The purpose of the present work is to understand the microstructure development and, particularly, to control the progress of recrystallisation in hot strip in the Al–Mg–Mn alloy AA 5454, which is typically used for the manufacture of structural automotive components. The chemical composition, together with the thermomechanical processing history of this material, has a strong influence on the microstructure of the product and the resulting properties as it is supplied to the customer. Electrical conductivity measurements, thermal analysis and electron microscopy have been carried out to characterise the evolution of precipitation state at various stages in the processing route. The conditions of the homogenisation heat treatment have been varied, and the effect on subsequent recrystallisation after hot rolling has been evaluated in both the as cast and rough rolled condition by optical microscopy techniques. Results indicate that the conditions of homogenisation heat treatment and roughing rolling are critical for the generation of a suitable recrystallised microstructure in AA 5454 hot strip. A new two stage homogenisation practice has been developed to expedite post-rolling recrystallisation in this alloy.

Computer-based alloy and process development requires integration of models for simulating the evolution of microstructure, microchemistry and crystallographic texture into process models of the thermo-mechanical production of Al sheet. The present paper focuses on recent developments in linking softening modules that simulate the progress of recovery and recrystallisation with the following texture changes to deformation and microchemistry models. The potential of such coupled simulations is illustrated by way of the thermo-mechanical processing of Al-Mg-Mn alloys. In particular, the progress of recrystallisation during coil cooling ("self-annealing") as well as the texture differences between production on a reversible rolling mill and a high-speed tandem line are explored.

In the last decade a significant market pull, especially from the automotive industry, has been observed for magnesium alloy sheet. Two main production routes for this material are currently under worldwide discussion; specifically, the classical route via DC casting plus hot rolling is competing with continuous casting techniques. A particularly flexible route for producing magnesium alloy sheet is provided by twin roll casting, followed by a hot rolling plus annealing sequence to the final gauge and temper.Early results and evaluations from trials with semiscale twin roll cast and hot rolled AZ31B are presented. Microstructures of as cast strip, as well as intermediate gauges and final thickness sheet, are discussed and related to production parameters such as casting speed and hot rolling process schemes. The promising mechanical properties which have been achieved are compared to those generated by a more conventional route and are discussed in terms of microstructural features, e.g., grain size and texture.

A through-process modeling study of texture development for AA5182 sheet production from hot rolling through cold rolling and annealing is reported. The thermomechanical process model was coupled to physics based microstructure models for deformation texture (GIA), work hardening (3IVM), nucleation, and recrystallization texture (StaRT). The simulations were run prior to any texture measurements, thus the results were fully predictive. The model overpredicts the Cube texture during hot rolling but properly predicts the terminal texture after multiple cold rolling with intermediate annealing. With a new concept of yield locus prediction the final predicted texture was fed into a FEM simulation of cup drawing, leading to good agreement with the measurement.

The previously presented Classical Nucleation and Growth model (ClaNG) of precipitation in aluminum alloys [1] was extended to describe simultaneous nucleation, growth and coarsening of several types of spherical precipitates for different heat treatments. It predicts the precipitation kinetics during annealing 1xxx, 5xxx and in particluar 3xxx series alloys. In order to describe the mentioned alloy systems with respect to phase diagram and latent heat over the whole range of temperature and concentrations, the model utilizes the commercial Gibbs energy minimizer ChemApp (GTT Technologies, Herzogenrath, Germany) [2] and thermodynamic databases which enables to calculate the chemical driving forces and equilibrium compositions. The main advantage of this strategy is no restriction to a special alloy system.

The principles of conventional and modem research and development methods applied in the Aluminium industry is presented. The new approach of integrated material and process modeling is described and discussed. Two examples are given for : i) The simulation of flow stress and strength evolution during rolling and annealing and ii) The simulation of texture evolution during sheet production, including the influence of material constitution and ingot homogenization variations on through process effects and final properties.

The texture evolution of commercially produced AA3103 has been investigated by means of plane strain compression (PSC) tests over a range of temperatures (250-400 degreesC), strains and strain rates which cover the range of conditions typical of industrial hot tandem rolling. In particular, the behaviour of Cube grains present prior to deformation has been studied. The centre plane textures have been measured and modelled with different approaches - a Taylor type model considering grain interaction (GIA), a viscoplastic self consistent model (VPSC) and a rate sensitive relaxed constraints (RCRS) Taylor model. Non-octahedral slip systems in different combinations with variations of resolved shear stresses were considered. Among the different predictions, the GIA model came closest to describing the Cube texture decrease. All simulation results lead to the indirect conclusion, that non-octahedral slip systems are not active at 250 C, but may be active at 400 C. It is shown that the activity of these slip systems based on single crystal experiments of Al-1%Mn should be adjusted for application to AA3103 polycrystals. The rate dependency of the textures was better reflected by the rate insensitive formulated, but flow stress dependent GIA model than by the two rate sensitive models. Detailed temperature dependencies of the plane strain texture components were not satisfactorily predicted by any of the models.

The authors present a model to describe the precipitation/dissolution kinetics during homogenization of multi-component alloys. The model is capable of describing the simultaneous nucleation, growth and coarsening of several types of spherical precipitates. To calculate the equilibrium phases, the chemical driving forces and the equilibrium concentrations, the commercial Gibbs energy minimizer ChemApp [1] (GTT Technologies, Herzogenrath, Germany) and a thermodynamic database (COST 507 [2]) have been embedded. The main advantage of this strategy is that there is nearly no restriction on a special alloy system. In contrast to simulations, taking into account local diffusion fluxes on a spatial grid (e.g. Cahn-Hilliard simulations), the present algorithm is fast and therefore of particular interest for industrial applications.

One of the challanges for the next generation of approaches for modelling of properties of aluminium sheet materials is to incorporate chemistry, in particular microchemistry effects. The paper describes and summarizes novel approaches - models and simulation techniques- for microstructure modelling with special focus on microchemistry effects. Methods and some selected results of coupling microchemistry effects to models for prediction of properties which are dedicated to be embedded into themomechanical models of rolling and annealing are presented. Ways to go in future to meet the vision of computer-based design of new alloys are discussed.

A kinetic model, based on the approach given by Cahn and Hilliard, was applied for a simulation of the formation and growth of Guinier–Preston (GP) zones in Al–Cu alloys. Thermodynamic data were provided by the CALPHAD method for the metastable phases to warrant for simulation of real Al–Cu alloys. The classic Cahn–Hilliard equation was extended to include the anisotropic elastic strain energy, which determines the morphology of GP zones. The model was applied to a simulation of spinodal decomposition cCu0=0.06 and precipitation of an alloy with concentration cCu0=0.02 which is beneath the maximum solubility. The latter simulation was carried out assuming pre-existing nuclei since the Cahn–Hilliard model does not include any mechanism for nucleation. The results (length and thickness of the precipitates) were in good agreement with experiments.

Two different experimental techniques, small angle neutron scattering and transmission electron microscopy studies were used to determine the size distribution of Guinier-Preston (GP) zones in binary A1-4 wt% Cu. Both methods complement each other with respect to the resolution of the two sizes, diameter and thickness, of the platelets. Both diameter and thickness followed during ageing at T = 423 K a power law behaviour with an exponent of n = 2 for short ageing times, when GP I-zones exist. At longer ageing times the precipitation process was dominated by the nucleation and growth of GP II-zones.

The nucleation and the growth period during the formation of GP-zones in an AlCu alloy is modelled with two different approaches. Both models are based on realistic thermodynamic data, complete each other and allow prediction of phase transformation kinetics scaled to real time.

In the current investigation the decomposed zones of AlCu-alloys (GP-zones) were simulated on the basis of the non-linear Cahn-Hilliard-Equation (CHE), an extended form of the diffusion-equations (2nd Fick's law). The CHE provides a basis to calculate the diffusion-field between precipitates. Originally, it was used in a linear form to describe spinodal decomposition analytically, but in this work the autors attempt to model the whole process of phase-transition, including coarsening, by means of a numerical solution of the CHE in two dimensions.