Daan M. Maijer

Modeling of microporosity formation in A356 aluminum alloy casting

A numerical model for predicting microporosity formation in aluminum castings has been developed, which describes the redistribution of hydrogen between solid and liquid phases, the transport of hydrogen in liquid by diffusion, and Darcy flow in the mushy zone. For simulating the nucleation of hydrogen pores, the initial pore radius is assumed to be a function of the secondary dendrite arm spacing, whereas pore growth is based on the assumption that hydrogen activity within the pore and the liquid are in equilibrium. One of the key features of the model is that it uses a two-stage approach for porosity prediction. In the first stage, the volume fraction of porosity is calculated based on the reduced pressure, whereas, in the second stage, at fractions solid greater than the liquid encapsulation point, the fraction porosity is calculated based on the volume of liquid trapped within the continuous solid network, which is estimated using a correlation based on the Niyama parameter. The porosity model is used in conjunction with a thermal model solved using the commercial finite-element package ABAQUS. The parameters influencing the formation of microporosity are discussed including a means to describe the supersaturation of hydrogen necessary for pore nucleation. The model has been applied to examine the evolution of porosity in a series of experimental samples cast using unmodified A356 in which the initial hydrogen content was varied from 0.048 to 0.137 (cc/100 g). A comparison between the model predictions and the experimental measurements indicates good agreement in terms of the variation in porosity with distance from the chill and the variation resulting from initial hydrogen content.

Mathematical model of deformation and microstructural evolution during hot rolling of aluminium alloy 5083

Abstract A mathematical model to predict the through thickness temperature, strain and strain rate distributions during hot rolling and the subsequent microstructure evolution was developed using the commercial finite element package ABAQUS. Microstructure evolution predictions included the amount of recrystallisation through the thickness of the sheet based on its thermomechanical history during rolling and thermal history after rolling. The equations used to predict the microstructure evolution were based on semiempirical relationships found in the literature for a 5083 aluminium alloy. Validation of the model predictions was done using comprehensive experimental measurements which were conducted using the Corus research multimill, a pilot scale experimental rolling facility, in Ijmuiden, The Netherlands. The results indicate that the through thickness temperature and strain distribution predictions for the rolling operation are reasonable. Hence, the boundary conditions used in the finite element model adequately represent the interface heat transfer and friction conditions. Microstructure predictions using the literature based equations significantly underestimate the amount of recrystallisation occurring in the sheet. A sensitivity analysis indicates that the recrystallisation kinetics are extremely sensitive to the fitting parameters used in the microstructure equation, and that the gradient in the recrystallisation kinetics is the result of the temperature gradient experienced by the specimen during deformation.

Simulation of microporosity in A356 aluminium alloy castings

Abstract A numerical model for predicting microporosity formation in aluminium castings has been developed, in which the redistribution of hydrogen between solid and liquid phases, and Darcy flow in the mushy zone were taken into account. For simulating the nucleation and growth of hydrogen pores, the pore radius is assumed to be a function of the secondary dendrite arm spacing, whereas the pore growth assumes equilibrium hydrogen activity between pore, liquid and solid. One of the key features of the model is that it uses a two-stage approach to porosity prediction. In the first stage, the volume fraction of porosity is calculated based on the reduced pressure, whereas, in the second stage, at fractions solid greater than grain coalescence f SC, an increment in the fraction porosity is calculated based on the volume of liquid trapped within a continuous solid network, which is estimated using a correlation based on the Niyama criterion. The model has been implemented within the commercial software package, ABAQUS, which has been used as a platform to solve the thermal field. The numerical model has been applied to a simple cylinder-shaped test casting, and the simulated results have been evaluated by comparing with the experimental results.

Analytical solution of the tooling/workpiece contact interface shape during a flow forming operation

Flow forming involves complicated tooling/workpiece interactions. Purely analytical models of the tool contact area are difficult to formulate, resulting in numerical approaches that are case-specific. Provided are the details of an analytical model that describes the steady-state tooling/workpiece contact area allowing for easy modification of the dominant geometric variables. The assumptions made in formulating this analytical model are validated with experimental results attained from physical modelling. The analysis procedure can be extended to other rotary forming operations such as metal spinning, shear forming, thread rolling and crankshaft fillet rolling.

Analysis and modelling of a rotary forming process for cast aluminium alloy A356

Abstract Spinning of a common aluminium automotive casting alloy A356 (Al–7Si–0.3Mg) at elevated temperatures has been investigated experimentally with a novel industrial-scale apparatus. This has permitted the implementation of a fully coupled thermomechanical finite element model aimed at quantifying the processing history (stress, strain, strain-rate and temperature) and predicting the final geometry. The geometric predictions of this model have been compared directly to the geometry of the workpieces obtained experimentally. This study is novel in regards to both the size and shape of the component as well as the constitutive material representation employed. The model predictions are in reasonable agreement with experimental results for small deformations, but errors increase for large deformation conditions. The model has also enabled the characterization of the mechanical state which leads to a common spinning defect. Suggestions for improving the accuracy and robustness of the model to provide a predictive tool for industry are discussed.

The effect of water ejection and water incursion on the evolution of thermal field during the start-up phase of the direct chill casting process

Abstract A comprehensive mathematical model has been developed to describe heat transfer during the start-up phase of the direct chill casting process. The model, based on the commercial finite element package ABAQUS, includes primary cooling to the mould, secondary cooling to water and ingot base cooling. The algorithm used to account for secondary cooling to the water includes boiling curves that are a function of surface temperature, water flow rate and position relative to the point of water impingement. In addition, the secondary cooling algorithm accounts for water ejection, which can occur at low water flow rates (low heat extraction rates). The algorithm used to describe ingot base cooling, includes the drop in contact heat transfer due to base deformation (butt curl) and also the increase in heat transfer due to water incursion between the ingot base and the bottom block. The model has been verified against temperature measurements obtained from two 711×1680 mm AA5182 ingots, cast under different conditions (non-typical “cold” practice and non-typical “hot” practice). Ingot base deflection data has also been obtained for the two test castings. Comparison of the model predictions with the data collected from the embedded thermocouples indicates that a 2-D longitudinal model is capable of describing the flow of heat in the early stages of the casting process in a region close to the centre of the rolling face. A sensitivity analysis completed with the model has clearly identified the link between ingot base cooling and secondary water-cooling heat transfer during the start-up phase.

Application of a mathematical model to multipass hot deformation of aluminium alloy AA5083

Abstract Sheet metal forming operations are used extensively in industry to alter the shape of the metal through plastic deformation. A critical step in the sheet manufacturing process is hot rolling which reduces the thickness of the ingot and can significantly impact the final sheet properties based on the microstructure evolution during this operation. A two-dimensional mathematical model was developed and experimentally validated to simulate deformation and microstructure evolution during multipass hot rolling for an AA5083 aluminium alloy. The details of model development and experimental validation can be found in earlier work. In this article, the application of the validated model to further understand and optimise the material stored energy and ensuing microstructure during multipass hot rolling is described. Specifically, the model was employed to examine the effect of changing the number of rolling passes as well as strain partitioning during multipass rolling on the material stored energy and the resulting microstructure. Results indicate that the number of passes has a significant effect on the stored energy which increases as the number of passes increases. In addition within a multipass rolling schedule the way in which the strain is partitioned is also shown to have an effect on the stored energy with a decreasing strain/pass schedule providing the highest material stored energy after rolling is complete. In contrast an increasing strain/pass schedule provides the lowest stored energy in the material after rolling. This overall effect is attributed to the differences in strip temperature as the lowest exit temperature strip has the highest stored energy. The model was further utilised to generate operational curves to predict the material stored energy and subsequent recrystallisation under different rolling conditions, namely at different interpass times and total strains for various start deformation temperatures.

Development of a 3D Filling Model of Low-Pressure Die-Cast Aluminum Alloy Wheels

A two-phase computational fluid dynamics model of the low-pressure die-cast process for the production of A356 aluminum alloy wheels has been developed to predict the flow conditions during die filling. The filling model represents a 36-deg section of a production wheel, and was developed within the commercial finite-volume package, ANSYS CFX, assuming isothermal conditions. To fully understand the behavior of the free surface, a novel technique was developed to approximate the vent resistances as they impact on the development of a backpressure within the die cavity. The filling model was first validated against experimental data, and then was used to investigate the effects of venting conditions and pressure curves during die filling. It was found that vent resistance and vent location strongly affected die filling time, free surface topography, and air entrainment for a given pressure fill-curve. With regard to the pressure curve, the model revealed a strong relation between the pressure curve and the flow behavior in the hub, which is an area prone to defect formation.

Numerical determination of permeability of Al–Cu alloys using 3D geometry from X-ray microtomography

Abstract In the present study, the permeabilities of Al–15·5 wt-%Cu and Al–19·5 wt-%Cu have been determined numerically by solving the full Navier–Stokes equations on 3D geometries of interdendritic channels obtained by X-ray microtomography. The samples for X-ray microtomography imaging were sectioned from different heights in directionally solidified cylinders in order to obtain a variety of microstructures for each composition. The flow has been solved using a second order accurate finite volume method approach. The effect of employing uniform or unstructured mesh on the calculated permeability has been studied. The Marching–Cube triangulation method was used to extract accurate surfaces for unstructured volume meshing. Calculated values of permeability for this range of solid fractions show partial agreement with previous experiments. Observed deviations are analysed and ultimately attributed to experimental error associated with the difficulties of measuring permeability.

Effect of chill cooling conditions on cooling rate, microstructure and casting/chill interfacial heat transfer coefficient for sand cast A319 alloy

Abstract A combination of experiments and numerical analyses were used to examine the cooling conditions, solidification microstructure and interfacial heat transfer in A319 cast in a chilled wedge format. Both solid copper chills and water cooled chills, with and without a delay in water cooling, were examined in the study. Various chill preheats were also included. The goal of the investigation is to explore methods of limiting heat transfer during solidification directly beside the chill and increasing cooling rates during solidification away from the chill. Within the range of conditions examined in the study, chill preheat was found to have only a small effect on cooling rates between 5 and 50 mm from the chill/casting interface, pour superheat a moderate effect and water cooling a significant effect. In comparison to the results for the solid chill, the solidification time at 5 mm with water cooling applied at the beginning of mould filling is reduced from 56 to 15 s and at 50 mm from 588 to 93·5 s. Furthermore, the average cooling rate during solidification is increased from 1·9 to 7·06°C s−1 at 5 mm and from 0·18 to 1·13°C s−1 at 50 mm. At 50 mm, for example, the increased cooling rate achieved with water translates into a reduction in secondary dendrite arm spacing from 40 to 25 μm or ∼40%. Delaying the water cooling by 10 s facilitated slow cooling rates at 5 mm (similar to those achieved with a solid chill) and high cooling rates 50 mm from the chill. A temperature based correlation was found to be suitable for characterising the behaviour of the interfacial heat transfer coefficient in the solid shill castings, whereas a time based correlation was needed for the water cooled castings.