To Si or not to Si

- That is the question.

The hot rolling process has a significant influence on the formability of Al-Mg-Si sheets (6xxx alloys) used in automotive bodywork. AMAG researchers collaborating with scientific partner institutes have identified a physical cause for this connection in the key AA6016 alloy: pure silicon precipitates grow at grain boundaries during and after hot rolling. If they grow too large, they do not disappear completely during solution heat treatment and reduce the material’s formability. To prevent this, AMAG uses a specifically developed material model that pre-calculates the growth of silicon precipitates. This computer-based model rapidly optimizes hot rolling pass schedules to achieve the best possible formability in sheets for the automotive industry.This successful development was built on interdisciplinary collaboration between representatives from AMAG and its scientific partners. While microscopy specialists from the Institute of Electron Microscopy and Nanoanalysis at Graz University of Technology (TU Graz) made the silicon precipitates “visible”, simulation experts from AMAG and the Institute of Materials Science and Technology at Graz University of Technology (TU Wien) developed a growth model for Si precipitates, and technologists from AMAG and metallurgists from the Institute of Nonferrous Metallurgy at MU Leoben contribute expertise in alloys and processes. Some results of this collaborative research were presented at the EUROMAT conference in Frankfurt [1]. The combined use of material models, microstructure analysis and metallurgical expertise saved industrial scale trial-and-error processes and helped shorten the development period.

Technological significance

Figure 1 shows a comparison between three microstructures of an AA6016 automotive alloy following hot rolling with three different process routes: A, B and C.The differences in the distribution and size of pure silicon precipitates, shown here in black, are clear to see. The microstructure produced by process route A, in the left-hand image, contains coarse precipitates with a diameter of up to 2.0 micrometers as well as numerous fine precipitates with a diameter of less than several hundred nanometers. Process route B produces a microstructure with no coarse precipitates but with a higher density of small precipitates. By contrast, process route C suppressed the small precipitates, leaving only coarse precipitates. The manifestation of Si precipitates depends on the process step in which they formed. In process route A, precipitation formation occurs following homogenization; in process route B, it occurs following the plate pass, and in process route C it occurs after coil pass. When it comes to formability, such as drawing down sheet metal parts, it is vital that the precipitates following hot rolling are not too large because they cannot be dissolved down during solution heat treatment.However, a certain size of precipitate is beneficial, because coarse precipitates serve as nucleation sites for recrystallization after cold rolling and therefore help to achieve consistent mechanical properties in the sheet - to be precise, through elevated r-values.

si-oder-nicht-si-abbildung-1 — **Figure 1:** Microstructures of an AA6016 automotive alloy following hot rolling with different process routes. In this images, produced using an electron microscope, silicone precipitates are shown in black.

Scientific background

Whether or not pure silicon precipitates form in an Al-Mg-Si alloy is primarily determined by the ratio of Mg to Si. Pure silicon is only thermodynamically stable starting at an Mg-Si ratio of one to two. It is therefore not found in most alloys in the 6xxx family, which have a higher ratio.In these cases, beta Mg₂Si dominates in the thermodynamic equilibrium. The 6016 alloy, which is widely used in the European automotive industry has a relatively low Mg:Si ratio of 1:3. The higher proportion of silicon accelerates the hardening process and has a positive impact on fracture strain. Until now, however, the scientific community had not identified that this can lead to the growth of silicone precipitates during hot rolling.

One possible reason for this blind spot is that silicon precipitates are not easily identifiable with traditional electron microscopy methods. Ultimately, everything hinges on proper sample preparation.Figure 2 shows SEM images taken with the secondary electron (SE) detector and the energy dispersive spectroscopy (EDS) detector. Iron-containing primary phases, pure Si precipitates and isolated Mg₂Si characterize the image of this typical microstructure. It is clear to see that the largest silicon precipitates form at the grain boundary. Not only is this notable - because the Mg₂Si precipitates that typically appear in the hot rolling of Al-Mg-Si alloys predominantly form in intermetallic phases that contain iron and less often at grain boundaries - it is also technologically significant. Precipitates at grain boundaries and in grain bulk grow at different speeds or, more precisely, there is a difference in the driving force and diffusion paths of the dissolved elements that promote growth.

si-oder-nicht-si-abbildung-2 — **Figure 2:** Microstructure images of a hot-rolled, annealed AA6016 alloy taken with a scanning electron microscope (SEM): (a) The image from the secondary electron (SE) detector shows the Si precipitates in black and iron-containing prim

Diffusion_EN — **Figure 3:** Schematic depiction of the diffusion path in the growth of Si precipitates at the grain boundary. The Si atoms migrate from the super-saturated matrix directly to the grain boundary and then along the grain boundary to the ne

Figure 3 is a simplified depiction of the diffusion paths for precipitates at the grain boundary and in the grain bulk. In the case of grain boundary precipitates, a solute atom initially takes the shortest route to the grain boundary and then moves along the grain boundary until it reaches the next precipitation. The speed of diffusion along the grain boundary is so high that the time required to move along the grain boundary is negligible, which makes up for the detour via the grain boundary.

AMAG simulation model for the collector plate mechanism

This is known as the “collector plate mechanism” in specialist literature and imagines the grain boundary as a plate that collects the solute atoms from the super-saturated matrix. The R&D department at AMAG has produced a physical model of this collector-plate mechanism and adapted it to depict the issue of Si precipitate formation along the AA6016 process chain.

Kühlrate_EN — **Figure 4:** Calculated phase fraction of silicon precipitates within the grain (intragranular) and at the grain boundary (intergranular) as a function of the cooling rate.

The model recalculates hot rolling pass schedules and forecasts the precipitation status following hot rolling. The temperature profile is decisive in this context. The longer the cooling period, the more time is available for precipitate formation. If the cooling is sufficiently rapid, it can suppress precipitate formation. The calculations show, however, that this critical cooling rate for Si precipitates at grain boundaries is significantly higher than for Si precipitates that form in the grain bulk.Figure 4 shows the phase fraction of Si precipitates at the grain boundary (intergranular) and in the grain bulk (intragranular) depending on the cooling rate. While precipitates in the grain bulk can be suppressed at cooling rates of 0.1 K/s and higher, suppression of grain boundary precipitates is only possible at a cooling rate of 10-100 K/s. The picture changes once again at lower cooling rates of 0.01 K/s and below. In this case, precipitates in the grain bulk grow faster than those at the grain boundary. Another important factor in controlling Si precipitates, almost as significant than the cooling rate, is the grain size.

Korngrenze_EN — **Figure 5:** Calculated phase fraction of silicon precipitates in the grain bulk (intragranular) and at the grain boundary (intergranular) depending on the grain boundary density.

The influence of grain boundary density on the phase fraction of Si precipitates at the grain boundary and in the grain bulk is shown in Figure 5.At a fixed cooling rate, the model forecasts a direct proportional relationship between the phase fraction and grain boundary density. Lower grain sizes means higher grain boundary density and, as a result, shorter diffusion paths to the nearest grain boundary. Grain size and grain boundary density do not influence precipitates in the grain bulk because these are, by definition, sufficiently distant from the grain boundary.Consequently, process optimizations not only need to consider the temperature changes but also the development of the grain structure. Grain sizes and shapes change continuously during hot rolling due to deformation and the recrystallization between individual rolling passes. Without materials modeling, this complex interrelation between processing techniques and microstructure would be very difficult to forecast.

Benefits for customers: Microstructure modeling throughout the value chain and implementation in industrial practice

Material models are digital reproductions of real-life materials. Over the years, AMAG has increasingly come to depend on material models to support its process and alloy development activities. The areas of application range from precipitation formation and recrystallization to damage and strength, and are continuously being extended. AMAG’s activities in this context focus on physical models that not only allow for quantitative forecasts but also contribute to a better understanding of the fundamental mechanisms. In addition to commercial software and databases, AMAG researchers are working with scientific partners to develop customized solutions, such as for Si precipitates in AA6016. Experimental microstructure characterization is highly important. Proper validation and calibration increases the predictive power and makes models more robust. For the past year, AMAG has used DAMASK software, a highly advanced modeling method that describes the development of three-dimensional microstructures. If computer analysis has shown a better production method, AMAG has worked hard to implement it as effectively as possible in industrial practice. The company’s integrated Ranshofen site has an important part to play in this because it enables AMAG to monitor alloy composition and all casting, rolling and heat treatment parameters.

References

[1] G. Falkinger, A. Thum, R. Kahlenberg, M. Theissing, S. Mitsche, S. Pogatscher: ‘Modelling the concurrent growth of inter- and intragranular Si-Diamond precipitates during slow cooling of the alloy AA16’ EUROMAT, Frankfurt, 2023.

Back