Multiphysical characterization of FSW of aluminum electrical busbars with copper ends
- 116 Downloads
This work investigates the benefits of having an aluminum (Al) busbar with welded copper (Cu) ends, and evaluates the force relaxation phenomena of a pre-loaded bolt joint on Cu versus Al, under cyclic thermal loading. The results show a force relaxation rate 50% lower in the Cu-bolted joint compared with the one in Al. The core of this research is the weldability analysis of Al-Cu butt joints made by friction stir welding (FSW). The materials are AA1050 H14/24 and Cu OF 04 with thickness of 6 mm. Temperature monitoring during the FSW cycle emphasize how heat generation depends mostly on local internal viscoplastic deformation. Tensile, bending, and microhardness tests were used to establish the mechanical properties. Optical microscope and scanning electron microscopy were used to characterize the microstructure. Joining mechanisms in the weld were investigated using energy-dispersive X-ray spectroscopy. The FSW resulted in 85% tensile strength efficiency compared to the Al base material, and 97% electrical conductivity efficiency compared to an ideal bimetallic component made of the same materials with no contact resistance. Electrical resistance of the FSW is 200 times lower than the electrical contact resistance between the Al-Cu materials while under high compressive force.
KeywordsFriction stir welding Aluminum Copper Busbar Electrical resistance Temperature Microstructure Intermetallic compounds Mechanical properties
Various industries constantly strive to improve their competitiveness, with higher performance products made with efficient processes with low environmental impact. Most of the developments are supported by material optimization demanding advanced solution for manufacturing, where joining of dissimilar engineering materials is typically the most challenging process. On this scope, the friction stir welding (FSW)  is a modern solid-state material joining technique, which opens up possibilities in manufacturing joints of dissimilar materials that are difficult, or even impossible, to do with conventional fusion welding methods. Due to high forces, the process is fully mechanized enabling high productivity for high production series, lowering the depended on skill requirements and thus cost of operation . In FSW, the heat is mostly generated by friction dissipation during the internal plastic deformation, and thus the materials do not reach their melting temperatures as the local heat generation is reduced to zero as the increasing temperature tends to melting [3, 4]. This means that when FSW of dissimilar materials, the local temperature is the one enabling the local viscoplastic deformation imposed by the rigid tool, clamping conditions, and process parameters. This same phenomenon is later emphasized in this paper, based on the results from the monitoring of the temperatures during the FSW of the Al-Cu joints. Good quality FSW bimetallic joints are now feasible for Al to steel [5, 6, 7] and Al to Cu . Avoiding common problems associated with fusion welding of dissimilar metals, such as mismatch of fusion temperature, formation of extensive brittle intermetallic compounds (IMC’s), gas solubility, and high distortion, and residual stress , the FSW opened up new possibilities in the design optimization and manufacturing of various products. In particular, electrical applications such as the busbars can get significant benefits from combining cheaper and lighter material, such as Al, with Cu that has lower electrical resistance, with more stable mechanical properties and corrosion resistance in a wider temperature range.
Busbars are conductive strips or bars used for short distance high current power transference. In recent years, the material choice for busbars has been changing from Cu to Al. This is due to the lower price and higher conductivity of the Al when considered on a per kilogram basis . The decrease in the direct cost and weight of the busbar is attractive, but perhaps short-sighted. High clamping forces are required for busbar connections to minimize the contact resistance between the busbars and other components . These high forces, along with the temperature, change the connection experience, while in operation slowly deform the pressed material, which in turn lowers the clamping forces. High thermal expansion and oxidation of the Al further degrade the electrical connection . To avoid component failure, the connection therefore needs to be retightened periodically. This results in maintenance costs, and increases the life cycle price of the Al busbar.
Elevated temperatures affect Al alloys commonly used for busbars such as AA6101-T4 more than high conductive Cu alloys like Cu-OF [12, 13]. The thermal expansion coefficient is also higher for Al than for Cu . These differences between the two materials might explain the higher maintenance associated with Al busbars. A potential solution to this problem is to use bimetallic busbars made from both materials, Al and Cu. A busbar mainly composed of Al but with Cu ends has the advantage of being cheaper and lighter than a monolithic Cu busbar but avoids the increase in maintenance cost associated with monolithic Al busbars. Additionally, Cu ends are more suitable for clamped connections with other Cu parts in an electrical system than Al/Cu- or Al/Al-bolted connections [15, 16]. For this solution to be viable, these bimetallic busbars need to be manufactured efficiently.
Difference in fusion temperatures, as between Cu and Al, is not a problem for the FSW, but several other issues remain contributing for the Al-Cu made by FSW to be a challenging joint, still demanding research on the influence of the FSW conditions on the joint properties. Among these issues are the different deformation behaviors, formation of IMCs even at low temperatures, and differences in physical properties and their evolution with temperature, promoting asymmetry in the flow of material and heat during the FSW . The IMCs are defined as solid phases containing two or more metallic elements, with optionally one or more non-metallic elements, whose crystal structure differs from that of the other constituents . They are generally very stable, brittle, and with a high fusion temperature, causing a problem in the welding of dissimilar materials, both in fusion and solid-state welding. In fusion welding, they are generally formed during the solidification of the welding pool. In FSW, they form under high pressures and intense plastic deformation  under the heating of the welded joint. In fusion welding of dissimilar materials, the amount of intermetallic compounds is so that it compromises the weld in almost every case and renders the welding method useless for many material combinations such as Al and Cu .
2 Evaluation of the clamping force relaxation
2.1 Experimental conditions and methods
Summary of the results of the clamping force relaxation test detailing the clamping force at different cycles as well as the average rate in clamping force between cycles 100–200 for each material
F @300 cycles
F @800 cycles
190 °C/60 °C
− 4.2 N
− 8.1 N
Plot of the clamping force versus time for both the AA6101-T4 and the Cu-OF-04. The detail focuses on the period of about 13 min, between the 50th and the 60th cycles. The relaxation rates exhibited by the Al component are about the double of the Cu end busbar component. The relaxation rates of the clamping force are higher at start of the test, when higher clamping forces are present. This phenomenon is probably induced by differences in yield strength at elevated temperatures and thermal expansion coefficients between the Al- and Cu-based materials. The AA6061 is probably overaged and loses its mechanical resistance in this range of temperature operation.
The threshold for preventive maintenance of the bolted clamping force will depend on the operation conditions and inherent interfacial electrical resistance for different levels of clamping force. These resistance values will be presented later in this paper. But immediately, whatever are the threshold criteria for maintenance operation, Cu ends for the busbar, will have a very significant benefit in increasing the preventive maintenance interval and overall safety of the electrical system operation.
3 Weldability analysis of the Al-Cu plate system
3.1 Experimental conditions and methods
The aluminum selected for the busbars was a rolled plate of AA1050-H14/H24, which is a highly conductive (σAA1050 ≅ 61% IACS) and commercial pure aluminum material (Al ≥ 99.5%). The Cu plate chosen is a high-purity, oxygen-free, non-phosphorus-deoxidized Cu alloy that does not contain any vacuum-evaporating elements. Generally, it is simply known as oxygen-free copper or Cu-OF (σCu-OF ≅ 100% IACS). The particular classification used during this work was Cu-OF-04. The thickness of all plates was 6 mm. The plates were prepared to be welded along the rolling direction (RD) and were cut with dimensions 250 mm (RD) × 60 mm. The hardness measured at top surface for the Al and Cu plates were HV05AA1050 ≅ 40 and HV05Cu-OF ≅ 88, respectively.
Welding conditions and parameters
Sixteen thermocouples K-type with Ø 0.8 mm were positioned in drilled holes with Ø1.0 mm at mid-plate thickness depth, i.e., 3 mm. Thermal paste was used for improved thermal conductivity and epoxy adhesive was applied to keep them stable. The thermocouples were positioned as represented in Fig. 6a. The strategical position was so that the thermocouples on each side were at same distance from the center of the stirred zone. The thermocouples were placed inside these holes. During the weld process, some of the thermocouples, especially on the Al side, were pushed away by the flash formed during the weld and their data is not considered for further evaluation (Fig. 9a, c).
Specimens for optical microscopy were polished using diamond paste down to 1 μm. Keller’s solution was used for the etching of the Al side while the Cu was etched using 100 mL of distilled water, 4 mL of saturated sodium chloric, 2 g of potassium dichromate, and 5 mL sulfuric acid. Optical micrographs were made with a Nikon Epiphot 200 microscope. SEM was made using a Zeiss Ultra 55 field emission scanning electron microscope. EDS line analysis was done using the same equipment.
Microhardness measurements of a cross section of the weld were made using a CSM microcombi tester. Four hundred fifty-one indentations were made with an indentation load of 0.5 N. The measurement matrix covered a 20 mm by 5 mm area containing the different weld zones and base materials present in the weld. The Oliver and Pharr  measurement method was used to determine the equivalent hardness Vickers of the indentations.
Mechanical testing (tensile and bending)
Electrical resistance testing
The contact electrical resistance between the Al and Cu base materials was also measured with no weld joint, using different contact forces. This research action enables to compare it with the welded condition, and to evaluate the impact of the clamping force relaxation of bolted joints in the operation of the busbars. The Al and Cu base materials were cut into samples having the same 5 mm × 6 mm cross section as the weld specimens and their contacting ends were milled. Then they were clamped together using a vice and the electrical resistance over the contact measured for low, medium, and high forces. The lowest force was so that the samples would stay in place but could easily be moved by hand. Medium force was about 25 kN, the level of force measured at the end of the clamping force relaxation of bolted joints (Fig. 4). The high force was about 40 kN, corresponding to the considered maximum clamping force of bolted joints.
4 Analysis of results
4.1 Temperature monitoring
Most of the tool is over the Al plate, due to the lateral 1.4 mm offset into the Al side;
The maximum temperatures reached in the Cu side are significantly higher than in Al side, especially for the thermocouples closer to the center of the stirred zone. As an example, the maximum temperature in Cu at 15.8 mm distance from the center of the stirred zone is 293.7 °C while the maximum temperature in Al at the same distance from the stirred zone is 219.5 °C.
These facts emphasize that differently from what some authors consider in their heat generation models [4, 25], the bulk of the heat is not generated due to the friction dissipation at the sliding interfaces, but in the internal energy dissipation inherent to the viscoplastic deformation of the material flow imposed by the tool geometry. As a matter of fact emphasized by the present results, the heat generated in the Cu is significantly higher than the heat generated in the Al, because for the Cu to reach the viscoplasticity, it dissipates more heat energy than the Al, because the Cu toughness is higher than the Al toughness. Moreover, the higher conductivity of the Cu reduces this temperature difference, as it homogenizes the thermal gradient faster than the aluminum. So, the monitored temperature difference is even more relevant due to the difference in thermal conductivity.
4.2 Microstructure analysis: OM, SEM, and EDS
Different Al-Cu interaction patterns can be identified such as intercalated lamellae in Fig. 11(2) and Fig. 12(c). This morphology usually consists of two or more IMC phases and the formation is deeply influenced by process parameters such as rotational speed and lateral tool offset as reported by Galvão et al.  and Liu et al. . Possible composite-like structures composed of Cu, Cu-rich, or IMC particles dispersed in an Al or Al-rich matrix are emphasized in Fig. 12(a). Interface layers of Al-Cu are focused in Fig. 11(4) and Fig. 12(b). The darker gray layer next to the bulk Al in Fig. 12(b) was measured, and found to have a thickness of 6 μm.
The evolution of the chemical composition obtained via EDS, at two Al-Cu interfaces with distinct patterns, is presented in Fig. 13 and 14. Based on the EDS, analysis is not possible to confirm the presence of IMC, but from the results of the chemical composition in Fig. 13, maybe Cu3Al2 (δ) and Al2O3 can be present. Also in Fig. 13, at the Cu base metal, from 28 to 42 μm, there is a solid solution of Al in Cu, similar to the one also reported by Galvão et al. . The line analysis in Fig. 14 depicts a mixed structure with multiple islands of Al and Cu along a distance from 70 to 250 μm. Thus, the multilayered structure at the Cu “tongue” near the root of the weld joint zone is well interlocked. The chemical composition changes in multilayered structure from 100% Al to about 70% Cu. This indicates that the weld is potentially strong mechanically. Again, although the EDS analysis is not conclusive in the identification of IMC, considering the maximum solubility of Al in Cu of about 19.7% of at. content , the transition interface near the Cu-rich zone may include the α phase peritectoid intermetallic compound AlCu4 (with at. Cu content of about 77%).
4.3 Hardness testing
4.4 Mechanical testing: tensile and bending loaded conditions
Tensile properties, with standard deviation, of the FSW joint, including the performance factor global efficiency to tensile strength (GETS)
90.54 ± 2
64.09 ± 2.9
90.14 ± 4.3
26.80 ± 4.8
4.32 ± 1.6
0.85 ± 0.1
Bending properties of the FSW joint, with the root and the face of the weld joint under tensile condition, including the performance factor global efficiency to bending (GEB)
Side under tensile
0.408 ± 0.02
4.5 Electrical resistance of the joint: welded and compressed not-welded conditions
Electrical resistance properties of the FSW joint versus a perfect electrical resistless joint
Joint resistance (μΩ)
Proportional increase (%)
0.55 ± 0.1
2.9 ± 5
0.97 ± 0.05
Contact electrical resistance (without weld joint) between 6 mm × 5 mm cross section surfaces of Cu-OF-04 and AA1050-H14/24 while subjected to various compressive contact force levels
Contact force (levels)
Contact resistance (μΩ)
~ 5 N (low)
8000 ± 2000
~ 25 kN (medium)
340 ± 30
~ 40 kN (high)
110 ± 30
The monitoring of the temperature disclosed that even though most of the tool is over the Al plate, due to the lateral 1.4 mm offset into the Al side, the maximum temperatures reached in the Cu side are significantly higher than in Al side. For example, at 15.8 mm distance from the center of the stirred zone, maximum temperature in Cu is 293.7 °C, while the maximum temperature in Al is 219.5 °C. These facts emphasize that the bulk of the heat is not generated due to the friction dissipation at the sliding interfaces, but in the internal energy dissipation inherent to the viscoplastic deformation of the material flow imposed by the tool geometry;
Metallurgical investigation of the joints shows an intense mixture of materials with large amounts of multilayered structures, both Al-matrix composite and intercalated lamellae. The Cu involves the aluminum by two “tongues” near the face and root, in a quasi “U” shape. Most of these microstructural features are possible to identify in the pattern of the hardness field;
In terms of mechanical properties compared to the Al component, the efficiency to tensile strength was 85%, and efficiency to bending was 41%;
FSW joints produced show a negligible electrical resistance compared to the resistance between clamped base materials. The electrical resistance of the FSW joint is 200 times lower than the contact resistance between the base materials while under the highest tested clamping force, of about 40 kN;
A significant operational benefit of Al busbar with FSWelded Cu ends was proved. The electrical contact resistance increases significantly with the drop of the clamping force and the Al ends present a significantly higher rate of relaxation of the clamping force compared with the Cu ends, e.g., two times faster for ΔF/cycle@100–200 first thermal cycles.
Open access funding provided by Aalto University.
- 1.Thomas W (1991) "“Friction stir butt welding,"” International Patent Application no.PCT/GB92/0220.Google Scholar
- 2.Lohwasser D, Chen Z (2009) Friction stir welding: from basics to applications. Elsevier, AmsterdamGoogle Scholar
- 12.W. S. Loewenthal and D. L. Ellis, "“Fabrication of GRCop-84 rocket thrust chambers,"” 2005.Google Scholar
- 13.Davis JR, Davis JR (1993) Aluminum and aluminum alloys. ASM international, ClevelandGoogle Scholar
- 15.Jackson R (1982) Electrical performance of aluminium/copper bolted joints. In: IEE Proceedings C (Generation, Transmission and Distribution), pp 177–184Google Scholar
- 18.Schulze GE (2013) Metallphysik: Ein Lehrbuch. Springer-Verlag, BerlinGoogle Scholar
- 21.Ólafsson D (2017) "“Friction stir welding of aluminum - copper"”. Aalto Master thesis.Google Scholar
- 24.Farrell T (2012) Chapter 2 - Measurement techniques. In: The Handbook of electrical resistivity: new materials and pressure effects by G. Dyos, The institution of Engineering and Technology, pp 11–24Google Scholar
Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.