This paper investigates the critical plunger velocity in high-pressure die casting during the slow phase of the piston motion and how it can be determined with computational fluid dynamics (CFD) in open source software. The melt-air system is modelled via an Eulerian volume-of-fluid approach, treating the air as a compressible perfect gas. The turbulence is treated via a Reynolds-averaged Navier Stokes (RANS) approach that uses the Menter SST k-ω model. Two different strategies for mesh motion are presented and compared against each other. The solver is validated via analytical models and empirical data. A method is then presented to determine the optimal velocity using a two-dimensional (2D) mesh. As a second step, it is then discussed how the results are in line with those obtained for an actual, industrially relevant, three-dimensional (3D) geometry that also includes the ingate system of the die.

This paper uses computational fluid dynamics (CFD), in the form of the OpenFOAM software package, to investigate the forces on the salt core in high-pressure die casting (HPDC) when being exposed to the impact of the inflowing melt in the die filling stage, with particular respect to the moment of first impact-commonly known as slamming. The melt-air system is modelled via an Eulerian volume-of-fluid approach, treating the air as a compressible perfect gas. The turbulence is treated via a Reynolds-averaged Navier Stokes (RANS) approach. The RNG k-epsilon and the Menter SST k-omega models are both evaluated, with the use of the latter ultimately being adopted for batch computations. A study of the effect of the Courant number, with a view to establishing mesh independence, indicates that meshes which are finer, and time steps that are smaller, than those previously employed for HPDC simulations are required to capture the effect of slamming on the core properly, with respect to existing analytical models and empirical measurements. As a second step, it is then discussed what response should be expected when this force, with its spike-like morphology and small force-time integral, impacts the core. It is found that the displacement of the core due to the spike in the force is so small that, even though the force is high in value, the bending stress inside the core remains below the critical limit for fracture. It can therefore be concluded that, when assuming homogeneous crack-free material conditions, the spike in the force is not failure-critical.

This paper investigates the fluid-structure interaction (FSI) that would be expected to occur when a lost core deforms in high-pressure die casting. A two-phase compressible Volume of Fluid approach is used to model the fluid. The turbulence contribution to the Navier-Stokes equations is accounted for by using the Reynolds-averaged Navier Stokes (RANS) Menter SST k−ω model, whilst an isotropic linear elastic model is assumed for the core material itself. The computed results for the core deformation were compared to those obtained for test bodies manufactured by high-pressure die casting, and good agreement was found. An interesting and surprising feature of both the experimental and theoretical results was that the core was found to bend in the direction opposite to that expected from intuition and to that obtained by an earlier model that did not use FSI.

In this work, the implementation of three turbulence models inside the open source C++ computational fluid dynamics (CFD) library OpenFOAM were tested in 2D and 3D to determine the viability of salt cores in high pressure die casting. A finite-volume and volume of fluid approach was used to model the two-phase flow of molten metal and air, with the latter being treated as compressible. Encouragingly, it is found that, although the choice of turbulence model seems to affect the dispersion of the two-phase interface, the force acting at the surface of the salt core depends only very weakly on the turbulence model used. The results were also compared against those obtained using the commercially available and widely-used casting software MAGMA(5).