FORMABILITY OF 6XXX ALUMINUM SHEETS FOR AUTOMOTIVE APPLICATIONS
Exploring the link between microstructure and formability?
Exterior body materials for automotive applications must meet high standards across various disciplines, including surface quality, corrosion resistance, weldability, adhesive bonding capability, and strength. However, in the press shop, the primary requirement is that the sheets can be successfully formed, and the subsequent hemming process results in a flawless hem edge. Hemming is a joining process where the outer and inner parts of a component, such as a car hood, are joined together (Figure 1).
In the AluReport (Issue 03/2023) , various methods used by AMAG to characterize the formability of automotive sheet metal were discussed. In addition to standard laboratory tests such as forming limit analysis, bulge testing, Erichsen cupping, hole expansion, and various bending and hemming tests, a tool was commissioned in collaboration with the Research Group Tools & Forming at Graz University of Technology. This tool enables the testing of sheet performance under industrial conditions. Consequently, the effects of process optimizations on formability can be directly examined using a complex outer body component.
This article aims to demonstrate which microstructural components influence formability and how they can be optimally adjusted through chemistry and processing. Essentially, the key question is:
What levers does AMAG, as a supplier of aluminum sheets, have to improve the formability of the material?
To answer this question, it is crucial to understand the failure mechanisms that occur during forming and the processes that take place at the microstructural level. Deep drawing, stretch forming, and bending are essential processes in sheet metal forming. In pure deep drawing, there is no reduction in sheet thickness, whereas in stretch forming, the deformation occurs entirely through a thickness reduction. The process of forming complex body parts is commonly referred to as deep drawing, although it usually involves a combination of both stretch forming and deep drawing. For simplicity, the term "deep drawing" will also be used here.In deep drawing, failure is characterized by necking due to significant material thinning caused by plastic instability. Conversely, during bending, fracture occurs without prior necking once a critical local stress is exceeded, such as near large intermetallic particles. The failure mechanisms and microstructural influences differ in each case. Therefore, both cases will be considered separately in the following sections.
What kind of microstructure is required to achieve material with good deep drawing properties?
When a car body part is deep drawn, the material strengthens due to the formation of dislocations. The strain hardening rate dσ/dε is defined as the first derivative of the flow stress σ with respect to the true strain ε. If the strain hardening rate is high, the deformation proceeds in a stable manner. This means that when there is a local reduction in cross-section during forming, the affected area hardens again, which inhibits further deformation in that region until a uniform cross-section is restored. In general, however, the strain hardening rate decreases with increasing deformation. For example, in a tensile test, once the uniform elongation is reached, necking occurs, followed by fracture of the material. Therefore, a high and slowly decreasing strain hardening rate is essential to ensure a stable forming process and to delay the onset of necking. [1, 2]
The success of deep drawing a car body part is closely linked to the material’s strain hardening behavior, which involves the formation, rearrangement, and annihilation of dislocations during deformation. For instance, 3xxx series alloys can rearrange dislocations into cell-like structures through strong dynamic recovery processes, facilitated by easier cross-slip of dislocations (Figure 2a). In contrast, 6xxx series alloys exhibit planar slip behavior, where dislocations concentrate on specific slip planes (Figure 2b). This leads to a more rapid onset of plastic instability and material failure. [3]
One way to improve the formability of 6xxx-series alloys is to delay dynamic recovery to prevent a rapid decrease in the strain hardening rate. [4]A useful method for analyzing strain hardening behavior is the Kocks-Mecking plot, in which the strain hardening rate is plotted against the increase in flow stress. [5]
The slope of the Kocks-Mecking plot is directly related to dynamic recovery - meaning the steeper the slope, the more pronounced the dynamic recovery [4, 6], as shown in Figure 3. Solute atoms in the aluminum matrix, such as Si, Mg, or Cu, can influence the strain hardening behavior by reducing dynamic recovery. [7-10] This effect is explained either by a reduction in stacking fault energy with increasing solute content [11], or by the influence of solute atoms segregated at dislocation cores on local recovery events. [12]
To demonstrate the influence of solute atoms on strain hardening behavior, Kocks-Mecking plots of a 6016 alloy were recorded after various solution annealing times (15 s - 300 s). The results are compared in Figure 3. As the solution annealing time increases, the slope of the plot decreases due to the rising number of dissolved Si and Mg atoms in the aluminum matrix.
The gradually decreasing strain hardening rate induced by longer solution annealing times is expected to have a beneficial effect on formability. [4]In addition to strain hardening behavior, plastic anisotropy (r-value) is an important parameter for deep drawability, as it indicates the material’s resistance to thinning during deformation. The r-value, defined as the ratio of the true width strain to the true thickness strain, is typically determined in tensile tests within the range of 2% to 20% true strain. Aluminum generally has an r-value below 1, meaning the flow resistance in the thickness direction is lower than in the plane of the sheet. Consequently, aluminum is more prone to thinning compared to deep-drawing steels, which have r-values well above 1.
In addition to a high average r-value (rm), a low planar anisotropy (Δr) is particularly advantageous for cold forming, as it leads to more uniform flow of the sheet in all directions. [13] The r-value is closely related to the texture of the sheet, i.e., the presence of preferred crystallographic orientations. [14] To achieve a low Δr and a high rm, a weakly developed texture is desirable. In sheet production, this is primarily achieved through appropriate temperature control during hot rolling. [14-16]
What microstructure is required to obtain material suitable for hemming?
In the hemming process, outer and inner body panels are joined by a bending operation. The outer panel is bent around the inner panel, and no cracking must occur on the outer surface. Failure during hemming can be divided into four stages: (1) orange peel formation due to grain rotation, (2) formation of ridges parallel to the bending axis, (3) deepening of the ridges, and (4) failure due to shear fracture. [13]The formation and deepening of ridges result from the development of slip bands and, later, shear bands. The shear fracture follows the typical stages of ductile fracture: nucleation, growth, and coalescence of voids.
Figure 4 shows the formation of shear bands (a), which ultimately result in a ductile shear fracture (b), as well as void formation at primary phases (c). A key factor influencing hemmability is the orientation of the grains, i.e., the texture of the sheet. Numerous studies suggest that a pronounced cube texture has a beneficial effect on hemmability, as this orientation hinders the formation of shear bands. [15, 18-21]In addition to texture, large primary phases negatively affect hemmability, as they serve as sites for stress concentration and void formation. [22, 23]Furthermore, the material’s strength should be as low as possible during the hemming process. In particular, hemmability of 6xxx-series alloys decreases with increasing strength due to long storage times. To demonstrate the influence of texture on hemmability, test materials 6016-V1 and 6016-V2 - both with identical chemical composition but processed with different hot rolling parameters - were examined. The strength and elongation of the materials are comparable, but significant differences are observed in their r-values (see Table 1). Due to its lower Δr-value and higher rm-value, variant 6016-V1 exhibits better deep drawability compared to variant 6016-V2.
| Material | Rp0.2 [MPa] | Rm [MPa] | Ag [%] | A80 [%] | r0 | r90 | r90 | Δr | rm | n5 |
|---|---|---|---|---|---|---|---|---|---|---|
| 6016-V1 | 114 | 239 | 23 | 28 | 0,70 | 0,75 | 0,55 | 0,18 | 0,64 | 0,30 |
| 6016-V2 | 114 | 235 | 23 | 27 | 0,68 | 0,65 | 0,43 | 0,24 | 0,55 | 0,30 |
Table 1: Mechanical properties of materials 6016-V1 and 6016-V2 after 30 days of natural aging, measured at 90° to the rolling direction as the average of three measurements. The r-values (r0, r90 and r45 )hrough tensile tests conducted at 0°, 90°, and 45° to the rolling direction, respectively. The planar anisotropy Δr was calculated using Δr = (r0 -2r45+r90)/2 and the average r-value rm using rm = (r0 +2r45+r90)/4.
However, when examining the hem edges after 180° bending in Figure 5, material 6016-V2 shows better results. The amount of primary phases, grain size, and strength are identical in both variants. The decisive difference lies in the sheet texture, particularly in the proportion of the cube texture component. Material 6016-V1 has a cube texture fraction of 14 ± 0.3%, while 6016-V2 has a fraction of 17 ± 1%. These seemingly small differences lead to significant variations in r-value and hemmability. These results also clearly show that there are conflicting requirements regarding texture:
- For good deep drawability, the goal is a random texture to achieve a low Δr-value.
- For good hemmability, the goal is a high proportion of cube texture.
Therefore, the process parameters and chemical composition must be optimally adjusted depending on the specific application to ensure the best possible formability.
Customer Benefit:
Sources:
[1] A. K. Ghosh, Journal of Engineering Materials and Technology 1977, 99, 264.[2] A. K. Sachdev, MTA 1990, 21, 165.[3] R. E. Sanders, in Formability of metallic materials-2000 A.D (Eds: J. R. Newby, B. A. Niemeier), ASTM STP 753, p. 35-59. American Society for Testing and Materials; ASTM International. Philadelphia, Pa., W. Conshohocken, Pa. 1982.[4] W. J. Poole, J. D. Embury, D. J. Lloyd, in Fundamentals of aluminium metallurgy: Production, processing and applications (Eds: R. Lumley), p. 307-344. Woodhead Pub. Philadelphia, PA 2011.[5] U. F. Kocks, H. Mecking, Progress in Materials Science 2003, 48, 171.[6] J. D. Embury, W. J. Poole, D. J. Lloyd, MSF 2006, 519-521, 71.[7] C. G. Schmidt, A. K. Miller, Acta Metallurgica 1982, 30, 615.[8] Ø. Ryen, B. Holmedal, O. Nijs, E. Nes, E. Sjölander, H.-E. Ekström, Metall and Mat Trans A 2006, 37, 1999.[9] Y. Chen, M. Weyland, C. R. Hutchinson, Acta Materialia 2013, 61, 5877.[10] K. Marthinsen, E. Nes, Materials Science and Technology 2001, 17, 376.[11] A. S. Krausz, K. Krausz, Unified constitutive laws of plastic deformation. Academic Press, San Diego 1996.[12] A. Deschamps, Y. Bréchet, C. J. Necker, S. Saimoto, J. D. Embury, Materials Science and Engineering: A 1996, 207, 143.[13] Anwendungstechnologie Aluminium, 3rd ed., VDI-Buch (Eds: F. Ostermann). Springer Vieweg, Berlin, Heidelberg 2014.[14] O. Engler, J. Hirsch, Materials Science and Engineering: A 2002, 336, 249.[15] X. Zhan, J. Tang, W. Tu, W. Liu, L. Cao, L. Ye, X. Yuan, G. Cao, Y. Zhang, L. Ma, Journal of Materials Research and Technology 2024, 32, 318.[16] J. Zhao, X. Shi, M. Liu, Z. Liu, B. Luo, K. Yu, Q. Li, J. Ran, P. Zhao, Journal of Alloys and Compounds 2024, 970, 172525.[17] Z. Li, Z. Zhang, G. Zhou, P. Zhao, Z. Jia, W. J. Poole, Materials Science and Engineering: A 2021, 814, 141199.[18] W. Muhammad, A. P. Brahme, U. Ali, J. Hirsch, O. Engler, H. Aretz, J. Kang, R. K. Mishra, K. Inal, Materials Science and Engineering: A 2019, 754, 161.[19] W. Muhammad, J. Kang, A. P. Brahme, U. Ali, J. Hirsch, H.-J. Brinkman, O. Engler, R. K. Mishra, K. Inal, Materials Science and Engineering: A 2019, 753, 179.[20] H. Takeda, A. Hibino, K. Takata, Mater. Trans. 2010, 51, 614.[21] K. Zhang, Q. He, J. H. Rao, Y. Wang, R. Zhang, X. Yuan, W. Feng, A. Huang, Journal of Alloys and Compounds 2021, 853, 157081.[22] D. J. Lloyd, in Automotive alloys 1999: Proceedings of the symposium held during the 1999 TMS annual meeting in San Diego, California, February 28 - March 4, 1999 (Eds: S. K. Das), p. 211-221. TMS. Warrendale, Pa. 1999.[23] W. M. Garrison, N. R. Moody, Journal of Physics and Chemistry of Solids 1987, 48, 1035.
