t e$ is the liquid backfill rate. Solidification cracking appears at , otherwise no cracking is observed.

To further reveal the relationship between cracking susceptibility and laser processing parameters, the thermal and solidification behaviors of the molten pool are simulated using FEM. Figure 9a and b show the temperature field of the molten pool. Apparently, the simulated molten pool has a good fit with the corresponding OM image (Fig. 9c). This illustrates that the simulation results reflect the thermal behavior of the molten pool in the LPBF processing of 2024 alloys.

FIG. 9.

(a) The simulated temperature field of the molten pool, (b) the magnified temperature field of the molten pool, (c) comparison of the simulated temperature field and the cross-section of a molten pool. (d) $L^{2} Ṫ^{1.6}$ of the LPBF-processed 2024 alloys at different laser powers and scan speeds. As $L^{2} Ṫ^{1.6}$ is larger than a critical value of 0.83, solidification cracks are formed in the mushy zone.

The length of the mushy zone L is the distance between the isothermal curves of T = 776 K and T = 876 K along the centerline direction of a molten pool (Fig. 9b). In this zone, liquid and solid coexist and the dendrites form a complex network that hinders liquid backfill. Based on the simulation results, L and the average cooling rate $Ṫ$ are derived from the LPBF-processed 2024 alloys at different laser processing parameters, as listed in Table 2.

Table 2.

Mushy Zone Lengths and Cooling Rates of the 2024 Alloy Specimens Fabricated at Different Scan Speeds and Laser Powers

Samples	Mushy zone, length L (m)	Cooling rate, $Ṫ$ (K/s)	Samples	Mushy zone, length L (m)	Cooling rate, $Ṫ$ (K/s)
A	3.43 × 10⁻⁶	2.42 × 10⁶	G	4.51 × 10⁻⁶	7.46 × 10⁶
B	3.13 × 10⁻⁶	4.48 × 10⁶	H	5.47 × 10⁻⁶	11.3 × 10⁶
C	3.35 × 10⁻⁶	9.84 × 10⁶	I	6.43 × 10⁻⁶	1.24 × 10⁶
D	3.73 × 10⁻⁶	19.2 × 10⁶	J	8.01 × 10⁻⁶	3.41 × 10⁶
E	4.87 × 10⁻⁶	1.78 × 10⁶	K	7.51 × 10⁻⁶	5.12 × 10⁶
F	4.26 × 10⁻⁶	2.87 × 10⁶	L	7.12 × 10⁻⁶	9.09 × 10⁶

The solidification shrinkage rate and the liquid backfill rate are given by the following equations: $V o l u m e t r i c s h r i n k a g e = \dot{𝜀} \cdot V_{v o l} = \dot{𝜀} \cdot L \cdot S,$ (2) $V o l u m e t r i c f l o w r a t e = V_{z} \cdot S,$ (3)

where $V_{v o l}$ is the volume of the mushy zone, S is the cross-sectional area of the mushy zone, $V_{z}$ is the liquid flow rate along the backfill direction, L is the length of the mushy zone, and $\dot{𝜀}$ is the strain rate. The SCS can be obtained by inserting Equations (2) and (3) into Equation (1), $S C S = \frac{\dot{𝜀} \cdot L \cdot S}{V_{z} \cdot S} = \frac{\dot{𝜀} \cdot L}{V_{z}},$ (4)

where $\dot{𝜀}$ is given by $\dot{𝜀} = \frac{1}{3} φ (f_{s}) β Ṫ,$ (5)

where β is the thermal expansion coefficient of 2.77 × 10⁻⁵, and $φ (f_{s})$ is a function of $f_{s}$ . Since solidification cracking occurs at the stage of complete solidification, $φ (f_{s}) \approx φ (1) = 1$ . $V_{z}$ can be calculated from Darcy's law,⁴² $f_{l} V_{z} = \frac{- K}{η} \frac{Δ P}{Δ z},$ (6)

where $f_{l}$ is the liquid fraction, $Δ P$ is the liquid pressure drop, $η$ is the melt viscosity of 2.81 × 10⁻³ Pa·S, and $Δ z$ is the distance of liquid flow, which is equal to L. The permeability (K) is determined approximately by the Carman–Kozeny equation⁴³: $K = \frac{{λ_{2}}^{2}}{180} \frac{{f_{l}}^{3}}{{f_{s}}^{2}},$ (7)

where $λ_{2}$ , the arm spacing of secondary dendrite, is related to the cooling rate $Ṫ$ . It can be derived as follows: $λ_{2} = A Ṫ^{- n},$ (8)

where the parameter n depends on the heat flow during the LPBF process, evaluated as 0.43.⁴⁴ A is a parameter depending on the alloy composition, which can be obtained from the following equation⁴⁵: $A = 143.73 x^{- 0.356} \times 1 0^{- 6},$ (9)

where x is the mass fraction of Cu and A = 8.33 × 10⁻⁵. The pressure drop ( $Δ P$ ) equals to the stress ( $σ$ ) obtained from the creep law⁴⁶: $Δ P = σ = σ_{0} exp (γ f_{s}) exp (\frac{m Q}{R T}) {\dot{𝜀}}^{m},$ (10)

where $σ_{0} = 4.5$ Pa and $γ = 10.2$ are determined by the material. Q is the activation energy of 1.6 × 10⁵ J·mol⁻¹, R is the gas constant of 8.314 J·mol⁻¹·K⁻¹, and m is the sensitivity coefficient of strain rate (0.26). Substituting Equations (5)–(10) into yields: $L^{2} Ṫ^{1.6} < M (f_{s}),$ (11)

where $M (f_{s}) = \frac{3^{1 - m} \cdot A^{2}}{180 η β^{1 - m}} σ_{0} exp (γ f_{s}) exp (\frac{m Q}{R T}) \frac{{(1 - f_{s})}^{2}}{{f_{s}}^{2}}$ .

When $f_{s} = 0.98$ , a minimum $M (f_{s})$ is calculated to be 0.83. When $L^{2} Ṫ^{1.6} < 0.83$ , no solidification cracks are formed in the LPBF-processed 2024 alloys. The calculated $L^{2} Ṫ^{1.6}$ values of the 2024 alloys at different laser processing parameters are given in Figure 9d. Samples A, B, E, F, and I show a $L^{2} Ṫ^{1.6}$ value <0.83. In these samples, few solidification cracks form. The theoretical calculations agree well with the experimental observations (Figs. 3, 6b, and 9d), indicating that this calculation method accurately reflects the SCS of the 2024 alloys processed at different laser processing parameters.

Effect of laser processing parameters on mechanical properties

Figure 10 shows the microhardness and tensile properties of samples A, B, and C fabricated at different scan speeds. As the scan speed increases from 100 to 400 mm/s, the average microhardness of the LPBF-processed 2024 alloys changes only insignificantly. The microhardness of samples A, B, and C are 110 ± 4 HV_0.2, 119 ± 5 HV_0.2, and 117 ± 6 HV_0.2, respectively. It is worth noting that the hardness deviation of the specimens gradually increases with increasing scan speed. This is due to a high probability that the hardness indents touch on defects for these samples, resulting in distinct hardness fluctuations.

FIG. 10.

Mechanical properties of the LPBF-processed 2024 alloy samples fabricated at different scan speeds and a laser power of 200 W: (a) microhardness and ultimate tensile strength, (b) engineering tensile stress–strain curves. EL, elongation; UTS, ultimate tensile strength.

As the scan speed increases from 100 to 400 mm/s, the ultimate tensile strength (UTS) of the 2024 alloys significantly decreases (Fig. 10b). At a scan speed of 100 mm/s, the LPBF-processed 2024 specimens reach a UTS of 300 ± 15 MPa and an elongation of ∼3%. When the scan speed increases to 200 and 400 mm/s, the UTS of the 2024 specimens are 142 ± 7 and 74 ± 4 MPa, decreasing by ∼53% and ∼75%, respectively, whereas the elongation decreases to <1%. This is closely related to the formation of solidification cracks in these LPBF-processed 2024 alloys (Figs. 3 and 6b), where premature failure occurs during tensile deformation.^47,48

The fractography of the LPBF-processed 2024 alloys shows features of intergranular fracture (Fig. 11). Apparent columnar grains and a few dimples are observed on the fracture of sample A (Fig. 11a, b), indicating a ductile–brittle fracture. Close examination shows that the fracture cracks propagate along the columnar grain boundaries, implying that the columnar grain boundaries in the LPBF-processed 2024 alloys induce the initiation and propagation of cracks under tensile stresses.³⁶ Coarse dendrites are observed on the fracture surface of sample C (Fig. 11d). In this case, the growth rate and the liquid backfill rate are much smaller than dendrite shrinkage at the final stage of solidification, resulting in the formation of solidification cracks.^34,49 During tensile testing, these microsized solidification cracks propagate, leading to brittle fracture of the as-fabricated specimens.

FIG. 11.

Fractography of the LPBF-processed 2024 alloys fabricated at different scan speeds and a laser power of 200 W: (a, b) 100 mm/s, (c, d) 400 mm/s.

Conclusions

In the present work, 2024 alloys have been fabricated through LPBF at laser powers of 200–400 W and scan speeds of 100–800 mm/s. The effect of laser processing parameters on the densification behavior and the mechanical properties of the LPBF-processed 2024 alloys has been investigated. The following conclusions can be drawn: 1.

The development of porosity and cracking in the LPBF-processed 2024 alloys are closely correlated to the laser processing parameters. The density of pores and cracks increases with increasing laser power, while the number of lack-of-fusion defects increases with increasing scan speed. The LPBF-processed 2024 alloys reach a high manufacturing quality with ∼0.2% porosity and without obvious cracks at a laser power of 200 W and a scan speed of 100 mm/s.

Based on finite element simulation and theoretical calculation of the SCS, the formation of cracks can be accurately predicted for the 2024 alloys manufactured at different processing parameters. There are no obvious solidification cracks in the LPBF-processed 2024 alloys when the liquid backflow rate is larger than the solidification shrinkage rate at $L^{2} Ṫ^{1.6}$ <0.83.

At optimal processing parameters of 200 W and 100 mm/s, the LPBF-processed 2024 alloy specimens exhibit a hardness of 110 ± 4 HV_0.2, a UTS of 300 ± 15 MPa, and an elongation of ∼3% with ductile–brittle fracture characteristics.

Footnotes

Authors' Contributions

L.X.: Formal analysis, methodology, and writing—review and editing. Q.L.: Data curation, formal analysis, methodology, and writing—original draft. D.G.: Conceptualization, validation, and writing—review and editing. S.C. and H.Z.: Investigation, and formal analysis. I.K., B.S., K.G.P., and J.E.: Writing—review and editing.

Author Disclosure Statement

The authors declare that they have no known competitive economic interests or personal relationships that could affect the work reported in this article.

Funding Information

This work was supported by the National Key Research and Development Program (Grant No. 2019YFE0107000); The Basic Strengthening Program (Grant No. 2019-JCJQ-JJ-331); The National Natural Science Foundation of China (Grant Nos. 51735005, 52205382); The 15th Batch of “Six Talents Peaks” Innovative Talents Team Program (Grant No. TD-GDZB- 001); The Equipment Pre-research Field Fund (No. JZX7Y20210263400301); The Austrian Federal Ministry of Education, Science, and Research and the Chinese Ministry of Science and Technology are acknowledged for support in the frame of the Austrian-Sino Collaboration Program, Austria.

J.E. and B.S. are grateful for the support from the European Research Council under the Advanced Grant “INTELHYB-Next generation of complex metallic materials in intelligent hybrid structures” (Grant No. ERC-2013-ADG-340025) and under the ERC Proof-of-Concept Grant TriboMetGlass (Grant No. ERC-2019-PoC-862485), as well as the Austrian Science Fund (FWF) (Grant No. I3937-N36). K.G.P. gratefully acknowledges financial support from the Guangdong Natural Science Foundation (Grant No. 2020A1515011242), the High-End Foreign Experts Recruitment Program (Grant Nos. G2021163004L and G2021181007L), and the Guangdong International Science and Technology Cooperation Program (Grant No. 2021A0505050002).

References

, Shi

, Poprawe

, et al. Material-structure-performance integrated laser-metal additive manufacturing. Science, 2021; 372(6545):abg1487; doi: 10.1126/science.abg1487

, Guo

, Gu

, et al. Microstructure development, tribological property and underlying mechanism of laser additive manufactured submicro-TiB2 reinforced Al-based composites. J Alloys Compd, 2020; 819:152980; doi: 10.1016/j.jallcom.2019.152980

Tang

, Ummethala

, Suryanarayana

, et al. Additive manufacturing of aluminum-based metal matrix composites—A review. Adv Eng Mater, 2021; 23(7):1–17; doi: 10.1002/adem.202100053

, Ding

, Zhang

, et al. In-situ synthesis of aluminum matrix nanocomposites by selective laser melting of carbon nanotubes modified Al-Mg-Sc-Zr alloys. J Alloys Compd, 2022; 891:162047; doi: 10.1016/j.jallcom.2021.162047

Wang

, Muñiz-Lerma

, Sánchez-Mata

, et al. Microstructure and mechanical properties of stainless steel 316L vertical struts manufactured by laser powder bed fusion process. Mater Sci Eng A, 2018; 736:27–40; doi: 10.1016/j.msea.2018.08.069

Suryawanshi

, Prashanth

, Ramamurty

. Mechanical behavior of selective laser melted 316L stainless steel. Mater Sci Eng A, 2017; 696:113–121; doi: 10.1016/j.msea.2017.04.058

Pauly

, Löber

, Petters

, et al. Processing metallic glasses by selective laser melting. Mater Today, 2013; 16(1–2):37–41; doi: 10.1016/j.mattod.2013.01.018

Brika

, Letenneur

, Dion

, et al. Influence of particle morphology and size distribution on the powder flowability and laser powder bed fusion manufacturability of Ti-6Al-4V alloy. Addit Manuf, 2020; 31:100929; doi: 10.1016/j.addma.2019.100929

Singh

, Hameed

, Ummethala

, et al. Selective laser manufacturing of Ti-based alloys and composites: Impact of process parameters, application trends, and future prospects. Mater Today Adv, 2020; 8:100097; doi: 10.1016/j.mtadv.2020.100097

10.

Bose

, Ke

, Sahasrabudhe

, et al. Additive manufacturing of biomaterials. Prog Mater Sci, 2018; 93:45–111; doi: 10.1016/j.pmatsci.2017.08.003

11.

Gokcekaya

, Ishimoto

, Hibino

, et al. Unique crystallographic texture formation in inconel 718 by laser powder bed fusion and its effect on mechanical anisotropy. Acta Mater, 2021; 212:116876; doi: 10.1016/j.actamat.2021.116876

12.

Elahinia

, Hashemi

, Tabesh

, et al. Manufacturing and processing of NiTi implants: A review. Prog Mater Sci, 2012; 57(5):911–946; doi: 10.1016/j.pmatsci.2011.11.001

13.

Aboulkhair

, Simonelli

, Parry

, et al. 3D printing of aluminium alloys: Additive manufacturing of aluminium alloys using selective laser melting. Prog Mater Sci, 2019; 106:100578; doi: 10.1016/j.pmatsci.2019.100578

14.

Karimi

, Suryanarayana

, Okulov

, et al. Selective laser melting of Ti6Al4V: Effect of laser re-melting. Mater Sci Eng A, 2021; 805:140558; doi: 10.1016/j.msea.2020.140558

15.

Olakanmi

, Cochrane

, Dalgarno

. A review on selective laser sintering/melting (SLS/SLM) of aluminium alloy powders: Processing, microstructure, and properties. Prog Mater Sci, 2015; 74:401–477; doi: 10.1016/j.pmatsci.2015.03.002

16.

Thijs

, Kempen

, Kruth

, et al. Fine-structured aluminium products with controllable texture by selective laser melting of pre-alloyed AlSi10Mg powder. Acta Mater, 2013; 61(5):1809–1819; doi: 10.1016/J.ACTAMAT.2012.11.052

17.

, Ji

, Chen

, et al. Selective laser melting of nano-TiB2decorated AlSi10Mg alloy with high fracture strength and ductility. Acta Mater, 2017; 129:183–193; doi: 10.1016/j.actamat.2017.02.062

18.

Prashanth

, Scudino

, Klauss

, et al. Microstructure and mechanical properties of Al-12Si produced by selective laser melting: Effect of heat treatment. Mater Sci Eng A, 2014; 590:153–160; doi: 10.1016/j.msea.2013.10.023

19.

Wang

, Wang

. Advance on wrought aluminium alloys used for aeronautic and astronautic industry. Light Alloy Fabr Technol, 2013; 41(8):1007; doi: 10.13979/j.1007-7235.2013.08.001

20.

Mair

, Goettgens

, Rainer

, et al. Laser powder bed fusion of nano-CaB6 decorated 2024 aluminum alloy. J Alloys Compd, 2021; 863:158714; doi: 10.1016/j.jallcom.2021.158714

21.

Kou

. A criterion for cracking during solidification. Acta Mater, 2015; 88:366–374; doi: 10.1016/j.actamat.2015.01.034

22.

Zhang

, Zhu

, Qi

, et al. Selective laser melting of high strength Al-Cu-Mg alloys: Processing, microstructure and mechanical properties. Mater Sci Eng A, 2016; 656:47–54; doi: 10.1016/j.msea.2015.12.101

23.

Nie

, Zhang

, Zhu

, et al. Analysis of processing parameters and characteristics of selective laser melted high strength Al-Cu-Mg alloys: From single tracks to cubic samples. J Mater Process Technol, 2018; 256:69–77; doi: 10.1016/j.jmatprotec.2018.01.030

24.

Tan

, Liu

, Fan

, et al. Effect of processing parameters on the densification of an additively manufactured 2024 Al alloy. J Mater Sci Technol, 2020; 58:34–45; doi: 10.1016/j.jmst.2020.03.070

25.

Wang

, Gammer

, Brenne

, et al. Microstructure and mechanical properties of a heat-treatable Al-3.5Cu-1.5Mg-1Si alloy produced by selective laser melting. Mater Sci Eng A, 2018; 711:562–570; doi: 10.1016/j.msea.2017.11.063

26.

Karg

MCH

, Rasch

, Schmidt

, et al. Laser alloying advantages by dry coating metallic powder mixtures with SiOx nanoparticles. Nanomaterials, 2018; 8(10):862; doi: 10.3390/nano8100862

27.

Koutny

, Palousek

, Pantelejev

, et al. Influence of scanning strategies on processing of aluminum alloy EN AW 2618 using selective laser melting. Materials (Basel), 2018; 11(2):298; doi: 10.3390/ma11020298

28.

Prashanth

, Shakur Shahabi

, Attar

, et al. Production of high strength Al 85 Nd 8 Ni 5 Co 2 alloy by selective laser melting. Addit Manuf, 2015; 6:1–5; doi: 10.1016/j.addma.2015.01.001

29.

Prashanth

, Wang

. Additive manufacturing: Alloy design and process innovations. Materials (Basel), 2020; 13(3):3–4; doi: 10.3390/ma13030542

30.

Prashanth

. Design of next-generation alloys for additive manufacturing. Mater Des Process Commun, 2019; 1(4):19–22; doi: 10.1002/mdp2.50

31.

. Laser Additive Manufacturing (AM): Classification, Processing Philosophy, and Metallurgical Mechanisms. Springer: Berlin;. 2015.

32.

, Zhu

, Zhang

, et al. Selective laser melting of Al7050 powder: Melting mode transition and comparison of the characteristics between the keyhole and conduction mode. Mater Des, 2017; 135:257–266; doi: 10.1016/j.matdes.2017.09.014

33.

King

, Barth

, Castillo

, et al. Observation of keyhole-mode laser melting in laser powder-bed fusion additive manufacturing. J Mater Process Technol, 2014; 214(12):2915–2925; doi: 10.1016/j.jmatprotec.2014.06.005

34.

Stopyra

, Gruber

, Smolina

, et al. Laser powder bed fusion of AA7075 alloy: Influence of process parameters on porosity and hot cracking. Addit Manuf, 2020; 35:101270; doi: 10.1016/j.addma.2020.101270

35.

Prashanth

, Scudino

, Maity

, et al. Is the energy density a reliable parameter for materials synthesis by selective laser melting?. Mater Res Lett, 2017; 5(6):386–390; doi: 10.1080/21663831.2017.1299808

36.

Nie

, Chen

, Qi

, et al. Effect of defocusing distance on laser powder bed fusion of high strength Al–Cu–Mg–Mn alloy. Virtual Phys Prototyp, 2020; 15(3):325–339; doi: 10.1080/17452759.2020.1760895

37.

Martin

, Yahata

, Hundley

, et al. 3D printing of high-strength aluminium alloys. Nature, 2017; 549(7672):365–369; doi: 10.1038/nature23894

38.

Guo

, Zhang

, Yang

, et al. Cracking mechanism of hastelloy X superalloy during directed energy deposition additive manufacturing. Addit Manuf, 2022; 55:102792; doi: 10.1016/j.addma.2022.102792

39.

Lopez-Botello

, Martinez-Hernandez

, Ramírez

, et al. Two-dimensional simulation of grain structure growth within selective laser melted AA-2024. Mater Des, 2017; 113:369–376; doi: 10.1016/j.matdes.2016.10.031

40.

Nagira

, Yamashita

, Kamai

, et al. Time-resolved X-ray imaging of solidification cracking for Al-Cu alloy at the weld crater. Mater Charact, 2020; 167:110469; doi: 10.1016/j.matchar.2020.110469

41.

Sheikhi

, Malek Ghaini

, Assadi

. Solidification crack initiation and propagation in pulsed laser welding of wrought heat treatable aluminium alloy. Sci Technol Weld Join, 2014; 19(3):250–255; doi: 10.1179/1362171813Y.0000000190

42.

Rappaz

, Drezet

, Gremaud

. A new hot-tearing criterion. Metall Mater Trans A Phys Metall Mater Sci, 1999; 30(2):449–455; doi: 10.1007/s11661-999-0334-z

43.

Ahmad

, Combeau

, Desbiolles

, et al. Numerical simulation of macrosegregation: A comparison between finite volume method and finite element method predictions and a confrontation with experiments. Metall Mater Trans A Phys Metall Mater Sci, 1998; 29(2):617–630; doi: 10.1007/s11661-998-0143-9

44.

, Nie

, Qi

, et al. Cracking criterion for high strength Al–Cu alloys fabricated by selective laser melting. Addit Manuf, 2021; 37:101709; doi: 10.1016/j.addma.2020.101709

45.

Brice

, Dennis

. Cooling rate determination in additively manufactured aluminum alloy 2219. Metall Mater Trans A Phys Metall Mater Sci, 2015; 46(5):2304–2308; doi: 10.1007/s11661-015-2775-x

46.

Coniglio

, Cross

. Mechanisms for solidification crack initiation and growth in aluminum welding. Metall Mater Trans A Phys Metall Mater Sci, 2009; 40(11):2718–2728; doi: 10.1007/s11661-009-9964-4

47.

Wang

, Xie

, Li

, et al. Premature failure of an additively manufactured material. NPG Asia Mater, 2020; 12(1):30; doi: 10.1038/s41427-020-0212-0

48.

Wang

, Tang

, Scudino

, et al. Additive manufacturing of a martensitic Co–Cr–Mo alloy: Towards circumventing the strength–ductility trade-off. Addit Manuf, 2021; 37:1–14; doi: 10.1016/j.addma.2020.101725

49.

Tang

, Panwisawas

, Ghoussoub

, et al. Alloys-by-design: Application to new superalloys for additive manufacturing. Acta Mater, 2021; 202:417–436; doi: 10.1016/j.actamat.2020.09.023

Samples	Laser power (W)	Scan speed (mm/s)	Samples	Laser power (W)	Scan speed (mm/s)
A	200	100	G	300	400
B	200	200	H	300	800
C	200	400	I	400	100
D	200	800	J	400	200
E	300	100	K	400	400
F	300	200	L	400	800

Samples	Laser power (W)	Scan speed (mm/s)	Samples	Laser power (W)	Scan speed (mm/s)
A	200	100	G	300	400
B	200	200	H	300	800
C	200	400	I	400	100
D	200	800	J	400	200
E	300	100	K	400	400
F	300	200	L	400	800

Samples	Laser power (W)	Scan speed (mm/s)	Samples	Laser power (W)	Scan speed (mm/s)
A	200	100	G	300	400
B	200	200	H	300	800
C	200	400	I	400	100
D	200	800	J	400	200
E	300	100	K	400	400
F	300	200	L	400	800