On determining lost core viability in high-pressure die casting using Computational Continuum Mechanics

Sammanfattning: The subject of this thesis work is to investigate whether Computational ContinuumMechanics (CCM) can serve as a valuable tool for the casting engineerto determine a priori whether a housing concept with inlying geometries thatso far only exists in Computer Aided Design (CAD) will have the desired coolingperformance and will be manufacturable with an acceptable number ofrejects.As of spring 2019, no application in serial production of lost cores, i.e.cores that are destroyed during deforming, in high-pressure die casting isknown. The reason for this is believed to be the absence of an engineeringtool that can tell upfront whether a concept of casting and process combinedwill be viable. This thesis aims to ll precisely that void by presenting, implementingand testing a CCM model inside the OpenFOAM toolbox in order todetermine upfront whether a design of a housing will be manufacturable withlost cores. The two-phase ow of air and melt is modeled with the volumeof-uid-concept. Turbulence modeling is done via the Reynolds-Averaged-Navier-Stokes (RANS) approach, mostly using the Menter SST k-omega-model.An isotropic linear elastic model was assumed for the solid mechanics.Industrial operators and managers like short and easy to grasp conclusions. As it, however, turned out during the process of this research project, there is no clear and easy answer to the question whether salt cores in high-pressuredie casting are a viable concept and will lead to sound castings. First of all, it was proven that housings made with lost cores can improve the heat transfer capabilities of castings. It was possible to produce castingswith cores up to an impact velocity of 30 ms^-1. The impact velocity wasfound to be the most decisive parameter. But the reader should bear in mindthat this limit is only valid for the given setup. Each conguration has to betested with the introduced model separately. The slamming events at rstimpact of the melt were found to be not failure-critical if crack-free coresare used. It was also found that the approach of evaluating only the peakforce does not go far enough. Eects later in the process may have a moreimportant impact due to larger force-time integrals. Also, dierent from theoriginal assumptions, the heat transferred from the melt to the core maynot be neglected even though filling times are below 0.1 s. Dening generalnumerical constraints for conditions under which salt cores are a viabletechnology is very dicult as geometry alterations play an important roletoo. This underscores the power and usefulness of the presented model evenIfurther as the engineer is now capable of testing each setup individually.It soon became clear that a fully comprehensive model is still for futureresearchers to develop. It was found that it is not benecial to attach the shotsleeve to the casting model with currently available open-source CFD technology.The presented strategy in this thesis together with the developed CCMtools can therefore provide a powerful tool for the casting or CAD-engineerto decide case by case whether a concept for a casting will be producible ornot. The tools range from a limited CFD approach for evaluating only theforces to a fully coupled FSI methodology describing the core deformationover time. All models have been tested and validated with high-pressure diecasting experiments and are in line with previously published ndings withdeviations of 5-10 % at maximum.