Lighter, faster, cheaper, and better- the old NASA mantra- has become the demand of all die casting customers. Achieving these goals has become a challenge because they require using a minimum of material in each application. This has driven the die casting industry to chase the thinnest wall castings possible. However, it is not always obvious whether a given thin wall section will fill or not.
RCM Industries has always relied on computer simulation to aid in designing thin wall castings, but historical approaches to predicting casting fill do not always work. The traditional approach involved looking at low temperature areas during filling. Areas that fill with cold metal are assumed to be the likely to not fill. The problem with this method is that it equally weights all low temperature regions. This means that the analysis cannot indicate whether or not an area will fill. It can only indicate where the casting will not fill if it does not fill. Knowing this is great for fixing in-process filling problems, but RCM wanted to take the next step and predict poor filling conditions before making any castings. RCM need to be able to tell their customers which areas of the casting can be a minimum thickness of 0.03, for example, and which areas should be thicker.
By working with MAGMA Foundry Technologies and University of Alabama researchers, an improved model was created to provide a more quantitative filling prediction of the fill risk. The new model recognizes that both the temperature and velocity should be considered to predict poor filling in die casting .
RCM Industries provided experimental data to evaluate the model. The experiments were based on an existing casting in production that is notoriously difficult to fill. Using the maximum fast shot plunger velocity of 150 in/s just fills the casting. Without the intensifier assisting, the overflows would not fill. To provide non-fill data RCM ran two additional cases where the maximum plunger velocity was 125 in/s and 100 in/s. Fast shot plunger velocities below 100 in/s were not possible because the poor fill resulted in large biscuits and stuck castings.
In the 125 in/s casting, poor fill conditions are clearly present. Using the fluidity model, MAGMASOFT can show areas most likely to suffer from these conditions. A comparison of the real casting with the MAGMASOFT prediction is shown in Figure 1. For this case the scale was set to show the color areas where the model predicts insufficient fluid life to fill the local area of the casting. Areas that are white should fill. Any colored areas indicate an increased likelihood of poor fill. The blue color correlates with the worst fill.
Figure 1: Comparison of simulation (top) and casting (bottom) for the 125 in/s case. White indicates good filling and blue colors correspond to worst filling condition.
Gross non-fill was clearly seen on the casting produced with a plunger velocity of 100 in/s. A direct comparison with the new filling model is shown in Figure 2. The model currently does not account for flow stopping. Instead, it estimates when the flow is likely to be disturbed by premature solidification. By comparing with the experimental casting from RCM, it became clear that the new model does best at identifying the onset of poor filling. This allows RCM to identify the hardest locations to fill on a given casting.
Figure 2: Comparison of simulation (top) with casting (bottom) for 100 in/s case. White indicates good filling and blue corresponds to worst filling. This industrial trial of the improved fill prediction at RCM has provided insight into what conditions cause poor filling in die castings.
For additional information on this study or further detail on how RCM Industries, Inc. may assist you in providing solutions to your part design challenges, please contact us at [email protected] or (847) 455-1950.
 Monroe, C., & Monroe, A. (2012). Predicting Flow Lengths in Die Casting Including Heat Advection. 2012 Die Casting Congress and Exposition (pp. T12-092). Indianapolis, IN: NADCA.
 Noble, A., Monroe, C., & Monroe, A. (2013). Predicting Flow Lengths in Die Casting in 3D. 2013 Die Casting Congress and Tabletop (pp. T13-101). Lousiville, KY: NADCA.