Obtaining High-Strength Laser Welds of Aircraft-Grade Aluminium Alloys
Optimal laser welding modes were found that ensure the production of welds without defects in the form of open porosity, undercuts, cracks in the weld and the heat-affected zone. The postprocessing of welded joints obtained under optimal laser welding conditions was carried out on the basis of heat treatment (quenching + artificial aging). It is shown that, by changing the heat treatment regimes, it is possible to control the mechanical parameters of welded joint: strength and plasticity of the samples.
For the first time for welded joints obtained by laser welding and optimal post-processing for aluminum-lithium thermally processed alloys, mechanical characteristics were achieved comparable to the value for the alloy in delivery condition.
DOI: 10.22184/1993-7296.FRos.2019.13.4.356.366
Introduction
Currently, when creating aerospace technology, new high-strength aluminium alloys are being introduced.
New high-strength thermally strengthened, deformable alloys of various systems have been created, for example: Al-Mg-Li, Al-Cu-Mg-Li, Al-Cu-Li, Al-Mg. High mechanical signals are alloys of the Al-Cu-Li system, alloys of the Al-Mg-Li system have medium strength, are ultra-light and corrosion-resistant [1–3]. AMg6 alloy is widespread in aircraft industry. It has good corrosion resistance and is well processed by cutting and pressure.
In order to replace the rivet joint and reduce the weight of the structure, technologies for welding various data are actively developed: friction stir welding, laser welding, laser welding with filler wire, electron beam welding, welding with a floating and non-consumable electrode.
One of the promising methods of welding is laser welding. This makes it possible to obtain all the necessary results, which ensures high accuracy, high welding speed, low heat transfer, high flexibility and automation.
Laser welding of promising heat strengthened aluminium alloys for use in welding structures causes certain difficulties. Welds of these alloys, low welding properties. The tensile strength of the weld is 0.6–0.85 of the strength of the base material. In this case, it is necessary to consider it established that in order to increase the strength of the weld of heat-resistant aluminium alloys, additional mechanical and heat treatment of the weld should be carried out [4–7].
The following alloys were chosen as aluminium alloys: A5M, AMg6, 1420; 1424, 1441, B‑1461, B‑1469. It is necessary to note the presence of additional complexity, due to the fact that the main alloy. In this case, the base alloy must be softened, and its mechanical characteristics must be restored during the heat treatment process.
In this work, to continue the work [4–6] to improve the mechanical properties, welds, applied complex approaches, including laser welding in the optimal mode and after processing (quenching, intermediate plastic deformation, artificial aging) of welds. Mark A5M, AMg6, 1420, 1424 and 1441 and B‑1469 and respectively: Al, Al-Mg, Al-Mg-Li, Al-Cu-Mg-Li and Al-Cu-Li. For all stages of heat treatment, the influence of the chemical composition is determined, i. e. the main alloying elements: Mg and Cu, for determining the strength and microstructure obtained using optical and electron microscopes, analysis for the presence of chemical elements and phase composition using X‑ray phase analysis.
Methods and study techniques
Laser welding (LS) of aluminium alloys with a thickness of 1.6 mm was carried out on the Siberia‑1 automated laser technological complex, developed at ITAM SB RAS. Laser radiation was focused on the surface using a ZnSe lens with a focal length of 254 mm. Inert gas (helium) was used to protect the weld. Oxide film with a thickness of 0.15–0.20 mm. Immediately before welding, the edges of samples that have been cleaned to shine with a metal scraper.
The strength of welds was measured on a Zwick / Roell Z100 machine. The fact that the welding sample was not uniform in length, for comparison parts (it was determined by the movement of the moving beam). The macro- and microstructure of the welds were examined with an Olympus LEXT OLS3000 optical microscope. The determination of the chemical composition of the weld and the base alloy was carried out using an EVO MA 15 scanning electron microscope (Carl Zeiss, Germany) and an energy dispersive X‑ray spectrometer (Oxford Instruments X–Max 80 mm2, United Kingdom). Heat treatment was carried out in a Carbolite chamber furnace. Preparation of thin sections of samples was carried out on automatic cutting and polishing machines for sample preparation.
The study of the phase composition of the samples was carried out by powder X‑ray diffraction. Diffractograms were recorded in two modes: the reflection (D8 Advance diffractometer) and the clearance, which was carried out at the station of the Siberian Centre for Synchrotron Radiation VEPP 3 “Diffractometry with the energy of 33.7 keV”.
At the initial stage, optimal laser welding conditions were determined for aluminium alloy sheets with a thickness of 1–2 mm, during which no external weld defects were observed in the form of incomplete penetration of pores and cracks. The radiation power range was 2–3.5 kW, the radiation displacement speed was 0.6–4 m / min, the focus was deepened 1–2 mm deep from the top border – a dagger penetration mode.
Upon completion of the welding process, all samples were sent for cutting. Samples for testing the strength were made according to GOST 1497–84.
Part of the dumbbells was subjected to heat treatment: quenching in water or quenching in water and artificial aging – under various conditions. Heating before quenching was in the temperature range 300–550 °C, artificial aging in the temperature range 100–220 °C.
The influence of each thermal procedure on the change in mechanical characteristics, microstructure, distribution of elements was investigated. To compare the effect of heat treatment on the strength properties of the alloy and the weld, samples of the base alloy without a weld were also heat treated.
Main results of the experiments
Laser welding. Fig. 1 shows the strength of samples with a weld and without thermally non-reinforced A5M alloy. The strength of the samples within the statistical variation is equal to the initial strength of the alloy (in a wide range of variation of the velocity of the laser beam).
Fig. 2 shows the stress-strain relationship for the AMg6, 1424 alloy and the weld. As can be seen from Fig. 2, for the AMg6 alloy, equal-strength welds are obtained, the ratio of the tensile strength of the weld in relation to the main alloy is 0.95. For alloy 1424, the ultimate strength of the weld is 0.8 of the strength of the main alloy. This result characterizes the fundamental difference between thermally strengthened alloys and thermally strengthened alloys. For strengthened alloys, the situation is completely different, the strength drops significantly, and the mechanism depends on the quenching method, be it thermal or mechanical quenching. Table 1 shows the main mechanical characteristics of the samples with a welded seam and without a seam of the investigated heat-strengthened alloys, where σв0 is the tensile strength, σ0.2 is the yield strength, δ ductility. The notation k1,2,3 used is the ratio of these values of the weld / initial alloy, respectively.
The reason for such a sharp decrease in strength is due to the peculiarities of the structure of the material. The XRF results show that the basis of the solid solution of the initial thermally strengthened alloys is the α1-Al phase with the cubic structure Fm3m, with the inclusion, depending on the chemical composition, of the alloying elements of the basic strengthening intermetallic phases: δ′(Al3Li) and the ternary phases T1 (Al2CuLi) and S1 (Al2MgLi).
In the process of welding there is a fundamental change in the microstructure of the material in the weld. Fig. 3 and 4 show typical photographs of thin sections of the cross-section of the weld and main alloy of the samples obtained using a scanning electron microscope in the mode of back-scattered electrons. Images of thin sections obtained after etching. Uneven etching of the surface of the thin section indicates the uneven distribution of the main alloying elements in dendritic cells.
We note the presence of both general patterns in changing the microstructure and mechanical characteristics of the systems Al-Mg-Li, (alloy 1424) and Al-Cu-Li (alloy 1469) during the formation of the weld, and the presence of specific features caused by the difference in the interaction of the alloying elements Mg and Cu with solid solution Al. It is common to reduce the strength of the weld to the level of 0.55–0.75 from the strength of the starting alloy (see Table 1) and the formation of a large number of agglomerates inside the weld, i. e. concentration areas of alloying elements. This circumstance may be due to non-equilibrium crystallization with the release of the alloying elements of the intermetallic phases. At the implemented cooling rates of the weld, diffusion in the solid phase does not have time to pass, while in the liquid it can be quite complete, i. e. the composition of the crystals released from the liquid then slightly change. Their composition is mainly determined by the composition of the liquid phase at the moment when they formed. In this case, the crystallization begins at the maximum temperature of Ti with the release of crystals of the composition αi-Al, containing alloying elements that maximize the melting-crystallization temperature. As the temperature decreases, crystallization begins with the release of αj-Al crystals, which is determined by the residual composition of the liquid metal. As a result, in systems of the eutectic and peritectic types under such conditions of crystallization, anomalously oversaturated solid solutions of alloying elements and impurities are formed unevenly distributed over the volume of dendrites of the solid solution [8]. Furthermore, the phases in non-equilibrium eutectics and primary intermetallic compounds may be present in the cast metal.
For the Al-Mg-Li system, the XRPA results (Fig. 5.) show that the basis of the solid solution of the starting alloy is the α1-Al phase with the cubic structure Fm3m, with the inclusion of the main quenching intermetallic phase δ′(Al3Li) and the ternary phase S1(Al2MgLi), which forms chains of dark agglomerates, which are located mainly on the boundaries of dendritic grains for the alloy (see Fig. 4.b). Special measurements have shown that in a solid solution of an alloy inside a dendritic grain, particles of about 25–60 nm in size are δ′(Al3Li), and 100–300 nm S1(Al2MgLi).
The melting process in the weld destroys the microstructure of the original alloy. In the process of solidification of the melt in the weld, the formation of the ternary phase S1(Al2MgLi) occurs as a result of the peritectic reaction, and the particles of this phase are randomly arranged in a solid solution (see Fig. 4a). The phase δ′(Al3Li) is absent in the weld (see Fig. 5), which causes a decrease in the strength of the weld (see Table 1).
It is noted that as a result of separation of the ternary phase S1(Al2MgLi), the solid solution is depleted in magnesium, which leads to a decrease in the lattice period.
For the Al-Cu-Li system in the process of crystallization of the weld, the peripheral zones of the dendritic branches are enriched with elements that lower the melting point of aluminium, in particular, with copper. In addition, the cast metal may contain phases that are included in non-equilibrium eutectics, and intermetallic phases that interact with aluminium in a eutectic reaction are located along the boundaries of the dendritic cells. In our case, the formation of T1(Al2CuLi) phase agglomerates at the grain boundaries of the weld makes them contrast (see Fig. 3). Diffusion of copper from a solid solution leads to the formation of a pronounced dendritic structure with a cluster of intermetallic particles of the phase T1(Al2CuLi) at the dendrite boundary. X‑ray diffraction, including using synchrotron radiation, revealed a large number of T1(Al2CuLi) phase reflections (Fig. 6) in the weld, which causes a high reliability of phase identification in the weld. Namely, the diffusion of copper from the solid solution and, as a result, the localization of the quenching phase T1(Al2CuLi) at the border of dendritic grains causes a decrease in the strength of the weld of the 1469 alloy from 560 MPa for the alloy to the level of 306 MPa for the seam.
Post-processing as a quenching process
To achieve the maximum strength of thermally strengthened alloys, it is necessary with the help of regulated heating to dissolve impurity phases to obtain some intermediate non-equilibrium structure, which corresponds to the initial stages of decomposition of the supersaturated solid solution during which quenching phases are formed in the solid solution.
After quenching in the alloy of the Al-Mg-Li system, a significant dissolution of metastable S1(Al2MgLi) and partial dissolution of the quenching phase δ′(Al3Li) inside the dendrite occurred (see Fig. 4 g). This circumstance causes a decrease in the alloy strength after quenching from 512 MPa to 371 MPa. As a result of quenching, a significant dissolution of the metastable phase S1(Al2MgLi) was recorded in the weld (see. Fig. 4, c). The quenching phase δ′(Al3Li) is practically absent in it. Note that the amounts of phase S1(Al2MgLi) in the seam were at the sensitivity threshold of the method used by the XRF, which caused the signal to be recorded at the noise level or even the absence of a reflection of this phase on radiographs.
The alloying components found in the intermetallic phases (in particular, in the ternary phase S1(Al2MgLi) were completely or partially dissolved in aluminium.
Thus, as a result of quenching, an extremely nonequilibrium state was formed – a supersaturated solid solution of Mg and Li alloying elements in aluminium. In the alloy and the weld, it was possible to achieve a levelling of the mechanical characteristics and to obtain, at the optimum quenching temperature T = 480–540 °C, the maximum homogenization of the composition of the entire sample.
Given that the strength characteristics of the weld without TT and samples after quenching are close (σв = 370–390 MPa), it can be concluded that the triple phase S1(Al2MgLi) has a weak effect on strength, but the relative elongation increases. Given you water is consistent with the works [9–10], in which the effect of S1(Al2MgLi) on the relative elongation is also shown. However, the quenching process at high temperatures T = 480–540 °C is necessary to create a homogeneous supersaturated solid solution in the entire product (i. e., in the seam and in the initial alloy), in which the reinforcing phase δ′(Al3Li) is effectively formed during the aging process. that causes a significant increase in the strength characteristics to the level of the main alloy.
The quenching of Al-Cu-Li system samples also determines the alignment of the mechanical properties of the weld and the original alloy. The alloy strength after quenching decreases from 557 MPa to 385 MPa, however, unlike Mg alloyed alloys, the strength of the weld doped with Cu increases from 306 MPa to 384 MPa, that is, multidirectional processes occur in them. In the solid solution of the alloy, partial quenching phases dissolve, which is manifested in a change in the microstructure (see Fig. 3). At the same time, quenching leads to homogenization and the formation of a saturated solid solution in the weld. The copper concentration in the solid solution increases from 0.37 to 1.2% atomic%, which is determined by the dissolution of the phase T1(Al2CuLi) in white agglomerates at the boundaries of dendritic grains.
When homogenizing, the alloying elements are evenly distributed in the solid solution, which, in principle, opens up the possibility of forming the main reinforcing phases T1(Al2CuLi), δ′(Al3Li) of the Al-Cu-Li system. For the Al-Cu-Li system, carrying out a heat treatment procedure in the form of quenching increases the tensile strength of the weld by 80 MPa, which is apparently due to the formation of the T1(Al2CuLi) phase in a solid solution in the form of particles 10–30 nm in size, and the nanostructure of the welded weld and alloy after quenching steel are almost the same. The presence of the phase T1(Al2CuLi) after quenching is confirmed by measurements of diffraction on reflection (see Fig. 6).
Post-processing in the form of quenching and artificial aging
Artificial aging of samples of the Al-Mg-Li system alloy allows for the separation of the main quenching intermetallic phase δ′(Al3Li) both in the weld and in the alloy, which is confirmed by X‑ray diffraction by measuring and recording reflexes of this phase (see Fig.5) and measuring structures at the nanoscale. When this occurs, the ternary phase S1(Al2MgLi) is localized at the boundary of dendritic grains, partial in the weld and almost complete in the alloy (see Fig. 4 e, f).
From the data in Fig. 7, 8: the dependence of the stress σ on deformation δ under tension shows a significant effect of artificial aging, both on strength and plasticity of samples for Al-Mg-Li system alloys. With an increase in the temperature of aging from 120 to 170 °C, the values of strength σB increase from 420 to 495 MPa. Note that the strength of the weld was only 330 MPa. At the same time, the material plasticity significantly decreased. In the delivery mode, the relative elongation was δ = 18%, after quenching, the value δ = 14% and after aging it decreased with increasing temperature to the level δ = 4%.
Thus, choosing the heat treatment mode, it is possible to obtain samples, both with high plasticity and with great strength.
Optimization of heat treatment modes for heat-strengthened alloys of Al-Mg-Li and Al-Cu-Li systems allowed us to create samples of permanent joints obtained by laser welding with mechanical characteristics close to or equal to the original alloy in the delivery condition. The main parameters for the investigated alloys are given in table 2.
Conclusions
A complex technology has been developed for creating permanent welds for modern high-strength, thermally strengthened aluminium alloys, which includes laser welding and subsequent special heat treatment of samples.
Optimal laser welding modes were found that ensure the production of welds without defects in the form of open porosity, undercuts, cracks in the weld and the heat-affected zone.
The postprocessing of welds obtained under optimal laser welding conditions was carried out on the basis of heat treatment (quenching + artificial aging). It is shown that, by changing the heat treatment regimes, it is possible to control the mechanical parameters of the permanent welds being created: strength and plasticity of the samples.
For the first time for welds obtained by laser welding and optimal post-processing for aluminium-lithium thermally processed alloys, mechanical characteristics were achieved comparable to the value for the alloy in delivery condition.